Citation
MHC control of mercury-induced autoimmunity in mice

Material Information

Title:
MHC control of mercury-induced autoimmunity in mice
Creator:
Hanley, Gregory Alan, 1965-
Publication Date:
Language:
English
Physical Description:
vii, 78 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Antibodies ( jstor )
Antigens ( jstor )
Autoantibodies ( jstor )
Autoimmunity ( jstor )
B lymphocytes ( jstor )
Bone marrow ( jstor )
Epitopes ( jstor )
Haplotypes ( jstor )
Mice ( jstor )
Molecules ( jstor )
Autoimmunity -- chemically induced ( mesh )
Autoimmunity -- genetics ( mesh )
Department of Veterinary Medicine thesis Ph.D ( mesh )
Dissertations, Academic -- College of Veterinary Medicine -- Department of Veterinary Medicine -- UF ( mesh )
Down-Regulation ( mesh )
Genes, MHC Class II ( mesh )
Major Histocompatibility Complex -- genetics ( mesh )
Major Histocompatibility Complex -- physiology ( mesh )
Mercury -- pharmacology ( mesh )
Mercury -- toxicity ( mesh )
Mice ( mesh )
Research ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Bibliography: leaves 69-77.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Gregory Alan Hanley.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
029109464 ( ALEPH )
50988812 ( OCLC )

Downloads

This item has the following downloads:


Full Text










MHC CONTROL OF MERCURY-INDUCED AUTOIVMUMTY IN MICE













By

GREGORY ALAN HAN~LEY

















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998
























To Kim, my beloved spouse and greatest supporter. Without her encouragement and understanding, I would have never completed this work.














ACKNOWLEDGMENTS


I would like to express my extreme gratitude to my mentor, Dr. Eric Sobel, for his endless support. I thank him for the innumerable hours he devoted to critical discussions. I also wish to thank the three other members of my graduate advisory committee. To Dr. Schiffenbauer, I extend my sincere appreciation for his intellectual and monetary support. I wish to thank Dr. Roberts who was instrumental in directing me down this path. To Dr. Davis, I extend my sincere appreciation for his support in this endeavor, as well as, his continued guidance in my career development.

Several people have provided me with tremendous technical support, and without their help I would have been unable to complete this research. I owe a special thanks to Ed Butfiloski, not only was he an invaluable resource, but he also was instrumental in maintaining my sanity during this pursuit. I would also like to thank Raquel Baert for her technical assistance and scholarly discussions.

Thus work was supported by NRSA Grant 5 T RR07001 and NIH K08 AR 01934.














TABLE OF CONTENTS




ACKNOW LEDGMENTS .............................................. iii

A B STR A C T ........................................................ vi

CHAPTERS

1 BACKGROUND AND SIGNIFICANCE ................................. 1
Introduction .................................................... 1
R ole of T C ells .................................................. 4
C ytokines ...................................................... 5
B Cells / Antibodies / MIHC ...................................... 8
F ibrillarin ...... ........ ..... ... .... .... ... .... .. ..... ... ....... 9

2 CLASS II HAPLOTYPE DIFFERENTIALLY REGULATES IMMUNE
RESPONSE IN HgCl2-TREATED MICE ............................ 12
Introduction ................................................... 12
M aterials and M ethods ........................................... 14
M ice . . . . . . 14
H gC12 Treatm ent ......................................... 15
Indirect Immunofluorescence ................................ 16
A ntibodies .............................................. 16
Im m unizations ........................................... 17
Anti-HEL ELISA ......................................... 17
Flow Cytom etry .......................................... 18
Statistical Analysis ........................................ 19
R esu lts . . . .. . . . 19
F1 Mice Between a Resistant and Sensitive Haplotype Were Resistant to
AN oA Induction .................................... 19
The Ab Gene Product Conferred Resistance in F, Mice ............. 22
Expression of I-E Had No Effect on Susceptibility to HgCI2 ......... 25
Responsiveness to Exogenous Antigen Was Inherited Co-dominantly in F1 M i ce ............................................. 27
D iscussion .................................................... 27


iv










3 RESISTANCE TO HgCl2-1NDUCED AUTOIMMUNITY IN HAPLOTYPE
HETEROZYGOUS MICE IS AN INTRINSIC PROPERTY OF B CELLS... 34 Introduction ................................................... 34
M aterials and M ethods ........................................... 36
M ice .. .. . . 36
In V ivo Treatm ents ........................................ 37
Flow Cytom etry .......................................... 37
Indirect Immunofluorescence ................................ 38
Allotype-Specific ELISA ................................... 39
R esults . .. . 39
MHC-Restricted T Cell Help Was Required to Produce ANoAs ...... 39
Absence of ANoA Production in Haplotype-Heterozygous Mice Was Not Due to a Difference in Thymic Education ................. 41
Allogeneic Mixed Chimeras Developed ANoA Titers in Response to HgCl2 .............................................. 41
Lack of ANoA in Non-Responsive Mixed Chimeras Was Unrelated to B
Cell Com position ................................... 43
B Cells of Both B6.TC and B6.SJL Origin Were Functional in Group 3
M ixed Chim eras .................................... 44
Lack of HgCl2-Induced ANoA Response by Haplotype-Heterozygous Mice Was Not Due to Absence of Anti-Fibrillarin-Specific I-ARestricted T Cell Help ................................ 48
B Cell Composition Was Not a Determining Factor in Responsiveness
. . . 4 8
Both Bonors Provided Functional B Cells ....................... 50
Absence of ANoA Production in Haplotype-Heterozygous Mice Was Not Due to the Presence of I-Ab-Restricted T Cells ............. 50
D iscussion .................................................... 52

4 DISCUSSION/CONCLUSION ....................................... 58

LIST OF REFERENCES .............................................. 69

BIOGRAPHICAL SKETCH ............................................ 78









V














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy



MHC CONTROL OF MERCURY-INDUCED AUTOIMMUNITY IN MICE By

GREGORY ALAN HANLEY

May 1998

Chairperson: Stephen M. Roberts
Major Department: Veterinary Medicine

One of the most striking features of exposure to low doses of mercury in mice is the high-titer haplotype-linked anti-nucleolar (ANoA) autoantibody response. Mice of the H-2' haplotype have been high responders while H-2b mice have been low. This pattern has been attributed to the class II molecule itself, but the poor response ofF1 crosses between high and low responders raised the possibility that the anti-fibrillarin specificity was actually due to a closely linked dominant negative gene. To test the role of class II explicitly, F1 crosses between congenic B6.SJL (H-2s) and C57BL/6 (H-2b) mice with a targeted deletion of I-Apb were generated, creating mice heterozygous for all MHC loci, but expressing only I-As. In comparison with B6. SJL mice, no diminution of titers was found, proving that I-A itself was responsible for susceptibility and I-Ab for downregulation. Unlike I-A, expression of the I-E class II molecule could not vi








downregulate the response in an otherwise susceptible mouse. Using adoptive transfer techniques we have examined several mechanisms by which the resistant haplotype could be downregulating the anti-fibrillarin response in F1 (s/b) mice. Similar to other autoimmune models, mercury induced autoimmunity requires cognate MHC-restricted T cell help. The absence of autoantibody production in F1 mice was not due to either a difference in thymic education nor to the absence of anti-fibrillarin-specific T cell help. These results suggest a complicated role for class II in the regulation of a novel, environmentally-induced autoimmune response.































vii













CHAPTER 1
BACKGROUND AND SIGNIFICANCE


Introduction


Autoimmune disease occurs when the body mounts an immune response against self This immune response can involve either humoral and/or cellular immune pathways. Currently, the cause of autoimmunity is not known but both environmental as well as genetic factors appear to be important. The prevalence of autoimmune diseases has been estimated to be approximately 3% (1).

Autoimmune diseases can be classified as systemic or organ-specific. Examples of organ-specific autoimmune diseases include: type I diabetes, myasthenia gravis, Graves disease and multiple sclerosis. The systemic autoimmune diseases are represented by rheumatoid arthritis, scleroderma and systemic lupus erythematosus (SLE).

SLE is a multisystem disorder characterized by the presence of a variety of

autoantibodies that are responsible for many of its clinical manifestations. The prevalence rate, based on the total population, is one in 2000, but one in 700 for women between the ages of 20 and 64 years and one in 245 for black women in the same age group (2). The etiology of SLE is unknown but many factors have been identified that influence SLE. Genetic susceptibility is strongly suggested because of the high concordance rate between monozygotic twins. Certain MIC haplotypes are more frequently associated with the


1








2

disease as well. The fact that women are much more susceptible to SLE points to hormonal factors playing a role. Finally, several environmental factors have been associated with SLE flare ups.

Experimental animal models have been widely used to study autoimmune diseases. Some of the models develop spontaneous autoimmunity like Type I diabetes in NOD mice and BB rats. Spontaneous disease similar to SLE develops in (NZB x NZW)F1 mice. Induced models of autoimmune disease require manipulation of the animal which in turn causes an immune response to self antigens such as antigen-induced orchitis and uveoretinitis.

A second type of induced autoimmune disease can be elicited in animals without actually injecting either the autoantigen itself or a cross reactive antigen. For example, exposure to Hg2+, an environmental toxin, in certain strains of rodents results in autoimmune disease. Interestingly, humans also have been reported to develop membranous glomerulonephritis upon exposure to mercury in an industrial setting (3). Some of the characteristics of this disease are remarkably similar to human SLE.

HgC12-induced autoimmunity was first reported in mice by Robinson et al. in 1984

(4). Within one week after the start of treatment with HgC12, outbred Swiss ICR mice produced antinuclear antibodies (ANA) of the IgG subclass. Further analysis of the ANAs revealed a nucleolar pattern (antinucleolar antibodies; ANoA). Inbred A/J mice, on the other hand, failed to respond to HgC12.

Since this initial discovery, the link between susceptibility and genetics has been

investigated by several groups. Initially the link was established between ANoA induction








3

and the mouse major histocompatibility complex (MHC) (5-7). Testing of numerous MHC-congenic strains led to the following ranking of susceptibility based on haplotype: d < k < b < r < m < f < q < s (6). Further studies involving intra-H-2 recombinant haplotypes revealed that the I region of the MHC, specifically the I-A region, controlled the response (5,6). The other I region locus, I-E, has been shown to either decrease the response or have no effect on the ability of HgCl2 to induce ANoAs (6-8).

The response was shown to be co-dominantly inherited by intercrossing susceptible and resistant strains of mice (9). For example, crosses between susceptible SJL and resistant B6 mice resulted in an intermediate response with 41% of the progeny developing ANoAs (9). Other gene(s) outside the MEC have also been shown to affect susceptibility. The influence of these background genes was observed when comparing mice of the H-2" haplotype on an A background to those on a B 10 or a B6 background (5,9). Non-MEC genes on the A background enhanced susceptibility while those present on the B background decreased responsiveness. Within the B background, B6 were more suppressive than B 10.

Not all of the effects seen with HgCl2 administration are as tightly controlled by the MHC as the ANoA response. For example, A/J mice (H-2k) did develop antichromatin (ACA) and antihistone (AHA) antibodies despite being resistant to ANoA induction (7). The immune complex deposition seen in this model also does not appear to be under the influence of the H-2 complex. Balb/c (H-2d) mice developed renal mesangial and vessel wall immune complexes despite being completely resistant to ANoA induction (9,10). When only ANoA susceptible mice (H-2s'qP) were examined for renal IgG








4
deposition the results were similar in that only the H-2' mice showed significant immune complex deposition (11,12).

Tissue concentrations of mercury chloride has been examined in several strains of inbred mice (13). Most HgCl2 accumulates in the kidney primarily, with the liver being the second largest pool followed by the gastrointestinal tract, skin, spleen and testicles. The levels reached a steady state in the blood and liver by 4 weeks with the spleen and liver reaching steady state at 8 weeks. Analysis of thymi revealed steadily increasing concentrations throughout the 12 week treatment period. Interestingly, the authors also noted that when comparing the H-2-congenic susceptible B 10. S (H-2s) and resistant B 1O.D2 (H-2d) a correlation was detected between susceptibility to ANoA and accumulation of Hg2" within the spleen (13). Autometallography has revealed that within the tissues of the immune system mercury chloride accumulated in macrophages but was not found in non-phagocytic antigen presenting cells (follicular dendritic cells, interdigitating cells and B cells). At the subcellular level mercury was localized in the lysosomes of the macrophages (14).


Role of T Cells


The requirement for T cells in this model was shown by Hultman et al. using the nude (athymic) mutation (15). SJL/N mice homozygous for the nude mutation (nu/nu) failed to develop neither the ANoA nor the systemic immune complex deposits seen in littermates heterozygous for the nude mutation (nu/+). Furthermore, the specific subset of T cells required was elucidated by noting a lack of response in euthymic SJL/N mice given anti-CD4 monoclonal antibody in addition to the mercury chloride treatment.








5

When in vivo T cell responses to HgC12 were studied using flow cytometry in

susceptible A. SW mice, a long lasting increase in the number of T cells dominated by the CD4+ subset was observed which was absent in resistant A.TH mice (16). In contrast, van Vliet et al. (17) reported that both H-2' and H-2' mice expressed enhanced splenic numbers of T cells expressing the activation marker CD45RBIOCD4+.

Similar effects were seen when the ability of mercury to stimulate cells in vitro was studied. In vitro treatment of lymphocytes with HgCI2 resulted in a significant increase in DNA synthesis in A. SW mice while only minimal increases were seen in DBA/2 mice (5). Anti-CD4 antibody again completely abrogated the response suggesting the importance of T helper cells in this model. Interestingly, mercury-exposed lymphocytes from the low responder mice became high responders, as measured by [3H]Thymidine uptake, after the removal of excess mercury by washing (18). This activation of T cells in vitro led Jiang et al.(19) to examine whether mercury acted as a T cell mitogen or as a superantigen. They found a biased usage of TCR Vp subsets by CD4' cells which they interpreted as a superantigen effect.


Cytokines


Considering the effects mercury has on T cells, cytokines must play an important role in HgC12-induced autoimmunity. In vitro, the continuous presence of mercury induced interleukin-2 (1L-2) and interferon-y (y-IFN) but not IL-4 production from both high and low responder mice. In contrast, pretreating the cells with mercury and then washing it away resulted in production of JL-4 from both groups (18).








6

Cytokine profiles of mice treated with mercury has given somewhat conflicting results. An analysis of IL-4 mRNA from two H-2-congenic strains revealed a strong increase in production in splenic CD4' T cells of the H-2' mice while those of H-2d mice showed only a weak increase (17). This result, along with similar results seen in the rat model of HgC12-induced autoimmunity (20,21), led Goldman et al. (22) to propose that susceptibility relies on the initiation of a TH2 response while resistance correlates predominantly with a THI response.

Further evidence of this dichotomy was provided by the fact that HgCl2-induced a striking increase, up to 30-fold, of the level of IgE in susceptible A. SW but not in resistant B6, DBA/2 or B1O.D2 mice (17,23). The serum IgE levels are important because they, as well as, the increases seen in class II expression are both highly dependent on the TH2 cytokine IL-4 (24). The connection between IgE and 1L-4 was further strengthened in this model by Ochel et al. (25) who showed that treating susceptible A. SW mice with anti1L-4 antibody completely abrogated the increase in total IgE. Surprisingly, this treatment did not prevent ANoA production although it did influence the pattern of IgG subclass distribution.

Using antibody response to sheep erythrocytes (sRBC) in mercury treated H-2congenic strains Doth et al. (26) also proposed a preferential activation of either THI or TH2 cells based on haplotype. The response to sRBC antigens is normal in treated B 10.S mice while it is depressed in B1O.D2 animals. They showed, using monoclonal antibodies against y-IFN, that the suppression seen in B 10.D2 mice was mediated by y-IFN, a THI cytokine. Furthermore, treatment of B 1O.S mice with recombinant y-IFN suppressed their








7
response to the sRBC antigens, yet did not prevent the induction of ANoAs nor the immune complex glomerulonephritis.

An argument against the TH1/TH2 dichotomy in this model was proposed by Johansson et al. (16). They found that A.SW mice showed a modest increase of y-IFN and IL-4 producing cells while H-2-congenic A.TH mice showed no increase in cytokine producing cells. Remarkably, the susceptible SJL strain, despite being severely deficient in TH2-promoting CD4 NIK1.1' T cells, increased their number of y-IFN producing cells. This indicated that a predominantly TH2 response is not necessary for the induction of autoimmunity by mercury.

Recent studies conducted by Pollard et al. (27) provided the strongest evidence to date against the idea that a THI response corresponds with resistance while a TH2 response is required for autoimmunity induction. Using H-2s mice deficient by gene knockout for IL-4 and y-IFN, they showed that all of the features of autoimmunity are controlled by y-IFN, a THI cytokine. TL-4-1" mice did develop ANoAs while y-IFN/ were completely resistant. They concluded that the requirement for y-IFN suggests that antigen dose is the limiting factor in autoimmunity induction.

The idea of a dose response relationship was examined by Hultman et al. (28). They found a positive correlation between the dose of HgCl2, the total body burden of mercury, and the degree of autoimmunity expressed as the serum ANoA titer. The minimum observed adverse effect level (MOAEL) was 1.25 mg HgCI2 / liter drinking water (1.25 ppm). At this dose approximately 50% of the mice responded. No ANoAs were seen in mice given 0.625 ppm while 100% responded to 5 ppm. They estimated,








8

taking into account the absorption of inorganic mercury by the gastrointestinal tract, that the MOAEL is 7-14 jgg Hg/ kg bodyweight. This is much less than the standard dose of

1.6 mg/kg used in the literature.


B Cells / Antibodies / MHC


Mercury also has pronounced effects on B cells as evidenced by the increases seen in immunoglobulin and class II surface expression. For example, SJL mice developed splenic cell hyperplasia with a transiently increased number of cells secreting IgM, IgG and IgGI immunoglobulins, as well as increased serum immunoglobulin concentrations. Resistant B6 mice also showed similar increases although they were more short lived (29). A significant increase in anti-TNP producing cells seen in SJL mice but not in B6 mice provided evidence for polyclonal B cell activation in the former. H-2-congenic strains also show the differential effects on B cells of mercury administration. B1O.S mice showed significantly increased numbers of Ig producing splenic B cells of the IgGI (30-fold), IgG2a (7-fold) and IgE (5-fold) while B1O.D2 mice in contrast did not show any significant increases (17).

Mercury also enhances class II expression on B cells from both ANoA resistant B 1O.D2 and ANoA susceptible B 10. S mice although significantly higher levels were observed in the B1O.S mice (17). Treatment with an anti-IL-4 antibody completely prevented the increases seen in both strains. In addition to mercury influencing class II expression, class II molecules are required for some of the effects induced by mercury. In an in vitro system using monoclonal antibodies against class II, Hu et al. (30) were able to








9

completely abrogate the cytokine production and proliferation induced by mercury. The requirement for class II molecules for the induction of ANoAs in vivo has not been investigated. In chapter 2, we formally address this question using mice whose MHC I-A gene has been deleted.


Fibrillarin


The nucleolar autoantigen against which HgCl2-induced autoantibodies are

directed has been identified as fibrillarin, a 34 kDa U3 ribonucleoprotein involved in the first step ofpreribosomal RNA processing (31,32). It is interesting that up to 58% of human scleroderma patients spontaneously produce autoantibodies against the same U3 ribonucleoprotein (33,34). Several groups have compared the reactivities toward fibrillarin of mercury-induced ANoAs with human scleroderma sera and found that they were indistinguishable (22,34). Both groups of autoantibodies recognize the full length protein but lose reactivity, in all but the highest titrated human sera, if either the N- or Cregions were cleaved (35).

In an attempt to determine what the T cell antigen is, bulk T cells from HgCl2treated B 10. S mice were challenged with a variety of different self proteins that contained minute amounts of mercury. The T cells reacted to all the proteins in an anamnestic fashion (36). Bulk T cells obtained after one week of treatment also reacted anamnestically to Hg-complexed fibrillarin, whereas after eight weeks a T cell response against untreated fibrillarin predominated. These results suggested determinant spreading of T cell specificity. Consistent with this, analysis of T cell hybridomas established from








10

mercury-treated H-2' mice revealed two types of CD4' T cells: one that specifically recognized Hg-complexed fibrillarin and another that reacted to untreated fibrillarin (37). Interestingly, both the Hg-induced and native fibrillarin determinants were presented without the addition of fibrillarin when spleen cells from animals treated with mercury were used as antigen presenting cells. In contrast, spleen cells from untreated mice could activate fibrillarin-specific T cells only if exogenous fibrillarin was added. This indicated that HgCl2 can induce abnormal presentation by antigen presenting cells of a self protein.

In an attempt to understand the mechanism whereby mercury elicits an

autoantibody response that specifically targets fibrillarin, Pollard et al. (38) studied the antigenicity and molecular properties of fibrillarin from cells undergoing mercury-induced death. Under nonreducing SDS-PAGE conditions, fibrillarin from mercury treated nuclei showed aberrant migration as evidenced by its change in migration from 34 kDa to 32 kDa. This modified fibrillarin also lost its B cell antigenicity as measured by both indirect immunofluorescence and immunoprecipitation. If either one or both of the two cysteines which fibrillarin contains were mutated to an alanine, the aberrant migration was abolished. This combined with the fact that iodoacetamide, which alkylates cysteine residues, also abolished the aberrant migration led the authors to propose that fibrillarin forms a disulfide bond which then appears as the 32 kDa band. They concluded that unmodified fibrillarin is the B cell antigen and mercury-modified fibrillarin is the T cell antigen.

The mechanism whereby mercury can induce autoimmune disease is a complicated one. To gain a better understanding of the genetics behind the response we first








11

investigated whether the class IIll-A molecule itself or a closely linked gene(s) was a requisite for susceptibility. We also sought to determine what effects, if any, the other class II molecule, I-E, had on susceptibility. In our initial studies we unexpectedly found that susceptibility was not codominantly inherited as previously reported. This led us to perform adoptive transfer experiments to elucidate the mechanisms whereby the resistant haplotype could down-regulate the response. In addition, we used these chimeric mice to investigate the cellular interactions between B and T cells in this model.














CHAPTER 2
CLASS II HAPLOTYPE DIFFERENTIALLY REGULATES IMMUNE RESPONSE IN HgCl2-TREATED MICE


Introduction


Subtoxic doses of HgCl2 in animals induce an autoimmune disease characterized by autoantibody production and variable amounts of immune complex glomerulopathy (22). In mice, for example, the glomerulopathy is relatively mild and characterized by mesangial deposits of IgG and C3 (10). Other features, such as splenic hyperplasia with increased serum immunoglobulins (especially IgG and IgE) have also been described (23,29). However, the most striking finding in mice is the antinucleolar antibodies (ANoA) elicited in susceptible strains as soon as one week after the start of treatment (4). The principal nucleolar autoantigen has been identified as fibrillarin, a U3 ribonucleoprotein involved in the first step ofpreribosomsal RNA processing (31,39). Interestingly, this same specificity is seen in patients suffering from systemic sclerosis and other connective tissue disorders (33,34). Even more remarkably, a comparison of the specificity of antifibrillarin autoantibodies between human and murine origin showed similar, if not identical conformational epitopes (40).

The mechanism by which HgCl2 causes autoimmunity is unknown, but the fact that a human scleroderma-associated autoantibody specificity can be produced and that certain



12








13

toxic exposures have resulted in scleroderma-like illnesses (41,42) makes it an important and intriguing model. Both MHC class II-linked loci and unknown non-MHC loci govern susceptibility, with mice of the H-2s haplotype being high responders, while those of the H-2' haplotype are resistant or low responders (5). Non-H-2 genes in H-2S mice did not prevent the disease but had a pronounced effect on the antinucleolar antibody (ANoA) titers (12). By using intra-H-2 recombinants, it has been shown that responsiveness could be mapped to the I-A region of the MIC. The other class II locus, I-E, was shown to either suppress (6) or have no effect on the immune response (7). The class II-linked susceptibility gene has previously been reported to be dominant, although the disease was attenuated in F, animals when compared to homozygous animals (5,9,43). The fact that most of these studies used non-MHC-congenic strains, however, makes it difficult to interpret the effects of non-MuC genes and I-E expression on the ANoA response.

To investigate the control of the ANoA response by the I-A region to HgCl in the absence of non-MC loci differences we studied two MHC-congenic strains of mice on the B6 background. In addition to the susceptible H-2' haplotype, we chose to study H-2b rather than the more commonly used H-2d haplotype because of the availability of animals on the B6 background which do not express class II due to a targeted disruption of the IApb chain. In addition, the H-2' and H-2b haplotypes have both lost functional expression of I-E, allowing the effects of I-A and I-E to be isolated. In our initial studies, we unexpectedly found that F, progeny between B6. SJL (H-2s) and B6 (H-2b) mice were highly resistant to HgCl2 treatment and suggested that this might have been due to a dominant negative gene linked to I-A". Such a possibility is not without precedent.








14

Recently, the clinical manifestation of arthritis in the autoimmune model of collageninduced arthritis was shown not to be exclusively limited to a particular MHC class II antigen but may be due to linked genes which participate in intracellular loading and selection of antigenic peptides (44). To distinguish between these two possibilities we generated mice on the B6 background which expressed I-A! and carried the genes linked to the I-Ab region but did not express I-Ab. These mice were as susceptible to the induction of ANoA as H-2' homozygous mice, despite carrying the genes linked to I-Ab. We also examined the effects of the other class II molecule, I-E, in our system by developing B6. SJL mice which carried the I-E~d transgene (45), allowing the functional expression of I-E on the H-2s background. Contrary to a previously published report and to our experience with I-A, the expression of I-E did not alter the ANoA response. Thus, our experiments demonstrate that it is the expression of the resistant H-2 allele (I-Ab) itself that down-regulated the ANoA response ofF1 crosses between resistant and sensitive haplotypes and strongly suggest that it is the H-2' molecule itself that is specifically required. These results have important implications for understanding the mechanism by which HgCl2 exposure causes a specific autoimmune syndrome.


Materials and Methods


Mice


C57BL/6J (B6; I-Ab, I-E-, Ighb), C57BL/6J.SJL-H-2s (B6.SJL; I-A!, I-E', Ighb), C57BL/6J-Igha Thyl- Gpi' (B6.TC; I-Ab, I-E, Igh), A.BY/SnJ, A.SW/SnJ, SJL/J and Balb/cJ mice were obtained directly or from breeders purchased from The Jackson








15

Laboratory (Bar Harbor, ME). Tg(H-2I:Ea)Bri39 (B6.I-Ed) mice were a gift from R. L. Brinster (45). CD2Tm (MHC Class II deficient, B6 background) were purchased from Taconic (Germantown, NY). These mice lack any cell surface expression of I-A due to the insertion of a loss of function mutation into the I-Ab gene (46). B6.SJL.I-Ed mice were developed by crossing B6.SJL with B6.I-End and then backcrossing to B6.SJL. The B6. SJL.I-Ed mice were then selected using FACS analysis. These mice expressed I-E molecules consisting of a P-chain of the b allele and an a-chain of the d allele (47,48). B6.I-A^ and B6.I-A'" mice were developed by intercrossing CD2m( B6.I-A"/ ) with B6 and the resultant F1 bred to B6.SJL. The MHC class II expression was verified using FACS analysis. To rule out an allotype effect on susceptibility B6.SJL.Igha mice were developed by intercrossing B6.SJL and(B6.SJL x B6.TC)F1. FACS analysis was used to select progeny which were I-As and Igh. These mice were then intercrossed and the progeny selected which were B6.SJL.Igha using FACS and the strain was established from a single breeding pair. All mice were housed in AAALAC approved facilities in compliance with all applicable federal, state and local laws (Table 2-1).


Hg_2 Treatment


HgCl2 (Sigma Chemical Co., St. Louis, MO) was prepared in sterile, pyrogen-free phosphate buffered saline (PBS). Mice were injected subcutaneously at a dose of 1.5 mg/kg three times weekly unless noted otherwise.








16

Indirect Immunofluorescence


Sera from mice was tested for the presence of ANoA by indirect

immunofluorescence using commercially prepared mouse frozen kidney slides (Sanofi, Chaska, MN). The slides were incubated with sera diluted 1:50 in PBS for 30 minutes at room temperature. For nonallotype-specific ANoA. the slides were incubated for 30 min at room temperature with FITC-conjugated goat anti-mouse IgG (Fc fragment specific, Jackson ImmunoResearch, West Grove, PA) diluted 1:50 in PBS. For the allotypespecific ANoA determination slides, were incubated with rabbit anti-mouse IgG2aa or IgG2ab (Nordic Labs, Capistrano Beach, CA) followed by FITC-conjugated goat antirabbit IgG (Jackson ImmunoResearch, West Grove, PA). These reagents had been pretitrated by ELISA against an allotype-nonspecific rabbit anti-IgG2a antibody (Nordic Labs) to produce equivalent sensitivities. Antinucleolar staining was evidenced by an intense homogeneous staining of the nucleoli (Fig. 2-1).


Antibodies


MK-S4 (HB4; murine IgG2b anti-I-Apsf'u (49)) and 14-4-4S (HB32; murine IgG2a K anti-I-Ea kd (50)) were purchased from American Type Tissue Culture (Rockville, MD). D3-137.5 (murine IgG2a anti-I-Apb antigen (51)) was originally obtained from Tonkonogy. D77, originally developed against yeast fibrillarin but cross-reactive with fibrillarin of rat and human origin (52), was a kind gift of Dr. John Aris (University of Florida, Gainesville, FL). Aliquots of stock mAb were prepared from overgrown cell culture supernatant








17

which was affinity purified on a protein G column. Monoclonal antibodies were labeled either with biotin hydrazide (53) or fluorescein isothiocyanate (FITC) (54), as needed. Immunizations


Mice were immunized by an intraperitoneal injection of 50 jig of hen egg lysozyme (HEL; Sigma Chemical Co., St. Louis, MO) in 100 l of CFA (Difco Laboratories, Detroit, MI) diluted 1:1 with sterile PBS. Fourteen days later the mice were boosted with the same antigen in IFA. Mice were then bled via tail vein 8 weeks after the secondary immunization and the sera stored at -200C until analyzed by ELISA.


Anti-HEL ELISA


Determination of antigen-specific antibody levels were determined using an

indirect ELISA. Immulon@ 2 microtiter plates (Dynatech Laboratories, Inc., Chantilly, VA) were coated overnight at 40C with HEL at 10ig/ml in bicarbonate buffer (0.1M, pH

8.2). Between all incubation steps microplates were washed three times with boratebuffered saline (BBS: 25 mM Na2B407' 10 H20, 75 mM NaCl, 100 mM H3BO3, pH 8.4) containing 0.05% Tween 20. After blocking with BBS containing 0.5% BSA and 0.4% Tween 20, serial dilutions of serum samples diluted in the blocking solution were incubated overnight at 40C. Either biotinylated goat anti-mouse IgG (Fc fragment specific) or biotinylated goat anti-mouse IgM (mu chain specific)(Jackson Immunoresearch, West Grove, PA) was added for one hour at room temperature. The indicator system consisted of ExtrAvidin alkaline phosphatase (Sigma Chemical Co., St.








18

Louis, MO) and the substrate Sigma 104 phosphatase substrate (p-nitrophenol phosphate, disodium, hexahydrate). The substrate turnover was determined by the difference between the OD40s and OD620 on a BioTek Instruments, Inc. (Winoski, VT) Ceres 9000 microplate reader. The concentration of antigen-specific IgG and IgM is reported in equivalent dilution factors of standardized reference B6. SJL sera. This is defined by the formula: EDF = (dilution of a standard reference sera that gives the equivalent OD of the test serum) X 104.


Flow Cytometry


Approximately 200 pl of tail vein blood was collected into heparinized tubes. PBMCs were isolated using Lympholyte M (Cerdarlane, Hornby, Ontario, Canada) density gradients. The cells were then collected into PBS, supplemented with 3% FCS and 0.1% NaN3. For cell surface staining, saturating amounts of 14-4-4S, D3-137.5biotin or MK-S4-FITC were used as the first step. The second step consisted of incubating with FITC-conjugated goat anti-mouse IgG (Fc fragment specific) or phycoerythrin (PE)-conjugated streptavidin (Jackson Immunoresearch, West Grove, PA). The cells were then washed three times in PBS and fixed with an equal volume of 2% paraformaldehyde. List mode data were acquired on a FACScan flow cytometer (Becton Dickinson, Palo Alto, CA) using PC Lysis software. Dead cells were excluded by forward and side scatter gating. List mode files were then analyzed using Lysis II software.








19

Statistical Analysis


The difference between the number of mice showing an ANoA between B6.I-A' and B6.I-A!" was analyzed using Fischers exact test. A t-test was used to compare baseline I-A! MFI (mean fluorescent intensity) values between these two groups. Differences between MFIs of the other groups was tested by an ANOVA (analysis of variance) followed by Dunnett's multiple comparison procedure. Ap value of <0.05 was considered to be significant.


Results


F Mice Between a Resistant and Sensitive Haplotype Were Resistant to ANoA Induction


In the course of our initial experiments with the HgCI2 model of induced

autoimmunity, we examined the ANoA response in an F1 cross between the susceptible B6.SJL (H-2'; Ighb) and the resistant B6.TC (H-2b: Igha) strains. As expected, B6.SJL (n=18) and B6.TC (n=13) were 100% and 0% susceptible, respectively. A representative example is shown in Fig. 2-1, where a serum sample from a B6.SJL mouse treated for 5 weeks with HgCI2 shows a distinct nucleolar pattern. However, the vast majority (42/45) ofF, intercrosses between these two strains were also resistant to HgCl2 treatment (Table 2-1). The resistance of the F, mice was surprising. Although the exact mechanism for HgC12-induced autoimmunity is unknown, the fact that the class II molecules are codominantly expressed led us to expect that the expression of I-A! would be sufficient to confer susceptibility. To make sure this resistance was not due to the allotype difference,








20

B6.SJL.Igha (H-2s, Igha) mice were produced and found to be 100% susceptible to ANoA formation. This showed that the Igh locus did not have an effect on susceptibility. Finally, to rule out a dose effect in the F1 crosses, groups of 4-6 (B6.SJL x B6.TC)F1 were dosed with either 3.0, 4.5, 7.5 or 10 mg/kg of HgCl2 for 3 weeks with no mice developing an ANoA titer (data not shown).

Because it has been reported that the B6 background produces lower titer ANoA responses than other H-2s strains, other crosses were tested. Initially, mice on the A background were used. Consistent with previous studies, A. SW (H-2') mice were 100% (3/3) susceptible while A.BY (H-2b) mice were resistant (0/4). Next, F1 crosses between these two strains were generated. Despite the more susceptible A background, only twenty percent (1/5) of (A.BY X A. SW)F1 mice showed ANoA positivity after treatment. To rule out a maternal effect in F1 mice as previously suggested (9), the same strain was produced except that the A. SW strain was used as the female parent. This cross, (A. SW x A.BY)F,, resulted in only 11% (1/9) of the animals responding. In further attempts to find an F, cross between resistant (H-2b) and sensitive (H-2s) mice in which the susceptibility gene(s) was dominant, the following crosses were tested: (A. SW X B6)FI, (B6 X SJL/J)Fl and (SJL/J X B6)F1. None of these animals developed an ANoA titer after mercury treatment.

The specificity of the autoantibody response was verified by comparing an

immunoblot of sera from a B6.SJL mouse. This serum sample which reacted with a 34kDa nucleolar protein shown to be fibrillarin by mAb D77 (data not shown).








21

































Figure 2-1 Serum from a B6.SJL mouse treated for 5 weeks with HgC12 on mouse
frozen kidney slides and stained with FITC-conjugated goat anti-mouse
IgG as a second step showing intense staining of nucleoli.








22

Table 2-1 Congenic and Fx mice: serum ANoAa after 5 weeks of treatment with HgC2b Strain Class II I-A I-Ec IgG1 nd %positive"
Haplotype A6 A, E5 E, allotype
A.BY b b b (b b) a 4 0
A.SW s s s (s s) a 3 100
(A.BY x A.SW)F1 s/b s/b s/b (s/b s/b) a 5 20
(A.SW x A.BY)F, s/b s/b s/b (s/b s/b) a 9 11
(A.SW x B6)F, s/b s/b s/b (s/b s/b) a/b 8 0
(B6 x SJL/J)F1 s/b s/b s/b (s/b s/b) b 6 0
(SJL/J x B6)F1 s/b s/b s/b (s/b s/b) b 6 0
B6.SJL s s s (s s) b 18 100
B6.TC b b b (b b) a 13 0
(B6.SJL x B6.TC)Fj s/b s/b s/b (s/b s/b) a/b 45 7
B6.SJL.Igh s s s (s s) a 9 100
aANoA, antinucleolar antibodies
bl.5 mg HgCl2/kg s.c. three times per week
calleles written in parenthesis denote nonexpression on the cell surface of the E E, molecule "Number of mice
'Fraction of positive mice


The Ab Gene Product Conferred Resistance in F, Mice


Because of the marked resistance of all (s/b) F1 mice tested and the possibility that

this effect was being mediated by non-class II molecules, the role of class II was tested

directly. Taking advantage of the availability of B6 mice with targeted deletions of the IAb gene, mice heterozygous for this induced mutation were mated with B6.SJL mice to

generate two groups of mice within the littermates. One group possessed the genes linked

to I-Ab without expressing the molecule itself, while the other possessed these same genes

but also expressed I-Ab. Those mice that expressed I-Ab (I-A") were resistant (1/14) to

the development of ANoA, as expected from our earlier results, despite expressing the

susceptible I-A' molecule. However, those that possessed both I-A' and I-Ab-linked

genes, but not I-Ab expression itself (I-A'), were as susceptible (17/20) to ANoA








23

formation as B6. SJL homozygous mice (Table 2-2). This clearly shows that it is the I-Ab molecules themselves and not a linked gene which conferred resistance. As a negative control, no mice (0/8) developed ANoA after treatment with PBS (Table 2-2).



Table 2-2 Effect of Ab deletion on ANoA" in HgCl,-treated(B6.SJL x B6.I-AbI) F1 mice Class II nc positived
Haplotypeb HgCl,' NaCV HgCl2 NaCl
s/b 14 4 79 0
s/- 20 4 85 0
aANoA, antinucleolar antibodies measured after 5 weeks of treatment 'F1 mice either did (s/b) or did not (s/-) express I-Ab but were heterozygous at all the MHC loci
cNumber of mice
"Fraction of positive mice
"1.5 mg HgCl2/kg s.c. three times per week fO. 1 ml NaCl s.c. three times per week
*Significant difference between (s/b) and (s/-) mice (p < 0.05 using Fischer exact test)


To rule out the possibility that the differential susceptibility between I-A" and IV' was due solely to increased expression of I-A! and not a negative regulatory influence by I-Ab, we examined I-A levels by haplotype-specific flow cytometry. Baseline I-A expression was similar whether or not the mouse carried the disrupted gene (MFI = 45 vs. 49; p = 0.1185). MHC class II expression on PBMCs was similarly increased in both strains when treated with HgC12 (Fig. 2-2). I-A expression was increased by 44% (p<0.05) in the mice which carried the disrupted allele, while it was increased by 48% (p<0.05) in those mice which expressed both haplotypes (s/b). The increase in I-Ab expression in the mice which carried that gene was 48% (p<0.05). No significant increase in class II expression was seen in those animals treated with PBS.








24






A B
170
80 v 160
A 150
70 A y 140

A v 13060- A v 120

50- A A 110

40 9010
0 A 80

I I II I I I I
Predose PBS Hg Predose PBS Hg Predose PBS Hg

B6.I-As/- B6.I-As/b B6.I-As/b



Figure 2-2 Effect of HgCl2 on class II expression in(B6.SJL x B6.I-Abt')F1 mice. (A) Predose I-As
expression was similar in B6.I-A" and B6.I-A!" mice. I-A' expression was increased by 44%
in the mice which carried the disrupted allele, while it was increased by 48% in those mice
which expressed both haplotypes (s/b). (B) The increase in I-Ab expression in the mice which
carried that gene was 48%. No significant increase in class II expression was seen in those
animals treated with PBS. *p_ 0.05








25
Expression of I-E Had No Effect on Susceptibility to HgCI,


It has previously been reported in a number of autoimmune models that I-E

expression can down-regulate autoantibody titers. Intra-MHC recombinant strains have given conflicting results in the HgCl2 model, but interpretation has been hampered by the difficulty in isolating the effects of I-E expression. To test this more explicitly and to see if competing expression by another class II molecule is a general property in the downregulation of the anti-nucleolar response to mercury, B6.SJL mice expressing the I-Ead transgene were bred. Restoring the ability to express I-E in otherwise susceptible mice had no effect on the ability to respond to HgCl2. All (5/5) B6.SJL.I-E.a mice retained the ability to generate an ANoA titer upon HgCl2 treatment (Table 2-3). Mean I-E expression in these mice (MFI = 280.7) was similar to Balb/c mice (MFI = 282.1) which normally express the I-E gene product. I-As expression between B6.SJL and B6.SJL.I-Ead mice was also similar (MFI = 267.9 vs. 253.2) (Fig. 2-3). The breeding scheme used to generate the B6.SJL.I-Ea mice also produced (B6.SJL x B6.I-Ed )BC, mice which either did (n=14) or did not (n=13) express the transgene. It has been suggested in the BxSB autoimmune model that I-E expression could block autoantibody production by producing an I-E-derived peptide that binds to I-Ab with such affinity that it competes with the putative autoantigen-derived peptides for presentation (6). Starting with the same transgenic parent strain as was used in the present experiments, we have also confirmed this observation in the BxSB model (unpublished data). If this is a general property in autoimmunity, it might be expected that I-E expression would convert an otherwise








26

resistant HgCl2-treated s/b heterozygous mouse into a sensitive one. Despite normal

levels of expression of I-E, however, all of these mice were resistant (Table 2-1).




Table 2-3 Effect of I-E on ANoAa response Strain Class II A I-E nb positivec
Haplotype AB A, Expression
B6.SJL.I-Ead s s s + 5 100
(B6.SJL x B6.I-Ead)F1 s/b s/b s/b + 14 0
(B6.SJL x B6)FI d s/b s/b s/b 13 0
aANoA, antinucleolar antibodies measured after 5 weeks of treatment with 1.5 mg HgCl2/kg s.c. three times per week bNumber of mice
'Fraction of positive mice
dNon-transgenic littermate of(B6. SJL x B6.I-Ead)F1




A Mean I-E Expression B Mean I-As Expression

SB6.SJL.I-E B6.SJL
282.1 m 267.9 M1

1 lb lb ib 15 b
FL1-Height FL1-Height


SBalb/c I ]B6.SJL.I-E
8253.2 M1
1 lb 1b2 ib 0 ib ib 2 lb 4
FL-Height FL1-Heght

Figure 2-3 Class II expression in B6.SJL.I-E d mice. (A) Mean I-E expression
was similar to Balb/c mice. (B) Mean I-A' expression was not
affected by I-E expression. Single-parameter fluorescence histograms.
Increasing fluorescence intensity is plotted on the x-axis in log fluorescent units versus cell number on the y-axis The numbers
represent the MFI (mean fluorescent intensity) of the cells expressing
the antigen of interest.








27

Responsiveness to Exogenous Antigen Was Inherited Co-dominantly in F1 Mice


The lack of ANoA's in our F1 mice led us to evaluate their responses to an

exogenous antigen (HEL). As expected, mice of the b haplotype (B6.TC) responded poorly while the B6. SJL mice mounted an intermediate titer. An intermediate response was seen in the F1 cross (Figure 2-4). Balb/c mice responded with a titer approximately 10 fold higher than any of the mice on the B6 background (data not shown). Thus it appears that the responsiveness in our F, mice to BEL was inherited in a co-dominant fashion.


Discussion


The mercury chloride response in rodents is a complicated one and classification of responder and non-responder strains depends on the outcome being measured. For example, by ANoA titer, the b haplotype is a non-responder. However, by measuring the weights of draining popliteal lymph nodes of (DBA/2 X C57BL/10)F1 mice exposed to HgCI2, b haplotype has been classified as intermediate responders.

The anti-fibrillarin response by susceptible mice, however, is very reproducible and specific and therefore more amenable to mechanistic dissection. As a first step, it is crucial to convincingly demonstrate that the anti-fibrillarin response is a property of the I-A molecule itself and not a closely linked gene. Results in the literature thus far have been suggestive but not conclusive. Robinson et al. (5) used two unrelated strains (A. SW and C57BL/10) to generate the Fj's. These mice were 57-100% susceptible to developing








28

ANoA, albeit of a lower titer. This led the authors to conclude that the response was controlled by the H-2 region and one or more unlinked genes. Recently, Hultman et al.

(9) found that 75% of male and 11% of female mercury treated (SJL x B6)F1 mice showed a clumpy IgG ANoA pattern with a low titer. These differences were attributed to the non-H-2 genes. For example, the A genetic background seemed to enhance susceptibility to the HgCl2 treatment, while the B 10 background seems to strengthen the resistance. They also suggested that the B6 non-H-2 background genes are more suppressive on ANoA expression than equivalent genes from the B 10 strain. A similar effect of non-H-2 genes was seen by Hultman et al. (12) when comparing the ANoA response in several H2' strains. B10.S mice showed a much lower titer than either A.SW or SJL mice after HgC12 treatment. This was again attributed to the non-H-2 genes.




A Anti-HEL IgG B Anti-HEL IgM
3000
1600
2000 1200

800
1000
S400
0- *
-0
I I I I I
B6.SJLB6.TC F1 B6.SJLB6.TC F1



Figure 2-4 Anti-HEL antibody levels 10 weeks after immunization. The response
to HEL was inherited co-dominantly in(B6.SJL X B6.TC)F1 mice.
B6.SJL and (B6.SJL X B6.TC)F, mice respond to HEL while B6.TC do
not. (A) IgG (B) IgM; EDF +/- SEM








29

In the process of studying the cellular control of the mercury-induced antifibrillarin response, we examined the susceptibility of a number of F, mice in our own laboratory. Contrary to previously published results summarized above, our F, mice seldom produced a measurable ANoA titer most likely due to the strong suppressive effects seen when using mice on the B6 background. This lead us to re-examine the role of class II expression. Several genes linked to class II have been shown to participate in antigen processing and presentation (44,55,56). The possibility existed that the genes linked to the I-Ab molecule could be altering the processing and/or presentation of fibrillarin resulting in a peptide which no longer is immunogenic. If this were the case then mice which express I-A! and carry the I-Ab linked genes without displaying the I-Ab molecule itself would be expected to be resistant.

Since the ANoA response was not dominant in our F, mice we compared the

responses to BEL in which responsiveness should be inherited in a co-dominant fashion. Mce of the b and s haplotype are classified as non-responders to BEL when measured by PFC assays, although SJL (H-2') mice do mount an intermediate serum titer (57). An analysis of anti-EEL serum titers revealed that B6.TC mice responded poorly, as expected, while both while B6.SJL and (B6.SJL x B6.TC)F1 mice were intermediate responders. It appears that there is a difference in the inheritability of responsiveness between the HgCl2-induced ANoA response and that to a conventional exogenous antigen.

To investigate whether the resistance of F1 mice is due to expression of the I-AY molecule or a linked gene, F, mice on the B6 background were produced which lacked








30

expression of the I-Ab molecule due to disruption of the MHC class H I_,% b gene (46). By using congenic strains which differed at only the H-2 region, we eliminated the effects of non-MHC-linked genes. Because they are of a low-responder background, the B6 strains also allowed us to more carefully isolate the effects of the MHC. Finally, the B6 background permitted us to take advantage of the availability of class II transgenics and targeted deletions. Our strategy was to cross the CD2T (B6.I-A") to a wild type B6.IAb/b to obtain B6.I-Ab/ mice. These mice were then bred to a B6.SJL to obtain the two strains used for analysis: B6.I-A and B6.I-A'. The incidence of ANoA in those mice which expressed the I-Ab gene product was significantly decreased showing that the I-Ab molecule itself and not a linked gene was responsible. To rule out the possibility that the mice carrying the disrupted allele could be expressing a higher density of I-As molecules thus leading to the increased susceptibility, we compared the predose mean I-A! expression between the 2 strains. The MEI of I-A expression for the mice carrying the disrupted allele (I-As) was not significantly different from the I-Ab mice.

The possibility exists that the I-A b gene product could pair with the I-A,'

molecule resulting in an haplotype-mismatched class II molecule (Ab / A/). Martin et al. (58) showed that the inefficient assembly of haplotype-mismatched class II polypeptides results from their inability to compete with the matched pairs. Different class II MHC a3 chain combinations vary greatly in their efficiency of cell-surface expression based on the allelic origin of the a and 3 amino-terminal regions (59). If this pair was being expressed we would have expected to see an increase in the I-A8 expression in the mice carrying the disrupted allele (I-A'), since the mAb MK-S4 recognizes the 3 chain of I-A!, however no difference was detected, and we believe this is an unlikely possibility.








31

Another possibility is that the HgC12 had a differential effect on the two class II molecules. If the HgC12 had caused a greater increase in the expression of the I-Ab molecule in those mice which expressed it, this could lead to a decreased presentation of antigen by the susceptible allele (I-A), perhaps falling below the threshold required for T cell activation. This is not likely since both class II molecules showed similar increases in expression upon HgCl2 treatment.

To examine directly the effect of I-E expression in otherwise susceptible mice, we generated mice on a B6 background expressing the I-A! and the I-E molecules. Like other H-2" mice, B6.SJL mice do not express I-E due to a deletion in the I-E. gene (60). To remedy this, we crossed B6. SJL mice with a B6 mouse carrying the I-E transgene (Tg(H21:Ea)Bri39) (45). After backcrossing to B6.SJL mice, the progeny which expressed I-A! and I-E were selected. These mice remained 100% susceptible to ANoA formation despite expressing the I-E molecule. Our results do not support the observation by Mircheva et al. (6) that the expression of I-E suppresses the induction of ANoAs. Hultman et al. (7) was also not able to confirm the suppression. Both of theses studies examined the effect of I-E by using intra-H-2 recombinant strains, which meant that the conclusions were based on inference from a number of comparisons of strains. We believe that our system of using the introduction of a transgene to restore expression offers an advantage over random recombinatorial events which can carry along other genes in the process.

The possibility exists that these mice do not express I-E at physiological levels. If I-E were being expressed at excessively high levels, antigen presentation by the I-A








32

molecule could be diminished. Conversely, low-level expression of I-E could result in little physiologic effect. To rule out either of these artifacts, we compared mean I-E expression between our mice and a strain which normally expresses I-E (Balb/c). There was no difference in the mean expression between the two strains. The expression of I-A was also similar between B6.SJL and the B6.SJL.I-E.d mice. Moreover, expression of IE was also physiologic in that it was expressed on B but not T cells (data not shown).

Merino et al. (47) showed that the I-Ed transgene protected BxSB mice against SLE. They proposed that overexpression the I-E a chain resulted in the generation of excessive amounts of a peptide displaying a high affinity for the I-Ab molecule thereby preventing autoantigen-derived peptides from being presented. This phenomenon has also been seen in mice homozygous for the lpr mutation (48). To investigate the possibility that this might work in reverse to promote autoimmunity in some cases, we generated the otherwise resistant (B6.SJL X B6.I-Ead)Fi mice which either did or did not express I-E. Theoretically, this could block I-Ab as a functional competitor and enhance an I-A!mediated response. However, neither strain developed ANoA after treatment with HgC12. Thus, the expression of I-E does not promote the development of ANoA in a strain which is otherwise resistant. The results are clear, although we cannot formally exclude the possibility that a peptide derived from I-E was even more effective in blocking I-A. However, if this were the case then we would have expected I-E to have decreased HgC12induced susceptibility in the B6.SJL.I-Ed mice.

Taken together, our results conclusively demonstrate for the first time that HgCl2 responsiveness is regulated by the I-A haplotype itself and not a closely linked gene.








33

However, the ability to downregulate the immune response by coexpression in a resistant haplotype in an otherwise sensitive strain is a complicated one. Antigen competetion, proposed to be important in other autoimmune models (47,48), does not appear to have a dominant role in the HgC12-induced response.













CHAPTER 3
RESISTANCE TO HgCl2-INDUCED AUTOIMMUNITY IN HAPLOTYPE
HETEROZYGOUS MICE IS AN INTRINSIC PROPERTY OF B CELLS Introduction


Systemic lupus erythematosus is a complicated systemic autoimmune disease with a multigenic mode of inheritance interacting with a potentially complex array of generally unknown environmental factors. Many of these features are also seen in the mercury chloride model of murine autoimmunity, which offers a unique opportunity to study the reproducible interaction between genetics and a simple, inorganic environmental agent. Susceptibility to mercury chloride in mice is characterized by splenic hyperplasia, hypergammaglobulinemia and autoantibody production (4,29), all of which appear as soon as one week after the start of treatment (4). The principle autoantibody target is fibrillarin, a U3 ribonucleoprotein found in the nucleolus and involved in the first step of preribosomal processing (31,39). The antifibrillarin specificity is also seen in patients with scleroderma and other connective tissue disorders (33,34). Recently, a comparison of the specificity of the human and murine autoantibodies showed that they recognize similar, if not identical, epitopes (40) and suggested an antigen-driven response in both.

Similar to human autoimmune disease, both MHC and non-MHC genes have been shown to have a pronounced effect on susceptibility (12). As defined by the presence of


34








35

antifibrillarin antibody, mice of the H-2' haplotype are high responders, while those of the H-2b and H-2 d haplotypes are nonresponders (5). In contrast, non-MHC genes affect primarily antibody titers but not specificity (12). Through the use of intra-H-2 recombinants, the presence of the HgCl2-induced antifibrillarin antibody response has been mapped specifically to the I-A region (6). Previous studies in our laboratory have shown that F, animals between MHC-congenic susceptible H-2s and resistant H-2b mice to be resistant to HgCl2-induced ANoAs (8). This is surprising, since the I-A molecule is codominantly expressed in the F, mice, and in other autoimmune models heterozygosity of class II either enhances autoimmunity or modestly affects antibody titers (61,62). Therefore, the profound resistance to HgCl2-induced ANoAs seen in all haplotypeheterozygous mice tested raised the possibility that resistance was mediated by a dominant gene linked to I-Ab. However, further experiments demonstrated that resistance was, in fact, caused by co-expression of the I-Ab molecule itself, and that this outcome was not merely the result of lower expression of the susceptible I-As haplotype on the otherwise responsive B cells (8).

We have utilized adoptive transfer experiments to explore further the mechanisms by which the resistant haplotype down regulated the HgCl2-induced antifibrillarin antibody response. Our experiments demonstrated that resistance was due neither to I-Ab-mediated alteration of T cell repertoire nor to inadequate I-AS-restricted T cell help but was an intrinsic property of the resistant haplotype-heterozygous B cells. These results suggest the presence of a novel mechanism of regulation of a potentially autoreactive immune response.








36


Materials and Methods

Nce


C57BL/6J (B6; I-Ab, Ighb), C57BL/6J.SJL (B6.SJL; I-A!, Ighb), C57BL/6J-Igha Thyla Gpia (B6.TC; I-Ab, Igha) mice were originally obtained from The Jackson Laboratory (Bar Harbor, ME) and maintained in our breeding facility.d B6.SJL.Igha mice were developed by intercrossing B6.SJL and(B6.SJL x B6.TC)F1. FACS analysis was used to select progeny which were I-A! and Igh. These mice were then intercrossed and the progeny selected which were B6.SJL.Igh using FACS and the strain was established from a single breeding pair. All mice were housed in AAALAC approved facilities in compliance with all applicable federal, state and local laws. Preparation of Chimeras


Preparation of chimeras was as previously described (63). Recipient mice were provided with Septra (1% v/v) treated water the day prior to irradiation. On the day prior to bone marrow transfer, the mice were treated with two doses of 525 rad of yradiation (Gammacell 40, Atomic Energy of Canada, Ltd., Ottawa, Canada) separated by 3-4 hours. The transfer involved i.v. reconstitution with a total of 10' bone marrow cells from age- and sex-matched donors. Bone marrow cells were depleted of mature T cells by incubation at 4o C for 30 min with a mixture of anti-mouse T-cell serum (Cedarlane Laboratories, Hornby, Ontario, Canada), 172-4 (rat IgM anti-CD4 (64)), and 31 M (rat IgM anti-CD8 (65)) followed by treatment with C at 370C for 1 h (Low-Tox Guinea Pig








37

C, Cedarlane Laboratories, Hornby, Ontario, Canada). To prevent graft rejection the B6.SJL mice receiving(B6.SJL x B6.TC)F1 cells were given 0.1 mg i.p. of MmTI (mouse IgG2a anti-CD90.2 (66)) at the time of transfer. 172-4 and 31 M were obtained from Dr. David Harris (University of Arizona, Tuscon, AZ) and prepared from overgrown cell culture supernatant which was affinity purified on a protein G column. In Vivo Treatments


HgCl2 (Sigma Chemical Co., St. Louis, MO) was prepared in sterile, pyrogen-free phosphate buffered saline (PBS). Mice were injected subcutaneously at a dose of 1.5 mg/kg three times weekly after graft acceptance was verified using flow cytometry. Mice were immunized intraperitoneally with 100 pg of human IgG (HGG; Sigma Chemical Co., St. Louis, MO) emulsified 1:1 in CFA (Difco Laboratories, Detroit, MI). Booster immunizations consisted of a second intraperitoneal injection of 100 pg of HGG in sterile, pyrogen-free PBS.


Flow Cytometry


Approximately 5-6 weeks after bone marrow transfer 200 pl of tail vein blood was collected into heparinized tubes. PBMCs were isolated using Lympholyte M (Cerdarlane Laboratories, Hornby, Ontario, Canada) density gradients. The cells were then collected into PBS, supplemented with 3% FCS and 0.1% NaN3. For cell surface staining, saturating amounts of biotinylated D3-137.5 (mouse IgG2a anti-I-Ab)(51) and fluoresceinated TIB 92 (mouse IgG2a anti-I-Ab)(67) or biotinylated AF6-78.25 (mouse








38

IgG1 anti-IgMb)(68) and fluoresceinated DS-1 (mouse anti-IgMa)(69) or biotinylated HIS51 (mouse IgG2a anti-CD90.1) and fluoresceinated 30-H12 (rat IgG2b anti-CD90.2) were used as the first step. The second step consisted of incubating with phycoerythrin

(PE)-conjugated streptavidin (Jackson ImmunoResearch, West Grove, PA). The cells were then washed three times in PBS and fixed with an equal volume of 2% paraformaldehyde. List mode data was acquired on a FACScan flow cytometer (Becton Dickinson, Palo Alto, CA) using PC Lysis software. Dead cells were excluded by forward and side scatter gating. List mode files were then analyzed using Lysis II software. Monoclonal antibodies were labeled either with biotin hydrazide (53) or fluorescein isothiocyanate (FITC) (54), as needed. At the study termination the same procedures were used with the exception that splenocytes were used for cell staining.


Indirect Immunofluorescence


Sera from mice collected 5 weeks after the initiation of HgCl2 was tested for the presence of ANoA by indirect immunofluorescence using commercially prepared mouse frozen kidney slides (Sanofi, Chaska, MN). The slides were incubated with sera diluted 1:50 in PBS for 30 minutes at room temperature. For nonallotype-specific ANoA the slides were incubated for 30 min at room temperature with FITC-conjugated goat antimouse IgG (Fc fragment specific, Jackson ImmunoResearch, West Grove, PA) diluted 1:50 in PBS. For the allotype-specific ANoA determination slides, were incubated with rabbit anti-mouse IgG2aa or IgG2a b (Nordic Labs, Capistrano Beach, CA) followed by FITC-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA).








39

These reagents had been pre-titrated by ELISA against an allotype-nonspecific rabbit antiIgG2a antibody (Nordic Labs) to produce equivalent sensitivities. Antinucleolar staining was evidenced by intense homogeneous staining of the nucleoli.


Allotype-Specific ELISA


The protocols used for measurement of allotype-specific serum total IgM and

IgG2a were minor modifications of previously described procedures (63). For serum total IgM, samples were developed with affinity-purified donkey anti-mouse IgM (Jackson ImmunoResearch, West Grove, PA). IgMa and IgMb were measured using DS-1 (69) and AF6-78.25 (68) respectively. Serum total IgG2a was determined using rabbit anti-mouse IgG2a (Nordic Labs, Capistrano Beach, CA) while allotype-specific IgG2a was measured using either rabbit anti-mouse IgG2aa or IgG2ab (Nordic Labs, Capistrano Beach, CA). These reagents had been pre-titrated by ELISA against the allotype-nonspecific rabbit anti-IgG2a antibody (Nordic Labs) to produce equivalent sensitivities. The rabbit antibodies were detected using alkaline phosphatase-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA). Allotype-specific IgG2a anti-HGG was measured using the same protocol except that HGG was used as the first step.


Results


MHC-Restricted T Cell Help Was Required to Produce ANoAs


Exposure to HgC12 induces a wide range of physiologic effects in mice, including the release of large amounts of cytokines, especially the Th2 cytokine IL-4. It was








40

therefore possible that the antifibrillarin response could result from non-MHC-restricted interactions in genetically susceptible mice. To test this possibility, B6.TC mice were reconstituted with T cell-depleted B6.SJL marrow. Because intrathymic positive selection of T cells is mediated by the radioresistant thymic cortical epithelial cells (70), in these mice all CD4+ T cells would be positively selected to interact with the host I-Ab and not the donor I-As haplotype. Therefore, the B6.SJL-derived B cells would not be expected to receive MIHIC-restricted help. In contrast, central tolerance is mediated by bone marrow-derived dendritic cells although thymic epithelial cells can make a significant contribution (71), and therefore the animals would be tolerant to both haplotypes. Flow cytometric analysis of peripheral blood lymphocytes showed that all of the B cells were of donor B6.SJL origin. Surprisingly, despite the presence of these susceptible B cells and the massive activation of cytokines following HgCl2 administration none of the animals developed ANoAs (Table 3-1). The addition of B6.TC cells in the inoculum did not alter the outcome. In marked contrast, syngeneically reconstituted B6.SJL mice responded well to mercury chloride, demonstrating that lack of responsiveness was not an artifact of radiation. We therefore conclude that the antifibrillarin specificity is mediated by I-As restricted T cell help.








41

Table 3-1 MHC-restricted T cell help was required to produce ANoAsa
Dono Host nb ANoA Positive (%)

B6.SJL & B6.TC B6.TC 4 0
B6.SJL & B6.TC B6.SJL 5 100
B6.SJL B6.SJL 7 100
B6.SJL B6.TC 5 0
aANoA, antinucleolar antibodies
'Number of mice
Traction of positive mice


Absence of ANoA Production in Haplotype-Heterozygous Mice Was Not Due to a Difference in Thymic Education


Although the above experiments showed the importance of T-B cell collaboration it is possible that I-Ab expression in the host and/or donor affected thymic education and therefore eliminated a population of responsive T cells. This led us to investigate the responses in (B6.SJL x B6.TC)Fl hosts. When (B6.SJL x B6.TC)Fi mice were used as hosts and reconstituted with B6.SJL, they all (12/12) responded to HgCl2. When the same hosts were used but the donor changed to B6.TC, none (0/6) of the mice produced ANoAs (Table 3-2). Therefore, the donor haplotype determined whether or not autoantibodies are produced when developing cells were positively selected on both haplotypes.


Allogeneic Mixed Chimeras Developed ANoA Titers in Response to HgCl2


In the rat model of mercury chloride-induced autoimmunity, allogeneic mixed microchimerism has been able to induce a state of tolerance (72). In contrast, other models have shown the potential for epitope spreading once tolerance is broken (73,74).








42

Therefore, it was of interest to determine whether the co-existence of B cells sensitive and resistant to the effects of HgCl2 could influence each other.

To test these possibilities, we lethally irradiated (B6. SJL x B6.TC)F1 mice and

transferred T cell-depleted bone marrow from the following: 1) B6.SJL alone; 2) B6.TC alone; 3) a combination ofB6.TC & B6.SJL; and 4) (B6.SJL x B6.TC)F1 alone. In all four groups, T cells would be positively selected by either I-As or I-Ab. After confirmation of mixed chimerism in group 3 (see below), the mice were treated with HgCl2 and the results are shown in Table 3-2. As expected, (B6.SJL x B6.TC)Fl mice reconstituted with only B6.SJL bone marrow responded readily to HgCl2 while mice given only resistant B6.TC bone marrow failed to develop an ANoA antibody titer. In contrast, syngeneic reconstitution with resistant (B6. SJL x B6.TC)F1 bone marrow resulted in a very poor ANoA response. These results essentially duplicated our experience in analogous non-chimeric HgC12-treated B6.SJL, B6.TC, and (B6.SJL x B6.TC)F, mice (8) and again demonstrated that radiation did not affect this outcome.

Strikingly, mice given a combination of resistant and susceptible bone marrow had an intermediate result, with nearly 50% of the mice responding (Table 3-2). By allotypespecific ANoA, all of the responders produced exclusively b allotype autoantibody and therefore of donor B6.SJL origin (data not shown). This is not due to an intrinsic property of the a allotype, for the a allotype B6. SJL-Igha strain responded equally well to mercury (data not shown). Thus, the presence of ANoA in these mice resulted from the loss of tolerance to fibrillarin in I-A' B cells.








43

Table 3-2 ANoA response in chimeric (B6.TC x B6.SJL)Ft host mice
Group Source of Donor Bone Marrow nb ANoA Positive (%)C
1 B6.SJL 12 100
2 B6.TC 6 0
3 B6.SJL & B6.TC 16 43.8
4 (B6.SJLxB6.TC)F1 17 11.8

'ANoA, antinucleolar antibodies
'Number of mice
'Fraction of positive mice


Lack of ANoA in Non-Responsive Mixed Chimeras Was Unrelated to B Cell Composition


Of interest, the response rate of 44% in the mixed chimeras (Group 3) was significantly different than either the 100% seen in Group 1 (p







44

B Cells of Both B6.TC and B6.SJL Origin Were Functional in Group 3 Mixed Chimeras


It was possible that the presence of a and b allotype B cells in the mixed chimeras influenced the development of antinucleolar antibodies through antigen non-specific mechanisms. For example, in some combinations, allotype-specific suppression of immune responses has been seen (75). This possibility was minimized by the use of allotypeheterozygous host mice. However, to assess further the functionality of B cells of both B6.SJL and B6.TC origin, allotype-specific total IgM and IgG2a ELISAs were performed. In addition, the mice were also assayed for their response to immunization with a T cell dependent antigen, HGG. As shown in Figure 3-2, total IgM and IgG2a of both allotypes were present. Interestingly, despite the predominance of IgM of B6.SJL origin, IgG2a was much better balanced between the two donors. Moreover, there was a good IgG2a response to a T cell dependent antigen by B cells of both B6.SJL and B6.TC origin (Figure 3-3A), and both ANoA positive and negative mice responded equally well (data not shown).



Lack of HgCl2-Induced ANoA Response by Haplotype-Heterozygous Mice Was Not Due
to Absence of Anti-Fibrillarin-Specific I-AS-Restricted T Cell Help


To determine whether or not the intermediate response seen in the mixed chimeras and in HgCl2-resistant (B6.SJL x B6.TC)F1 mice was due to a lack of I-A'-restricted T cell help, we reconstituted (B6.SJL x B6.TC)F1 mice with a combination of syngeneic (B6.SJL x B6.TC)F1 and B6.SJL bone marrow. Fortyfive percent (5/11) of these mice produced









45








B6.TC + B6.SJL ---> F1


A. Class II Expression B. Surface IgM

100- 100
90- -90
80 X 80
70 X 0 70
60 -60
50- 50
40 X X 40
30 30
20 20
10 O 10
0 II I I 0
pos neg pos neg
ANoA



Figure 3-1 Immunofluorescence on peripheral blood lymphocytes of parental
into F, chimeras. Data are presented as percent of cells expressing
class II. A, Class II expression; B, surface IgM.










46








B6.TC+B6.SJL --->Fl


A. Serum IgM B. Serum IgG2a

100- 100
90- 8 90
80 80
70 X 70
o
60 60
b50 50
40 -40
30 30
20 X 20
O
10 10

0 I-I I I 0
pos neg ANoA pos neg




Figure 3-2 Allotype-specific total IgM and IgG2a ELISA data of parental into F1
chimeras. Data are presented as a percentage of total isotype. A, Serum
IgM; B, serum IgG2a.










47








Anfi-HGG liter


ANB6.TC+B6.SJL --- >F1 B. B6. SJL +F I--- >FI
100- -100
go- 90
go- X X 0 80
70- 70
x
60- x X O 60
A 50o 0 50

40 40 ~
30- 30
0
20- 20
10- 10
0- -0
pas neg ANoA pos neg




Figure 3-3 Allotype-specific anti-HGG ELISA data. Data are presented as a
percentage of total isotype. AB6.TC +B6.SJL --- > F, mice;
B, B6. SJL + F --- >F, mice.








48

ANoAs upon HgCl2 treatment. The fact that some of the mice responded meant that T cells were available in which tolerance to fibrillarin was broken. Allotypic analysis revealed that all of the autoantibodies were of the b allotype. It seems likely that they arose from the B6.SJL donor since the B cells from the (B6.SJL x B6.TC)Fi mice would have produced autoantibodies of either a or b allotype and no a allotype autoantibodies were detected even at a 1:10 dilution (data not shown). Thus, despite the presence of T cells which provided help in an I-A'-restricted fashion, those B cells which co-expressed both haplotypes did not produce autoantibodies. B Cell Composition Was Not a Determining Factor in Responsiveness


Figure 3-4 shows the B cell makeup of the B6.SJL + (B6.SJL x B6.TC)F1 ---> (B6.SJL x B6.TC)Fj mice. Flow cytometric analysis of class II expression showed these mice to be very well balanced (%I-A"b range: 33-54%). The predominance of IgM of the b allotype (61-82%) is expected considering that approximately one-half of the B cells from the (B6.SJL x B6.TC)Fl donor would express IgM of the b allotype. Similar to the parental into (B6.SJL x B6.TC)F1 mice no correlation was noted between the B cell makeup and responsiveness (Figure 3-4).


Both Bonors Provided Functional B Cells


The apparent lack of response from the a allotype B cells led us to examine

whether these B cells were functionally equivalent. Figure 3-5 shows that IgM and IgG2a of both allotypes were produced although the composition is somewhat skewed to the b allotype. Again, this is an expected finding since the cells from the (B6.SJL x B6.TC)F,









49







B6.SJL + F --> F1


A. Class II Expression B. Surface IgM

100- 100
90- -90
80 X 80
70 70
X O
60 O 60
50- 50
40 40
30 30
20 20
10- 10
0 I I I 0
pos neg ANoA pos neg





Figure 3-4 Immunofluorescence on peripheral blood lymphocytes ofB6.SJL +
F1 ---> FI chimeras. Data are presented as percent of cells
expressing class II. A, Class II expression; B, surface IgM.








50

mice would produce either a or b allotype. Similarly, B cells of the a allotype responded equally well when the response to a T cell dependent antigen (HGG) was analyzed in an allotypic fashion (Figure 3-3B). A comparison between positive and negative mice revealed no significant differences. Taken together, we conclude that the inability of haplotype-heterozygous mice to respond to HgCl2 with an ANoA response is an intrinsic property of the B cells.


Absence of ANoA Production in Haplotype-Heterozygous Mice Was Not Due to the
Presence of I-Ab-Restricted T Cells


It was possible that the non-responsiveness of the mixed chimeric as well as the syngeneically reconstituted (B6.SJL x B6.TC)F, mice was due to a negative regulatory effect of I-Ab -restricted T cells. To test for this possibility we lethally irradiated B6. SJL mice and transferred T-cell-depleted bone marrow from (B6.SJL x B6.TC)Fi mice. To ensure no carry over of T cells educated in an (B6.SJL x B6.TC)F1 mouse we treated the recipients at the time of transfer with 0.1 mg i.p. of MmT1 (mouse IgG2a anti-CD90.2

(66)). The T cells in these mice are unable to provide help through I-Ab since they developed in an I-A 'expressing host. Flow cytometric analysis confirmed that all the B cells in these mice expressed both haplotypes. Interestingly, none (0/11) of the mice produced ANoAs following HgCl2 treatment. The resistance of these mice could not have been due to negative influences by I-Ab -restricted T cells.








51







B6.SJL + FI ---> F1


A. Serum IgM B. Senrum IgGC2a

100- 100
90 90
8
80 x 80
70 -8 70

60 60
X
x 0
.0 X O0
S50- O 50

40 X 40
30 30

20 20

10 10

0 I I I 0
pos neg ANoA pos neg



Figure 3-5 Allotype-specific total IgM and IgG2a ELISA data ofB6.SJL + F1
---> F1 chimeras. Data are presented as a percentage of total
isotype. A, Serum IgM; B, serum IgG2a.








52

Discussion


HgCl2 treatment in mice induces a wide range of physiological responses including marked increases in IL-4 production (20,25). The requirement for T cells in this model was shown by the fact that nude mice on the H-2' background fail to develop autoantibodies (15). Splenic CD4+ T cells of HgCl2-treated H-2' mice have been shown to have a strong increase in IL-4 mRNA, whereas those of H-2d mice showed only a weak increase (17). Interestingly, treatment of H-2' mice with anti-IL-4 monoclonal antibody did not prevent ANoA production although it changed the isotype profile of the autoantibodies (25). Despite the possibility for non-cognate help via cytokines, we have shown that MHC-restricted T cell help was required to produce ANoAs. The B cells were not merely responding to cytokines in the environment but required specific MHCrestricted signals from T cells.

Cognate interactions are also required in several other models of autoimmune

disease. Double parental-into-F1 chimeras were used in the graft-vs-host model to show that autoantibodies were derived nearly entirely from B cells receiving direct alloreactive T cell help (51). Bone-marrow chimeras were also used to show that autoantibody production in lpr mice was restricted to those B cells that received T cell help (76). Thus, in three very different murine models of autoimmunity, in vivo autoantibody responses were MHC-restricted.

When the difference in responsiveness between B6.SJL ---> B6.TC and B6.SJL

---> B6.SJL mice was examined it became apparent that the possibility existed for significant differences in the T cell repertoire. Using bone marrow transfers involving








53

class II deficient mice Glimcher et al (77) showed that it is the radioresistant host thymic epithelial cells which mediate positive selection. In our model it was possible that lack of expression of I-A! on the radioresistant thymic population eliminated responsive T cells. The use of (B6.SJL x B6.TC)F1 mice as hosts allowed for positive selection on both I-A! and I-Ab and led us to conclude that responsiveness is dependent on the donor haplotype when T cells for both haplotypes are positively selected.

Mixed chimerism has been shown to prevent autoimmunity in several systems.

Nonobese diabetic mice were protected from diabetes when they were made chimeric with diabetes resistant bone marrow (78,79). It appeared that the autoimmune potential of the NOD cells was restrained. However, when the amount of resistant cells was decreased, a low incidence of insulinitis was seen. Using the rat model of mercury-induced autoimmunity Delaney et al. (72) showed that the presence of resistant bone marrow cells (microchimerism) converted an otherwise sensitive rat strain to a resistant one. If similar to the rat model, it would be expected that the presence of immune competent cells of resistant origin would prevent B6.SJL-derived B cells from responding to HgC12 with an antifibrillarin response. Our experiments differed somewhat in that we used a resistant (B6.SJL x B6.TC)F, mouse which had been coinfused with a combination of resistant and sensitive bone marrow. However, B6.SJL alone as the bone marrow donor completely restored the ability to produce ANoAs upon treatment with HgC12. Despite containing a significant amount of resistant cells we still obtained an intermediate response.

The presence of autoantibodies could play a role in spreading autoimmunity from a dominant epitope to previously cryptic epitopes (80). Presumably, binding of antibody can alter antigen processing, revealing new epitopes. Several other investigators have








54

shown that is it possible to break T cell tolerance to self antigens by coimmunizing mice with self and foreign antigens which in turn generates cross-reactive B cells that can elicit an autoimmune T cell response to previously cryptic self determinants on the autoantigen (73,74,81,82). These forms of epitope spreading do not appear to be occurring in our model. The production of autoantibodies by the B6.SJL B cells in the B6.SJL + (B6.SJL x B6.TC)Fj ---> (B6.SJL x B6.TC)Fl and the B6.SJL + B6.TC ---> (B6.SJL x B6.TC)Fj

mice did not result in the loss of tolerance to fibrillarin by previously resistant B cells.

Several other differences exists between our model and that used by Delaney (72). To induce chimerism without myeloablation the rats were transiently treated with the immunosuppressive agent FK506. Some protection from the manifestations of HgC12induced autoimmune disease was seen in the rats receiving FK506 alone. The authors acknowledged that the results do suggest that transient immunosuppression was an important component of the protection. We used a complete myeloablation procedure which allowed us to forego any immunosuppressive treatment. Another important difference lies in the fact that in the rat model regulatory T cells appear which render the rats resistant to additional HgC12 injections and can confer resistance to naive rats (83). Regulatory T cells have not been identified in the mouse model and the animals do not become resistant to further treatment.

The intermediate response seen in group 3 (Table 3-2) mice was a surprising finding. It was possible that the ANoA negative mice failed to respond because the number of potentially responsive B cells of I-A! origin had been reduced by the presence of resistant B cells of I-Ab origin. This mechanism appears unlikely in as much as the B








55

cell composition of the chimeras had no relationship to the development of an ANoA response. Another antigen non-specific mechanism which could have accounted for the lack of response involves allotype suppression. For example, in studying chronic graft-vshost disease Morris et al. found that in allotype-heterozygous recipients, the autoantibodies were preferentially made by those host cells that expressed the donor allotype, whereas those host B cells that expressed nondonor allotype were relatively suppressed. In allotype-homozygous recipients, the donor cells frequently suppressed the host allotype completely (75). To minimize the possibility that this phenomenon could be occurring in our mice we used allotype-heterozygous mice as hosts. Allotypic analysis of spontaneous as well as antigen specific antibodies demonstrated good participation by both allotypes. Therefore, allotype suppression was not a factor in our mixed chimeras.

A more likely possibility for the poor responsiveness of haplotype-heterozyous mice was lack of specific T cell help. By substituting (B6.SJL x B6.TC)F, for B6.TC as the resistant donor in mixed chimeras also receiving susceptible B6.SJL bone marrow, we provided a mechanism to verify the presence of I-A-restricted T cell help in individual chimeric mice. In those mice with a positive ANoA, T cells capable of helping antifibrillarin-expressing B cells of I-As origin must be present, and these activated T cells, particularly with their reduced stimulation threshold, should also be capable of interacting with antifibrillarin-expressing B cells of I-A' origin. Moreover, by using haplotypeheterozygous mice as co-donors, the overall expression of the resistant I-Ab haplotype was reduced in cells of donor origin. Despite this, there was no increase in ANoA response rate. More surprisingly, in the ANoA positive chimeras, this specificity was limited to the








56

b allotype, indicating that the haplotype-heterozygous B cells failed to participate in this response. These results strongly suggest that lack of responsiveness is an intrinsic regulatory property of B cells.

The presence of I-Ab restricted T cell help could have accounted for the

intermediate response seen in the mixed chimeras. The (B6.SJL x B6.TC)F1 mice that served as hosts provided an environment in which T cells would be selected to interact with I-Ab expressed on the B cells from the B6.TC hosts. By transferring (B6.SJL x B6.TC)F1 bone marrow into B6.SJL mice, there would be little potential for positive selection of T cells on I-Ab. Despite the absence of these cells, there was no ANoA response to HgCI2. This again points to an intrinsic defect in (B6.SJL x B6.TC)F1 B cells.

One possibility for this intrinsic property is the concept of MHC-guided processing leading to determinant capture (84). This hypothesis states that when an antigen is taken up by APCs and begins unfolding, different MHC molecules can compete for determinants. Once it is bound by class II, the antigen is then trimmed down to its final size while the remainder of the antigen including cryptic epitopes is discarded. An example of this phenomenon was seen in the autoimmune disease insulin-dependent diabetes mellitus (85). The response to the subdominant ANOD-restricted determinant of HEL disappears when NOD mice were made transgenic by introduction of the Efa. The responsivness was restored when scission of the HEL separated this determinant from its adjoining, competitively dominant, Ed-restricted determinant. This suggested that the Ed molecule bound and protected its dominant determinant on a long peptide while captured neighboring determinants were lost during proteolysis. In our mice, 1-Ab could effectively








57

be binding fibrillarin in the (B6.SJL x B6.TC)F1 cells, thereby preventing I-As from presenting fibrillarin and receiving T cell help.

Recently the molecular and antigenic properties of mercury-modified fibrillarin has been examined. The exposure offibrillarin both in vivo and in vitro caused a change in its migration under non-reducing SDS-PAGE and resulted in a loss of reactivity to autoantibodies. Mutation of the cysteines in fibrillarin resulted in a loss of mercuryinduced modification. The authors concluded that unmodified fibrillarin is the B cell antigen while the T cell antigen appears to be mercury-modified fibrillarin (38). Therefore, if our model is correct, the presence of I-Ab affected the processing of Hgmodified fibrillarin in the context of I-As and provide a unique opportunity for testing the role of antigen competition in an important environmental model of induced autoimmunity.













CHAPTER 4
DISCUS SION/CONCLUSION


Humans are continually exposed to mercury via various routes resulting in high levels in skin, nails, hair and the kidneys. Many exposures are a result of mercury being widely distributed in the environment. The major natural source of mercury is the degassing of the earths crust, and is estimated to produce between 2700 to 6000 tons per year. Approximately 10,000 tons of mercury are mined each year. Even industrial activities not directly employing mercury or mercury products give rise to substantial amounts. Fossil fuels, for example, may contain up to I ppm resulting in 5000 tons per year being emitted from burning coal, natural gas and from refining petroleum products

(86). Human are also exposed to mercury in the form of pharmaceuticals, cosmetics and dental amalgams.

On the basis of toxicologic characteristics, there are three forms of mercury; elemental, inorganic, and organic compounds. Both organic and inorganic forms of mercury undergo environmental transformation. Metallic mercury may be oxidized to inorganic divalent mercury, particularly in the presence of organic material. Divalent mercury may, in turn, be reduced to metallic mercury when reducing conditions are present. Anaerobic bacteria are capable of methylating the divalent form resulting in dimethyl mercury. In tissues, methyl mercury undergoes biotransformation to divalent mercury compounds (86).


58








59

Metallic or elemental mercury volatilizes to mercury vapor at ambient

temperatures, and most human exposure is by inhalation. Acute exposure can result in corrosive bronchitis and interstitial pneumonia. Long term exposure to mercury vapor targets the central nervous system producing a triad of signs including; increased excitability, tremors and gingivitis. Renal effects are sometimes seen with chronic mercury vapor exposure and may be similar to that which occurs following inorganic mercury exposure. The brain is also a target organ for methyl mercury. Environmental exposures, such as eating mercury containing fish can result in neurotoxic effects. These include parathesia, ataxia, neurasthenia, vision and/or hearing loss, tremors and possibly coma followed by death (86).

Inorganic mercury salts may be present as divalent (mercuric) or monovalent

(mercurous). Ingestion of high doses of this form of mercury causes corrosive ulceration and necrosis of the gastrointestinal tract followed by circulatory collapse. If this initial insult is survived, renal failure occurs within 24 hours due to necrosis of the proximal tubular epithelium.

Chronic low-dose exposure to inorganic mercury can cause immunological aberrations in both humans and rabbits (27). Immunosuppressive effects have been reported in mice resulting in impaired host resistance (87). The immunoactivating properties of inorganic mercury in mice have been widely studied and can be divided into three major pathological sequelae: lymphoproliferation, hypergammaglobulinemia and the development of autoimmunity (27). Lymphoproliferation is classified as an MHCindependent effect while autoinmunity development is tightly restricted by haplotype.








60

Currently, the link between MHC haplotype and susceptibility to HgCl2 induced ANoAs has remained unexplained. The experiments conducted for this dissertation were begun with the idea of examining just how the MHC haplotype determines the response to HgCl2. Through the use of H-2-recombinants it has been shown that responsiveness could be mapped to the I-A region of the MHC (6). Mice of the H-2' haplotype are high responders while those of the H-2b haplotype are resistant (5). In an F, cross between two different haplotypes, the products of both alleles are expressed on the same cell, so expression is said to be co-dominant. Therefore, an F, (s/b) mouse should be susceptible to ANoA induction. This result has been reported in the literature although the response was somewhat attenuated (5,9,43). When we generated s/b haplotype mice using H-2congenic founder strains, we surprisingly found that the mice were resistant. This led us to examine how the resistant haplotype was exerting its effects.

It was possible that a gene linked to the resistant haplotype and not the I-A

molecule itself that could be preventing the response. To this end we generated F, mice between resistant and susceptible strains in which the resistant class II genes were eliminated. These mice did develop ANoAs which showed that the downregulation was a result of the resistant allele itself and not a linked gene.

We also directly examined what effects, if any, that the other class II molecule, IE, had on susceptibility. Previous studies in the urine mercury model have shown expression to either suppress (6) or have no effect (88) on susceptibility. The protection afforded by I-E is not without precedent. Expression of I-E in lupus prone B6.lpr mice significantly reduced the production of autoantibodies as well as lowering the spleen and








61

lymph node weights (48). Using bone marrow chimeras in which I-E-positive and I-Enegative B cells co-existed, it was determined that the amelioration of signs resulted from a direct effect of I-E on the B cell This phenomenon may be a universal feature of autoimmune disease on the H-2B background. In an unrelated autoimmune mouse model, BXSB, which also bears the H2b haplotype, expression of I-E prevented the development of SLE (47), and a potential mechanism was proposed. It was shown that the a chain of the I-E molecule generates a peptide with high affinity to I-Ab thereby competing with the pathogenic autoantigen-derived peptides for presentation to B cells. Interestingly, if this same mechanism were important in the HgCl2 model, we would predict just the opposite results, in our (B6.SJL X B6.I-E)F1 mice. Since the that lack of response in s/b mice was due to expression of I-Ab, the I-E a-derived peptide could have preferentially blocked I-Ab and enhanced the I-A-mediated ANoA response. However, we found no enhancing effect on autoimmunity by co-expression of I-E on F, mice. One possibility for this lack of effect was that an I-E-derived peptide might also have an equal or even greater affinity for I-A. Against this possibility was our results that co-expression of I-E on the H-2' homozygous background did not attenuate the autoimmune response, as would be expected based on the experience with the BXSB and lpr models of autoimmunity. Therefore, the lack of autoantibody production in these F, mice led us to investigate other mechanisms whereby I-Ab could be down regulating the response.

For these studies, we turned to bone marrow chimeras. Adoptive bone marrow transfer is an important tool that can be utlitized to dissect the cellular interactions in an adaptive immune response by manipulating the environment in which T cell development occurs thereby modifying the T cell repertoire. By taking advantage of the differences in








62

positive and negative selection seen between host and donor cells we showed that MHC restricted T cell help was required to produce ANoAs. This result was not surprising but was instrumental in interpreting future results. The next possibility we examined was whether or not the differences in T cell repertoire could account for the differences in susceptibility. By using an F1 cross between the two donor strains we eliminated the T cell differences yet still saw differences in responsiveness.

The intermediate response seen when resistant and sensitive bone marrow was coinfused into a haplotype heterozygous host was surprising. Several possibilities could have accounted for this. The first possibility, dilution of responsive B cells, was ruled out by noting a lack of any correlation between B cell composition and responsiveness. Secondly, we needed to rule out that allotype suppression in which autoantibodies are preferentially produced by those B cells expressing the same haplotype as the host was occurring in these chimeras. We attempted to eliminate this possibility by using F, hosts. We also verified that allotype suppression was not an important mechanism in the chimeras by observing a brisk response by both allotypes when immunized with an exogenous antigen. The final possibility we examined that may have led to the intermediate result involved the question of T cell help. By using F, rather than B6 mice as the resistant co-donor and observing an anti-nucleolar response by the B cells of donor B6.SJL origin, we were assured that activated T cells capable of providing I-A-restricted help to antifibrillarin B cells were present. Despite the presence of adequate T cell help that should have been able to interact with (s/b)F B cells, no increase in susceptibility was observed. The fact that these mice produced ANoA only of the b allotype combined with








63

the observation that no response was seen when I-Ab restricted T cells were eliminated led us to conclude that the resistance of the F1 B cells was due to an intrinsic defect of the B cells themselves.

Several theories have been proposed as to how mercury can lead to autoimmunity. It has been suggested that mercury may alter self antigen by complexing with the antigen itself. As mentioned above, divalent mercury is a highly reactive molecule with a propensity to bind sulfhydryl groups of proteins and nonprotein thiols. Reactivity has also been found towards hydroxyl, carboxyl and phosphoryl groups albeit at a lower affinity

(89). These modified self antigens may then be recognized by the immune system as foreign and in turn induce an immune response capable of recognizing the modified self antigen.

During T cell development, self tolerance is only induced to efficiently-presented dominant epitopes but not to cryptic ones (90). Therefore, potentially autoreactive T cells which recognize these cryptic epitopes that would arise by random rearrangement of the T cell receptor would not be eliminated by negative selection. The dominant epitope presented is influenced by a number of factors and is incompletely understood. Factors that have been described include the haplotype, protein structure and/or conformation. Changes in any of these factors could theoretically result in novel cleavage products. This phenomenon has been observed for many of the autoantigens recognized by SLE patients. Casciola-Rosen et al. (91) showed that during apoptosis the lupus autoantigens cluster and are concentrated in the surface blebs of apoptotic cells. This could result in the autoantigens becoming substrates for enzymes not present within the cell resulting in new








64

cleavage products. A follow-up study by the same group (90) showed that several of the autoantigens targeted in scleroderma are susceptible to reactive oxygen species in a metaldependent manner. This cleavage generates unique epitopes resulting in autoantibody formation. They propose that the repetitive episodes of ischemia-reperfusion, which occurs in these patients as a result of vasomotor instability of their arterioles, are responsible for the generation of the reactive oxygen species. In contrast, fibrillarin was not fragmented under identical conditions, thereby suggesting an alternative mechanism was responsible for development of this specificity in a subset of scleroderma patients.

Pollard et al. (38) proposed a mechanism whereby mercury can induce antifibrillarin autoantibodies. They showed that mercury-induced cell death was associated with a loss of fibrillarin antigenicity and modification of the molecular properties of fibrillarin. By mutating the cysteines in fibrillarin to alanines they confirmed that HgCl2 exposure in vitro leads to a disulfide bonded form of fibrillarin which is poorly recognized by autoantibodies using either indirect immunofluorescence or immunoprecipitation. This suggests that the mercury-induced change involves a conformational structure change that is no longer detected by autoantibodies.

The above mechanisms do not take into account the tight link between

susceptibility to autoimmunity and the mouse MHC. We propose a model which elaborates on both the modified self-antigen theory and takes into account the lack of response seen in haplotype-heterozygous mice.

During T cell development, fibrillarin is processed and the immunodominant peptide(s) is presented. T cells which recognize this peptide are deleted resulting in








65

tolerance to fibrillarin. On exposure to HgCl2, a disulfide bond forms which affects the immunodominant peptide normally presented by I-As, resulting in the presentation of a cryptic epitope. This cryptic peptide could potentially take a number of forms. The production of a mercury-modified disulfide link could protect the area around that peptide from being digested and the resultant cryptic peptide would be a normal sequence derived from another portion of the fibrillarin molecule. This would be a remote effect. It could also result from a conformational change that exposes a buried sequence to digestion. Mercury-modified disulfide could provide a completely novel peptide sequence centered around the cysteine-Hg-cysteine sequence. Arguing against this possibility is the fact that no mercury was detected in the glomerular basement membrane in HgC12-treated rabbits using autoradiography (92). Another possibility is that the disulfide residue could be rereduced during processing but one of the amino acids in the binding groove has Hg-linked side chain that alters the immunogeneicity. The net result, then, would be the expression of a self-epitope not previously encountered. The immune system would recognize this as foreign, resulting in autoimmunity. In the case of a b haplotype mouse, the disulfide bond doesn't interfere with the processing and presentation of its immunodominant peptide. No cryptic epitopes are revealed and hence no autoimmunity to fibrillarin develops.

Our experiments with haplotype heterozygous mice adds some interesting potential insights into antigen processing and the expression of cryptic epitopes. We propose that the cryptic epitope presented by I-As is either the same or overlaps with the peptide normally presented by I-Ab. When an s/b haplotype mouse is exposed to mercury, a competition for access to that part of fibrillarin that could be bound by either of the two








66

class II alleles will result. If the I-Ab molecule has a greater avidity for this peptide than IAs, no new peptide is seen in the context of I-A! and no autoimmunity results.

There is a precedent for the idea that different class II molecules actually compete for the same or overlapping peptide. Sercarz et al, (84) showed that competition takes place between I-A and I-E in the mouse. They coined the phrase "NMC guided processing" to describe the process where the MHC class II molecule would encounter a tightly folded protein antigen in one of the acidic compartments of the endosomallysosomal system soon after preliminary processing had succeeded in partially unfolding it, making one or a few determinants available. The dominant determinant would be the one that initially won the competition for binding to the MC and provided the binding was stable, the residues bound within the MHC would be protected from further proteolysis or binding by a different MUG molecule. Thus far, this has been described in in vitro models. Our experiments in the mercury chloride-induced model would be the first demonstration of its potential importance in vivo.

Our investigations of mercury-induced autoimmunity in mice have stimulated the need for future investigations addressing the many questions spawned by our initial research into this area. Several major questions need to be addressed to either support or refute our model. Firstly, if in fact a competition is ongoing between MUG molecules in haplotype heterozygous mice, will increasing the dose of mercury which in turn should provide a greater pool of modified fibrillarin result in these animals now becoming susceptible? This is a fairly straightforward question to address. The second and perhaps more fundamental question to be addressed is the relationship between the T cell and B








67

cell epitopes. As previously discussed, the T cell epitope has been proposed to be mercury-modified fibrillarin, while the B cell epitope is a conformational determinant lost by exposure to mercury (38). This absence of an antibody response to mercury-modified fibrillarin is a surprising one. One possible explanation is that binding of the mercurymodified epitope by surface Ig protects that portion of the molecule from antigen processing. The net result is that no mercury-induced cryptic determinants are presented by those B cells with specificity for the mercury modification. While reasonable, this raises yet another issue: how then is it possible for a specific immune response to develop if the B cells that can bind to native fibrillarin do not have a mercury-modified T-cell epitope to present? One possibility is that the mercury either directly or indirectly causes an aggregation of fibrillarin in which some molecules are mercury-modified and some are still in the native conformation. When a B cell internalizes this aggregate specifically via its surface Ig receptor, which recognizes native fribrillarin, it processes the mercury modified sites within the aggregate and presents it to T cells, which recognize this modified form as foreign. While this hypothesis readily explains the potential for T-B interactions in the generation of a Hg-induced response to native fibrillarin it re-opens the question as to why there is no response to Hg-modified fibrillarin. It would seem unlikely that surface Ig could bind every Hg-modified residue in an aggregate. Failure to do so would allow the B cell specific for Hg-modified fibrillarin to present the cryptic epitope and thereby receive T cell help.

We have shown that it is the I-Ab molecule that confers resistance to HgCI2induced autoimmunity in F, (s/b) mice. Through the use of bone marrow chimeric animals








68

we determined that this effect is not mediated by a change in the T-cell repertoire. It appears that theAMC may be exerting its effects by manipulating the antigen processing/presentation pathway. To address this, one would need to elute and sequence the peptides that are actually being presented on the surface of the antigen presentation cells. Similarly, to address the apparent paradox between the T cell and B cell epitopes one would need to obtain the same information for both the T cell and the B cell receptors.












LIST OF REFERENCES


1. Theofilopoulos, A. N. 1993. Molecular Pathology of Autoimmunity. In The Molecular Pathology ofAutoimmune Diseases. C.A. Bona, K.A. Siminovitch, M. Zanetti and A.N. Theofilopoulos, eds. Harwood Academic Publishers, Chur, Switzerland, p. 1.

2. Condemi, J. J. 1992. The autoimmune diseases. JAMA 268:2882.

3. Tubbs, R. R., G. N. Gephardt, J. T. McMahon, M. C. Pohl, D. G. Vidt, S. A.
Barenberg, and R. Valenzuela. 1982. Membranous glomerulonephritis associated with industrial mercury exposure. Study of pathogenetic mechanisms. Am. J Clin. Pathol. 77:409.

4. Robinson, C. J., A. A. Abraham, and T. Balazs. 1984. Induction of anti-nuclear antibodies by mercuric chloride in mice. Clin. Exp. Immunol. 58:300.

5. Robinson, C. J., T. Balazs, and I. K. Egorov. 1986. Mercuric chloride-, gold sodium thiomalate-, and D-penicillamine-induced antinuclear antibodies in mice. Toxicol. Appl. Pharmacol. 86:159.

6. Mirtcheva, J., C. Pfeiffer, J. A. De Bruijn, F. Jacquesmart, and E. Gleichmann. 1989. Immunological alterations inducible by mercury compounds. III. H-2A acts as an immune response and H-2E as an immune "suppression" locus for HgCl2-induced antinucleolar autoantibodies. Eur. J 1mmunol. 19:2257.

7. Hultman, P., L. J. Bell, S. Enestrom, and K. M. Pollard. 1992. Murine susceptibility to mercury. I. Autoantibody profiles and systemic immune deposits in inbred, congenic, and intra-H-2 recombinant strains. Clin. Immunol. Immunopathol. 65:98.

8. Hanley, G. A., J. Schiffenbauer, and E. S. Sobel. 1997. Class II haplotype differentially regulates immune response in HgCl2-treated mice. Clin. Immunol. Immunopathol. 84:328.

9. Hultman, P., L. J. Bell, S. Enestrom, and K. M. Pollard. 1993. Murine susceptibility to mercury. II. autoantibody profiles and renal immune deposits in hybrid, backcross, and H-2d congenic mice. Clin. Immunol. Immunopathol. 68:9.

10. Hultman, P. and S. Enestrom. 1988. Mercury induced antinuclear antibodies in mice: characterization and correlation with renal immune complex deposits. Clin. Exp. Immunol. 71:269.


69








70

11. Robinson, C. J., H. J. White, and N. R. Rose. 1997. Murine strain differences in response to mercuric chloride: antinucleolar antibodies production does not correlate with renal immune complex deposition. Clin. Immunol. Immunopathol. 83:127.

12. Hultman, P., S. J. Turley, S. Enestrom, U. Lindh, and K. M. Pollard. 1996. Murine Genotype Influences the Specificity, Magnitude and Persistence of Murine Mercury-Induced Autoimmunity. J. Autoimmunity 9:139.

13. Griem, P., E. Scholz, M. Turfeld, D. Zander, U. Wiesner, L. Dunemann, and E. Gleichmann. 1997. Strain differences in tissue concentrations of mercury in inbred mice treated with mercuric chloride. Toxicol. Appl. Pharmacol. 144:163.

14. Christensen, M. M. 1996. Histochemical localization of autometallographically detectable mercury in tissues of the immune system from mice exposed to mercuric chloride. Histochem. J. 28:217.

15. Hultman, P., U. Johansson, and F. Dagnaes-Hansen. 1995. Murine Mercury-Induced Autoimmunity: the Role of T-Helper Cells. J. Autoimmunity 8:809.

16. Johansson, U., B. Sander, and P. Hultman. 1997. Effects of the murine genotype on T cell activation and cytokine production in murine mercury-induced autoimmunity. J. Autoimmun. 10:347.

17. van Vliet, E., M. Uhrberg, C. Stein, and E. Gleichmann. 1993. MHC control of IL-4-dependent enhancement of B cell Ia expression and Ig class switching in mice treated with mercuric chloride. Int. Arch. Allergy Immunol. 101:392.

18. Hu, H., M. Abedi Valugerdi, and G. Moller. 1997. Pretreatment of lymphocytes with mercury in vitro induces a response in T cells from genetically determined low-responders and a shift of the interleukin profile. Immunology 90:198.

19. Jiang, Y. and G. Moller. 1996. In vitro effects of HgCl2 on murine lymphocytes. II. Selective activation of T cells expressing certain V beta TCR. Int. Immunol. 8:1729.

20. Gillespie, K. M., F. J. Qasim, L. M. Tibbatts, S. Thiru, D. B. Oliveira, and P. W.
Mathieson. 1995. Interleukin-4 gene expression in mercury-induced autoimmunity. Scand. J. Immunol. 41:268.

21. Prigent, P., A. Saoudi, C. Pannetier, P. Graber, J. Y. Bonnefoy, P. Druet, and F. Hirsch. 1995. Mercuric chloride, a chemical responsible for T helper cell (Th)2-mediated autoimmunity in brown Norway rats, directly triggers T cells to produce interleukin-4. J Clin. Invest. 96:1484.








71

22. Goldman, M., P. Druet, and E. Gleichmann. 1991. TH2 cells in systemic autoimmunity: insights from allogeneic diseases and chemically-induced autoimmunity. Immunol. Today 12:223.

23. Pietsch, P., H. W. Vohr, K. Degitz, and E. Gleichmann. 1989. Immunological alterations inducible by mercury compounds. II. HgCl2 and gold sodium thiomalate enhance serum IgE and IgG concentrations in susceptible mouse strains. Int. Arch. Allergy Appl. Immunol. 90:47.

24. Lundgren, M., U. Persson, P. Larsson, C. Magnusson, C. I. Smith, L. Hammarstrom, and E. Severinson. 1989. Interleukin 4 induces synthesis of IgE and IgG4 in human B cells. Eur. J Immunol. 19:1311.

25. Ochel, M., H. W. Vohr, C. Pfeiffer, and E. Gleichmann. 1991. IL-4 is required for the IgE and IgG1 increase and IgG1 autoantibody formation in mice treated with mercuric chloride. J. Immunol. 146:3006.

26. Doth, M., M. Fricke, F. Nicoletti, G. Garotta, M. L. Van Velthuysen, J. A. Bruijn, and E. Gleichmann. 1997. Genetic differences in immune reactivity to mercuric chloride (HgCl2): immunosuppression of H-2d mice is mediated by interferon-gamma (IFN-gamma). Clin. Exp. Immunol. 109:149.

27. Pollard, K. M. 1997. Induction of Autoimmunity by Mercuric Chloride. Fifth International Workshop on the Molecular and Cell Biology ofAutoantibodies and Autoimmunity 24.(Abstract)

28. Hultman, P. and S. Enestrom. 1992. Dose-response studies in murine mercury-induced autoimmunity and immune-complex disease. Toxicol. Appl. Pharmacol. 113:199.

29. Hultman, P. and S. Enestrom. 1989. Mercury induced B-cell activation and antinuclear antibodies in mice. J. Clin. Lab. Immunol. 28:143.

30. Hu, H., G. Moller, M. Abedi Valugerdi, U. Johansson, B. Sander, and P. Hultman. 1997. Major histocompatibility complex class II antigens are required for both cytokine production and proliferation induced by mercuric chloride in vitro Effects of the murine genotype on T cell activation and cytokine production in murine mercury-induced autoimmunity. J. Autoimmun. 10:347.

31. Reuter, R., G. Tessars, H. W. Vohr, E. Gleichmann, and R. Luhrmann. 1989. Mercuric chloride induces autoantibodies against U3 small nuclear ribonucleoprotein in susceptible mice. Proc. Natl. Acad Sci. U S. A. 86:237.








72

32. Hultman, P., S. Enestrom, K. M. Pollard, and E. M. Tan. 1989. Anti-fibrillarin autoantibodies in mercury-treated mice. Clin. Exp. Immunol. 78:470.

33. Okano, Y., V. D. Steen, and T. A. Medsger. 1992. Autoantibody to U3 nucleolar ribonucleoprotein (fibrillarin) in patients with systemic sclerosis. Arthritis Rheum. 35:95.

34. Kasturi, K. N., A. Hatakeyama, H. Spiera, and C. A. Bona. 1995. Antifibrillarin autoantibodies present in systemic sclerosis and other connective tissue diseases interact with similar epitopes. J. Exp. Med 181:1027.

35. Takeuchi, K., S. J. Turley, E. M. Tan, and K. M. Pollard. 1995. Analysis of the autoantibody response to fibrillarin in human disease and murine models of autoimmunity. J. Immunol. 154:961.

36. Kubicka Muranyi, M., J. Kremer, N. Rottmann, B. Lubben, R. Albers, N. Bloksma, R. Luhrmann, and E. Gleichmann. 1996. Murine systemic autoimmune disease induced by mercuric chloride: T helper cells reacting to self proteins. Int. Arch. Allergy Immunol. 109:11.

37. Kubicka Muranyi, M., P. Griem, B. Lubben, N. Rottmann, R. Luhrmann, and E. Gleichmann. 1995. Mercuric-chloride-induced autoimmunity in mice involves up-regulated presentation by spleen cells of altered and unaltered nucleolar self antigen. nt. Arch. Allergy Immunol. 108:1.

38. Pollard, K. M., D. K. Lee, C. A. Casiano, M. Bluthner, M. M. Johnston, and E. M.
Tan. 1997. The autoimmunity-inducing xenobiotic mercury interacts with the autoantigen fibrillarin and modifies its molecular and antigenic properties. J. Immunol. 158:3521.

39. Kass, S., K. Tyc, J. A. Steitz, and B. Sollner Webb. 1990. The U3 small nucleolar ribonucleoprotein functions in the first step of preribosomal RNA processing. Cell 60:897.

40. Lubben, B., N. Rottmann, M. Kubicka Muranyi, E. Gleichmann, and R. Luhrmann. 1994. The specificity of disease-associated anti-fibrillarin autoantibodies compared with that of HgCl2-induced autoantibodies. Mol. Biol. Rep. 20:63.

41. Bencko, V., V. Wagner, M. Wagnerova, and V. Ondrejcak. 1990. Immunological profiles in workers occupationally exposed to inorganic mercury. J. Hyg. Epidemiol. Microbiol. Immunol. 34:9.

42. Tubbs, R. R., G. N. Gephardt, J. T. McMahon, M. C. Pohl, D. G. Vidt, S. A.
Barenberg, and R. Valenzuela. 1982. Membranous glomerulonephritis associated with industrial mercury exposure. Study of pathogenetic mechanisms. Am. J. Clin. Pathol. 77:409.








73

43. Stiller Winkler, R., T. Radaszkiewicz, and E. Gleichmann. 1988. Immunopathological signs in mice treated with mercury compounds--I. Identification by the popliteal lymph node assay of responder and nonresponder strains. Int. J. Immunopharmacol. 10:475.

44. Walter, W., M. Loos, and M. J. Maeurer. 1996. H2-M polymorphism in mice susceptible to collagen-induced arthritis involves the peptide binding groove. Immunogenetics 44:19.

45. Widera, G., L. C. Burkly, C. A. Pinkert, E. C. Bottger, C. Cowing, R. D. Palmiter, R.
L. Brinster, and R. A. Flavell. 1987. Transgenic mice selectively lacking MHC class II (I-E) antigen expression on B cells: an in vivo approach to investigate Ia gene function. Cell 51:175.

46. Grusby, M. J., R. S. Johnson, V. E. Papaioannou, and L. H. Glimcher. 1991.
Depletion of CD4+ T cells in major histocompatibility complex class II-deficient mice. Science 253:1417.

47. Merino, R., M. Iwamoto, L. Fossati, P. Muniesa, K. Araki, S. Takahashi, J. Huarte, K. Yamamura, J. D. Vassalli, and S. Izui. 1993. Prevention of systemic lupus erythematosus in autoimmune BXSB mice by a transgene encoding I-E alpha chain. J Exp. Med. 178:1189.

48. Creech, E. A., D. Nakul Aquaronne, E. A. Reap, R. L. Cheek, P. A. Wolthusen, P. L.
Cohen, and R. A. Eisenberg. 1996. MHC genes modify systemic autoimmune disease. The role of the I-E locus. J lmmunol. 156:812.

49. Kappler, J. W., B. Skidmore, J. White, and P. Marrack. 1981. Antigen-inducible, H-2-restricted, interleukin-2-producing T cell hybridomas. Lack of independent antigen and H-2 recognition. J. Exp. Med. 153:1198.

50. Ozato, K., N. Mayer, and D. H. Sachs. 1980. Hybridoma cell lines secreting monoclonal antibodies to mouse H-2 and Ia antigens. J. Immunol. 124:533.

51. Morris, S. C., R. L. Cheek, P. L. Cohen, and R. A. Eisenberg. 1990. Autoantibodies in
chronic graft versus host result from cognate T-B interactions. J. Exp. Med 171:503.

52. Aris, J. P. and G. Blobel. 1988. Identification and characterization of a yeast nucleolar protein that is similar to a rat liver nucleolar protein. J. Cell Biol. 107:17.

53. O'Shannessy, D. J., M. J. Dobersen, and R. H. Quarles. 1984. A novel procedure for labeling immunoglobulins by conjugation to oligosaccharide moieties. Immunol. Lett. 8:273.








74

54. Goding, J. W. 1976. Conjugation of antibodies with fluorochromes: modifications to the standard methods. J Immunol. Methods 13:215.

55. Ceman, S., R. A. Rudersdorf, J. M. Petersen, and R. DeMars. 1995. DMA and DMB are the only genes in the class II region of the human MHC needed for class II-associated antigen processing. J. Immunol. 154:2545.

56. Riberdy, J. M. and P. Cresswell. 1992. The antigen-processing mutant T2 suggests a role for MHC-linked genes in class II antigen presentation [published erratum appears in J Immunol 1992 Aug 1;149(3):1113]. J. Immunol. 148:2586.

57. Hill, S. W., D. E. Kipp, I. Melchers, J. A. Frelinger, and E. E. Sercarz. 1980. Multiple
H-2 and non-H-2 genes controlling the antilysozyme response: alternative gene constellations can lead to responsiveness. Eur. J. Immunol. 10:384.

58. Martin, J., B. Y. Wei, and C. S. David. 1993. Expression of MHC class II A alpha k transgene. Pairing with endogenous beta-chain is dependent upon affinity and competition. Transplantation 56:1233.

59. Layet, C. and R. N. Germain. 1991. Invariant chain promotes egress of poorly expressed, haplotype-mismatched class II major histocompatibility complex A alpha A beta dimers from the endoplasmic reticulum/cis-Golgi compartment. Proc. Natl. Acad Sci. U S. A. 88:2346.

60. Mathis, D. J., C. Benoist, V. E. Williams, M. Kanter, and H. O. McDevitt. 1983.
Several mechanisms can account for defective E alpha gene expression in different mouse haplotypes. Proc. Natl. Acad. Sci. U. S. A. 80:273.

61. Nygard, N. R., D. M. McCarthy, J. Schiffenbauer, and B. D. Schwartz. 1993. Mixed haplotypes and autoimmunity. Immunol. Today 14:53.

62. Hirose, S., G. Ueda, K. Noguchi, T. Okada, I. Sekigawa, H. Sato, and T. Shirai. 1986. Requirement of H-2 heterozygosity for autoimmunity in (NZB X NZW)F1 hybrid mice. Eur. J. Immunol. 16:1631.

63. Sobel, E. S., T. Katagiri, K. Katagiri, S. C. Morris, P. L. Cohen, and R. A. Eisenberg.
1991. An intrinsic B cell defect is required for the production of autoantibodies in the lpr model of murine systemic autoimmunity. J Exp. Med 173:1441.

64. Ceredig, R., J. W. Lowenthal, M. Nabholz, and H. R. MacDonald. 1985. Expression of interleukin-2 receptors as a differentiation marker on intrathymic stem cells. Nature 314:98.








75

65. Sarmiento, M., A. L. Glasebrook, and F. W. Fitch. 1980. IgG or IgM monoclonal antibodies reactive with different determinants on the molecular complex bearing Lyt 2 antigen block T cell-mediated cytolysis in the absence of complement. J. Immunol. 125:2665.

66. Kremmer, E., S. Thierfelder, U. Kummer, R. Lederer, and J. Mysliwietz. 1989. Neutralization of immunosuppression by antibodies against variable as well as constant regions of monoclonal anti-Thy-1 xenoantibodies and their ability to be suppressed by initial T cell depletion. Transplantation 47:641.

67. Oi, V. T., P. P. Jones, J. W. Goding, and L. A. Herzenberg. 1978. Properties of
monoclonal antibodies to mouse Ig allotypes, H-2, and Ia antigens. Curr. Top. Microbiol. Immunol. 81:115.

68. Stall, A. M. and M. R. Loken. 1984. Allotypic specificities ofmurine IgD and IgM recognized by monoclonal antibodies. J Immunol. 132:787.

69. Sieckmann, D. G., A. M. Stall, and B. Subbarao. 1991. A mouse monoclonal antibody specific for an allotypic determinant of the Igha allele of murine IgM: genetic and functional analysis of Igh-6a epitopes using anti-IgM monoclonal antibodies. Hybridoma 10:121.

70. Cosgrove, D., S. H. Chan, C. Waltzinger, C. Benoist, and D. Mathis. 1992. The thymic compartment responsible for positive selection of CD4+ T cells. Int. Immunol. 4:707.

71. Sprent, J., H. Kosaka, and E. K. Gao. 1992. T cell tolerance after bone marrow transplantation in mice. Bone Marrow Transplant. 10 Suppl 1:5.

72. Delaney, C. P., N. Murase, M. Chen Woan, J. J. Fung, T. E. Starzl, and A. J.
Demetris. 1996. Allogeneic hematolymphoid microchimerism and prevention of autoimmune disease in the rat. A relationship between allo- and autoimmunity. J. Clin. Invest. 97:217.

73. Mamula, M. J., S. Fatenejad, and J. Craft. 1994. B cells process and present lupus autoantigens that initiate autoimmune T cell responses. J. Immunol. 152:1453.

74. Mamula, M. J., R. H. Lin, C. A. Janeway,Jr., and J. A. Hardin. 1992. Breaking T cell
tolerance with foreign and self co-immunogens. A study of autoimmune B and T cell epitopes of cytochrome c. J Immunol. 149:789.

75. Morris, S. C., R. L. Cheek, P. L. Cohen, and R. A. Eisenberg. 1990. Allotype-specific
immunoregulation of autoantibody production by host B cells in chronic graft-versus host disease. J Immunol. 144:916.








76

76. Sobel, E. S., V. N. Kakkanaiah, M. Kakkanaiah, R. L. Cheek, P. L. Cohen, and R. A.
Eisenberg. 1994. T-B collaboration for autoantibody production in lpr mice is cognate and MHC-restricted. J Immunol. 152:6011.

77. Markowitz, J. S., H. Auchincloss,Jr., M. J. Grusby, and L. H. Glimcher. 1993. Class
II-positive hematopoietic cells cannot mediate positive selection of CD4+ T lymphocytes in class II-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 90:2779.

78. Li, H., C. L. Kaufman, S. S. Boggs, P. C. Johnson, K. D. Patrene, and S. T. Ildstad.
1996. Mixed allogeneic chimerism induced by a sublethal approach prevents autoimmune diabetes and reverses insulitis in nonobese diabetic (NOD) mice. J. Immunol. 156:380.

79. Mathieu, C., K. Casteels, R. Bouillon, and M. Waer. 1997. Protection against autoimmune diabetes in mixed bone marrow chimeras: mechanisms involved. J. Immunol. 158:1453.

80. Satoh, M., A. K. Ajmani, L. Stojanov, J. J. Langdon, T. Ogasawara, J. Wang, M. A.
Dooley, H. B. Richards, J. B. Winfield, T. H. Carter, and W. H. Reeves. 1996. Autoantibodies that stabilize the molecular interaction of Ku antigen with DNA-dependent protein kinase catalytic subunit. Clin. Exp. Immunol. 105:460.

81. Lin, R. H., M. J. Mamula, J. A. Hardin, and C. A. Janeway,Jr. 1991. Induction of
autoreactive B cells allows priming of autoreactive T cells. J. Exp. Med 173:1433.

82. Fatenejad, S., M. J. Mamula, and J. Craft. 1993. Role ofintermolecular/intrastructural B- and T-cell determinants in the diversification of autoantibodies to ribonucleoprotein particles. Proc. Natl. Acad. Sci. U. S. A. 90:12010.

83. Bowman, C., D. W. Mason, C. D. Pusey, and C. M. Lockwood. 1984. Autoregulation of autoantibody synthesis in mercuric chloride nephritis in the Brown Norway rat. I. A role for T suppressor cells. Eur. J Immunol. 14:464.

84. Sercarz, E. E., P. V. Lehmann, A. Ametani, G. Benichou, A. Miller, and K. Moudgil. 1993. Dominance and crypticity of T cell antigenic determinants. Annu. Rev. Immunol. 11:729.

85. Deng, H., R. Apple, M. Clare Salzler, S. Trembleau, D. Mathis, L. Adorini, and E. Sercarz. 1993. Determinant capture as a possible mechanism of protection afforded by major histocompatibility complex class II molecules in autoimmune disease. J Exp. Med. 178:1675.

86. Goyer, R. A. 1996. Toxic Effects of Metals. In Casarett andDoull's Toxicology. Fifth ed. C.D. Klaassen, ed. McGraw-Hill, New York, p. 691.








77

87. Ellermann Eriksen, S., M. M. Christensen, and S. C. Mogensen. 1994. Effect of mercuric chloride on macrophage-mediated resistance mechanisms against infection with herpes simplex virus type 2. Toxicology 93:269.

88. Vladutiu, A. O. 1991. Treatment of autoimmune diseases with antibodies to class II major histocompatibility complex antigens. Clin. Immunol. Immunopathol. 61:1.

89. Utschig, L. M., J. G. Wright, and T. V. O'Halloran. 1993. Biochemical and spectroscopic probes of mercury(II) coordination environments in proteins. Methods Enzymol. 226:71.

90. Casciola Rosen, L., F. Wigley, and A. Rosen. 1997. Scleroderma autoantigens are uniquely fragmented by metal-catalyzed oxidation reactions: implications for pathogenesis. J. Exp. Med 185:71.

91. Casciola Rosen, L. A., G. Anhalt, and A. Rosen. 1994. Autoantigens targeted in systemic lupus erythematosus are clustered in two populations of surface structures on apoptotic keratinocytes [see comments]. J. Exp. Med 179:1317.

92. Druet, P., A. Bernard, F. Hirsch, J. J. Weening, P. Gengoux, P. Mahieu, and S. Birkeland. 1982. Immunologically mediated glomerulonephritis induced by heavy metals. Arch. Toxicol. 50:187.














BIOGRAPICAL SKETCH


Gregory Alan Hanley was born in Batavia, New York, November 24, 1965. He received a Bachelor of Science degree from Geneseo State, a branch of the State University of New York. He then attended the University of Florida, College of Veterinary Medicine and was awarded the Doctor of Veterinary Medicine in 1993. After graduation Dr. Hanley received an NIH postdoctoral fellowship to pursue advanced training in the veterinary speciality of laboratory animal medicine and training in a basic science leading to a Ph.D. Dr. Hanley's focus during his studies concentrated on the induction of autoimnmunity in mice using mercuric chloride, the results of which are included in this dissertation.






















78








I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Stephen M. Robrs hi
Associate Professor of Veterinary Medicine

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Eric S. Sobel, Cochair
Assistant Professor of Medicine

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.



Frofe sor of Veterinary Medicine

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Joli)S chi ffenbauer
gsociate Professor of Medicine

This dissertation was submitted to the Graduate Faculty of the College of
Veterinary Medicine and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.

May, 1998 Dean, College of Vete~iiary) Medicine


Dean, Graduate School




Full Text

PAGE 1

0+& &21752/ 2) 0(5&85<,1'8&(' $872,0081,7< ,1 0,&( %\ *5(*25< $/$1 +$1/(< $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

7R .LP P\ EHORYHG VSRXVH DQG JUHDWHVW VXSSRUWHU :LWKRXW KHU HQFRXUDJHPHQW DQG XQGHUVWDQGLQJ ZRXOG KDYH QHYHU FRPSOHWHG WKLV ZRUN

PAGE 3

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

PAGE 4

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

PAGE 5

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

PAGE 6

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ 0+& &21752/ 2) 0(5&85<,1'8&(' $872,0081,7< ,1 0,&( %\ *5(*25< $/$1 +$1/(< 0D\ &KDLUSHUVRQ 6WHSKHQ 0 5REHUWV 0DMRU 'HSDUWPHQW 9HWHULQDU\ 0HGLFLQH 2QH RI WKH PRVW VWULNLQJ IHDWXUHV RI H[SRVXUH WR ORZ GRVHV RI PHUFXU\ LQ PLFH LV WKH KLJKWLWHU KDSORW\SHOLQNHG DQWLQXFOHRODU $1R$f DXWRDQWLERG\ UHVSRQVH 0LFH RI WKH +6 KDSORW\SH KDYH EHHQ KLJK UHVSRQGHUV ZKLOH +E PLFH KDYH EHHQ ORZ 7KLV SDWWHUQ KDV EHHQ DWWULEXWHG WR WKH FODVV ,, PROHFXOH LWVHOI EXW WKH SRRU UHVSRQVH RI )M FURVVHV EHWZHHQ KLJK DQG ORZ UHVSRQGHUV UDLVHG WKH SRVVLELOLW\ WKDW WKH DQWLILEULOODULQ VSHFLILFLW\ ZDV DFWXDOO\ GXH WR D FORVHO\ OLQNHG GRPLQDQW QHJDWLYH JHQH 7R WHVW WKH UROH RI FODVV ,, H[SOLFLWO\ )M FURVVHV EHWZHHQ FRQJHQLH %6-/ +6f DQG &%/ +Ef PLFH ZLWK D WDUJHWHG GHOHWLRQ RI ,$SE ZHUH JHQHUDWHG FUHDWLQJ PLFH KHWHUR]\JRXV IRU DOO 0+& ORFL EXW H[SUHVVLQJ RQO\ ,$V ,Q FRPSDULVRQ ZLWK % 6-/ PLFH QR GLPLQXWLRQ RI WLWHUV ZDV IRXQG SURYLQJ WKDW ,$V LWVHOI ZDV UHVSRQVLEOH IRU VXVFHSWLELOLW\ DQG ,$E IRU GRZQUHJXODWLRQ 8QOLNH ,$ H[SUHVVLRQ RI WKH ,( FODVV ,, PROHFXOH FRXOG QRW YL

PAGE 7

GRZQUHJXODWH WKH UHVSRQVH LQ DQ RWKHUZLVH VXVFHSWLEOH PRXVH 8VLQJ DGRSWLYH WUDQVIHU WHFKQLTXHV ZH KDYH H[DPLQHG VHYHUDO PHFKDQLVPV E\ ZKLFK WKH UHVLVWDQW KDSORW\SH FRXOG EH GRZQUHJXODWLQJ WKH DQWLILEULOODULQ UHVSRQVH LQ ) VEf PLFH 6LPLODU WR RWKHU DXWRLPPXQH PRGHOV PHUFXU\ LQGXFHG DXWRLPPXQLW\ UHTXLUHV FRJQDWH 0+&UHVWULFWHG 7 FHOO KHOS 7KH DEVHQFH RI DXWRDQWLERG\ SURGXFWLRQ LQ ) PLFH ZDV QRW GXH WR HLWKHU D GLIIHUHQFH LQ WK\PLF HGXFDWLRQ QRU WR WKH DEVHQFH RI DQWLILEULOODULQVSHFLILF 7 FHOO KHOS 7KHVH UHVXOWV VXJJHVW D FRPSOLFDWHG UROH IRU FODVV ,, LQ WKH UHJXODWLRQ RI D QRYHO HQYLURQPHQWDOO\LQGXFHG DXWRLPPXQH UHVSRQVH YX

PAGE 8

&+$37(5 %$&.*5281' $1' 6,*1,),&$1&( ,QWURGXFWLRQ $XWRLPPXQH GLVHDVH RFFXUV ZKHQ WKH ERG\ PRXQWV DQ LPPXQH UHVSRQVH DJDLQVW VHOI 7KLV LPPXQH UHVSRQVH FDQ LQYROYH HLWKHU KXPRUDO DQGRU FHOOXODU LPPXQH SDWKZD\V &XUUHQWO\ WKH FDXVH RI DXWRLPPXQLW\ LV QRW NQRZQ EXW ERWK HQYLURQPHQWDO DV ZHOO DV JHQHWLF IDFWRUV DSSHDU WR EH LPSRUWDQW 7KH SUHYDOHQFH RI DXWRLPPXQH GLVHDVHV KDV EHHQ HVWLPDWHG WR EH DSSUR[LPDWHO\ b f $XWRLPPXQH GLVHDVHV FDQ EH FODVVLILHG DV V\VWHPLF RU RUJDQVSHFLILF ([DPSOHV RI RUJDQVSHFLILF DXWRLPPXQH GLVHDVHV LQFOXGH W\SH GLDEHWHV P\DVWKHQLD JUDYLV *UDYHV GLVHDVH DQG PXOWLSOH VFOHURVLV 7KH V\VWHPLF DXWRLPPXQH GLVHDVHV DUH UHSUHVHQWHG E\ UKHXPDWRLG DUWKULWLV VFOHURGHUPD DQG V\VWHPLF OXSXV HU\WKHPDWRVXV 6/(f 6/( LV D PXOWLV\VWHP GLVRUGHU FKDUDFWHUL]HG E\ WKH SUHVHQFH RI D YDULHW\ RI DXWRDQWLERGLHV WKDW DUH UHVSRQVLEOH IRU PDQ\ RI LWV FOLQLFDO PDQLIHVWDWLRQV 7KH SUHYDOHQFH UDWH EDVHG RQ WKH WRWDO SRSXODWLRQ LV RQH LQ EXW RQH LQ IRU ZRPHQ EHWZHHQ WKH DJHV RI DQG \HDUV DQG RQH LQ IRU EODFN ZRPHQ LQ WKH VDPH DJH JURXS f 7KH HWLRORJ\ RI 6/( LV XQNQRZQ EXW PDQ\ IDFWRUV KDYH EHHQ LGHQWLILHG WKDW LQIOXHQFH 6/( *HQHWLF VXVFHSWLELOLW\ LV VWURQJO\ VXJJHVWHG EHFDXVH RI WKH KLJK FRQFRUGDQFH UDWH EHWZHHQ PRQR]\JRWLF WZLQV &HUWDLQ 0+& KDSORW\SHV DUH PRUH IUHTXHQWO\ DVVRFLDWHG ZLWK WKH

PAGE 9

GLVHDVH DV ZHOO 7KH IDFW WKDW ZRPHQ DUH PXFK PRUH VXVFHSWLEOH WR 6/( SRLQWV WR KRUPRQDO IDFWRUV SOD\LQJ D UROH )LQDOO\ VHYHUDO HQYLURQPHQWDO IDFWRUV KDYH EHHQ DVVRFLDWHG ZLWK 6/( IODUH XSV ([SHULPHQWDO DQLPDO PRGHOV KDYH EHHQ ZLGHO\ XVHG WR VWXG\ DXWRLPPXQH GLVHDVHV 6RPH RI WKH PRGHOV GHYHORS VSRQWDQHRXV DXWRLPPXQLW\ OLNH 7\SH GLDEHWHV LQ 12' PLFH DQG %% UDWV 6SRQWDQHRXV GLVHDVH VLPLODU WR 6/( GHYHORSV LQ 1=% [ 1=:f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f 6RPH RI WKH FKDUDFWHULVWLFV RI WKLV GLVHDVH DUH UHPDUNDEO\ VLPLODU WR KXPDQ 6/( +J&OLQGXFHG DXWRLPPXQLW\ ZDV ILUVW UHSRUWHG LQ PLFH E\ 5RELQVRQ HW DO LQ f :LWKLQ RQH ZHHN DIWHU WKH VWDUW RI WUHDWPHQW ZLWK +J&O RXWEUHG 6ZLVV ,&5 PLFH SURGXFHG DQWLQXFOHDU DQWLERGLHV $1$f RI WKH ,J* VXEFODVV )XUWKHU DQDO\VLV RI WKH $1$V UHYHDOHG D QXFOHRODU SDWWHUQ DQWLQXFOHRODU DQWLERGLHV $1R$f ,QEUHG $PLFH RQ WKH RWKHU KDQG IDLOHG WR UHVSRQG WR +J&O 6LQFH WKLV LQLWLDO GLVFRYHU\ WKH OLQN EHWZHHQ VXVFHSWLELOLW\ DQG JHQHWLFV KDV EHHQ LQYHVWLJDWHG E\ VHYHUDO JURXSV ,QLWLDOO\ WKH OLQN ZDV HVWDEOLVKHG EHWZHHQ $1R$ LQGXFWLRQ

PAGE 10

DQG WKH PRXVH PDMRU KLVWRFRPSDWLELOLW\ FRPSOH[ 0+&f f 7HVWLQJ RI QXPHURXV 0+&FRQJHQLF VWUDLQV OHG WR WKH IROORZLQJ UDQNLQJ RI VXVFHSWLELOLW\ EDVHG RQ KDSORW\SH G NEUPITV f )XUWKHU VWXGLHV LQYROYLQJ LQWUD+ UHFRPELQDQW KDSORW\SHV UHYHDOHG WKDW WKH UHJLRQ RI WKH 0+& VSHFLILFDOO\ WKH ,$ UHJLRQ FRQWUROOHG WKH UHVSRQVH f 7KH RWKHU UHJLRQ ORFXV ,( KDV EHHQ VKRZQ WR HLWKHU GHFUHDVH WKH UHVSRQVH RU KDYH QR HIIHFW RQ WKH DELOLW\ RI +J&O WR LQGXFH $1R$V f 7KH UHVSRQVH ZDV VKRZQ WR EH FRGRPLQDQWO\ LQKHULWHG E\ LQWHUFURVVLQJ VXVFHSWLEOH DQG UHVLVWDQW VWUDLQV RI PLFH f )RU H[DPSOH FURVVHV EHWZHHQ VXVFHSWLEOH 6-/ DQG UHVLVWDQW % PLFH UHVXOWHG LQ DQ LQWHUPHGLDWH UHVSRQVH ZLWK b RI WKH SURJHQ\ GHYHORSLQJ $1R$V f 2WKHU JHQHVf RXWVLGH WKH 0+& KDYH DOVR EHHQ VKRZQ WR DIIHFW VXVFHSWLELOLW\ 7KH LQIOXHQFH RI WKHVH EDFNJURXQG JHQHV ZDV REVHUYHG ZKHQ FRPSDULQJ PLFH RI WKH +6 KDSORW\SH RQ DQ $ EDFNJURXQG WR WKRVH RQ D %,2 RU D % EDFNJURXQG f 1RQ0+& JHQHV RQ WKH $ EDFNJURXQG HQKDQFHG VXVFHSWLELOLW\ ZKLOH WKRVH SUHVHQW RQ WKH % EDFNJURXQG GHFUHDVHG UHVSRQVLYHQHVV :LWKLQ WKH % EDFNJURXQG % ZHUH PRUH VXSSUHVVLYH WKDQ %,2 1RW DOO RI WKH HIIHFWV VHHQ ZLWK +J&O DGPLQLVWUDWLRQ DUH DV WLJKWO\ FRQWUROOHG E\ WKH 0+& DV WKH $1R$ UHVSRQVH )RU H[DPSOH $PLFH +Nf GLG GHYHORS DQWLFKURPDWLQ $&$f DQG DQWLKLVWRQH $+$f DQWLERGLHV GHVSLWH EHLQJ UHVLVWDQW WR $1R$ LQGXFWLRQ f 7KH LPPXQH FRPSOH[ GHSRVLWLRQ VHHQ LQ WKLV PRGHO DOVR GRHV QRW DSSHDU WR EH XQGHU WKH LQIOXHQFH RI WKH + FRPSOH[ %DOEF +Gf PLFH GHYHORSHG UHQDO PHVDQJLDO DQG YHVVHO ZDOO LPPXQH FRPSOH[HV GHVSLWH EHLQJ FRPSOHWHO\ UHVLVWDQW WR $1R$ LQGXFWLRQ f :KHQ RQO\ $1R$ VXVFHSWLEOH PLFH +VITSf ZHUH H[DPLQHG IRU UHQDO ,J*

PAGE 11

GHSRVLWLRQ WKH UHVXOWV ZHUH VLPLODU LQ WKDW RQO\ WKH +6 PLFH VKRZHG VLJQLILFDQW LPPXQH FRPSOH[ GHSRVLWLRQ f 7LVVXH FRQFHQWUDWLRQV RI PHUFXU\ FKORULGH KDV EHHQ H[DPLQHG LQ VHYHUDO VWUDLQV RI LQEUHG PLFH f 0RVW +J&O DFFXPXODWHV LQ WKH NLGQH\ SULPDULO\ ZLWK WKH OLYHU EHLQJ WKH VHFRQG ODUJHVW SRRO IROORZHG E\ WKH JDVWURLQWHVWLQDO WUDFW VNLQ VSOHHQ DQG WHVWLFOHV 7KH OHYHOV UHDFKHG D VWHDG\ VWDWH LQ WKH EORRG DQG OLYHU E\ ZHHNV ZLWK WKH VSOHHQ DQG OLYHU UHDFKLQJ VWHDG\ VWDWH DW ZHHNV $QDO\VLV RI WK\PL UHYHDOHG VWHDGLO\ LQFUHDVLQJ FRQFHQWUDWLRQV WKURXJKRXW WKH ZHHN WUHDWPHQW SHULRG ,QWHUHVWLQJO\ WKH DXWKRUV DOVR QRWHG WKDW ZKHQ FRPSDULQJ WKH +FRQJHQLF VXVFHSWLEOH %,26 +6f DQG UHVLVWDQW %' +Gf D FRUUHODWLRQ ZDV GHWHFWHG EHWZHHQ VXVFHSWLELOLW\ WR $1R$ DQG DFFXPXODWLRQ RI +J ZLWKLQ WKH VSOHHQ f $XWRPHWDOORJUDSK\ KDV UHYHDOHG WKDW ZLWKLQ WKH WLVVXHV RI WKH LPPXQH V\VWHP PHUFXU\ FKORULGH DFFXPXODWHG LQ PDFURSKDJHV EXW ZDV QRW IRXQG LQ QRQSKDJRF\WLF DQWLJHQ SUHVHQWLQJ FHOOV IROOLFXODU GHQGULWLF FHOOV LQWHUGLJLWDWLQJ FHOOV DQG % FHOOVf $W WKH VXEFHOOXODU OHYHO PHUFXU\ ZDV ORFDOL]HG LQ WKH O\VRVRPHV RI WKH PDFURSKDJHV f 5ROH RI 7 &HOOV 7KH UHTXLUHPHQW IRU 7 FHOOV LQ WKLV PRGHO ZDV VKRZQ E\ +XOWPDQ HW DO XVLQJ WKH QXGH DWK\PLFf PXWDWLRQ f 6-/1 PLFH KRPR]\JRXV IRU WKH QXGH PXWDWLRQ QXQXf IDLOHG WR GHYHORS QHLWKHU WKH $1R$ QRU WKH V\VWHPLF LPPXQH FRPSOH[ GHSRVLWV VHHQ LQ OLWWHUPDWHV KHWHUR]\JRXV IRU WKH QXGH PXWDWLRQ QXf )XUWKHUPRUH WKH VSHFLILF VXEVHW RI 7 FHOOV UHTXLUHG ZDV HOXFLGDWHG E\ QRWLQJ D ODFN RI UHVSRQVH LQ HXWK\PLF 6-/1 PLFH JLYHQ DQWL&' PRQRFORQDO DQWLERG\ LQ DGGLWLRQ WR WKH PHUFXU\ FKORULGH WUHDWPHQW

PAGE 12

:KHQ LQ YLYR 7 FHOO UHVSRQVHV WR +J&O ZHUH VWXGLHG XVLQJ IORZ F\WRPHWU\ LQ VXVFHSWLEOH $6: PLFH D ORQJ ODVWLQJ LQFUHDVH LQ WKH QXPEHU RI 7 FHOOV GRPLQDWHG E\ WKH &' VXEVHW ZDV REVHUYHG ZKLFK ZDV DEVHQW LQ UHVLVWDQW $7+ PLFH f ,Q FRQWUDVW YDQ 9OLHW HW DO f UHSRUWHG WKDW ERWK +6 DQG +G PLFH H[SUHVVHG HQKDQFHG VSOHQLF QXPEHUV RI 7 FHOOV H[SUHVVLQJ WKH DFWLYDWLRQ PDUNHU &'5%O&' 6LPLODU HIIHFWV ZHUH VHHQ ZKHQ WKH DELOLW\ RI PHUFXU\ WR VWLPXODWH FHOOV LQ YLWUR ZDV VWXGLHG ,Q YLWUR WUHDWPHQW RI O\PSKRF\WHV ZLWK +J&O UHVXOWHG LQ D VLJQLILFDQW LQFUHDVH LQ '1$ V\QWKHVLV LQ $6: PLFH ZKLOH RQO\ PLQLPDO LQFUHDVHV ZHUH VHHQ LQ '%$ PLFH f $QWL&' DQWLERG\ DJDLQ FRPSOHWHO\ DEURJDWHG WKH UHVSRQVH VXJJHVWLQJ WKH LPSRUWDQFH RI 7 KHOSHU FHOOV LQ WKLV PRGHO ,QWHUHVWLQJO\ PHUFXU\H[SRVHG O\PSKRF\WHV IURP WKH ORZ UHVSRQGHU PLFH EHFDPH KLJK UHVSRQGHUV DV PHDVXUHG E\ >+@7K\PLGLQH XSWDNH DIWHU WKH UHPRYDO RI H[FHVV PHUFXU\ E\ ZDVKLQJ f 7KLV DFWLYDWLRQ RI 7 FHOOV LQ YLWUR OHG -LDQJ HW DO f WR H[DPLQH ZKHWKHU PHUFXU\ DFWHG DV D 7 FHOO PLWRJHQ RU DV D VXSHUDQWLJHQ 7KH\ IRXQG D ELDVHG XVDJH RI 7&5 9S VXEVHWV E\ &' FHOOV ZKLFK WKH\ LQWHUSUHWHG DV D VXSHUDQWLJHQ HIIHFW &\WRNLQHV &RQVLGHULQJ WKH HIIHFWV PHUFXU\ KDV RQ 7 FHOOV F\WRNLQHV PXVW SOD\ DQ LPSRUWDQW UROH LQ +J&OLQGXFHG DXWRLPPXQLW\ ,Q YLWUR WKH FRQWLQXRXV SUHVHQFH RI PHUFXU\ LQGXFHG LQWHUOHXNLQ (/f DQG LQWHUIHURQ\ \,)1f EXW QRW ,/ SURGXFWLRQ IURP ERWK KLJK DQG ORZ UHVSRQGHU PLFH ,Q FRQWUDVW SUHWUHDWLQJ WKH FHOOV ZLWK PHUFXU\ DQG WKHQ ZDVKLQJ LW DZD\ UHVXOWHG LQ SURGXFWLRQ RI ,/ IURP ERWK JURXSV f

PAGE 13

&\WRNLQH SURILOHV RI PLFH WUHDWHG ZLWK PHUFXU\ KDV JLYHQ VRPHZKDW FRQIOLFWLQJ UHVXOWV $Q DQDO\VLV RI ,/ P51$ IURP WZR +FRQJHQLF VWUDLQV UHYHDOHG D VWURQJ LQFUHDVH LQ SURGXFWLRQ LQ VSOHQLF &' 7 FHOOV RI WKH +6 PLFH ZKLOH WKRVH RI +G PLFH VKRZHG RQO\ D ZHDN LQFUHDVH f 7KLV UHVXOW DORQJ ZLWK VLPLODU UHVXOWV VHHQ LQ WKH UDW PRGHO RI +J&OLQGXFHG DXWRLPPXQLW\ f OHG *ROGPDQ HW DO f WR SURSRVH WKDW VXVFHSWLELOLW\ UHOLHV RQ WKH LQLWLDWLRQ RI D 7+ UHVSRQVH ZKLOH UHVLVWDQFH FRUUHODWHV SUHGRPLQDQWO\ ZLWK D 7+ UHVSRQVH )XUWKHU HYLGHQFH RI WKLV GLFKRWRP\ ZDV SURYLGHG E\ WKH IDFW WKDW +J&OLQGXFHG D VWULNLQJ LQFUHDVH XS WR IROG RI WKH OHYHO RI ,J( LQ VXVFHSWLEOH $6: EXW QRW LQ UHVLVWDQW % '%$ RU %' PLFH f 7KH VHUXP ,J( OHYHOV DUH LPSRUWDQW EHFDXVH WKH\ DV ZHOO DV WKH LQFUHDVHV VHHQ LQ FODVV ,, H[SUHVVLRQ DUH ERWK KLJKO\ GHSHQGHQW RQ WKH 7+ F\WRNLQH ,/ f 7KH FRQQHFWLRQ EHWZHHQ ,J( DQG ,/ ZDV IXUWKHU VWUHQJWKHQHG LQ WKLV PRGHO E\ 2FKHO HW DO f ZKR VKRZHG WKDW WUHDWLQJ VXVFHSWLEOH $6: PLFH ZLWK DQWL ,/ DQWLERG\ FRPSOHWHO\ DEURJDWHG WKH LQFUHDVH LQ WRWDO ,J( 6XUSULVLQJO\ WKLV WUHDWPHQW GLG QRW SUHYHQW $1R$ SURGXFWLRQ DOWKRXJK LW GLG LQIOXHQFH WKH SDWWHUQ RI ,J* VXEFODVV GLVWULEXWLRQ 8VLQJ DQWLERG\ UHVSRQVH WR VKHHS HU\WKURF\WHV V5%&f LQ PHUFXU\ WUHDWHG + FRQJHQLF VWUDLQV 'RWK HW DO f DOVR SURSRVHG D SUHIHUHQWLDO DFWLYDWLRQ RI HLWKHU 7+ RU 7+ FHOOV EDVHG RQ KDSORW\SH 7KH UHVSRQVH WR V5%& DQWLJHQV LV QRUPDO LQ WUHDWHG %6 PLFH ZKLOH LW LV GHSUHVVHG LQ %' DQLPDOV 7KH\ VKRZHG XVLQJ PRQRFORQDO DQWLERGLHV DJDLQVW \,)1 WKDW WKH VXSSUHVVLRQ VHHQ LQ %' PLFH ZDV PHGLDWHG E\ \,)1 D 7+ F\WRNLQH )XUWKHUPRUH WUHDWPHQW RI %,2 6 PLFH ZLWK UHFRPELQDQW \,)1 VXSSUHVVHG WKHLU

PAGE 14

UHVSRQVH WR WKH V5%& DQWLJHQV \HW GLG QRW SUHYHQW WKH LQGXFWLRQ RI $1R$V QRU WKH LPPXQH FRPSOH[ JORPHUXORQHSKULWLV $Q DUJXPHQW DJDLQVW WKH 7+7+ GLFKRWRP\ LQ WKLV PRGHO ZDV SURSRVHG E\ -RKDQVVRQ HW DO f 7KH\ IRXQG WKDW $6: PLFH VKRZHG D PRGHVW LQFUHDVH RI \,)1 DQG 7/ SURGXFLQJ FHOOV ZKLOH +FRQJHQLF $7+ PLFH VKRZHG QR LQFUHDVH LQ F\WRNLQH SURGXFLQJ FHOOV 5HPDUNDEO\ WKH VXVFHSWLEOH 6-/ VWUDLQ GHVSLWH EHLQJ VHYHUHO\ GHILFLHQW LQ 7+SURPRWLQJ &' 1.8 7 FHOOV LQFUHDVHG WKHLU QXPEHU RI \,)1 SURGXFLQJ FHOOV 7KLV LQGLFDWHG WKDW D SUHGRPLQDQWO\ 7+ UHVSRQVH LV QRW QHFHVVDU\ IRU WKH LQGXFWLRQ RI DXWRLPPXQLW\ E\ PHUFXU\ 5HFHQW VWXGLHV FRQGXFWHG E\ 3ROODUG HW DO f SURYLGHG WKH VWURQJHVW HYLGHQFH WR GDWH DJDLQVW WKH LGHD WKDW D 7+ UHVSRQVH FRUUHVSRQGV ZLWK UHVLVWDQFH ZKLOH D 7+ UHVSRQVH LV UHTXLUHG IRU DXWRLPPXQLW\ LQGXFWLRQ 8VLQJ +6 PLFH GHILFLHQW E\ JHQH NQRFNRXW IRU ,/ DQG \,)1 WKH\ VKRZHG WKDW DOO RI WKH IHDWXUHV RI DXWRLPPXQLW\ DUH FRQWUROOHG E\ \,)1 D 7+ F\WRNLQH ,/nn PLFH GLG GHYHORS $1R$V ZKLOH \,)1nn ZHUH FRPSOHWHO\ UHVLVWDQW 7KH\ FRQFOXGHG WKDW WKH UHTXLUHPHQW IRU \,)1 VXJJHVWV WKDW DQWLJHQ GRVH LV WKH OLPLWLQJ IDFWRU LQ DXWRLPPXQLW\ LQGXFWLRQ 7KH LGHD RI D GRVH UHVSRQVH UHODWLRQVKLS ZDV H[DPLQHG E\ +XOWPDQ HW DO f 7KH\ IRXQG D SRVLWLYH FRUUHODWLRQ EHWZHHQ WKH GRVH RI +J&O WKH WRWDO ERG\ EXUGHQ RI PHUFXU\ DQG WKH GHJUHH RI DXWRLPPXQLW\ H[SUHVVHG DV WKH VHUXP $1R$ WLWHU 7KH PLQLPXP REVHUYHG DGYHUVH HIIHFW OHYHO 02$(/f ZDV PJ +J&O OLWHU GULQNLQJ ZDWHU SSPf $W WKLV GRVH DSSUR[LPDWHO\ b RI WKH PLFH UHVSRQGHG 1R $1R$V ZHUH VHHQ LQ PLFH JLYHQ SSP ZKLOH b UHVSRQGHG WR SSP 7KH\ HVWLPDWHG

PAGE 15

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f $ VLJQLILFDQW LQFUHDVH LQ DQWL713 SURGXFLQJ FHOOV VHHQ LQ 6-/ PLFH EXW QRW LQ % PLFH SURYLGHG HYLGHQFH IRU SRO\FORQDO % FHOO DFWLYDWLRQ LQ WKH IRUPHU +FRQJHQLF VWUDLQV DOVR VKRZ WKH GLIIHUHQWLDO HIIHFWV RQ % FHOOV RI PHUFXU\ DGPLQLVWUDWLRQ %6 PLFH VKRZHG VLJQLILFDQWO\ LQFUHDVHG QXPEHUV RI ,J SURGXFLQJ VSOHQLF % FHOOV RI WKH ,J*O IROGf ,J*D IROGf DQG ,J( IROGf ZKLOH %' PLFH LQ FRQWUDVW GLG QRW VKRZ DQ\ VLJQLILFDQW LQFUHDVHV f 0HUFXU\ DOVR HQKDQFHV FODVV ,, H[SUHVVLRQ RQ % FHOOV IURP ERWK $1R$ UHVLVWDQW %' DQG $1R$ VXVFHSWLEOH %,26 PLFH DOWKRXJK VLJQLILFDQWO\ KLJKHU OHYHOV ZHUH REVHUYHG LQ WKH %6 PLFH f 7UHDWPHQW ZLWK DQ DQWL,/ DQWLERG\ FRPSOHWHO\ SUHYHQWHG WKH LQFUHDVHV VHHQ LQ ERWK VWUDLQV ,Q DGGLWLRQ WR PHUFXU\ LQIOXHQFLQJ FODVV ,, H[SUHVVLRQ FODVV ,, PROHFXOHV DUH UHTXLUHG IRU VRPH RI WKH HIIHFWV LQGXFHG E\ PHUFXU\ ,Q DQ LQ YLWUR V\VWHP XVLQJ PRQRFORQDO DQWLERGLHV DJDLQVW FODVV ,, +X HW DO f ZHUH DEOH WR

PAGE 16

FRPSOHWHO\ DEURJDWH WKH F\WRNLQH SURGXFWLRQ DQG SUROLIHUDWLRQ LQGXFHG E\ PHUFXU\ 7KH UHTXLUHPHQW IRU FODVV ,, PROHFXOHV IRU WKH LQGXFWLRQ RI $1R$V LQ YLYR KDV QRW EHHQ LQYHVWLJDWHG ,Q FKDSWHU ZH IRUPDOO\ DGGUHVV WKLV TXHVWLRQ XVLQJ PLFH ZKRVH 0+& ,$ JHQH KDV EHHQ GHOHWHG )LEULOODULQ 7KH QXFOHRODU DXWRDQWLJHQ DJDLQVW ZKLFK +J&OLQGXFHG DXWRDQWLERGLHV DUH GLUHFWHG KDV EHHQ LGHQWLILHG DV ILEULOODULQ D N'D 8 ULERQXFOHRSURWHLQ LQYROYHG LQ WKH ILUVW VWHS RI SUHULERVRPDO 51$ SURFHVVLQJ f ,W LV LQWHUHVWLQJ WKDW XS WR b RI KXPDQ VFOHURGHUPD SDWLHQWV VSRQWDQHRXVO\ SURGXFH DXWRDQWLERGLHV DJDLQVW WKH VDPH 8 ULERQXFOHRSURWHLQ f 6HYHUDO JURXSV KDYH FRPSDUHG WKH UHDFWLYLWLHV WRZDUG ILEULOODULQ RI PHUFXU\LQGXFHG $1R$V ZLWK KXPDQ VFOHURGHUPD VHUD DQG IRXQG WKDW WKH\ ZHUH LQGLVWLQJXLVKDEOH f %RWK JURXSV RI DXWRDQWLERGLHV UHFRJQL]H WKH IXOO OHQJWK SURWHLQ EXW ORVH UHDFWLYLW\ LQ DOO EXW WKH KLJKHVW WLWUDWHG KXPDQ VHUD LI HLWKHU WKH 1 RU & UHJLRQV ZHUH FOHDYHG f ,Q DQ DWWHPSW WR GHWHUPLQH ZKDW WKH 7 FHOO DQWLJHQ LV EXON 7 FHOOV IURP +J&O WUHDWHG %6 PLFH ZHUH FKDOOHQJHG ZLWK D YDULHW\ RI GLIIHUHQW VHOI SURWHLQV WKDW FRQWDLQHG PLQXWH DPRXQWV RI PHUFXU\ 7KH 7 FHOOV UHDFWHG WR DOO WKH SURWHLQV LQ DQ DQDPQHVWLF IDVKLRQ f %XON 7 FHOOV REWDLQHG DIWHU RQH ZHHN RI WUHDWPHQW DOVR UHDFWHG DQDPQHVWLFDOO\ WR +JFRPSOH[HG ILEULOODULQ ZKHUHDV DIWHU HLJKW ZHHNV D 7 FHOO UHVSRQVH DJDLQVW XQWUHDWHG ILEULOODULQ SUHGRPLQDWHG 7KHVH UHVXOWV VXJJHVWHG GHWHUPLQDQW VSUHDGLQJ RI 7 FHOO VSHFLILFLW\ &RQVLVWHQW ZLWK WKLV DQDO\VLV RI 7 FHOO K\EULGRPDV HVWDEOLVKHG IURP

PAGE 17

PHUFXU\WUHDWHG +6 PLFH UHYHDOHG WZR W\SHV RI &' 7 FHOOV RQH WKDW VSHFLILFDOO\ UHFRJQL]HG +JFRPSOH[HG ILEULOODULQ DQG DQRWKHU WKDW UHDFWHG WR XQWUHDWHG ILEULOODULQ f ,QWHUHVWLQJO\ ERWK WKH +JLQGXFHG DQG QDWLYH ILEULOODULQ GHWHUPLQDQWV ZHUH SUHVHQWHG ZLWKRXW WKH DGGLWLRQ RI ILEULOODULQ ZKHQ VSOHHQ FHOOV IURP DQLPDOV WUHDWHG ZLWK PHUFXU\ ZHUH XVHG DV DQWLJHQ SUHVHQWLQJ FHOOV ,Q FRQWUDVW VSOHHQ FHOOV IURP XQWUHDWHG PLFH FRXOG DFWLYDWH ILEULOODULQVSHFLILF 7 FHOOV RQO\ LI H[RJHQRXV ILEULOODULQ ZDV DGGHG 7KLV LQGLFDWHG WKDW +J&O FDQ LQGXFH DEQRUPDO SUHVHQWDWLRQ E\ DQWLJHQ SUHVHQWLQJ FHOOV RI D VHOI SURWHLQ ,Q DQ DWWHPSW WR XQGHUVWDQG WKH PHFKDQLVP ZKHUHE\ PHUFXU\ HOLFLWV DQ DXWRDQWLERG\ UHVSRQVH WKDW VSHFLILFDOO\ WDUJHWV ILEULOODULQ 3ROODUG HWDO f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

PAGE 18

LQYHVWLJDWHG ZKHWKHU WKH FODVV ,,,$ PROHFXOH LWVHOI RU D FORVHO\ OLQNHG JHQHVf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

PAGE 19

&+$37(5 &/$66 ,, +$3/27<3( ',))(5(17,$//< 5(*8/$7(6 ,0081( 5(63216( ,1 +J&O75($7(' 0,&( ,QWURGXFWLRQ 6XEWR[LF GRVHV RI +J&O LQ DQLPDOV LQGXFH DQ DXWRLPPXQH GLVHDVH FKDUDFWHUL]HG E\ DXWRDQWLERG\ SURGXFWLRQ DQG YDULDEOH DPRXQWV RI LPPXQH FRPSOH[ JORPHUXORSDWK\ f ,Q PLFH IRU H[DPSOH WKH JORPHUXORSDWK\ LV UHODWLYHO\ PLOG DQG FKDUDFWHUL]HG E\ PHVDQJLDO GHSRVLWV RI ,J* DQG & f 2WKHU IHDWXUHV VXFK DV VSOHQLF K\SHUSODVLD ZLWK LQFUHDVHG VHUXP LPPXQRJOREXOLQV HVSHFLDOO\ ,J* DQG ,J(f KDYH DOVR EHHQ GHVFULEHG f +RZHYHU WKH PRVW VWULNLQJ ILQGLQJ LQ PLFH LV WKH DQWLQXFOHRODU DQWLERGLHV $1R$f HOLFLWHG LQ VXVFHSWLEOH VWUDLQV DV VRRQ DV RQH ZHHN DIWHU WKH VWDUW RI WUHDWPHQW f 7KH SULQFLSDO QXFOHRODU DXWRDQWLJHQ KDV EHHQ LGHQWLILHG DV ILEULOODULQ D 8 ULERQXFOHRSURWHLQ LQYROYHG LQ WKH ILUVW VWHS RI SUHULERVRPVDO 51$ SURFHVVLQJ f ,QWHUHVWLQJO\ WKLV VDPH VSHFLILFLW\ LV VHHQ LQ SDWLHQWV VXIIHULQJ IURP V\VWHPLF VFOHURVLV DQG RWKHU FRQQHFWLYH WLVVXH GLVRUGHUV f (YHQ PRUH UHPDUNDEO\ D FRPSDULVRQ RI WKH VSHFLILFLW\ RI DQWLILEULOODULQ DXWRDQWLERGLHV EHWZHHQ KXPDQ DQG PXULQH RULJLQ VKRZHG VLPLODU LI QRW LGHQWLFDO FRQIRUPDWLRQDO HSLWRSHV f 7KH PHFKDQLVP E\ ZKLFK +J&O FDXVHV DXWRLPPXQLW\ LV XQNQRZQ EXW WKH IDFW WKDW D KXPDQ VFOHURGHUPDDVVRFLDWHG DXWRDQWLERG\ VSHFLILFLW\ FDQ EH SURGXFHG DQG WKDW FHUWDLQ

PAGE 20

WR[LF H[SRVXUHV KDYH UHVXOWHG LQ VFOHURGHUPDOLNH LOOQHVVHV f PDNHV LW DQ LPSRUWDQW DQG LQWULJXLQJ PRGHO %RWK 0+& FODVV ,,OLQNHG ORFL DQG XQNQRZQ QRQ0+& ORFL JRYHUQ VXVFHSWLELOLW\ ZLWK PLFH RI WKH +6 KDSORW\SH EHLQJ KLJK UHVSRQGHUV ZKLOH WKRVH RI WKH +E KDSORW\SH DUH UHVLVWDQW RU ORZ UHVSRQGHUV f 1RQ+ JHQHV LQ +6 PLFH GLG QRW SUHYHQW WKH GLVHDVH EXW KDG D SURQRXQFHG HIIHFW RQ WKH DQWLQXFOHRODU DQWLERG\ $1R$f WLWHUV f %\ XVLQJ LQWUD+ UHFRPELQDQWV LW KDV EHHQ VKRZQ WKDW UHVSRQVLYHQHVV FRXOG EH PDSSHG WR WKH ,$ UHJLRQ RI WKH 0+& 7KH RWKHU FODVV ,, ORFXV ,( ZDV VKRZQ WR HLWKHU VXSSUHVV f RU KDYH QR HIIHFW RQ WKH LPPXQH UHVSRQVH f 7KH FODVV ,,OLQNHG VXVFHSWLELOLW\ JHQH KDV SUHYLRXVO\ EHHQ UHSRUWHG WR EH GRPLQDQW DOWKRXJK WKH GLVHDVH ZDV DWWHQXDWHG LQ ) DQLPDOV ZKHQ FRPSDUHG WR KRPR]\JRXV DQLPDOV f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f DQG % +Ef PLFH ZHUH KLJKO\ UHVLVWDQW WR +J&O WUHDWPHQW DQG VXJJHVWHG WKDW WKLV PLJKW KDYH EHHQ GXH WR D GRPLQDQW QHJDWLYH JHQH OLQNHG WR ;$E 6XFK D SRVVLELOLW\ LV QRW ZLWKRXW SUHFHGHQW

PAGE 21

5HFHQWO\ WKH FOLQLFDO PDQLIHVWDWLRQ RI DUWKULWLV LQ WKH DXWRLPPXQH PRGHO RI FROODJHQ LQGXFHG DUWKULWLV ZDV VKRZQ QRW WR EH H[FOXVLYHO\ OLPLWHG WR D SDUWLFXODU 0+& FODVV ,, DQWLJHQ EXW PD\ EH GXH WR OLQNHG JHQHV ZKLFK SDUWLFLSDWH LQ LQWUDFHOOXODU ORDGLQJ DQG VHOHFWLRQ RI DQWLJHQLF SHSWLGHV f 7R GLVWLQJXLVK EHWZHHQ WKHVH WZR SRVVLELOLWLHV ZH JHQHUDWHG PLFH RQ WKH % EDFNJURXQG ZKLFK H[SUHVVHG ,$ DQG FDUULHG WKH JHQHV OLQNHG WR WKH ,$E UHJLRQ EXW GLG QRW H[SUHVV ,$E 7KHVH PLFH ZHUH DV VXVFHSWLEOH WR WKH LQGXFWLRQ RI $1R$ DV +6 KRPR]\JRXV PLFH GHVSLWH FDUU\LQJ WKH JHQHV OLQNHG WR ,$E :H DOVR H[DPLQHG WKH HIIHFWV RI WKH RWKHU FODVV ,, PROHFXOH ,( LQ RXU V\VWHP E\ GHYHORSLQJ %6-/ PLFH ZKLFK FDUULHG WKH ,(DG WUDQVJHQH f DOORZLQJ WKH IXQFWLRQDO H[SUHVVLRQ RI ,( RQ WKH + EDFNJURXQG &RQWUDU\ WR D SUHYLRXVO\ SXEOLVKHG UHSRUW DQG WR RXU H[SHULHQFH ZLWK ,$ WKH H[SUHVVLRQ RI ,( GLG QRW DOWHU WKH $1R$ UHVSRQVH 7KXV RXU H[SHULPHQWV GHPRQVWUDWH WKDW LW LV WKH H[SUHVVLRQ RI WKH UHVLVWDQW + DOOHOH ,$Ef LWVHOI WKDW GRZQUHJXODWHG WKH $1R$ UHVSRQVH RI )M FURVVHV EHWZHHQ UHVLVWDQW DQG VHQVLWLYH KDSORW\SHV DQG VWURQJO\ VXJJHVW WKDW LW LV WKH +6 PROHFXOH LWVHOI WKDW LV VSHFLILFDOO\ UHTXLUHG 7KHVH UHVXOWV KDYH LPSRUWDQW LPSOLFDWLRQV IRU XQGHUVWDQGLQJ WKH PHFKDQLVP E\ ZKLFK +J&O H[SRVXUH FDXVHV D VSHFLILF DXWRLPPXQH V\QGURPH 0DWHULDOV DQG 0HWKRGV 0LFH &%/% ,$E ,(n ,JKEf &%/-6-/I %6-/ ,$ ,(n ,JKEf &%/-J"D 7K\,D *SLD %7& ,$E ,(n ,JKDf $%<6Q$6:6Q6-/DQG %DOEFPLFH ZHUH REWDLQHG GLUHFWO\ RU IURP EUHHGHUV SXUFKDVHG IURP 7KH -DFNVRQ

PAGE 22

/DERUDWRU\ %DU +DUERU 0(f 7J+,(Df%UL %,(DGf PLFH ZHUH D JLIW IURP 5 / %ULQVWHU f &'r1 0+& &ODVV ,, GHILFLHQW % EDFNJURXQGf ZHUH SXUFKDVHG IURP 7DFRQLF *HUPDQWRZQ 1
PAGE 23

,QGLUHFW ,PPXQRIOXRUHVFHQFH 6HUD IURP PLFH ZDV WHVWHG IRU WKH SUHVHQFH RI $1R$ E\ LQGLUHFW LPPXQRIOXRUHVFHQFH XVLQJ FRPPHUFLDOO\ SUHSDUHG PRXVH IUR]HQ NLGQH\ VOLGHV 6DQRIL &KDVND 01f 7KH VOLGHV ZHUH LQFXEDWHG ZLWK VHUD GLOXWHG LQ 3%6 IRU PLQXWHV DW URRP WHPSHUDWXUH )RU QRQDOORW\SHVSHFLILF $1R$ WKH VOLGHV ZHUH LQFXEDWHG IRU PLQ DW URRP WHPSHUDWXUH ZLWK ),7&FRQMXJDWHG JRDW DQWLPRXVH ,J* )F IUDJPHQW VSHFLILF -DFNVRQ ,PPXQR5HVHDUFK :HVW *URYH 3$f GLOXWHG LQ 3%6 )RU WKH DOORW\SH VSHFLILF $1R$ GHWHUPLQDWLRQ VOLGHV ZHUH LQFXEDWHG ZLWK UDEELW DQWLPRXVH ,J*DD RU ,J*DE 1RUGLF /DEV &DSLVWUDQR %HDFK &$f IROORZHG E\ ),7&FRQMXJDWHG JRDW DQWLn UDEELW ,J* -DFNVRQ ,PPXQR5HVHDUFK :HVW *URYH 3$f 7KHVH UHDJHQWV KDG EHHQ SUHn WLWUDWHG E\ (/,6$ DJDLQVW DQ DOORW\SHQRQVSHFLILF UDEELW DQWL,J*D DQWLERG\ 1RUGLF /DEVf WR SURGXFH HTXLYDOHQW VHQVLWLYLWLHV $QWLQXFOHRODU VWDLQLQJ ZDV HYLGHQFHG E\ DQ LQWHQVH KRPRJHQHRXV VWDLQLQJ RI WKH QXFOHROL )LJ f $QWLERGLHV 0.6 +% PXULQH ,J*E DQWL,$SVILX ff DQG 6 +% PXULQH ,J*D N DQWL,(D0 ff ZHUH SXUFKDVHG IURP $PHULFDQ 7\SH 7LVVXH &XOWXUH 5RFNYLOOH 0'f PXULQH ,J*D DQWL,$SE DQWLJHQ ff ZDV RULJLQDOO\ REWDLQHG IURP 7RQNRQRJ\ RULJLQDOO\ GHYHORSHG DJDLQVW \HDVW ILEULOODULQ EXW FURVVUHDFWLYH ZLWK ILEULOODULQ RI UDW DQG KXPDQ RULJLQ f ZDV D NLQG JLIW RI 'U -RKQ $ULV 8QLYHUVLW\ RI )ORULGD *DLQHVYLOOH )/f $OLTXRWV RI VWRFN P$E ZHUH SUHSDUHG IURP RYHUJURZQ FHOO FXOWXUH VXSHUQDWDQW

PAGE 24

ZKLFK ZDV DIILQLW\ SXULILHG RQ D SURWHLQ FROXPQ 0RQRFORQDO DQWLERGLHV ZHUH ODEHOHG HLWKHU ZLWK ELRWLQ K\GUD]LGH f RU IOXRUHVFHLQ LVRWKLRF\DQDWH ),7&f f DV QHHGHG ,PPXQL]DWLRQV 0LFH ZHUH LPPXQL]HG E\ DQ LQWUDSHULWRQHDO LQMHFWLRQ RI SJ RI KHQ HJJ O\VR]\PH +(/ 6LJPD &KHPLFDO &R 6W /RXLV 02f LQ SL RI &)$ 'LIFR /DERUDWRULHV 'HWURLW 0,f GLOXWHG ZLWK VWHULOH 3%6 )RXUWHHQ GD\V ODWHU WKH PLFH ZHUH ERRVWHG ZLWK WKH VDPH DQWLJHQ LQ ,) $ 0LFH ZHUH WKHQ EOHG YLD WDLO YHLQ ZHHNV DIWHU WKH VHFRQGDU\ LPPXQL]DWLRQ DQG WKH VHUD VWRUHG DW r& XQWLO DQDO\]HG E\ (/,6$ $QWL+(/ (/,6$ 'HWHUPLQDWLRQ RI DQWLJHQVSHFLILF DQWLERG\ OHYHOV ZHUH GHWHUPLQHG XVLQJ DQ LQGLUHFW (/,6$ ,PPXORQp PLFURWLWHU SODWHV '\QDWHFK /DERUDWRULHV ,QF &KDQWLOO\ 9$f ZHUH FRDWHG RYHUQLJKW DW r& ZLWK +(/ DW SJPO LQ ELFDUERQDWH EXIIHU 0 S+ f %HWZHHQ DOO LQFXEDWLRQ VWHSV PLFURSODWHV ZHUH ZDVKHG WKUHH WLPHV ZLWK ERUDWH EXIIHUHG VDOLQH %%6 P0 1D%n + P0 1D&O P0 +% S+ f FRQWDLQLQJ b 7ZHHQ $IWHU EORFNLQJ ZLWK %%6 FRQWDLQLQJ b %6$ DQG b 7ZHHQ VHULDO GLOXWLRQV RI VHUXP VDPSOHV GLOXWHG LQ WKH EORFNLQJ VROXWLRQ ZHUH LQFXEDWHG RYHUQLJKW DW r& (LWKHU ELRWLQ\ODWHG JRDW DQWLPRXVH ,J* )F IUDJPHQW VSHFLILFf RU ELRWLQ\ODWHG JRDW DQWLPRXVH ,J0 PX FKDLQ VSHFLILFf-DFNVRQ ,PPXQRUHVHDUFK :HVW *URYH 3$f ZDV DGGHG IRU RQH KRXU DW URRP WHPSHUDWXUH 7KH LQGLFDWRU V\VWHP FRQVLVWHG RI ([WU$YLGLQp DONDOLQH SKRVSKDWDVH 6LJPD &KHPLFDO &R 6W

PAGE 25

/RXLV 02f DQG WKH VXEVWUDWH 6LJPD p SKRVSKDWDVH VXEVWUDWH SQLWURSKHQRO SKRVSKDWH GLVRGLXP KH[DK\GUDWHf 7KH VXEVWUDWH WXUQRYHU ZDV GHWHUPLQHG E\ WKH GLIIHUHQFH EHWZHHQ WKH 2' DQG 2' RQ D %LR7HN ,QVWUXPHQWV ,QF :LQRVNL 97f &HUHV PLFURSODWH UHDGHU 7KH FRQFHQWUDWLRQ RI DQWLJHQVSHFLILF ,J* DQG ,J0 LV UHSRUWHG LQ HTXLYDOHQW GLOXWLRQ IDFWRUV RI VWDQGDUGL]HG UHIHUHQFH %6-/ VHUD 7KLV LV GHILQHG E\ WKH IRUPXOD (') GLOXWLRQ RI D VWDQGDUG UHIHUHQFH VHUD WKDW JLYHV WKH HTXLYDOHQW 2' RI WKH WHVW VHUXPf ; )ORZ &\WRPHWU\ $SSUR[LPDWHO\ SL RI WDLO YHLQ EORRG ZDV FROOHFWHG LQWR KHSDULQL]HG WXEHV 3%0&V ZHUH LVRODWHG XVLQJ /\PSKRO\WH 0 &HUGDUODQH +RUQE\ 2QWDULR &DQDGDf GHQVLW\ JUDGLHQWV 7KH FHOOV ZHUH WKHQ FROOHFWHG LQWR 3%6 VXSSOHPHQWHG ZLWK b )&6 DQG b 1D1 )RU FHOO VXUIDFH VWDLQLQJ VDWXUDWLQJ DPRXQWV RI 6 ELRWLQ RU 0.6),7& ZHUH XVHG DV WKH ILUVW VWHS 7KH VHFRQG VWHS FRQVLVWHG RI LQFXEDWLQJ ZLWK ),7&FRQMXJDWHG JRDW DQWLPRXVH ,J* )F IUDJPHQW VSHFLILFf RU SK\FRHU\WKULQ 3(fFRQMXJDWHG VWUHSWDYLGLQ -DFNVRQ ,PPXQRUHVHDUFK :HVW *URYH 3$f 7KH FHOOV ZHUH WKHQ ZDVKHG WKUHH WLPHV LQ 3%6 DQG IL[HG ZLWK DQ HTXDO YROXPH RI b SDUDIRUPDOGHK\GH /LVW PRGH GDWD ZHUH DFTXLUHG RQ D )$&6FDQ IORZ F\WRPHWHU %HFWRQ 'LFNLQVRQ 3DOR $OWR &$f XVLQJ 3& /\VLV VRIWZDUH 'HDG FHOOV ZHUH H[FOXGHG E\ IRUZDUG DQG VLGH VFDWWHU JDWLQJ /LVW PRGH ILOHV ZHUH WKHQ DQDO\]HG XVLQJ /\VLV ,, VRIWZDUH

PAGE 26

6WDWLVWLFDO $QDO\VLV 7KH GLIIHUHQFH EHWZHHQ WKH QXPEHU RI PLFH VKRZLQJ DQ $1R$ EHWZHHQ %,$VE DQG %,$n ZDV DQDO\]HG XVLQJ )LVFKHUV H[DFW WHVW $ WWHVW ZDV XVHG WR FRPSDUH EDVHOLQH ,$V 0), PHDQ IOXRUHVFHQW LQWHQVLW\f YDOXHV EHWZHHQ WKHVH WZR JURXSV 'LIIHUHQFHV EHWZHHQ 0),V RI WKH RWKHU JURXSV ZDV WHVWHG E\ DQ $129$ DQDO\VLV RI YDULDQFHf IROORZHG E\ 'XQQHWWnV PXOWLSOH FRPSDULVRQ SURFHGXUH $S YDOXH RI ZDV FRQVLGHUHG WR EH VLJQLILFDQW 5HVXOWV )M 0LFH %HWZHHQ D 5HVLVWDQW DQG 6HQVLWLYH +DSORWYSH :HUH 5HVLVWDQW WR $1R$ ,QGXFWLRQ ,Q WKH FRXUVH RI RXU LQLWLDO H[SHULPHQWV ZLWK WKH +J&O PRGHO RI LQGXFHG DXWRLPPXQLW\ ZH H[DPLQHG WKH $1R$ UHVSRQVH LQ DQ )[ FURVV EHWZHHQ WKH VXVFHSWLEOH %6-/ +6 ,JKEf DQG WKH UHVLVWDQW %7& +E ,JKDf VWUDLQV $V H[SHFWHG %6-/ Q f DQG %7& Q f ZHUH b DQG b VXVFHSWLEOH UHVSHFWLYHO\ $ UHSUHVHQWDWLYH H[DPSOH LV VKRZQ LQ )LJ ZKHUH D VHUXP VDPSOH IURP D %6-/ PRXVH WUHDWHG IRU ZHHNV ZLWK +J&O VKRZV D GLVWLQFW QXFOHRODU SDWWHUQ +RZHYHU WKH YDVW PDMRULW\ f RI )M LQWHUFURVVHV EHWZHHQ WKHVH WZR VWUDLQV ZHUH DOVR UHVLVWDQW WR +J&O WUHDWPHQW 7DEOH f 7KH UHVLVWDQFH RI WKH ) PLFH ZDV VXUSULVLQJ $OWKRXJK WKH H[DFW PHFKDQLVP IRU +J&OLQGXFHG DXWRLPPXQLW\ LV XQNQRZQ WKH IDFW WKDW WKH FODVV ,, PROHFXOHV DUH FR GRPLQDQWO\ H[SUHVVHG OHG XV WR H[SHFW WKDW WKH H[SUHVVLRQ RI ,$V ZRXOG EH VXIILFLHQW WR FRQIHU VXVFHSWLELOLW\ 7R PDNH VXUH WKLV UHVLVWDQFH ZDV QRW GXH WR WKH DOORW\SH GLIIHUHQFH

PAGE 27

%6-/,JKD +6 ,JKDf PLFH ZHUH SURGXFHG DQG IRXQG WR EH b VXVFHSWLEOH WR $1R$ IRUPDWLRQ 7KLV VKRZHG WKDW WKH ,JK ORFXV GLG QRW KDYH DQ HIIHFW RQ VXVFHSWLELOLW\ )LQDOO\ WR UXOH RXW D GRVH HIIHFW LQ WKH )[ FURVVHV JURXSV RI %6-/ [ %7&f) ZHUH GRVHG ZLWK HLWKHU RU PJNJ RI +J&O IRU ZHHNV ZLWK QR PLFH GHYHORSLQJ DQ $1R$ WLWHU GDWD QRW VKRZQf %HFDXVH LW KDV EHHQ UHSRUWHG WKDW WKH % EDFNJURXQG SURGXFHV ORZHU WLWHU $1R$ UHVSRQVHV WKDQ RWKHU +6 VWUDLQV RWKHU FURVVHV ZHUH WHVWHG ,QLWLDOO\ PLFH RQ WKH $ EDFNJURXQG ZHUH XVHG &RQVLVWHQW ZLWK SUHYLRXV VWXGLHV $6: +6f PLFH ZHUH b f VXVFHSWLEOH ZKLOH $% < +Ef PLFH ZHUH UHVLVWDQW f 1H[W )L FURVVHV EHWZHHQ WKHVH WZR VWUDLQV ZHUH JHQHUDWHG 'HVSLWH WKH PRUH VXVFHSWLEOH $ EDFNJURXQG RQO\ WZHQW\ SHUFHQW f RI $%< ; $6:f) PLFH VKRZHG $1R$ SRVLWLYLW\ DIWHU WUHDWPHQW 7R UXOH RXW D PDWHUQDO HIIHFW LQ )[ PLFH DV SUHYLRXVO\ VXJJHVWHG f WKH VDPH VWUDLQ ZDV SURGXFHG H[FHSW WKDW WKH $6: VWUDLQ ZDV XVHG DV WKH IHPDOH SDUHQW 7KLV FURVV $6: [ $%
PAGE 28

)LJXUH 6HUXP IURP D %6-/ PRXVH WUHDWHG IRU ZHHNV ZLWK +J&O RQ PRXVH IUR]HQ NLGQH\ VOLGHV DQG VWDLQHG ZLWK ),7&FRQMXJDWHG JRDW DQWLPRXVH ,J* DV D VHFRQG VWHS VKRZLQJ LQWHQVH VWDLQLQJ RI QXFOHROL

PAGE 29

7DEOH &RQJHQLH DQG )W PLFH VHUXP $1R$D DIWHU ZHHNV RI WUHDWPHQW ZLWK +J&O 6WUDLQ &ODVV ,, /$ ,(F ,J*O QG bSRVLWLYHH +DSORW\SH $ $f ( (f DOORW\SH E Ef D $%< E E E $6: V V V $% < [ $ 6:f); VE VE VE $6: [ $%
PAGE 30

IRUPDWLRQ DV %6-/ KRPR]\JRXV PLFH 7DEOH f 7KLV FOHDUO\ VKRZV WKDW LW LV WKH ,$E PROHFXOHV WKHPVHOYHV DQG QRW D OLQNHG JHQH ZKLFK FRQIHUUHG UHVLVWDQFH $V D QHJDWLYH FRQWURO QR PLFH f GHYHORSHG $1R$ DIWHU WUHDWPHQW ZLWK 3%6 7DEOH f 7DEOH (IIHFW RI $QE GHOHWLRQ RQ $1R$r LQ +J&OWUHDWHG% 6-/ [ %,$Eaf ) PLFH &ODVV ,, +DSORW\SHr QF +JF\ 1D&I bSRVLWLYHL +J&O 1D&O VE OJ V f$1R$ DQWLQXFOHRODU DQWLERGLHV PHDVXUHG DIWHU ZHHNV RI WUHDWPHQW PLFH HLWKHU GLG VEf RU GLG QRW Vf H[SUHVV ,$E EXW ZHUH KHWHUR]\JRXV DW DOO WKH 0+& ORFL A1XPEHU RI PLFH A)UDFWLRQ RI SRVLWLYH PLFH H PJ +J&ONJ VF WKUHH WLPHV SHU ZHHN I4 PO 1D&O VF WKUHH WLPHV SHU ZHHN A6LJQLILFDQW GLIIHUHQFH EHWZHHQ VEf DQG Vf PLFH S XVLQJ )LVFKHU H[DFW WHVWf 7R UXOH RXW WKH SRVVLELOLW\ WKDW WKH GLIIHUHQWLDO VXVFHSWLELOLW\ EHWZHHQ ,$VE DQG $n ZDV GXH VROHO\ WR LQFUHDVHG H[SUHVVLRQ RI ,$V DQG QRW D QHJDWLYH UHJXODWRU\ LQIOXHQFH E\ ,$E ZH H[DPLQHG ,$ OHYHOV E\ KDSORW\SHVSHFLILF IORZ F\WRPHWU\ %DVHOLQH ,$ H[SUHVVLRQ ZDV VLPLODU ZKHWKHU RU QRW WKH PRXVH FDUULHG WKH GLVUXSWHG JHQH 0), YV S f 0+& FODVV ,, H[SUHVVLRQ RQ 3%0&V ZDV VLPLODUO\ LQFUHDVHG LQ ERWK VWUDLQV ZKHQ WUHDWHG ZLWK +J&O )LJ f ,$ H[SUHVVLRQ ZDV LQFUHDVHG E\ b Sf LQ WKH PLFH ZKLFK FDUULHG WKH GLVUXSWHG DOOHOH ZKLOH LW ZDV LQFUHDVHG E\ b .f LQ WKRVH PLFH ZKLFK H[SUHVVHG ERWK KDSORW\SHV VEf 7KH LQFUHDVH LQ ,$E H[SUHVVLRQ LQ WKH PLFH ZKLFK FDUULHG WKDW JHQH ZDV b Sf 1R VLJQLILFDQW LQFUHDVH LQ FODVV ,, H[SUHVVLRQ ZDV VHHQ LQ WKRVH DQLPDOV WUHDWHG ZLWK 3%6

PAGE 31

$ % 8K %,$n VE )LJXUH (IIHFW RI +J&O RQ FODVV ,, H[SUHVVLRQ LQ%6-/ [ %,$A)M PLFH $f 3UHGRVH ,$V H[SUHVVLRQ ZDV VLPLODU LQ %,$VE DQG %,$6n PLFH ,$V H[SUHVVLRQ ZDV LQFUHDVHG E\ b LQ WKH PLFH ZKLFK FDUULHG WKH GLVUXSWHG DOOHOH ZKLOH LW ZDV LQFUHDVHG E\ b LQ WKRVH PLFH ZKLFK H[SUHVVHG ERWK KDSORW\SHV VEf %f 7KH LQFUHDVH LQ ,$E H[SUHVVLRQ LQ WKH PLFH ZKLFK FDUULHG WKDW JHQH ZDV b 1R VLJQLILFDQW LQFUHDVH LQ FODVV ,, H[SUHVVLRQ ZDV VHHQ LQ WKRVH DQLPDOV WUHDWHG ZLWK 3%6 rS

PAGE 32

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f %6-/,(DG PLFH UHWDLQHG WKH DELOLW\ WR JHQHUDWH DQ $1R$ WLWHU XSRQ +J&O WUHDWPHQW 7DEOH f 0HDQ ,( H[SUHVVLRQ LQ WKHVH PLFH 0), f ZDV VLPLODU WR %DOEF PLFH 0), f ZKLFK QRUPDOO\ H[SUHVV WKH ,( JHQH SURGXFW ,$V H[SUHVVLRQ EHWZHHQ %6-/ DQG %6-/,(DG PLFH ZDV DOVR VLPLODU 0), YV f )LJ f 7KH EUHHGLQJ VFKHPH XVHG WR JHQHUDWH WKH %6-/,(DG PLFH DOVR SURGXFHG %6-/ [ %,(DG f%&M PLFH ZKLFK HLWKHU GLG Q f RU GLG QRW Q f H[SUHVV WKH WUDQVJHQH ,W KDV EHHQ VXJJHVWHG LQ WKH %[6% DXWRLPPXQH PRGHO WKDW ,( H[SUHVVLRQ FRXOG EORFN DXWRDQWLERG\ SURGXFWLRQ E\ SURGXFLQJ DQ ,(GHULYHG SHSWLGH WKDW ELQGV WR ,$E ZLWK VXFK DIILQLW\ WKDW LW FRPSHWHV ZLWK WKH SXWDWLYH DXWRDQWLJHQGHULYHG SHSWLGHV IRU SUHVHQWDWLRQ f 6WDUWLQJ ZLWK WKH VDPH WUDQVJHQLF SDUHQW VWUDLQ DV ZDV XVHG LQ WKH SUHVHQW H[SHULPHQWV ZH KDYH DOVR FRQILUPHG WKLV REVHUYDWLRQ LQ WKH %[6% PRGHO XQSXEOLVKHG GDWDf ,I WKLV LV D JHQHUDO SURSHUW\ LQ DXWRLPPXQLW\ LW PLJKW EH H[SHFWHG WKDW ,( H[SUHVVLRQ ZRXOG FRQYHUW DQ RWKHUZLVH

PAGE 33

UHVLVWDQW +J&LWUHDWHG VE KHWHUR]\JRXV PRXVH LQWR D VHQVLWLYH RQH 'HVSLWH QRUPDO OHYHOV RI H[SUHVVLRQ RI ,( KRZHYHU DOO RI WKHVH PLFH ZHUH UHVLVWDQW 7DEOH f LDRLH ]M QHFW RU LH RQ YLYRD 6WUDLQ &ODVV ,, +DSORW\SH UHVSRQ L $IL VH ? $m ,( ([SUHVVLRQ QE bSRVLWLYHF %6-/,(.G V V V %6-/ [ %,(A)M VE VE VE %6-/[%f) A VE VE VE f$1R$ DQWLQXFOHRODU DQWLERGLHV PHDVXUHG DIWHU ZHHNV RI WUHDWPHQW ZLWK PJ +J&ONJ VF WKUHH WLPHV SHU ZHHN A1XPEHU RI PLFH 7UDFWLRQ RI SRVLWLYH PLFH ARQWUDQVJHQLF OLWWHUPDWH RI%6-/ [ %,(A) 0HDQ ,( ([SUHVVLRQ %6-/,( 9 n V )/+HLJKW fLE % 0HDQ ,$ ([SUHVVLRQ %6-/ ,< E f 7L7 IWS  )/ +HLJKW LE %DOEF LE n LE )/+HLJKW LE %6-/,( LE n LE LE )/+HLJKW nOE )LJXUH &ODVV ,, H[SUHVVLRQ LQ %6-/,(DG PLFH $f 0HDQ ,( H[SUHVVLRQ ZDV VLPLODU WR %DOEF PLFH %f 0HDQ ,$V H[SUHVVLRQ ZDV QRW DIIHFWHG E\ ,( H[SUHVVLRQ 6LQJOHSDUDPHWHU IOXRUHVFHQFH KLVWRJUDPV ,QFUHDVLQJ IOXRUHVFHQFH LQWHQVLW\ LV SORWWHG RQ WKH [D[LV LQ ORJ IOXRUHVFHQW XQLWV YHUVXV FHOO QXPEHU RQ WKH \D[LV 7KH QXPEHUV UHSUHVHQW WKH 0), PHDQ IOXRUHVFHQW LQWHQVLW\f RI WKH FHOOV H[SUHVVLQJ WKH DQWLJHQ RI LQWHUHVW

PAGE 34

5HVSRQVLYHQHVV WR ([RJHQRXV $QWLJHQ :DV ,QKHULWHG &RGRPLQDQWOY LQ ) 0LFH 7KH ODFN RI $1R$nV LQ RXU )W PLFH OHG XV WR HYDOXDWH WKHLU UHVSRQVHV WR DQ H[RJHQRXV DQWLJHQ +(/f $V H[SHFWHG PLFH RI WKH E KDSORW\SH % 7&f UHVSRQGHG SRRUO\ ZKLOH WKH %6-/ PLFH PRXQWHG DQ LQWHUPHGLDWH WLWHU $Q LQWHUPHGLDWH UHVSRQVH ZDV VHHQ LQ WKH )W FURVV )LJXUH f %DOEF PLFH UHVSRQGHG ZLWK D WLWHU DSSUR[LPDWHO\ IROG KLJKHU WKDQ DQ\ RI WKH PLFH RQ WKH % EDFNJURXQG GDWD QRW VKRZQf 7KXV LW DSSHDUV WKDW WKH UHVSRQVLYHQHVV LQ RXU )W PLFH WR +(/ ZDV LQKHULWHG LQ D FRGRPLQDQW IDVKLRQ 'LVFXVVLRQ 7KH PHUFXU\ FKORULGH UHVSRQVH LQ URGHQWV LV D FRPSOLFDWHG RQH DQG FODVVLILFDWLRQ RI UHVSRQGHU DQG QRQUHVSRQGHU VWUDLQV GHSHQGV RQ WKH RXWFRPH EHLQJ PHDVXUHG )RU H[DPSOH E\ $1R$ WLWHU WKH E KDSORW\SH LV D QRQUHVSRQGHU +RZHYHU E\ PHDVXULQJ WKH ZHLJKWV RI GUDLQLQJ SRSOLWHDO O\PSK QRGHV RI '%$ ; &%/f) PLFH H[SRVHG WR +J&O E KDSORW\SH KDV EHHQ FODVVLILHG DV LQWHUPHGLDWH UHVSRQGHUV 7KH DQWLILEULOODULQ UHVSRQVH E\ VXVFHSWLEOH PLFH KRZHYHU LV YHU\ UHSURGXFLEOH DQG VSHFLILF DQG WKHUHIRUH PRUH DPHQDEOH WR PHFKDQLVWLF GLVVHFWLRQ $V D ILUVW VWHS LW LV FUXFLDO WR FRQYLQFLQJO\ GHPRQVWUDWH WKDW WKH DQWLILEULOODULQ UHVSRQVH LV D SURSHUW\ RI WKH ,$ PROHFXOH LWVHOI DQG QRW D FORVHO\ OLQNHG JHQH 5HVXOWV LQ WKH OLWHUDWXUH WKXV IDU KDYH EHHQ VXJJHVWLYH EXW QRW FRQFOXVLYH 5RELQVRQ HW DOf XVHG WZR XQUHODWHG VWUDLQV $6: DQG &%/f WR JHQHUDWH WKH )V 7KHVH PLFH ZHUH b VXVFHSWLEOH WR GHYHORSLQJ

PAGE 35

$1R$ DOEHLW RI D ORZHU WLWHU 7KLV OHG WKH DXWKRUV WR FRQFOXGH WKDW WKH UHVSRQVH ZDV FRQWUROOHG E\ WKH + UHJLRQ DQG RQH RU PRUH XQOLQNHG JHQHV 5HFHQWO\ +XOWPDQ HW DO f IRXQG WKDW b RI PDOH DQG b RI IHPDOH PHUFXU\ WUHDWHG 6-/ [ %f)M PLFH VKRZHG D FOXPS\ ,J* $1R$ SDWWHUQ ZLWK D ORZ WLWHU 7KHVH GLIIHUHQFHV ZHUH DWWULEXWHG WR WKH QRQ+ JHQHV )RU H[DPSOH WKH $ JHQHWLF EDFNJURXQG VHHPHG WR HQKDQFH VXVFHSWLELOLW\ WR WKH +J&O WUHDWPHQW ZKLOH WKH %,2 EDFNJURXQG VHHPV WR VWUHQJWKHQ WKH UHVLVWDQFH 7KH\ DOVR VXJJHVWHG WKDW WKH % QRQ+ EDFNJURXQG JHQHV DUH PRUH VXSSUHVVLYH RQ $1R$ H[SUHVVLRQ WKDQ HTXLYDOHQW JHQHV IURP WKH % VWUDLQ $ VLPLODU HIIHFW RI QRQ+ JHQHV ZDV VHHQ E\ +XOWPDQ HW DO f ZKHQ FRPSDULQJ WKH $1R$ UHVSRQVH LQ VHYHUDO + V VWUDLQV %6 PLFH VKRZHG D PXFK ORZHU WLWHU WKDQ HLWKHU $6: RU 6-/ PLFH DIWHU +J&O WUHDWPHQW 7KLV ZDV DJDLQ DWWULEXWHG WR WKH QRQ+ JHQHV $ $QWL+(/ ,J* % $QWL+(/ ,J0 E Z %6-/%7& ), )LJXUH $QWL+(/ DQWLERG\ OHYHOV ZHHNV DIWHU LPPXQL]DWLRQ 7KH UHVSRQVH WR +(/ ZDV LQKHULWHG FRGRPLQDQWO\ LQ%6-/ ; %74)M PLFH %6-/ DQG %6-/ ; %74)L PLFH UHVSRQG WR +(/ ZKLOH %7& GR QRW $f ,J* %f ,J0 (') 6(0

PAGE 36

,Q WKH SURFHVV RI VWXG\LQJ WKH FHOOXODU FRQWURO RI WKH PHUFXU\LQGXFHG DQWL ILEULOODULQ UHVSRQVH ZH H[DPLQHG WKH VXVFHSWLELOLW\ RI D QXPEHU RI PLFH LQ RXU RZQ ODERUDWRU\ &RQWUDU\ WR SUHYLRXVO\ SXEOLVKHG UHVXOWV VXPPDUL]HG DERYH RXU )c PLFH VHOGRP SURGXFHG D PHDVXUDEOH $1R$ WLWHU PRVW OLNHO\ GXH WR WKH VWURQJ VXSSUHVVLYH HIIHFWV VHHQ ZKHQ XVLQJ PLFH RQ WKH % EDFNJURXQG 7KLV OHDG XV WR UHH[DPLQH WKH UROH RI FODVV ,, H[SUHVVLRQ 6HYHUDO JHQHV OLQNHG WR FODVV ,, KDYH EHHQ VKRZQ WR SDUWLFLSDWH LQ DQWLJHQ SURFHVVLQJ DQG SUHVHQWDWLRQ f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f PLFH GR PRXQW DQ LQWHUPHGLDWH VHUXP WLWHU f $Q DQDO\VLV RI DQWL+(/ VHUXP WLWHUV UHYHDOHG WKDW %7& PLFH UHVSRQGHG SRRUO\ DV H[SHFWHG ZKLOH ERWK ZKLOH %6-/ DQG %6-/ [ %74)M PLFH ZHUH LQWHUPHGLDWH UHVSRQGHUV ,W DSSHDUV WKDW WKHUH LV D GLIIHUHQFH LQ WKH LQKHULWDELOLW\ RI UHVSRQVLYHQHVV EHWZHHQ WKH +J&OLQGXFHG $1R$ UHVSRQVH DQG WKDW WR D FRQYHQWLRQDO H[RJHQRXV DQWLJHQ 7R LQYHVWLJDWH ZKHWKHU WKH UHVLVWDQFH RI )M PLFH LV GXH WR H[SUHVVLRQ RI WKH ,$E PROHFXOH RU D OLQNHG JHQH )W PLFH RQ WKH % EDFNJURXQG ZHUH SURGXFHG ZKLFK ODFNHG

PAGE 37

H[SUHVVLRQ RI WKH ,$E PROHFXOH GXH WR GLVUXSWLRQ RI WKH 0+& FODVV ,, ,$SE JHQH f %\ XVLQJ FRQJHQLH VWUDLQV ZKLFK GLIIHUHG DW RQO\ WKH + UHJLRQ ZH HOLPLQDWHG WKH HIIHFWV RI QRQ0+&OLQNHG JHQHV %HFDXVH WKH\ DUH RI D ORZUHVSRQGHU EDFNJURXQG WKH % VWUDLQV DOVR DOORZHG XV WR PRUH FDUHIXOO\ LVRODWH WKH HIIHFWV RI WKH 0+& )LQDOO\ WKH % EDFNJURXQG SHUPLWWHG XV WR WDNH DGYDQWDJH RI WKH DYDLODELOLW\ RI FODVV ,, WUDQVJHQLFV DQG WDUJHWHG GHOHWLRQV 2XU VWUDWHJ\ ZDV WR FURVV WKH &'r1 %,$nf WR D ZLOG W\SH %, $EE WR REWDLQ %,$E PLFH 7KHVH PLFH ZHUH WKHQ EUHG WR D %6-/ WR REWDLQ WKH WZR VWUDLQV XVHG IRU DQDO\VLV %,$VE DQG %,$n 7KH LQFLGHQFH RI $1R$ LQ WKRVH PLFH ZKLFK H[SUHVVHG WKH ,$E JHQH SURGXFW ZDV VLJQLILFDQWO\ GHFUHDVHG VKRZLQJ WKDW WKH ,$E PROHFXOH LWVHOI DQG QRW D OLQNHG JHQH ZDV UHVSRQVLEOH 7R UXOH RXW WKH SRVVLELOLW\ WKDW WKH PLFH FDUU\LQJ WKH GLVUXSWHG DOOHOH FRXOG EH H[SUHVVLQJ D KLJKHU GHQVLW\ RI ,$V PROHFXOHV WKXV OHDGLQJ WR WKH LQFUHDVHG VXVFHSWLELOLW\ ZH FRPSDUHG WKH SUHGRVH PHDQ ,$V H[SUHVVLRQ EHWZHHQ WKH VWUDLQV 7KH 0), RI ,$V H[SUHVVLRQ IRU WKH PLFH FDUU\LQJ WKH GLVUXSWHG DOOHOH ,$f ZDV QRW VLJQLILFDQWO\ GLIIHUHQW IURP WKH ,$VE PLFH 7KH SRVVLELOLW\ H[LVWV WKDW WKH ,$Mr JHQH SURGXFW FRXOG SDLU ZLWK WKH ,$SV PROHFXOH UHVXOWLQJ LQ DQ KDSORW\SHPLVPDWFKHG FODVV ,, PROHFXOH $DE $SVf 0DUWLQ HW DO f VKRZHG WKDW WKH LQHIILFLHQW DVVHPEO\ RI KDSORW\SHPLVPDWFKHG FODVV ,, SRO\SHSWLGHV UHVXOWV IURP WKHLU LQDELOLW\ WR FRPSHWH ZLWK WKH PDWFKHG SDLUV 'LIIHUHQW FODVV ,, 0+& D3 FKDLQ FRPELQDWLRQV YDU\ JUHDWO\ LQ WKHLU HIILFLHQF\ RI FHOOVXUIDFH H[SUHVVLRQ EDVHG RQ WKH DOOHOLF RULJLQ RI WKH D DQG 3 DPLQRWHUPLQDO UHJLRQV f ,I WKLV SDLU ZDV EHLQJ H[SUHVVHG ZH ZRXOG KDYH H[SHFWHG WR VHH DQ LQFUHDVH LQ WKH ,$V H[SUHVVLRQ LQ WKH PLFH FDUU\LQJ WKH GLVUXSWHG DOOHOH ,$Anf VLQFH WKH P$E 0.6 UHFRJQL]HV WKH 3 FKDLQ RI ,$V KRZHYHU QR GLIIHUHQFH ZDV GHWHFWHG DQG ZH EHOLHYH WKLV LV DQ XQOLNHO\ SRVVLELOLW\

PAGE 38

$QRWKHU SRVVLELOLW\ LV WKDW WKH +J&O KDG D GLIIHUHQWLDO HIIHFW RQ WKH WZR FODVV ,, PROHFXOHV ,I WKH +J&O KDG FDXVHG D JUHDWHU LQFUHDVH LQ WKH H[SUHVVLRQ RI WKH ,$E PROHFXOH LQ WKRVH PLFH ZKLFK H[SUHVVHG LW WKLV FRXOG OHDG WR D GHFUHDVHG SUHVHQWDWLRQ RI DQWLJHQ E\ WKH VXVFHSWLEOH DOOHOH ,$Vf SHUKDSV IDOOLQJ EHORZ WKH WKUHVKROG UHTXLUHG IRU 7 FHOO DFWLYDWLRQ 7KLV LV QRW OLNHO\ VLQFH ERWK FODVV ,, PROHFXOHV VKRZHG VLPLODU LQFUHDVHV LQ H[SUHVVLRQ XSRQ +J&O WUHDWPHQW 7R H[DPLQH GLUHFWO\ WKH HIIHFW RI ,( H[SUHVVLRQ LQ RWKHUZLVH VXVFHSWLEOH PLFH ZH JHQHUDWHG PLFH RQ D % EDFNJURXQG H[SUHVVLQJ WKH ,$V DQG WKH ,( PROHFXOHV /LNH RWKHU +6 PLFH %6-/ PLFH GR QRW H[SUHVV ,( GXH WR D GHOHWLRQ LQ WKH ,(f JHQH f 7R UHPHG\ WKLV ZH FURVVHG %6-/ PLFH ZLWK D % PRXVH FDUU\LQJ WKH ,( WUDQVJHQH 7J+ ,(Df%ULf f $IWHU EDFNFURVVLQJ WR %6-/ PLFH WKH SURJHQ\ ZKLFK H[SUHVVHG ,$V DQG ,( ZHUH VHOHFWHG 7KHVH PLFH UHPDLQHG b VXVFHSWLEOH WR $1R$ IRUPDWLRQ GHVSLWH H[SUHVVLQJ WKH ,( PROHFXOH 2XU UHVXOWV GR QRW VXSSRUW WKH REVHUYDWLRQ E\ 0LUFKHYD HW DO f WKDW WKH H[SUHVVLRQ RI ,( VXSSUHVVHV WKH LQGXFWLRQ RI $1R$V +XOWPDQ HW DO f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

PAGE 39

PROHFXOH FRXOG EH GLPLQLVKHG &RQYHUVHO\ ORZOHYHO H[SUHVVLRQ RI ,( FRXOG UHVXOW LQ OLWWOH SK\VLRORJLF HIIHFW 7R UXOH RXW HLWKHU RI WKHVH DUWLIDFWV ZH FRPSDUHG PHDQ ,( H[SUHVVLRQ EHWZHHQ RXU PLFH DQG D VWUDLQ ZKLFK QRUPDOO\ H[SUHVVHV ,( %DOEFf 7KHUH ZDV QR GLIIHUHQFH LQ WKH PHDQ H[SUHVVLRQ EHWZHHQ WKH WZR VWUDLQV 7KH H[SUHVVLRQ RI ,$ ZDV DOVR VLPLODU EHWZHHQ %6-/ DQG WKH %6-/,(DG PLFH 0RUHRYHU H[SUHVVLRQ RI ( ZDV DOVR SK\VLRORJLF LQ WKDW LW ZDV H[SUHVVHG RQ % EXW QRW 7 FHOOV GDWD QRW VKRZQf 0HULQR HW DO f VKRZHG WKDW WKH ,(DG WUDQVJHQH SURWHFWHG %[6% PLFH DJDLQVW 6/( 7KH\ SURSRVHG WKDW RYHUH[SUHVVLRQ WKH ,( D FKDLQ UHVXOWHG LQ WKH JHQHUDWLRQ RI H[FHVVLYH DPRXQWV RI D SHSWLGH GLVSOD\LQJ D KLJK DIILQLW\ IRU WKH ,$E PROHFXOH WKHUHE\ SUHYHQWLQJ DXWRDQWLJHQGHULYHG SHSWLGHV IURP EHLQJ SUHVHQWHG 7KLV SKHQRPHQRQ KDV DOVR EHHQ VHHQ LQ PLFH KRPR]\JRXV IRU WKH ,SU PXWDWLRQ f 7R LQYHVWLJDWH WKH SRVVLELOLW\ WKDW WKLV PLJKW ZRUN LQ UHYHUVH WR SURPRWH DXWRLPPXQLW\ LQ VRPH FDVHV ZH JHQHUDWHG WKH RWKHUZLVH UHVLVWDQW %6-/ ; %,(A)M PLFH ZKLFK HLWKHU GLG RU GLG QRW H[SUHVV ,( 7KHRUHWLFDOO\ WKLV FRXOG EORFN ,$E DV D IXQFWLRQDO FRPSHWLWRU DQG HQKDQFH DQ ,$V PHGLDWHG UHVSRQVH +RZHYHU QHLWKHU VWUDLQ GHYHORSHG $1R$ DIWHU WUHDWPHQW ZLWK +J&O 7KXV WKH H[SUHVVLRQ RI ,( GRHV QRW SURPRWH WKH GHYHORSPHQW RI $1R$ LQ D VWUDLQ ZKLFK LV RWKHUZLVH UHVLVWDQW 7KH UHVXOWV DUH FOHDU DOWKRXJK ZH FDQQRW IRUPDOO\ H[FOXGH WKH SRVVLELOLW\ WKDW D SHSWLGH GHULYHG IURP ,( ZDV HYHQ PRUH HIIHFWLYH LQ EORFNLQJ ,$V +RZHYHU LI WKLV ZHUH WKH FDVH WKHQ ZH ZRXOG KDYH H[SHFWHG ,( WR KDYH GHFUHDVHG +J&O LQGXFHG VXVFHSWLELOLW\ LQ WKH %6-/,(DG PLFH 7DNHQ WRJHWKHU RXU UHVXOWV FRQFOXVLYHO\ GHPRQVWUDWH IRU WKH ILUVW WLPH WKDW +J&O UHVSRQVLYHQHVV LV UHJXODWHG E\ WKH ,$ KDSORW\SH LWVHOI DQG QRW D FORVHO\ OLQNHG JHQH

PAGE 40

+RZHYHU WKH DELOLW\ WR GRZQUHJXODWH WKH LPPXQH UHVSRQVH E\ FRH[SUHVVLRQ LQ D UHVLVWDQW KDSORW\SH LQ DQ RWKHUZLVH VHQVLWLYH VWUDLQ LV D FRPSOLFDWHG RQH $QWLJHQ FRPSHWHWLRQ SURSRVHG WR EH LPSRUWDQW LQ RWKHU DXWRLPPXQH PRGHOV f GRHV QRW DSSHDU WR KDYH D GRPLQDQW UROH LQ WKH +J&OLQGXFHG UHVSRQVH

PAGE 41

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f DOO RI ZKLFK DSSHDU DV VRRQ DV RQH ZHHN DIWHU WKH VWDUW RI WUHDWPHQW f 7KH SULQFLSOH DXWRDQWLERG\ WDUJHW LV ILEULOODULQ D 8 ULERQXFOHRSURWHLQ IRXQG LQ WKH QXFOHROXV DQG LQYROYHG LQ WKH ILUVW VWHS RI SUHULERVRPDO SURFHVVLQJ f 7KH DQWLILEULOODULQ VSHFLILFLW\ LV DOVR VHHQ LQ SDWLHQWV ZLWK VFOHURGHUPD DQG RWKHU FRQQHFWLYH WLVVXH GLVRUGHUV f 5HFHQWO\ D FRPSDULVRQ RI WKH VSHFLILFLW\ RI WKH KXPDQ DQG PXULQH DXWRDQWLERGLHV VKRZHG WKDW WKH\ UHFRJQL]H VLPLODU LI QRW LGHQWLFDO HSLWRSHV f DQG VXJJHVWHG DQ DQWLJHQGULYHQ UHVSRQVH LQ ERWK 6LPLODU WR KXPDQ DXWRLPPXQH GLVHDVH ERWK 0+& DQG QRQ0+& JHQHV KDYH EHHQ VKRZQ WR KDYH D SURQRXQFHG HIIHFW RQ VXVFHSWLELOLW\ f $V GHILQHG E\ WKH SUHVHQFH RI

PAGE 42

DQWLILEULOODULQ DQWLERG\ PLFH RI WKH +6 KDSORW\SH DUH KLJK UHVSRQGHUV ZKLOH WKRVH RI WKH +E DQG +G KDSORW\SHV DUH QRQUHVSRQGHUV f ,Q FRQWUDVW QRQ0+& JHQHV DIIHFW SULPDULO\ DQWLERG\ WLWHUV EXW QRW VSHFLILFLW\ f 7KURXJK WKH XVH RI LQWUD+ UHFRPELQDQWV WKH SUHVHQFH RI WKH +J&OLQGXFHG DQWLILEULOODULQ DQWLERG\ UHVSRQVH KDV EHHQ PDSSHG VSHFLILFDOO\ WR WKH ,$ UHJLRQ f 3UHYLRXV VWXGLHV LQ RXU ODERUDWRU\ KDYH VKRZQ WKDW )M DQLPDOV EHWZHHQ 0+&FRQJHQLF VXVFHSWLEOH +6 DQG UHVLVWDQW +E PLFH WR EH UHVLVWDQW WR +J&OLQGXFHG $1R$V f 7KLV LV VXUSULVLQJ VLQFH WKH ,$ PROHFXOH LV FR GRPLQDQWO\ H[SUHVVHG LQ WKH )W PLFH DQG LQ RWKHU DXWRLPPXQH PRGHOV KHWHUR]\JRVLW\ RI FODVV ,, HLWKHU HQKDQFHV DXWRLPPXQLW\ RU PRGHVWO\ DIIHFWV DQWLERG\ WLWHUV f 7KHUHIRUH WKH SURIRXQG UHVLVWDQFH WR +J&OLQGXFHG $1R$V VHHQ LQ DOO KDSORW\SH KHWHUR]\JRXV PLFH WHVWHG UDLVHG WKH SRVVLELOLW\ WKDW UHVLVWDQFH ZDV PHGLDWHG E\ D GRPLQDQW JHQH OLQNHG WR ,$E +RZHYHU IXUWKHU H[SHULPHQWV GHPRQVWUDWHG WKDW UHVLVWDQFH ZDV LQ IDFW FDXVHG E\ FRH[SUHVVLRQ RI WKH ,$E PROHFXOH LWVHOI DQG WKDW WKLV RXWFRPH ZDV QRW PHUHO\ WKH UHVXOW RI ORZHU H[SUHVVLRQ RI WKH VXVFHSWLEOH ,$V KDSORW\SH RQ WKH RWKHUZLVH UHVSRQVLYH % FHOOV f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

PAGE 43

0LFH 0DWHULDOV DQG 0HWKRGV &%/% ,$E ,JKEf &%/-6-/ %6-/ ,$V ,JKEf &%/-J]r 7K\OD *SI %7& ,$E ,JKDf PLFH ZHUH RULJLQDOO\ REWDLQHG IURP 7KH -DFNVRQ /DERUDWRU\ %DU +DUERU 0(f DQG PDLQWDLQHG LQ RXU EUHHGLQJ IDFLOLW\G %6-/,JKD PLFH ZHUH GHYHORSHG E\ LQWHUFURVVLQJ %6-/ DQG%6-/ [ %7&f)@ )$&6 DQDO\VLV ZDV XVHG WR VHOHFW SURJHQ\ ZKLFK ZHUH ,$V DQG ,JKDE 7KHVH PLFH ZHUH WKHQ LQWHUFURVVHG DQG WKH SURJHQ\ VHOHFWHG ZKLFK ZHUH %6-/,JKD XVLQJ )$&6 DQG WKH VWUDLQ ZDV HVWDEOLVKHG IURP D VLQJOH EUHHGLQJ SDLU $OO PLFH ZHUH KRXVHG LQ $$$/$& DSSURYHG IDFLOLWLHV LQ FRPSOLDQFH ZLWK DOO DSSOLFDEOH IHGHUDO VWDWH DQG ORFDO ODZV 3UHSDUDWLRQ RI &KLPHUDV 3UHSDUDWLRQ RI FKLPHUDV ZDV DV SUHYLRXVO\ GHVFULEHG f 5HFLSLHQW PLFH ZHUH SURYLGHG ZLWK 6HSWUDp b YYf WUHDWHG ZDWHU WKH GD\ SULRU WR LUUDGLDWLRQ 2Q WKH GD\ SULRU WR ERQH PDUURZ WUDQVIHU WKH PLFH ZHUH WUHDWHG ZLWK WZR GRVHV RI UDG RI \ UDGLDWLRQ *DPPDFHOO $WRPLF (QHUJ\ RI &DQDGD /WG 2WWDZD &DQDGDf VHSDUDWHG E\ KRXUV 7KH WUDQVIHU LQYROYHG LY UHFRQVWLWXWLRQ ZLWK D WRWDO RI ERQH PDUURZ FHOOV IURP DJH DQG VH[PDWFKHG GRQRUV %RQH PDUURZ FHOOV ZHUH GHSOHWHG RI PDWXUH 7 FHOOV E\ LQFXEDWLRQ DW r & IRU PLQ ZLWK D PL[WXUH RI DQWLPRXVH 7FHOO VHUXP &HGDUODQH /DERUDWRULHV +RUQE\ 2QWDULR &DQDGDf UDW ,J0 DQWL&' ff DQG 0 UDW ,J0 DQWL&' ff IROORZHG E\ WUHDWPHQW ZLWK & DW r& IRU K /RZ7R[ *XLQHD 3LJ

PAGE 44

& &HGDUODQH /DERUDWRULHV +RUQE\ 2QWDULR &DQDGDf 7R SUHYHQW JUDIW UHMHFWLRQ WKH %6-/ PLFH UHFHLYLQJ%6-/ [ %74)M FHOOV ZHUH JLYHQ PJ LS RI0P7O PRXVH ,J*D DQWL&' ff DW WKH WLPH RI WUDQVIHU DQG 0 ZHUH REWDLQHG IURP 'U 'DYLG +DUULV 8QLYHUVLW\ RI $UL]RQD 7XVFRQ $=f DQG SUHSDUHG IURP RYHUJURZQ FHOO FXOWXUH VXSHUQDWDQW ZKLFK ZDV DIILQLW\ SXULILHG RQ D SURWHLQ FROXPQ ,Q 9LYR 7UHDWPHQWV +J&O 6LJPD &KHPLFDO &R 6W /RXLV 02f ZDV SUHSDUHG LQ VWHULOH S\URJHQIUHH SKRVSKDWH EXIIHUHG VDOLQH 3%6f 0LFH ZHUH LQMHFWHG VXEFXWDQHRXVO\ DW D GRVH RI PJNJ WKUHH WLPHV ZHHNO\ DIWHU JUDIW DFFHSWDQFH ZDV YHULILHG XVLQJ IORZ F\WRPHWU\ 0LFH ZHUH LPPXQL]HG LQWUDSHULWRQHDOO\ ZLWK SJ RI KXPDQ ,J* +** 6LJPD &KHPLFDO &R 6W /RXLV 02 f HPXOVLILHG LQ &)$ 'LIFR /DERUDWRULHV 'HWURLW 0,f %RRVWHU LPPXQL]DWLRQV FRQVLVWHG RI D VHFRQG LQWUDSHULWRQHDO LQMHFWLRQ RI SJ RI +** LQ VWHULOH S\URJHQIUHH 3%6 )ORZ &\WRPHWU\ $SSUR[LPDWHO\ ZHHNV DIWHU ERQH PDUURZ WUDQVIHU SL RI WDLO YHLQ EORRG ZDV FROOHFWHG LQWR KHSDULQL]HG WXEHV 3%0&V ZHUH LVRODWHG XVLQJ /\PSKRO\WH 0 &HUGDUODQH /DERUDWRULHV +RUQE\ 2QWDULR &DQDGDf GHQVLW\ JUDGLHQWV 7KH FHOOV ZHUH WKHQ FROOHFWHG LQWR 3%6 VXSSOHPHQWHG ZLWK b )&6 DQG b 1D1 )RU FHOO VXUIDFH VWDLQLQJ VDWXUDWLQJ DPRXQWV RI ELRWLQ\ODWHG PRXVH ,J*D DQWL,$Eff DQG IOXRUHVFHLQDWHG 7,% PRXVH ,J*D DQWL,$Eff RU ELRWLQ\ODWHG $) PRXVH

PAGE 45

,J*O DQWL,J0Eff DQG IOXRUHVFHLQDWHG '6 PRXVH DQWL,J0Dff RU ELRWLQ\ODWHG +,6 PRXVH ,J*D DQWL&'f DQG IOXRUHVFHLQDWHG + UDW ,J*E DQWL&'f ZHUH XVHG DV WKH ILUVW VWHS 7KH VHFRQG VWHS FRQVLVWHG RI LQFXEDWLQJ ZLWK SK\FRHU\WKULQ 3(fFRQMXJDWHG VWUHSWDYLGLQ -DFNVRQ ,PPXQR5HVHDUFK :HVW *URYH 3$f 7KH FHOOV ZHUH WKHQ ZDVKHG WKUHH WLPHV LQ 3%6 DQG IL[HG ZLWK DQ HTXDO YROXPH RI b SDUDIRUPDOGHK\GH /LVW PRGH GDWD ZDV DFTXLUHG RQ D )$&6FDQ IORZ F\WRPHWHU %HFWRQ 'LFNLQVRQ 3DOR $OWR &$f XVLQJ 3& /\VLV VRIWZDUH 'HDG FHOOV ZHUH H[FOXGHG E\ IRUZDUG DQG VLGH VFDWWHU JDWLQJ /LVW PRGH ILOHV ZHUH WKHQ DQDO\]HG XVLQJ /\VLV ,, VRIWZDUH 0RQRFORQDO DQWLERGLHV ZHUH ODEHOHG HLWKHU ZLWK ELRWLQ K\GUD]LGH f RU IOXRUHVFHLQ LVRWKLRF\DQDWH ),7&f f DV QHHGHG $W WKH VWXG\ WHUPLQDWLRQ WKH VDPH SURFHGXUHV ZHUH XVHG ZLWK WKH H[FHSWLRQ WKDW VSOHQRF\WHV ZHUH XVHG IRU FHOO VWDLQLQJ ,QGLUHFW ,PPXQRIOXRUHVFHQFH 6HUD IURP PLFH FROOHFWHG ZHHNV DIWHU WKH LQLWLDWLRQ RI +J&O ZDV WHVWHG IRU WKH SUHVHQFH RI $1R$ E\ LQGLUHFW LPPXQRIOXRUHVFHQFH XVLQJ FRPPHUFLDOO\ SUHSDUHG PRXVH IUR]HQ NLGQH\ VOLGHV 6DQRIL &KDVND 01f 7KH VOLGHV ZHUH LQFXEDWHG ZLWK VHUD GLOXWHG LQ 3%6 IRU PLQXWHV DW URRP WHPSHUDWXUH )RU QRQDOORW\SHVSHFLILF $1R$ WKH VOLGHV ZHUH LQFXEDWHG IRU PLQ DW URRP WHPSHUDWXUH ZLWK ),7&FRQMXJDWHG JRDW DQWLn PRXVH ,J* )F IUDJPHQW VSHFLILF -DFNVRQ ,PPXQR5HVHDUFK :HVW *URYH 3$f GLOXWHG LQ 3%6 )RU WKH DOORW\SHVSHFLILF $1R$ GHWHUPLQDWLRQ VOLGHV ZHUH LQFXEDWHG ZLWK UDEELW DQWLPRXVH ,J*DD RU ,J*DE 1RUGLF /DEV &DSLVWUDQR %HDFK &$f IROORZHG E\ ),7&FRQMXJDWHG JRDW DQWLUDEELW ,J* -DFNVRQ ,PPXQR5HVHDUFK :HVW *URYH 3$f

PAGE 46

7KHVH UHDJHQWV KDG EHHQ SUHWLWUDWHG E\ (/,6$ DJDLQVW DQ DOORW\SHQRQVSHFLILF UDEELW DQWL ,J*D DQWLERG\ 1RUGLF /DEVf WR SURGXFH HTXLYDOHQW VHQVLWLYLWLHV $QWLQXFOHRODU VWDLQLQJ ZDV HYLGHQFHG E\ LQWHQVH KRPRJHQHRXV VWDLQLQJ RI WKH QXFOHROL $OORW\SH6SHFLILF (/,6$ 7KH SURWRFROV XVHG IRU PHDVXUHPHQW RI DOORW\SHVSHFLILF VHUXP WRWDO ,J0 DQG ,J*D ZHUH PLQRU PRGLILFDWLRQV RI SUHYLRXVO\ GHVFULEHG SURFHGXUHV f )RU VHUXP WRWDO ,J0 VDPSOHV ZHUH GHYHORSHG ZLWK DIILQLW\SXULILHG GRQNH\ DQWLPRXVH ,J0 -DFNVRQ ,PPXQR5HVHDUFK :HVW *URYH 3$f ,J0D DQG ,J0E ZHUH PHDVXUHG XVLQJ '6 f DQG $) f UHVSHFWLYHO\ 6HUXP WRWDO ,J*D ZDV GHWHUPLQHG XVLQJ UDEELW DQWLPRXVH ,J*D 1RUGLF /DEV &DSLVWUDQR %HDFK &$f ZKLOH DOORW\SHVSHFLILF ,J*D ZDV PHDVXUHG XVLQJ HLWKHU UDEELW DQWLPRXVH ,J*DD RU ,J*DE 1RUGLF /DEV &DSLVWUDQR %HDFK &$f 7KHVH UHDJHQWV KDG EHHQ SUHWLWUDWHG E\ (/,6$ DJDLQVW WKH DOORW\SHQRQVSHFLILF UDEELW DQWL,J*D DQWLERG\ 1RUGLF /DEVf WR SURGXFH HTXLYDOHQW VHQVLWLYLWLHV 7KH UDEELW DQWLERGLHV ZHUH GHWHFWHG XVLQJ DONDOLQH SKRVSKDWDVHFRQMXJDWHG GRQNH\ DQWLUDEELW ,J* -DFNVRQ ,PPXQR5HVHDUFK :HVW *URYH 3$f $OORW\SHVSHFLILF ,J*D DQWL+** ZDV PHDVXUHG XVLQJ WKH VDPH SURWRFRO H[FHSW WKDW +** ZDV XVHG DV WKH ILUVW VWHS 5HVXOWV 0+&5HVWULFWHG 7 &HOO +HOS :DV 5HTXLUHG WR 3URGXFH $1R$V ([SRVXUH WR +J&O LQGXFHV D ZLGH UDQJH RI SK\VLRORJLF HIIHFWV LQ PLFH LQFOXGLQJ WKH UHOHDVH RI ODUJH DPRXQWV RI F\WRNLQHV HVSHFLDOO\ WKH 7K F\WRNLQH ,/ ,W ZDV

PAGE 47

WKHUHIRUH SRVVLEOH WKDW WKH DQWLILEULOODULQ UHVSRQVH FRXOG UHVXOW IURP QRQ0+&UHVWULFWHG LQWHUDFWLRQV LQ JHQHWLFDOO\ VXVFHSWLEOH PLFH 7R WHVW WKLV SRVVLELOLW\ %7& PLFH ZHUH UHFRQVWLWXWHG ZLWK 7 FHOOGHSOHWHG %6-/ PDUURZ %HFDXVH LQWUDWK\PLF SRVLWLYH VHOHFWLRQ RI 7 FHOOV LV PHGLDWHG E\ WKH UDGLRUHVLVWDQW WK\PLF FRUWLFDO HSLWKHOLDO FHOOV f LQ WKHVH PLFH DOO &' 7 FHOOV ZRXOG EH SRVLWLYHO\ VHOHFWHG WR LQWHUDFW ZLWK WKH KRVW ,$E DQG QRW WKH GRQRU ,$V KDSORW\SH 7KHUHIRUH WKH %6-/GHULYHG % FHOOV ZRXOG QRW EH H[SHFWHG WR UHFHLYH 0+&UHVWULFWHG KHOS ,Q FRQWUDVW FHQWUDO WROHUDQFH LV PHGLDWHG E\ ERQH PDUURZGHULYHG GHQGULWLF FHOOV DOWKRXJK WK\PLF HSLWKHOLDO FHOOV FDQ PDNH D VLJQLILFDQW FRQWULEXWLRQ f DQG WKHUHIRUH WKH DQLPDOV ZRXOG EH WROHUDQW WR ERWK KDSORW\SHV )ORZ F\WRPHWULF DQDO\VLV RI SHULSKHUDO EORRG O\PSKRF\WHV VKRZHG WKDW DOO RI WKH % FHOOV ZHUH RI GRQRU %6-/ RULJLQ 6XUSULVLQJO\ GHVSLWH WKH SUHVHQFH RI WKHVH VXVFHSWLEOH % FHOOV DQG WKH PDVVLYH DFWLYDWLRQ RI F\WRNLQHV IROORZLQJ +J&O DGPLQLVWUDWLRQ QRQH RI WKH DQLPDOV GHYHORSHG $1R$V 7DEOH f 7KH DGGLWLRQ RI %7& FHOOV LQ WKH LQRFXOXP GLG QRW DOWHU WKH RXWFRPH ,Q PDUNHG FRQWUDVW V\QJHQHLFDOO\ UHFRQVWLWXWHG %6-/ PLFH UHVSRQGHG ZHOO WR PHUFXU\ FKORULGH GHPRQVWUDWLQJ WKDW ODFN RI UHVSRQVLYHQHVV ZDV QRW DQ DUWLIDFW RI UDGLDWLRQ :H WKHUHIRUH FRQFOXGH WKDW WKH DQWLILEULOODULQ VSHFLILFLW\ LV PHGLDWHG E\ ,$V UHVWULFWHG 7 FHOO KHOS

PAGE 48

7DEOH 0+&UHVWULFWHG 7 FHOO KHOS ZDV UHTXLUHG WR SURGXFH $1R$Vr 'SQRU +RVW QE $1R$ 3RVLWLYH b< %6-/ t %7& %7& %6-/ t %7& %6-/ %6-/ %6-/ %6-/ %7& f$1R$ DQWLQXFOHRODU DQWLERGLHV GXPEHU RI PLFH 7UDFWLRQ RI SRVLWLYH PLFH $EVHQFH RI $1R$ 3URGXFWLRQ LQ +DSORW\SH+HWHUR]YJRXV 0LFH :DV 1RW 'XH WR D 'LIIHUHQFH LQ 7K\PLF (GXFDWLRQ $OWKRXJK WKH DERYH H[SHULPHQWV VKRZHG WKH LPSRUWDQFH RI 7% FHOO FROODERUDWLRQ LW LV SRVVLEOH WKDW ,$E H[SUHVVLRQ LQ WKH KRVW DQGRU GRQRU DIIHFWHG WK\PLF HGXFDWLRQ DQG WKHUHIRUH HOLPLQDWHG D SRSXODWLRQ RI UHVSRQVLYH 7 FHOOV 7KLV OHG XV WR LQYHVWLJDWH WKH UHVSRQVHV LQ %6-/ [ %74) KRVWV :KHQ %6-/ [ %7&f); PLFH ZHUH XVHG DV KRVWV DQG UHFRQVWLWXWHG ZLWK %6-/ WKH\ DOO f UHVSRQGHG WR +J&O :KHQ WKH VDPH KRVWV ZHUH XVHG EXW WKH GRQRU FKDQJHG WR %7& QRQH f RI WKH PLFH SURGXFHG $1R$V 7DEOH f 7KHUHIRUH WKH GRQRU KDSORW\SH GHWHUPLQHG ZKHWKHU RU QRW DXWRDQWLERGLHV DUH SURGXFHG ZKHQ GHYHORSLQJ FHOOV ZHUH SRVLWLYHO\ VHOHFWHG RQ ERWK KDSORW\SHV $OORJHQHLF 0L[HG &KLPHUDV 'HYHORSHG $1R$ 7LWHUV LQ 5HVSRQVH WR +J&8 ,Q WKH UDW PRGHO RI PHUFXU\ FKORULGHLQGXFHG DXWRLPPXQLW\ DOORJHQHLF PL[HG PLFURFKLPHULVP KDV EHHQ DEOH WR LQGXFH D VWDWH RI WROHUDQFH f ,Q FRQWUDVW RWKHU PRGHOV KDYH VKRZQ WKH SRWHQWLDO IRU HSLWRSH VSUHDGLQJ RQFH WROHUDQFH LV EURNHQ f

PAGE 49

7KHUHIRUH LW ZDV RI LQWHUHVW WR GHWHUPLQH ZKHWKHU WKH FRH[LVWHQFH RI% FHOOV VHQVLWLYH DQG UHVLVWDQW WR WKH HIIHFWV RI +J&O FRXOG LQIOXHQFH HDFK RWKHU 7R WHVW WKHVH SRVVLELOLWLHV ZH OHWKDOO\ LUUDGLDWHG %6-/ [ %7&f) PLFH DQG WUDQVIHUUHG 7 FHOOGHSOHWHG ERQH PDUURZ IURP WKH IROORZLQJ f %6-/ DORQH f %7& DORQH f D FRPELQDWLRQ RI %7& t %6-/ DQG f %6-/ [ %74)M DORQH ,Q DOO IRXU JURXSV 7 FHOOV ZRXOG EH SRVLWLYHO\ VHOHFWHG E\ HLWKHU ,$V RU ,$E $IWHU FRQILUPDWLRQ RI PL[HG FKLPHULVP LQ JURXS VHH EHORZf WKH PLFH ZHUH WUHDWHG ZLWK +J&O DQG WKH UHVXOWV DUH VKRZQ LQ 7DEOH $V H[SHFWHG %6-/ [ %7&f) PLFH UHFRQVWLWXWHG ZLWK RQO\ %6-/ ERQH PDUURZ UHVSRQGHG UHDGLO\ WR +J&O ZKLOH PLFH JLYHQ RQO\ UHVLVWDQW %7& ERQH PDUURZ IDLOHG WR GHYHORS DQ $1R$ DQWLERG\ WLWHU ,Q FRQWUDVW V\QJHQHLF UHFRQVWLWXWLRQ ZLWK UHVLVWDQW %6-/ [ %7&f)M ERQH PDUURZ UHVXOWHG LQ D YHU\ SRRU $1R$ UHVSRQVH 7KHVH UHVXOWV HVVHQWLDOO\ GXSOLFDWHG RXU H[SHULHQFH LQ DQDORJRXV QRQFKLPHULF +J&OWUHDWHG %6-/ %7& DQG %6-/ [ %7&f) PLFH f DQG DJDLQ GHPRQVWUDWHG WKDW UDGLDWLRQ GLG QRW DIIHFW WKLV RXWFRPH 6WULNLQJO\ PLFH JLYHQ D FRPELQDWLRQ RI UHVLVWDQW DQG VXVFHSWLEOH ERQH PDUURZ KDG DQ LQWHUPHGLDWH UHVXOW ZLWK QHDUO\ b RI WKH PLFH UHVSRQGLQJ 7DEOH f %\ DOORW\SH VSHFLILF $1R$ DOO RI WKH UHVSRQGHUV SURGXFHG H[FOXVLYHO\ E DOORW\SH DXWRDQWLERG\ DQG WKHUHIRUH RI GRQRU %6-/ RULJLQ GDWD QRW VKRZQf 7KLV LV QRW GXH WR DQ LQWULQVLF SURSHUW\ RI WKH D DOORW\SH IRU WKH D DOORW\SH %6-/,JKD VWUDLQ UHVSRQGHG HTXDOO\ ZHOO WR PHUFXU\ GDWD QRW VKRZQf 7KXV WKH SUHVHQFH RI $1R$ LQ WKHVH PLFH UHVXOWHG IURP WKH ORVV RI WROHUDQFH WR ILEULOODULQ LQ ,$V % FHOOV

PAGE 50

7DEOH $1R$D UHVSRQVH LQ FKLPHULF %7& [ %6-/M) KRVW PLFH *URXS 6RXUFH RI 'RQRU %RQH 0DUURZ a $1R$ 3RVLWLYH bfn %6-/ %7& %6-/ t %7& %6-/ [ %7&M) D$1R$ DQWLQXFOHRODU DQWLERGLHV A1XPEHU RI PLFH 7UDFWLRQ RI SRVLWLYH PLFH /DFN RI $1R$ LQ 1RQ5HVSRQVLYH 0L[HG &KLPHUDV :DV 8QUHODWHG WR % &HOO &RPSRVLWLRQ 2I LQWHUHVW WKH UHVSRQVH UDWH RI b LQ WKH PL[HG FKLPHUDV *URXS f ZDV VLJQLILFDQWO\ GLIIHUHQW WKDQ HLWKHU WKH b VHHQ LQ *URXS S E\ ]WHVW IRU SURSRUWLRQVf RU WKH b VHHQ LQ *URXS Sf DQG VXJJHVWHG WKDW WKH SUHVHQFH RI UHVLVWDQW FODVV ,, FRXOG H[HUW D QHJDWLYH LQIOXHQFH HYHQ ZKHQ QRW FRH[SUHVVHG RQ WKH FHOOV RWKHUZLVH FDSDEOH RI UHVSRQGLQJ 7R HYDOXDWH IRU WKH SRVVLELOLW\ WKDW WKLV HIIHFW ZDV PHUHO\ GXH WR GLOXWLRQ RI SRWHQWLDOO\ VXVFHSWLEOH ,$VEHDULQJ % FHOOV E\ UHVLVWDQW ,$E % FHOOV ZH FRPSDUHG WKH % FHOO PDNHXS RI WKH SRVLWLYH DQG QHJDWLYH PLFH 7KH UHVXOWV DUH VKRZQ LQ )LJXUH 7KH PHGLDQ b RI ,$V H[SUHVVLQJ % FHOOV ZDV b LQ WKRVH PLFH ZKLFK IDLOHG WR SURGXFH $1R$V ZKLOH LW ZDV b LQ WKRVH PLFH WKDW GLG )LJXUH $f (YHQ PLFH ZKLFK KDG DV IHZ DV b RI WKHLU % FHOOV IURP WKH VHQVLWLYH SDUHQW ZHUH DEOH WR SURGXFH DXWRDQWLERGLHV &RPSDUDEOH UHVXOWV ZHUH VHHQ ZKHQ DQDO\]HG E\ % FHOO DOORW\SH )LJXUH %f 7KHUHIRUH WKHVH GDWD VXJJHVW WKDW WKH GHFUHDVHG UHVSRQVLYHQHVV LQ WKH SUHVHQFH RI UHVLVWDQW % FHOOV ZDV QRW MXVW D GLOXWLRQDO HIIHFW

PAGE 51

% &HOOV RI %RWK %7& DQG %6-/ 2ULJLQ :HUH )XQFWLRQDO LQ *URXS 0L[HG &KLPHUDV ,W ZDV SRVVLEOH WKDW WKH SUHVHQFH RI D DQG E DOORW\SH % FHOOV LQ WKH PL[HG FKLPHUDV LQIOXHQFHG WKH GHYHORSPHQW RI DQWLQXFOHRODU DQWLERGLHV WKURXJK DQWLJHQ QRQVSHFLILF PHFKDQLVPV )RU H[DPSOH LQ VRPH FRPELQDWLRQV DOORW\SHVSHFLILF VXSSUHVVLRQ RI LPPXQH UHVSRQVHV KDV EHHQ VHHQ f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f DQG ERWK $1R$ SRVLWLYH DQG QHJDWLYH PLFH UHVSRQGHG HTXDOO\ ZHOO GDWD QRW VKRZQf /DFN RI +J&N,QGXFHG $1R$ 5HVSRQVH E\ +DSORW\SH+HWHUR]\JRXV 0LFH :DV 1RW 'XH WR $EVHQFH RI $QWL)LEULOODULQ6SHFLILF ,$V5HVWULFWHG 7 &HOO +HOS 7R GHWHUPLQH ZKHWKHU RU QRW WKH LQWHUPHGLDWH UHVSRQVH VHHQ LQ WKH PL[HG FKLPHUDV DQG LQ +J&OUHVLVWDQW %6-/ [ %74)M PLFH ZDV GXH WR D ODFN RI ,$VUHVWULFWHG 7 FHOO KHOS ZH UHFRQVWLWXWHG %6-/ [ %74)M PLFH ZLWK D FRPELQDWLRQ RI V\QJHQHLF %6-/ [ %74)L DQG %6-/ ERQH PDUURZ )RUW\ILYH SHUFHQW f RI WKHVH PLFH SURGXFHG

PAGE 52

b,$n %7& %6-/ f§! )/ $ &ODVV ,, ([SUHVVLRQ % 6XUIDFH ,J0 )LJXUH ,PPXQRIOXRUHVFHQFH RQ SHULSKHUDO EORRG O\PSKRF\WHV RI SDUHQWDO LQWR )M FKLPHUDV 'DWD DUH SUHVHQWHG DV SHUFHQW RI FHOOV H[SUHVVLQJ FODVV ,, $ &ODVV ,, H[SUHVVLRQ % VXUIDFH ,J0 b,J0

PAGE 53

%7& %6-/ f§!)M $ 6HUXP ,J0 % 6HUXP ,J*D R $1R$ SRV QHJ ; L ; ; R )LJXUH $OORW\SHVSHFLILF WRWDO ,J0 DQG ,J*D (/,6$ GDWD RI SDUHQWDO LQWR )M FKLPHUDV 'DWD DUH SUHVHQWHG DV D SHUFHQWDJH RI WRWDO LVRW\SH $ 6HUXP ,J0 % VHUXP ,J*D b,J*D

PAGE 54

$QWL+** WLWHU $%7&%6-/f§!)M %%•6-/)M a!) )LJXUH $OORW\SHVSHFLILF DQWL+** (/,6$ GDWD 'DWD DUH SUHVHQWHG DV D SHUFHQWDJH RI WRWDO LVRW\SH $ %7& %6-/ f§! )[ PLFH % %6-/ )Mf§!)PLFH b,J*D

PAGE 55

$1R$V XSRQ +J&O WUHDWPHQW 7KH IDFW WKDW VRPH RI WKH PLFH UHVSRQGHG PHDQW WKDW 7 FHOOV ZHUH DYDLODEOH LQ ZKLFK WROHUDQFH WR ILEULOODULQ ZDV EURNHQ $OORW\SLF DQDO\VLV UHYHDOHG WKDW DOO RI WKH DXWRDQWLERGLHV ZHUH RI WKH E DOORW\SH ,W VHHPV OLNHO\ WKDW WKH\ DURVH IURP WKH %6-/ GRQRU VLQFH WKH % FHOOV IURP WKH %6-/ [ %74)M PLFH ZRXOG KDYH SURGXFHG DXWRDQWLERGLHV RI HLWKHU D RU E DOORW\SH DQG QR D DOORW\SH DXWRDQWLERGLHV ZHUH GHWHFWHG HYHQ DW D GLOXWLRQ GDWD QRW VKRZQf 7KXV GHVSLWH WKH SUHVHQFH RI 7 FHOOV ZKLFK SURYLGHG KHOS LQ DQ ,$V UHVWULFWHG IDVKLRQ WKRVH % FHOOV ZKLFK FRH[SUHVVHG ERWK KDSORW\SHV GLG QRW SURGXFH DXWRDQWLERGLHV % &HOO &RPSRVLWLRQ :DV 1RW D 'HWHUPLQLQJ )DFWRU LQ 5HVSRQVLYHQHVV )LJXUH VKRZV WKH % FHOO PDNHXS RI WKH %6-/ %6-/ [ %7&f)M f§! %6-/ [ %74)M PLFH )ORZ F\WRPHWULF DQDO\VLV RI FODVV ,, H[SUHVVLRQ VKRZHG WKHVH PLFH WR EH YHU\ ZHOO EDODQFHG b,$VE UDQJH bf 7KH SUHGRPLQDQFH RI ,J0 RI WKH E DOORW\SH bf LV H[SHFWHG FRQVLGHULQJ WKDW DSSUR[LPDWHO\ RQHKDOI RI WKH % FHOOV IURP WKH %6-/ [ %74)L GRQRU ZRXOG H[SUHVV ,J0 RI WKH E DOORW\SH 6LPLODU WR WKH SDUHQWDO LQWR %6-/ [ %7&f) PLFH QR FRUUHODWLRQ ZDV QRWHG EHWZHHQ WKH % FHOO PDNHXS DQG UHVSRQVLYHQHVV )LJXUH f %RWK %RQRUV 3URYLGHG )XQFWLRQDO % &HOOV 7KH DSSDUHQW ODFN RI UHVSRQVH IURP WKH D DOORW\SH % FHOOV OHG XV WR H[DPLQH ZKHWKHU WKHVH % FHOOV ZHUH IXQFWLRQDOO\ HTXLYDOHQW )LJXUH VKRZV WKDW ,J0 DQG ,J*D RI ERWK DOORW\SHV ZHUH SURGXFHG DOWKRXJK WKH FRPSRVLWLRQ LV VRPHZKDW VNHZHG WR WKH E DOORW\SH $JDLQ WKLV LV DQ H[SHFWHG ILQGLQJ VLQFH WKH FHOOV IURP WKH %6-/ [ %7&f)M

PAGE 56

b,$n %6-/ )M f§!)M $ &ODVV ,, ([SUHVVLRQ % 6XUIDFH ,J0 $1R$ )LJXUH ,PPXQRIOXRUHVFHQFH RQ SHULSKHUDO EORRG O\PSKRF\WHV RI%6-/ ) f§! )M FKLPHUDV 'DWD DUH SUHVHQWHG DV SHUFHQW RI FHOOV H[SUHVVLQJ FODVV ,, $ &ODVV ,, H[SUHVVLRQ % VXUIDFH ,J0 b,J0

PAGE 57

PLFH ZRXOG SURGXFH HLWKHU D RU E DOORW\SH 6LPLODUO\ % FHOOV RI WKH D DOORW\SH UHVSRQGHG HTXDOO\ ZHOO ZKHQ WKH UHVSRQVH WR D 7 FHOO GHSHQGHQW DQWLJHQ +**f ZDV DQDO\]HG LQ DQ DOORW\SLF IDVKLRQ )LJXUH %f $ FRPSDULVRQ EHWZHHQ SRVLWLYH DQG QHJDWLYH PLFH UHYHDOHG QR VLJQLILFDQW GLIIHUHQFHV 7DNHQ WRJHWKHU ZH FRQFOXGH WKDW WKH LQDELOLW\ RI KDSORW\SHKHWHUR]\JRXV PLFH WR UHVSRQG WR +J&O ZLWK DQ $1R$ UHVSRQVH LV DQ LQWULQVLF SURSHUW\ RI WKH % FHOOV $EVHQFH RI $1R$ 3URGXFWLRQ LQ +DSORW\SH+HWHUR]\JRXV 0LFH :DV 1RW 'XH WR WKH 3UHVHQFH RI ,$E5HVWULFWHG 7 &HOOV ,W ZDV SRVVLEOH WKDW WKH QRQUHVSRQVLYHQHVV RI WKH PL[HG FKLPHULF DV ZHOO DV WKH V\QJHQHLFDOO\ UHFRQVWLWXWHG %6-/ [ %7&f) PLFH ZDV GXH WR D QHJDWLYH UHJXODWRU\ HIIHFW RI ,$E UHVWULFWHG 7 FHOOV 7R WHVW IRU WKLV SRVVLELOLW\ ZH OHWKDOO\ LUUDGLDWHG %6-/ PLFH DQG WUDQVIHUUHG 7FHOOGHSOHWHG ERQH PDUURZ IURP %6-/ [ %74)M PLFH 7R HQVXUH QR FDUU\ RYHU RI 7 FHOOV HGXFDWHG LQ DQ %6-/ [ %7&f)M PRXVH ZH WUHDWHG WKH UHFLSLHQWV DW WKH WLPH RI WUDQVIHU ZLWK PJ LS RI 0P7O PRXVH ,J*D DQWL&' ff 7KH 7 FHOOV LQ WKHVH PLFH DUH XQDEOH WR SURYLGH KHOS WKURXJK ,$E VLQFH WKH\ GHYHORSHG LQ DQ ,$ H[SUHVVLQJ KRVW )ORZ F\WRPHWULF DQDO\VLV FRQILUPHG WKDW DOO WKH % FHOOV LQ WKHVH PLFH H[SUHVVHG ERWK KDSORW\SHV ,QWHUHVWLQJO\ QRQH f RI WKH PLFH SURGXFHG $1R$V IROORZLQJ +J&O WUHDWPHQW 7KH UHVLVWDQFH RI WKHVH PLFH FRXOG QRW KDYH EHHQ GXH WR QHJDWLYH LQIOXHQFHV E\ ,$E UHVWULFWHG 7 FHOOV

PAGE 58

%6-/ )M f§! $ 6HPPOJ0 % 6HUXP ,J*D )LJXUH $OORW\SHVSHFLILF WRWDO ,J0 DQG ,J*D (/,6$ GDWD RI %6-/ )M f§! ) W FKLPHUDV 'DWD DUH SUHVHQWHG DV D SHUFHQWDJH RI WRWDO LVRW\SH $ 6HUXP ,J0 % VHUXP ,J*D b,J*D

PAGE 59

'LVFXVVLRQ +J&O WUHDWPHQW LQ PLFH LQGXFHV D ZLGH UDQJH RI SK\VLRORJLFDO UHVSRQVHV LQFOXGLQJ PDUNHG LQFUHDVHV LQ -/ SURGXFWLRQ f 7KH UHTXLUHPHQW IRU 7 FHOOV LQ WKLV PRGHO ZDV VKRZQ E\ WKH IDFW WKDW QXGH PLFH RQ WKH +6 EDFNJURXQG IDLO WR GHYHORS DXWRDQWLERGLHV f 6SOHQLF &' 7 FHOOV RI +J&OWUHDWHG +6 PLFH KDYH EHHQ VKRZQ WR KDYH D VWURQJ LQFUHDVH LQ ,/ P51$ ZKHUHDV WKRVH RI +G PLFH VKRZHG RQO\ D ZHDN LQFUHDVH f ,QWHUHVWLQJO\ WUHDWPHQW RI +6 PLFH ZLWK DQWL,/ PRQRFORQDO DQWLERG\ GLG QRW SUHYHQW $1R$ SURGXFWLRQ DOWKRXJK LW FKDQJHG WKH LVRW\SH SURILOH RI WKH DXWRDQWLERGLHV f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f %RQHPDUURZ FKLPHUDV ZHUH DOVR XVHG WR VKRZ WKDW DXWRDQWLERG\ SURGXFWLRQ LQ ,SU PLFH ZDV UHVWULFWHG WR WKRVH % FHOOV WKDW UHFHLYHG 7 FHOO KHOS f 7KXV LQ WKUHH YHU\ GLIIHUHQW PXULQH PRGHOV RI DXWRLPPXQLW\ LQ YLYR DXWRDQWLERG\ UHVSRQVHV ZHUH 0+&UHVWULFWHG :KHQ WKH GLIIHUHQFH LQ UHVSRQVLYHQHVV EHWZHHQ %6-/ f§! %7& DQG %6-/ f§! %6-/ PLFH ZDV H[DPLQHG LW EHFDPH DSSDUHQW WKDW WKH SRVVLELOLW\ H[LVWHG IRU VLJQLILFDQW GLIIHUHQFHV LQ WKH 7 FHOO UHSHUWRLUH 8VLQJ ERQH PDUURZ WUDQVIHUV LQYROYLQJ

PAGE 60

FODVV ,, GHILFLHQW PLFH *OLPFKHU HW DO f VKRZHG WKDW LW LV WKH UDGLRUHVLVWDQW KRVW WK\PLF HSLWKHOLDO FHOOV ZKLFK PHGLDWH SRVLWLYH VHOHFWLRQ ,Q RXU PRGHO LW ZDV SRVVLEOH WKDW ODFN RI H[SUHVVLRQ RI ,$V RQ WKH UDGLRUHVLVWDQW WK\PLF SRSXODWLRQ HOLPLQDWHG UHVSRQVLYH 7 FHOOV 7KH XVH RI %6-/ [ %7&f); PLFH DV KRVWV DOORZHG IRU SRVLWLYH VHOHFWLRQ RQ ERWK ,$V DQG ,$E DQG OHG XV WR FRQFOXGH WKDW UHVSRQVLYHQHVV LV GHSHQGHQW RQ WKH GRQRU KDSORW\SH ZKHQ 7 FHOOV IRU ERWK KDSORW\SHV DUH SRVLWLYHO\ VHOHFWHG 0L[HG FKLPHULVP KDV EHHQ VKRZQ WR SUHYHQW DXWRLPPXQLW\ LQ VHYHUDO V\VWHPV 1RQREHVH GLDEHWLF PLFH ZHUH SURWHFWHG IURP GLDEHWHV ZKHQ WKH\ ZHUH PDGH FKLPHULF ZLWK GLDEHWHV UHVLVWDQW ERQH PDUURZ f ,W DSSHDUHG WKDW WKH DXWRLPPXQH SRWHQWLDO RI WKH 12' FHOOV ZDV UHVWUDLQHG +RZHYHU ZKHQ WKH DPRXQW RI UHVLVWDQW FHOOV ZDV GHFUHDVHG D ORZ LQFLGHQFH RI LQVXOLQLWLV ZDV VHHQ 8VLQJ WKH UDW PRGHO RI PHUFXU\LQGXFHG DXWRLPPXQLW\ 'HODQH\ HW DO f VKRZHG WKDW WKH SUHVHQFH RI UHVLVWDQW ERQH PDUURZ FHOOV PLFURFKLPHULVPf FRQYHUWHG DQ RWKHUZLVH VHQVLWLYH UDW VWUDLQ WR D UHVLVWDQW RQH ,I VLPLODU WR WKH UDW PRGHO LW ZRXOG EH H[SHFWHG WKDW WKH SUHVHQFH RI LPPXQH FRPSHWHQW FHOOV RI UHVLVWDQW RULJLQ ZRXOG SUHYHQW %6-/GHULYHG % FHOOV IURP UHVSRQGLQJ WR +J&O ZLWK DQ DQWLILEULOODULQ UHVSRQVH 2XU H[SHULPHQWV GLIIHUHG VRPHZKDW LQ WKDW ZH XVHG D UHVLVWDQW %6-/ [ %7&f); PRXVH ZKLFK KDG EHHQ FRLQIXVHG ZLWK D FRPELQDWLRQ RI UHVLVWDQW DQG VHQVLWLYH ERQH PDUURZ +RZHYHU %6-/ DORQH DV WKH ERQH PDUURZ GRQRU FRPSOHWHO\ UHVWRUHG WKH DELOLW\ WR SURGXFH $1R$V XSRQ WUHDWPHQW ZLWK +J&O 'HVSLWH FRQWDLQLQJ D VLJQLILFDQW DPRXQW RI UHVLVWDQW FHOOV ZH VWLOO REWDLQHG DQ LQWHUPHGLDWH UHVSRQVH 7KH SUHVHQFH RI DXWRDQWLERGLHV FRXOG SOD\ D UROH LQ VSUHDGLQJ DXWRLPPXQLW\ IURP D GRPLQDQW HSLWRSH WR SUHYLRXVO\ FU\SWLF HSLWRSHV f 3UHVXPDEO\ ELQGLQJ RI DQWLERG\ FDQ DOWHU DQWLJHQ SURFHVVLQJ UHYHDOLQJ QHZ HSLWRSHV 6HYHUDO RWKHU LQYHVWLJDWRUV KDYH

PAGE 61

VKRZQ WKDW LV LW SRVVLEOH WR EUHDN 7 FHOO WROHUDQFH WR VHOI DQWLJHQV E\ FRLPPXQL]LQJ PLFH ZLWK VHOI DQG IRUHLJQ DQWLJHQV ZKLFK LQ WXUQ JHQHUDWHV FURVVUHDFWLYH % FHOOV WKDW FDQ HOLFLW DQ DXWRLPPXQH 7 FHOO UHVSRQVH WR SUHYLRXVO\ FU\SWLF VHOI GHWHUPLQDQWV RQ WKH DXWRDQWLJHQ f 7KHVH IRUPV RI HSLWRSH VSUHDGLQJ GR QRW DSSHDU WR EH RFFXUULQJ LQ RXU PRGHO 7KH SURGXFWLRQ RI DXWRDQWLERGLHV E\ WKH %6-/ % FHOOV LQ WKH %6-/ %6-/ [ %74) f§! %6-/ [ %74)M DQG WKH %6-/ %7& f§! %6-/ [ %74)M PLFH GLG QRW UHVXOW LQ WKH ORVV RI WROHUDQFH WR ILEULOODULQ E\ SUHYLRXVO\ UHVLVWDQW % FHOOV 6HYHUDO RWKHU GLIIHUHQFHV H[LVWV EHWZHHQ RXU PRGHO DQG WKDW XVHG E\ 'HODQH\ f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f 5HJXODWRU\ 7 FHOOV KDYH QRW EHHQ LGHQWLILHG LQ WKH PRXVH PRGHO DQG WKH DQLPDOV GR QRW EHFRPH UHVLVWDQW WR IXUWKHU WUHDWPHQW 7KH LQWHUPHGLDWH UHVSRQVH VHHQ LQ JURXS 7DEOH f PLFH ZDV D VXUSULVLQJ ILQGLQJ ,W ZDV SRVVLEOH WKDW WKH $1R$ QHJDWLYH PLFH IDLOHG WR UHVSRQG EHFDXVH WKH QXPEHU RI SRWHQWLDOO\ UHVSRQVLYH % FHOOV RI ,$V RULJLQ KDG EHHQ UHGXFHG E\ WKH SUHVHQFH RI UHVLVWDQW % FHOOV RI ,$E RULJLQ 7KLV PHFKDQLVP DSSHDUV XQOLNHO\ LQ DV PXFK DV WKH %

PAGE 62

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f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

PAGE 63

E DOORW\SH LQGLFDWLQJ WKDW WKH KDSORW\SHKHWHUR]\JRXV % FHOOV IDLOHG WR SDUWLFLSDWH LQ WKLV UHVSRQVH 7KHVH UHVXOWV VWURQJO\ VXJJHVW WKDW ODFN RI UHVSRQVLYHQHVV LV DQ LQWULQVLF UHJXODWRU\ SURSHUW\ RI % FHOOV 7KH SUHVHQFH RI ,$E UHVWULFWHG 7 FHOO KHOS FRXOG KDYH DFFRXQWHG IRU WKH LQWHUPHGLDWH UHVSRQVH VHHQ LQ WKH PL[HG FKLPHUDV 7KH %6-/ [ %74)M PLFH WKDW VHUYHG DV KRVWV SURYLGHG DQ HQYLURQPHQW LQ ZKLFK 7 FHOOV ZRXOG EH VHOHFWHG WR LQWHUDFW ZLWK ,$E H[SUHVVHG RQ WKH % FHOOV IURP WKH %7& KRVWV %\ WUDQVIHUULQJ %6-/ [ %7&f)M ERQH PDUURZ LQWR %6-/ PLFH WKHUH ZRXOG EH OLWWOH SRWHQWLDO IRU SRVLWLYH VHOHFWLRQ RI 7 FHOOV RQ ,$E 'HVSLWH WKH DEVHQFH RI WKHVH FHOOV WKHUH ZDV QR $1R$ UHVSRQVH WR +J&O 7KLV DJDLQ SRLQWV WR DQ LQWULQVLF GHIHFW LQ %6-/ [ %7&f) % FHOOV 2QH SRVVLELOLW\ IRU WKLV LQWULQVLF SURSHUW\ LV WKH FRQFHSW RI 0+&JXLGHG SURFHVVLQJ OHDGLQJ WR GHWHUPLQDQW FDSWXUH f 7KLV K\SRWKHVLV VWDWHV WKDW ZKHQ DQ DQWLJHQ LV WDNHQ XS E\ $3&V DQG EHJLQV XQIROGLQJ GLIIHUHQW 0+& PROHFXOHV FDQ FRPSHWH IRU GHWHUPLQDQWV 2QFH LW LV ERXQG E\ FODVV ,, WKH DQWLJHQ LV WKHQ WULPPHG GRZQ WR LWV ILQDO VL]H ZKLOH WKH UHPDLQGHU RI WKH DQWLJHQ LQFOXGLQJ FU\SWLF HSLWRSHV LV GLVFDUGHG $Q H[DPSOH RI WKLV SKHQRPHQRQ ZDV VHHQ LQ WKH DXWRLPPXQH GLVHDVH LQVXOLQGHSHQGHQW GLDEHWHV PHOOLWXV f 7KH UHVSRQVH WR WKH VXEGRPLQDQW $12'UHVWULFWHG GHWHUPLQDQW RI +(/ GLVDSSHDUV ZKHQ 12' PLFH ZHUH PDGH WUDQVJHQLF E\ LQWURGXFWLRQ RI WKH (DG 7KH UHVSRQVLYQHVV ZDV UHVWRUHG ZKHQ VFLVVLRQ RI WKH +(/ VHSDUDWHG WKLV GHWHUPLQDQW IURP LWV DGMRLQLQJ FRPSHWLWLYHO\ GRPLQDQW (GUHVWULFWHG GHWHUPLQDQW 7KLV VXJJHVWHG WKDW WKH (G PROHFXOH ERXQG DQG SURWHFWHG LWV GRPLQDQW GHWHUPLQDQW RQ D ORQJ SHSWLGH ZKLOH FDSWXUHG QHLJKERULQJ GHWHUPLQDQWV ZHUH ORVW GXULQJ SURWHRO\VLV ,Q RXU PLFH ,$E FRXOG HIIHFWLYHO\

PAGE 64

EH ELQGLQJ ILEULOODULQ LQ WKH %6-/ [ %7&f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f 7KHUHIRUH LI RXU PRGHO LV FRUUHFW WKH SUHVHQFH RI ,$E DIIHFWHG WKH SURFHVVLQJ RI +J PRGLILHG ILEULOODULQ LQ WKH FRQWH[W RI ,$ DQG SURYLGH D XQLTXH RSSRUWXQLW\ IRU WHVWLQJ WKH UROH RI DQWLJHQ FRPSHWLWLRQ LQ DQ LPSRUWDQW HQYLURQPHQWDO PRGHO RI LQGXFHG DXWRLPPXQLW\

PAGE 65

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f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f

PAGE 66

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f ,QRUJDQLF PHUFXU\ VDOWV PD\ EH SUHVHQW DV GLYDOHQW PHUFXULFf RU PRQRYDOHQW PHUFXURXVf ,QJHVWLRQ RI KLJK GRVHV RI WKLV IRUP RI PHUFXU\ FDXVHV FRUURVLYH XOFHUDWLRQ DQG QHFURVLV RI WKH JDVWURLQWHVWLQDO WUDFW IROORZHG E\ FLUFXODWRU\ FROODSVH ,I WKLV LQLWLDO LQVXOW LV VXUYLYHG UHQDO IDLOXUH RFFXUV ZLWKLQ KRXUV GXH WR QHFURVLV RI WKH SUR[LPDO WXEXODU HSLWKHOLXP &KURQLF ORZGRVH H[SRVXUH WR LQRUJDQLF PHUFXU\ FDQ FDXVH LPPXQRORJLFDO DEHUUDWLRQV LQ ERWK KXPDQV DQG UDEELWV f ,PPXQRVXSSUHVVLYH HIIHFWV KDYH EHHQ UHSRUWHG LQ PLFH UHVXOWLQJ LQ LPSDLUHG KRVW UHVLVWDQFH f 7KH LPPXQRDFWLYDWLQJ SURSHUWLHV RI LQRUJDQLF PHUFXU\ LQ PLFH KDYH EHHQ ZLGHO\ VWXGLHG DQG FDQ EH GLYLGHG LQWR WKUHH PDMRU SDWKRORJLFDO VHTXHODH O\PSKRSUROLIHUDWLRQ K\SHUJDPPDJOREXOLQHPLD DQG WKH GHYHORSPHQW RI DXWRLPPXQLW\ f /\PSKRSUROLIHUDWLRQ LV FODVVLILHG DV DQ 0+& LQGHSHQGHQW HIIHFW ZKLOH DXWRLPPXQLW\ GHYHORSPHQW LV WLJKWO\ UHVWULFWHG E\ KDSORW\SH

PAGE 67

&XUUHQWO\ WKH OLQN EHWZHHQ 0+& KDSORW\SH DQG VXVFHSWLELOLW\ WR +J&O LQGXFHG $1R$V KDV UHPDLQHG XQH[SODLQHG 7KH H[SHULPHQWV FRQGXFWHG IRU WKLV GLVVHUWDWLRQ ZHUH EHJXQ ZLWK WKH LGHD RI H[DPLQLQJ MXVW KRZ WKH 0+& KDSORW\SH GHWHUPLQHV WKH UHVSRQVH WR +J&O 7KURXJK WKH XVH RI +UHFRPELQDQWV LW KDV EHHQ VKRZQ WKDW UHVSRQVLYHQHVV FRXOG EH PDSSHG WR WKH ,$ UHJLRQ RI WKH 0+& f 0LFH RI WKH +6 KDSORW\SH DUH KLJK UHVSRQGHUV ZKLOH WKRVH RI WKH +E KDSORW\SH DUH UHVLVWDQW f ,Q DQ )M FURVV EHWZHHQ WZR GLIIHUHQW KDSORW\SHV WKH SURGXFWV RI ERWK DOOHOHV DUH H[SUHVVHG RQ WKH VDPH FHOO VR H[SUHVVLRQ LV VDLG WR EH FRGRPLQDQW 7KHUHIRUH DQ )M VEf PRXVH VKRXOG EH VXVFHSWLEOH WR $1R$ LQGXFWLRQ 7KLV UHVXOW KDV EHHQ UHSRUWHG LQ WKH OLWHUDWXUH DOWKRXJK WKH UHVSRQVH ZDV VRPHZKDW DWWHQXDWHG f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f RU KDYH QR HIIHFW f RQ VXVFHSWLELOLW\ 7KH SURWHFWLRQ DIIRUGHG E\ ,( LV QRW ZLWKRXW SUHFHGHQW ([SUHVVLRQ RI ,( LQ OXSXV SURQH %SU PLFH VLJQLILFDQWO\ UHGXFHG WKH SURGXFWLRQ RI DXWRDQWLERGLHV DV ZHOO DV ORZHULQJ WKH VSOHHQ DQG

PAGE 68

O\PSK QRGH ZHLJKWV f 8VLQJ ERQH PDUURZ FKLPHUDV LQ ZKLFK ,(SRVLWLYH DQG ,( QHJDWLYH % FHOOV FRH[LVWHG LW ZDV GHWHUPLQHG WKDW WKH DPHOLRUDWLRQ RI VLJQV UHVXOWHG IURP D GLUHFW HIIHFW RI ,( RQ WKH % FHOO 7KLV SKHQRPHQRQ PD\ EH D XQLYHUVDO IHDWXUH RI DXWRLPPXQH GLVHDVH RQ WKH +% EDFNJURXQG ,Q DQ XQUHODWHG DXWRLPPXQH PRXVH PRGHO %;6% ZKLFK DOVR EHDUV WKH +E KDSORW\SH H[SUHVVLRQ RI ,( SUHYHQWHG WKH GHYHORSPHQW RI 6/( f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

PAGE 69

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f)M % FHOOV QR LQFUHDVH LQ VXVFHSWLELOLW\ ZDV REVHUYHG 7KH IDFW WKDW WKHVH PLFH SURGXFHG $1R$ RQO\ RI WKH E DOORW\SH FRPELQHG ZLWK

PAGE 70

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f 7KHVH PRGLILHG VHOI DQWLJHQV PD\ WKHQ EH UHFRJQL]HG E\ WKH LPPXQH V\VWHP DV IRUHLJQ DQG LQ WXUQ LQGXFH DQ LPPXQH UHVSRQVH FDSDEOH RI UHFRJQL]LQJ WKH PRGLILHG VHOI DQWLJHQ 'XULQJ 7 FHOO GHYHORSPHQW VHOI WROHUDQFH LV RQO\ LQGXFHG WR HIILFLHQWO\SUHVHQWHG GRPLQDQW HSLWRSHV EXW QRW WR FU\SWLF RQHV f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f VKRZHG WKDW GXULQJ DSRSWRVLV WKH OXSXV DXWRDQWLJHQV FOXVWHU DQG DUH FRQFHQWUDWHG LQ WKH VXUIDFH EOHEV RI DSRSWRWLF FHOOV 7KLV FRXOG UHVXOW LQ WKH DXWRDQWLJHQV EHFRPLQJ VXEVWUDWHV IRU HQ]\PHV QRW SUHVHQW ZLWKLQ WKH FHOO UHVXOWLQJ LQ QHZ

PAGE 71

FOHDYDJH SURGXFWV $ IROORZXS VWXG\ E\ WKH VDPH JURXS f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f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f LV SUHVHQWHG 7 FHOOV ZKLFK UHFRJQL]H WKLV SHSWLGH DUH GHOHWHG UHVXOWLQJ LQ

PAGE 72

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f $QRWKHU SRVVLELOLW\ LV WKDW WKH GLVXOILGH UHVLGXH FRXOG EH UHn UHGXFHG GXULQJ SURFHVVLQJ EXW RQH RI WKH DPLQR DFLGV LQ WKH ELQGLQJ JURRYH KDV +JOLQNHG VLGH FKDLQ WKDW DOWHUV WKH LPPXQRJHQHLFLW\ 7KH QHW UHVXOW WKHQ ZRXOG EH WKH H[SUHVVLRQ RI D VHOIHSLWRSH QRW SUHYLRXVO\ HQFRXQWHUHG 7KH LPPXQH V\VWHP ZRXOG UHFRJQL]H WKLV DV IRUHLJQ UHVXOWLQJ LQ DXWRLPPXQLW\ ,Q WKH FDVH RI D E KDSORW\SH PRXVH WKH GLVXOILGH ERQG GRHVQf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

PAGE 73

FODVV ,, DOOHOHV ZLOO UHVXOW ,I WKH ,$E PROHFXOH KDV D JUHDWHU DYLGLW\ IRU WKLV SHSWLGH WKDQ $V QR QHZ SHSWLGH LV VHHQ LQ WKH FRQWH[W RI ,$V DQG QR DXWRLPPXQLW\ UHVXOWV 7KHUH LV D SUHFHGHQW IRU WKH LGHD WKDW GLIIHUHQW FODVV ,, PROHFXOHV DFWXDOO\ FRPSHWH IRU WKH VDPH RU RYHUODSSLQJ SHSWLGH 6HUFDU] HW DO f VKRZHG WKDW FRPSHWLWLRQ WDNHV SODFH EHWZHHQ ,$ DQG ,( LQ WKH PRXVH 7KH\ FRLQHG WKH SKUDVH f0+& JXLGHG SURFHVVLQJf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

PAGE 74

FHOO HSLWRSHV $V SUHYLRXVO\ GLVFXVVHG WKH 7 FHOO HSLWRSH KDV EHHQ SURSRVHG WR EH PHUFXU\PRGLILHG ILEULOODULQ ZKLOH WKH % FHOO HSLWRSH LV D FRQIRUPDWLRQDO GHWHUPLQDQW ORVW E\ H[SRVXUH WR PHUFXU\ f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f PLFH 7KURXJK WKH XVH RI ERQH PDUURZ FKLPHULF DQLPDOV

PAGE 75

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

PAGE 76

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

PAGE 77

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fPHGLDWHG DXWRLPPXQLW\ LQ EURZQ 1RUZD\ UDWV GLUHFWO\ WULJJHUV 7 FHOOV WR SURGXFH LQWHUOHXNLQ &OLQ ,QYHVW

PAGE 78

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f LPPXQRVXSSUHVVLRQ RI +G PLFH LV PHGLDWHG E\ LQWHUIHURQJDPPD ,)1JDPPDf &OLQ ([S ,PPXQRO 3ROODUG 0 ,QGXFWLRQ RI $XWRLPPXQLW\ E\ 0HUFXULF &KORULGH )LIWK ,QWHUQDWLRQDO :RUNVKRS RQ WKH 0ROHFXODU DQG &HOO %LRORJ\ RI $XWRDQWLERGLHV DQG $XWRLPPXQLW\ $EVWUDFWf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

PAGE 79

+XOWPDQ 3 6 (QHVWURP 0 3ROODUG DQG ( 0 7DQ $QWLILEULOODULQ DXWRDQWLERGLHV LQ PHUFXU\WUHDWHG PLFH &OLQ ([S ,PPXQRO 2NDQR < 9 6WHHQ DQG 7 $ 0HGVJHU $XWRDQWLERG\ WR 8 QXFOHRODU ULERQXFOHRSURWHLQ ILEULOODULQf LQ SDWLHQWV ZLWK V\VWHPLF VFOHURVLV $UWKULWLV 5KHXP .DVWXUL 1 $ +DWDNH\DPD + 6SLHUD DQG & $ %RQD $QWLILEULOODULQ DXWRDQWLERGLHV SUHVHQW LQ V\VWHPLF VFOHURVLV DQG RWKHU FRQQHFWLYH WLVVXH GLVHDVHV LQWHUDFW ZLWK VLPLODU HSLWRSHV ([S 0HG 7DNHXFKL 6 7XUOH\ ( 0 7DQ DQG 0 3ROODUG $QDO\VLV RI WKH DXWRDQWLERG\ UHVSRQVH WR ILEULOODULQ LQ KXPDQ GLVHDVH DQG PXULQH PRGHOV RI DXWRLPPXQLW\ ,PPXQRO .XELFND 0XUDQ\L 0 .UHPHU 1 5RWWPDQQ % /XEEHQ 5 $OEHUV 1 %ORNVPD 5 /XKUPDQQ DQG ( *OHLFKPDQQ 0XULQH V\VWHPLF DXWRLPPXQH GLVHDVH LQGXFHG E\ PHUFXULF FKORULGH 7 KHOSHU FHOOV UHDFWLQJ WR VHOI SURWHLQV ,QW $UFK $OOHUJ\ ,PPXQRO .XELFND 0XUDQ\L 0 3 *ULHP % /XEEHQ 1 5RWWPDQQ 5 /XKUPDQQ DQG ( *OHLFKPDQQ 0HUFXULFFKORULGHLQGXFHG DXWRLPPXQLW\ LQ PLFH LQYROYHV XSUHJXODWHG SUHVHQWDWLRQ E\ VSOHHQ FHOOV RI DOWHUHG DQG XQDOWHUHG QXFOHRODU VHOI DQWLJHQ ,QW $UFK $OOHUJ\ ,PPXQRO 3ROODUG 0 /HH & $ &DVLDQR 0 %OXWKQHU 0 0 -RKQVWRQ DQG ( 0 7DQ 7KH DXWRLPPXQLW\LQGXFLQJ [HQRELRWLF PHUFXU\ LQWHUDFWV ZLWK WKH DXWRDQWLJHQ ILEULOODULQ DQG PRGLILHV LWV PROHFXODU DQG DQWLJHQLF SURSHUWLHV ,PPXQRO .DVV 6 7\F $ 6WHLW] DQG % 6ROOQHU :HEE 7KH 8 VPDOO QXFOHRODU ULERQXFOHRSURWHLQ IXQFWLRQV LQ WKH ILUVW VWHS RI SUHULERVRPDO 51$ SURFHVVLQJ &HOO /XEEHQ % 1 5RWWPDQQ 0 .XELFND 0XUDQ\L ( *OHLFKPDQQ DQG 5 /XKUPDQQ 7KH VSHFLILFLW\ RI GLVHDVHDVVRFLDWHG DQWLILEULOODULQ DXWRDQWLERGLHV FRPSDUHG ZLWK WKDW RI +J&LQGXFHG DXWRDQWLERGLHV 0RO %LRO 5HS %HQFNR 9 9 :DJQHU 0 :DJQHURYD DQG 9 2QGUHMFDN ,PPXQRORJLFDO SURILOHV LQ ZRUNHUV RFFXSDWLRQDOO\ H[SRVHG WR LQRUJDQLF PHUFXU\ +\J (SLGHPLRO 0LFURELRO ,PPXQRO 7XEEV 5 5 1 *HSKDUGW 7 0F0DKRQ 0 & 3RKO 9LGW 6 $ %DUHQEHUJ DQG 5 9DOHQ]XHOD 0HPEUDQRXV JORPHUXORQHSKULWLV DVVRFLDWHG ZLWK LQGXVWULDO PHUFXU\ H[SRVXUH 6WXG\ RI SDWKRJHQHWLF PHFKDQLVPV $P &OLQ 3DWKRO

PAGE 80

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f DQWLJHQ H[SUHVVLRQ RQ % FHOOV DQ LQ YLYR DSSURDFK WR LQYHVWLJDWH OD JHQH IXQFWLRQ &HOO *UXVE\ 0 5 6 -RKQVRQ 9 ( 3DSDLRDQQRX DQG / + *OLPFKHU 'HSOHWLRQ RI &' 7 FHOOV LQ PDMRU KLVWRFRPSDWLELOLW\ FRPSOH[ FODVV ,,GHILFLHQW PLFH 6FLHQFH 0HULQR 5 0 ,ZDPRWR / )RVVDWL 3 0XQLHVD $UDNL 6 7DNDKDVKL +XDUWH
PAGE 81

*RGLQJ : &RQMXJDWLRQ RI DQWLERGLHV ZLWK IOXRURFKURPHV PRGLILFDWLRQV WR WKH VWDQGDUG PHWKRGV ,PPXQRO 0HWKRGV &HPDQ 6 5 $ 5XGHUVGRUI 0 3HWHUVHQ DQG 5 'H0DUV '0$ DQG '0% DUH WKH RQO\ JHQHV LQ WKH FODVV ,, UHJLRQ RI WKH KXPDQ 0+& QHHGHG IRU FODVV ,,DVVRFLDWHG DQWLJHQ SURFHVVLQJ ,PPXQRO 5LEHUG\ 0 DQG 3 &UHVVZHOO 7KH DQWLJHQSURFHVVLQJ PXWDQW 7 VXJJHVWV D UROH IRU 0+&OLQNHG JHQHV LQ FODVV ,, DQWLJHQ SUHVHQWDWLRQ >SXEOLVKHG HUUDWXP DSSHDUV LQ ,PPXQRO $XJ f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f) K\EULG PLFH (XU ,PPXQRO 6REHO ( 6 7 .DWDJLUL .DWDJLUL 6 & 0RUULV 3 / &RKHQ DQG 5 $ (LVHQEHUJ $Q LQWULQVLF % FHOO GHIHFW LV UHTXLUHG IRU WKH SURGXFWLRQ RI DXWRDQWLERGLHV LQ WKH OSU PRGHO RI PXULQH V\VWHPLF DXWRLPPXQLW\ ([S 0HG &HUHGLJ 5 : /RZHQWKDO 0 1DEKRO] DQG + 5 0DF'RQDOG ([SUHVVLRQ RI LQWHUOHXNLQ UHFHSWRUV DV D GLIIHUHQWLDWLRQ PDUNHU RQ LQWUDWK\PLF VWHP FHOOV 1DWXUH

PAGE 82

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

PAGE 83

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f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g 6HUFDU] ( ( 3 9 /HKPDQQ $ $PHWDQL %HQLFKRX $ 0LOOHU DQG 0RXGJLO 'RPLQDQFH DQG FU\SWLFLW\ RI 7 FHOO DQWLJHQLF GHWHUPLQDQWV $QQX 5HY ,PPXQRO 'HQJ + 5 $SSOH 0 &ODUH 6DO]OHU 6 7UHPEOHDX 0DWKLV / $GRULQL DQG ( 6HUFDU] 'HWHUPLQDQW FDSWXUH DV D SRVVLEOH PHFKDQLVP RI SURWHFWLRQ DIIRUGHG E\ PDMRU KLVWRFRPSDWLELOLW\ FRPSOH[ FODVV ,, PROHFXOHV LQ DXWRLPPXQH GLVHDVH ([S 0HG *R\HU 5 $ 7R[LF (IIHFWV RI 0HWDOV ,Q &DVDUHWW DQG 'RXOOnV 7R[LFRORJ\ )LIWK HG &' .ODDVVHQ HG 0F*UDZ+LOO 1HZ
PAGE 84

(OOHUPDQQ (ULNVHQ 6 0 0 &KULVWHQVHQ DQG 6 & 0RJHQVHQ (IIHFW RI PHUFXULF FKORULGH RQ PDFURSKDJHPHGLDWHG UHVLVWDQFH PHFKDQLVPV DJDLQVW LQIHFWLRQ ZLWK KHUSHV VLPSOH[ YLUXV W\SH 7R[LFRORJ\ 9ODGXWLX $ 2 7UHDWPHQW RI DXWRLPPXQH GLVHDVHV ZLWK DQWLERGLHV WR FODVV ,, PDMRU KLVWRFRPSDWLELOLW\ FRPSOH[ DQWLJHQV &OLQ ,PPXQRO ,PPXQRSDWKRO 8WVFKLJ / 0 :ULJKW DQG 7 9 2n+DOORUDQ %LRFKHPLFDO DQG VSHFWURVFRSLF SUREHV RI PHUFXU\,,f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

PAGE 85

%,2*5$3+,&$/ 6.(7&+ *UHJRU\ $ODQ +DQOH\ ZDV ERP LQ %DWDYLD 1HZ
PAGE 86

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fDYLV RI 9HWHULQDU\ 0HGLFLQH FHUWLI\ WKDW KDYH UHDG WKLV VWXG\ DQG WKDW LQ P\ RSLQLRQ LW FRQIRUPV WR DFFHSWDEOH VWDQGDUGV RI VFKRODUO\ SUHVHQWDWLRQ DQG LV IXOO\ DGHTXDWH LQ VFRSH DQG TXDOLW\ DV D GLVVHUWDWLRQ IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ -RHO 6FKLIIHQEDXHU V ‘ $VVRFLDWH 3URIHVVRU RI 0HGLFLQH 7KLV GLVVHUWDWLRQ ZDV VXEPLWWHG WR WKH *UDGXDWH )DFXOW\ RI WKH &ROOHJH RI 9HWHULQDU\ 0HGLFLQH DQG WR WKH *UDGXDWH 6FKRRO DQG ZDV DFFHSWHG DV SDUWLDO IXOILOOPHQW RI WKH UHTXLUHPHQWV IRU WKH GHJUHH RI 'RFWRU RI 3KLORVRSK\ 0D\ 'HDQ &ROOHJH RI 9HWHULQDU\ 0HGLFLQH 'HDQ *UDGXDWH 6FKRRO


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E9HFS5SK0_GPAGSS INGEST_TIME 2014-08-25T18:35:37Z PACKAGE AA00024957_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


MHC CONTROL OF MERCURY-INDUCED AUTOIMMUNITY IN MICE
By
GREGORY ALAN HANLEY
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1998

To Kim, my beloved spouse and greatest supporter. Without her encouragement
and understanding, I would have never completed this work.

ACKNOWLEDGMENTS
I would like to express my extreme gratitude to my mentor, Dr. Eric Sobel, for his
endless support. I thank him for the innumerable hours he devoted to critical discussions.
I also wish to thank the three other members of my graduate advisory committee. To Dr.
Schiffenbauer, I extend my sincere appreciation for his intellectual and monetary support.
I wish to thank Dr. Roberts who was instrumental in directing me down this path. To Dr.
Davis, I extend my sincere appreciation for his support in this endeavor, as well as, his
continued guidance in my career development.
Several people have provided me with tremendous technical support, and without
their help I would have been unable to complete this research. I owe a special thanks to
Ed Butfiloski, not only was he an invaluable resource, but he also was instrumental in
maintaining my sanity during this pursuit. I would also like to thank Raquel Baert for her
technical assistance and scholarly discussions.
Thus work was supported by NRSA Grant 5 T RR07001 and NIH K08 AR
01934.
m

TABLE OF CONTENTS
ACKNOWLEDGMENTS iii
ABSTRACT vi
CHAPTERS
1 BACKGROUND AND SIGNIFICANCE 1
Introduction I
Role of T Cells 4
Cytokines 5
B Cells / Antibodies / MHC 8
Fibrillarin 9
2 CLASS II HAPLOTYPE DIFFERENTIALLY REGULATES IMMUNE
RESPONSE IN HgCl2-TREATED MICE 12
Introduction 12
Materials and Methods 14
Mice 14
HgCl2 Treatment 15
Indirect Immunofluorescence 16
Antibodies 16
Immunizations 17
Anti-HEL ELISA 17
Flow Cytometry 18
Statistical Analysis 19
Results 19
Ft Mice Between a Resistant and Sensitive Haplotype Were Resistant to
ANoA Induction 19
The Ab Gene Product Conferred Resistance in Fi Mice 22
Expression of I-E Had No Effect on Susceptibility to HgCl2 25
Responsiveness to Exogenous Antigen Was Inherited Co-dominantly in Fj
Mice 27
Discussion 27
iv

3 RESISTANCE TO HgCl2-INDUCED AUTOIMMUNITY IN HAPLOTYPE
HETEROZYGOUS MICE IS AN INTRINSIC PROPERTY OF B CELLS ... 34
Introduction 34
Materials and Methods 36
Mice 36
In Vivo Treatments 37
Flow Cytometry 37
Indirect Immunofluorescence 38
Allotype-Specific ELISA 39
Results 39
MHC-Restricted T Cell Help Was Required to Produce ANoAs 39
Absence of ANoA Production in Haplotype-Heterozygous Mice Was Not
Due to a Difference in Thymic Education 41
Allogeneic Mixed Chimeras Developed ANoA Titers in Response to HgCl2
41
Lack of ANoA in Non-Responsive Mixed Chimeras Was Unrelated to B
Cell Composition 43
B Cells of Both B6.TC and B6.SJL Origin Were Functional in Group 3
Mixed Chimeras 44
Lack of HgCl2-Induced ANoA Response by Haplotype-Heterozygous Mice
Was Not Due to Absence of Anti-Fibrillarin-Specific I-As-
Restricted T Cell Help 48
B Cell Composition Was Not a Determining Factor in Responsiveness
48
Both Bonors Provided Functional B Cells 50
Absence of ANoA Production in Haplotype-Heterozygous Mice Was Not
Due to the Presence of I-Ab-Restricted T Cells 50
Discussion 52
4 DISCUSSION/CONCLUSION 58
LIST OF REFERENCES 69
BIOGRAPHICAL SKETCH 78
v

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
MHC CONTROL OF MERCURY-INDUCED AUTOIMMUNITY IN MICE
By
GREGORY ALAN HANLEY
May 1998
Chairperson: Stephen M. Roberts
Major Department: Veterinary Medicine
One of the most striking features of exposure to low doses of mercury in mice is
the high-titer haplotype-linked anti-nucleolar (ANoA) autoantibody response. Mice of the
H-2S haplotype have been high responders while H-2b mice have been low. This pattern
has been attributed to the class II molecule itself, but the poor response of Fj crosses
between high and low responders raised the possibility that the anti-fibrillarin specificity
was actually due to a closely linked dominant negative gene. To test the role of class II
explicitly, Fj crosses between congenie B6.SJL (H-2S) and C57BL/6 (H-2b) mice with a
targeted deletion of I-Apb were generated, creating mice heterozygous for all MHC loci,
but expressing only I-As. In comparison with B6 .SJL mice, no diminution of titers was
found, proving that I-As itself was responsible for susceptibility and I-Ab for
downregulation. Unlike I-A, expression of the I-E class II molecule could not
vi

downregulate the response in an otherwise susceptible mouse. Using adoptive transfer
techniques we have examined several mechanisms by which the resistant haplotype could
be downregulating the anti-fibrillarin response in F, (s/b) mice. Similar to other
autoimmune models, mercury induced autoimmunity requires cognate MHC-restricted T
cell help. The absence of autoantibody production in F, mice was not due to either a
difference in thymic education nor to the absence of anti-fibrillarin-specific T cell help.
These results suggest a complicated role for class II in the regulation of a novel,
environmentally-induced autoimmune response.
vu

CHAPTER 1
BACKGROUND AND SIGNIFICANCE
Introduction
Autoimmune disease occurs when the body mounts an immune response against
self. This immune response can involve either humoral and/or cellular immune pathways.
Currently, the cause of autoimmunity is not known but both environmental as well as
genetic factors appear to be important. The prevalence of autoimmune diseases has been
estimated to be approximately 3% (1).
Autoimmune diseases can be classified as systemic or organ-specific. Examples of
organ-specific autoimmune diseases include: type I diabetes, myasthenia gravis, Graves
disease and multiple sclerosis. The systemic autoimmune diseases are represented by
rheumatoid arthritis, scleroderma and systemic lupus erythematosus (SLE).
SLE is a multisystem disorder characterized by the presence of a variety of
autoantibodies that are responsible for many of its clinical manifestations. The prevalence
rate, based on the total population, is one in 2000, but one in 700 for women between the
ages of 20 and 64 years and one in 245 for black women in the same age group (2). The
etiology of SLE is unknown but many factors have been identified that influence SLE.
Genetic susceptibility is strongly suggested because of the high concordance rate between
monozygotic twins. Certain MHC haplotypes are more frequently associated with the
1

2
disease as well. The fact that women are much more susceptible to SLE points to
hormonal factors playing a role. Finally, several environmental factors have been
associated with SLE flare ups.
Experimental animal models have been widely used to study autoimmune diseases.
Some of the models develop spontaneous autoimmunity like Type I diabetes in NOD mice
and BB rats. Spontaneous disease similar to SLE develops in (NZB x NZW)F, mice.
Induced models of autoimmune disease require manipulation of the animal which in turn
causes an immune response to self antigens such as antigen-induced orchitis and
uveoretinitis.
A second type of induced autoimmune disease can be elicited in animals without
actually injecting either the autoantigen itself or a cross reactive antigen. For example,
exposure to Hg2+, an environmental toxin, in certain strains of rodents results in
autoimmune disease. Interestingly, humans also have been reported to develop
membranous glomerulonephritis upon exposure to mercury in an industrial setting (3).
Some of the characteristics of this disease are remarkably similar to human SLE.
HgCl2-induced autoimmunity was first reported in mice by Robinson et al. in 1984
(4). Within one week after the start of treatment with HgCl2, outbred Swiss ICR mice
produced antinuclear antibodies (ANA) of the IgG subclass. Further analysis of the ANAs
revealed a nucleolar pattern (antinucleolar antibodies; ANoA). Inbred A/J mice, on the
other hand, failed to respond to HgCl2.
Since this initial discovery, the link between susceptibility and genetics has been
investigated by several groups. Initially the link was established between ANoA induction

3
and the mouse major histocompatibility complex (MHC) (5-7). Testing of numerous
MHC-congenic strains led to the following ranking of susceptibility based on haplotype: d
haplotypes revealed that the I region of the MHC, specifically the I-A region, controlled
the response (5,6). The other I region locus, I-E, has been shown to either decrease the
response or have no effect on the ability of HgCl2 to induce ANoAs (6-8).
The response was shown to be co-dominantly inherited by intercrossing susceptible
and resistant strains of mice (9). For example, crosses between susceptible SJL and
resistant B6 mice resulted in an intermediate response with 41% of the progeny
developing ANoAs (9). Other gene(s) outside the MHC have also been shown to affect
susceptibility. The influence of these background genes was observed when comparing
mice of the H-2S haplotype on an A background to those on a BIO or a B6 background
(5.9). Non-MHC genes on the A background enhanced susceptibility while those present
on the B background decreased responsiveness. Within the B background, B6 were more
suppressive than BIO.
Not all of the effects seen with HgCl2 administration are as tightly controlled by
the MHC as the ANoA response. For example, A/J mice (H-2k) did develop
antichromatin (ACA) and antihistone (AHA) antibodies despite being resistant to ANoA
induction (7). The immune complex deposition seen in this model also does not appear to
be under the influence of the H-2 complex. Balb/c (H-2d) mice developed renal mesangial
and vessel wall immune complexes despite being completely resistant to ANoA induction
(9.10). When only ANoA susceptible mice (H-2s,f’q,p) were examined for renal IgG

4
deposition the results were similar in that only the H-2S mice showed significant immune
complex deposition (11,12).
Tissue concentrations of mercury chloride has been examined in several strains of
inbred mice (13). Most HgCl2 accumulates in the kidney primarily, with the liver being the
second largest pool followed by the gastrointestinal tract, skin, spleen and testicles. The
levels reached a steady state in the blood and liver by 4 weeks with the spleen and liver
reaching steady state at 8 weeks. Analysis of thymi revealed steadily increasing
concentrations throughout the 12 week treatment period. Interestingly, the authors also
noted that when comparing the H-2-congenic susceptible BIO.S (H-2S) and resistant
B10.D2 (H-2d) a correlation was detected between susceptibility to ANoA and
accumulation of Hg2+ within the spleen (13). Autometallography has revealed that within
the tissues of the immune system mercury chloride accumulated in macrophages but was
not found in non-phagocytic antigen presenting cells (follicular dendritic cells,
interdigitating cells and B cells). At the subcellular level mercury was localized in the
lysosomes of the macrophages (14).
Role of T Cells
The requirement for T cells in this model was shown by Hultman et al. using the
nude (athymic) mutation (15). SJL/N mice homozygous for the nude mutation (nu/nu)
failed to develop neither the ANoA nor the systemic immune complex deposits seen in
littermates heterozygous for the nude mutation (nu/+). Furthermore, the specific subset
of T cells required was elucidated by noting a lack of response in euthymic SJL/N mice
given anti-CD4 monoclonal antibody in addition to the mercury chloride treatment.

5
When in vivo T cell responses to HgCl2 were studied using flow cytometry in
susceptible A.SW mice, a long lasting increase in the number of T cells dominated by the
CD4+ subset was observed which was absent in resistant A.TH mice (16). In contrast, van
Vliet et al. (17) reported that both H-2S and H-2d mice expressed enhanced splenic
numbers of T cells expressing the activation marker CD45RBl0CD4+.
Similar effects were seen when the ability of mercury to stimulate cells in vitro was
studied. In vitro treatment of lymphocytes with HgCl2 resulted in a significant increase in
DNA synthesis in A.SW mice while only minimal increases were seen in DBA/2 mice (5).
Anti-CD4 antibody again completely abrogated the response suggesting the importance of
T helper cells in this model. Interestingly, mercury-exposed lymphocytes from the low
responder mice became high responders, as measured by [3H]Thymidine uptake, after the
removal of excess mercury by washing (18). This activation of T cells in vitro led Jiang et
al.( 19) to examine whether mercury acted as a T cell mitogen or as a superantigen. They
found a biased usage of TCR Vp subsets by CD4+ cells which they interpreted as a
superantigen effect.
Cytokines
Considering the effects mercury has on T cells, cytokines must play an important
role in HgCl2-induced autoimmunity. In vitro, the continuous presence of mercury
induced interleukin-2 (EL-2) and interferon-y (y-IFN) but not IL-4 production from both
high and low responder mice. In contrast, pretreating the cells with mercury and then
washing it away resulted in production of IL-4 from both groups (18).

6
Cytokine profiles of mice treated with mercury has given somewhat conflicting
results. An analysis of IL-4 mRNA from two H-2-congenic strains revealed a strong
increase in production in splenic CD4+ T cells of the H-2S mice while those of H-2d mice
showed only a weak increase (17). This result, along with similar results seen in the rat
model of HgCl2-induced autoimmunity (20,21), led Goldman et al. (22) to propose that
susceptibility relies on the initiation of a TH2 response while resistance correlates
predominantly with a TH1 response.
Further evidence of this dichotomy was provided by the fact that HgCl2-induced a
striking increase, up to 30-fold, of the level of IgE in susceptible A.SW but not in resistant
B6, DBA/2 or B10.D2 mice (17,23). The serum IgE levels are important because they, as
well as, the increases seen in class II expression are both highly dependent on the TH2
cytokine IL-4 (24). The connection between IgE and IL-4 was further strengthened in
this model by Ochel et al. (25) who showed that treating susceptible A.SW mice with anti-
IL-4 antibody completely abrogated the increase in total IgE. Surprisingly, this treatment
did not prevent ANoA production although it did influence the pattern of IgG subclass
distribution.
Using antibody response to sheep erythrocytes (sRBC) in mercury treated H-2-
congenic strains Doth et al. (26) also proposed a preferential activation of either TH1 or
TH2 cells based on haplotype. The response to sRBC antigens is normal in treated B10.S
mice while it is depressed in B10.D2 animals. They showed, using monoclonal antibodies
against y-IFN, that the suppression seen in B10.D2 mice was mediated by y-IFN, a TH1
cytokine. Furthermore, treatment of BIO. S mice with recombinant y-IFN suppressed their

7
response to the sRBC antigens, yet did not prevent the induction of ANoAs nor the
immune complex glomerulonephritis.
An argument against the TH1/TH2 dichotomy in this model was proposed by
Johansson et al. (16). They found that A.SW mice showed a modest increase of y-IFN
and IL-4 producing cells while H-2-congenic A.TH mice showed no increase in cytokine
producing cells. Remarkably, the susceptible SJL strain, despite being severely deficient in
TH2-promoting CD4+, NK1.U T cells, increased their number of y-IFN producing cells.
This indicated that a predominantly TH2 response is not necessary for the induction of
autoimmunity by mercury.
Recent studies conducted by Pollard et al. (27) provided the strongest evidence to
date against the idea that a TH1 response corresponds with resistance while a TH2
response is required for autoimmunity induction. Using H-2S mice deficient by gene
knockout for IL-4 and y-IFN, they showed that all of the features of autoimmunity are
controlled by y-IFN, a TH1 cytokine. TLA'1' mice did develop ANoAs while y-IFN"7'
were completely resistant. They concluded that the requirement for y-IFN suggests that
antigen dose is the limiting factor in autoimmunity induction.
The idea of a dose response relationship was examined by Hultman et al. (28).
They found a positive correlation between the dose of HgCl2, the total body burden of
mercury, and the degree of autoimmunity expressed as the serum ANoA titer. The
minimum observed adverse effect level (MOAEL) was 1.25 mg HgCl2 / liter drinking
water (1.25 ppm). At this dose approximately 50% of the mice responded. No ANoAs
were seen in mice given 0.625 ppm while 100% responded to 5 ppm. They estimated,

8
taking into account the absorption of inorganic mercury by the gastrointestinal tract, that
the MOAEL is 7-14 pg Hg/ kg bodyweight. This is much less than the standard dose of
1.6 mg/kg used in the literature.
B Cells / Antibodies / MHC
Mercury also has pronounced effects on B cells as evidenced by the increases seen
in immunoglobulin and class II surface expression. For example, SJL mice developed
splenic cell hyperplasia with a transiently increased number of cells secreting IgM, IgG and
IgGl immunoglobulins, as well as increased serum immunoglobulin concentrations.
Resistant B6 mice also showed similar increases although they were more short lived (29).
A significant increase in anti-TNP producing cells seen in SJL mice but not in B6 mice
provided evidence for polyclonal B cell activation in the former. H-2-congenic strains also
show the differential effects on B cells of mercury administration. B10.S mice showed
significantly increased numbers of Ig producing splenic B cells of the IgGl (30-fold),
IgG2a (7-fold) and IgE (5-fold) while B10.D2 mice in contrast did not show any
significant increases (17).
Mercury also enhances class II expression on B cells from both ANoA resistant
B10.D2 and ANoA susceptible BIO.S mice although significantly higher levels were
observed in the B10.S mice (17). Treatment with an anti-IL-4 antibody completely
prevented the increases seen in both strains. In addition to mercury influencing class II
expression, class II molecules are required for some of the effects induced by mercury. In
an in vitro system using monoclonal antibodies against class II, Hu et al. (30) were able to

9
completely abrogate the cytokine production and proliferation induced by mercury. The
requirement for class II molecules for the induction of ANoAs in vivo has not been
investigated. In chapter 2, we formally address this question using mice whose MHC I-A
gene has been deleted.
Fibrillarin
The nucleolar autoantigen against which HgCl2-induced autoantibodies are
directed has been identified as fibrillarin, a 34 kDa U3 ribonucleoprotein involved in the
first step of preribosomal RNA processing (31,32). It is interesting that up to 58% of
human scleroderma patients spontaneously produce autoantibodies against the same U3
ribonucleoprotein (33,34). Several groups have compared the reactivities toward
fibrillarin of mercury-induced ANoAs with human scleroderma sera and found that they
were indistinguishable (22,34). Both groups of autoantibodies recognize the full length
protein but lose reactivity, in all but the highest titrated human sera, if either the N- or C-
regions were cleaved (35).
In an attempt to determine what the T cell antigen is, bulk T cells from HgCl2-
treated B10.S mice were challenged with a variety of different self proteins that contained
minute amounts of mercury. The T cells reacted to all the proteins in an anamnestic
fashion (36). Bulk T cells obtained after one week of treatment also reacted
anamnestically to Hg-complexed fibrillarin, whereas after eight weeks a T cell response
against untreated fibrillarin predominated. These results suggested determinant spreading
of T cell specificity. Consistent with this, analysis of T cell hybridomas established from

10
mercury-treated H-2S mice revealed two types of CD4+ T cells: one that specifically
recognized Hg-complexed fibrillarin and another that reacted to untreated fibrillarin (37).
Interestingly, both the Hg-induced and native fibrillarin determinants were presented
without the addition of fibrillarin when spleen cells from animals treated with mercury
were used as antigen presenting cells. In contrast, spleen cells from untreated mice could
activate fibrillarin-specific T cells only if exogenous fibrillarin was added. This indicated
that HgCl2 can induce abnormal presentation by antigen presenting cells of a self protein.
In an attempt to understand the mechanism whereby mercury elicits an
autoantibody response that specifically targets fibrillarin, Pollard etal. (38) studied the
antigenicity and molecular properties of fibrillarin from cells undergoing mercury-induced
death. Under nonreducing SDS-PAGE conditions, fibrillarin from mercury treated nuclei
showed aberrant migration as evidenced by its change in migration from 34 kDa to 32
kDa. This modified fibrillarin also lost its B cell antigenicity as measured by both indirect
immunofluorescence and immunoprecipitation. If either one or both of the two cysteines
which fibrillarin contains were mutated to an alanine, the aberrant migration was
abolished. This combined with the fact that iodoacetamide, which alkylates cysteine
residues, also abolished the aberrant migration led the authors to propose that fibrillarin
forms a disulfide bond which then appears as the 32 kDa band. They concluded that
unmodified fibrillarin is the B cell antigen and mercury-modified fibrillarin is the T cell
antigen.
The mechanism whereby mercury can induce autoimmune disease is a complicated
one. To gain a better understanding of the genetics behind the response we first

11
investigated whether the class III-A molecule itself or a closely linked gene(s) was a
requisite for susceptibility. We also sought to determine what effects, if any, the other
class II molecule, I-E, had on susceptibility. In our initial studies we unexpectedly found
that susceptibility was not codominantly inherited as previously reported. This led us to
perform adoptive transfer experiments to elucidate the mechanisms whereby the resistant
haplotype could down-regulate the response. In addition, we used these chimeric mice to
investigate the cellular interactions between B and T cells in this model.

CHAPTER 2
CLASS II HAPLOTYPE DIFFERENTIALLY REGULATES IMMUNE RESPONSE
IN HgCl2-TREATED MICE
Introduction
Subtoxic doses of HgCl2 in animals induce an autoimmune disease characterized by
autoantibody production and variable amounts of immune complex glomerulopathy (22).
In mice, for example, the glomerulopathy is relatively mild and characterized by mesangial
deposits of IgG and C3 (10). Other features, such as splenic hyperplasia with increased
serum immunoglobulins (especially IgG and IgE) have also been described (23,29).
However, the most striking finding in mice is the antinucleolar antibodies (ANoA) elicited
in susceptible strains as soon as one week after the start of treatment (4). The principal
nucleolar autoantigen has been identified as fibrillarin, a U3 ribonucleoprotein involved in
the first step of preribosomsal RNA processing (31,39). Interestingly, this same specificity
is seen in patients suffering from systemic sclerosis and other connective tissue disorders
(33,34). Even more remarkably, a comparison of the specificity of antifibrillarin
autoantibodies between human and murine origin showed similar, if not identical
conformational epitopes (40).
The mechanism by which HgCl2 causes autoimmunity is unknown, but the fact that
a human scleroderma-associated autoantibody specificity can be produced and that certain
12

13
toxic exposures have resulted in scleroderma-like illnesses (41,42) makes it an important
and intriguing model. Both MHC class II-linked loci and unknown non-MHC loci govern
susceptibility, with mice of the H-2S haplotype being high responders, while those of the
H-2b haplotype are resistant or low responders (5). Non-H-2 genes in H-2S mice did not
prevent the disease but had a pronounced effect on the antinucleolar antibody (ANoA)
titers (12). By using intra-H-2 recombinants, it has been shown that responsiveness could
be mapped to the I-A region of the MHC. The other class II locus, I-E, was shown to
either suppress (6) or have no effect on the immune response (7). The class II-linked
susceptibility gene has previously been reported to be dominant, although the disease was
attenuated in F! animals when compared to homozygous animals (5,9,43). The fact that
most of these studies used non-MHC-congenic strains, however, makes it difficult to
interpret the effects of non-MHC genes and I-E expression on the ANoA response.
To investigate the control of the ANoA response by the I-A region to HgCl2 in the
absence of non-MHC loci differences we studied two MHC-congenic strains of mice on
the B6 background. In addition to the susceptible H-2S haplotype, we chose to study H-2b
rather than the more commonly used H-2d haplotype because of the availability of animals
on the B6 background which do not express class II due to a targeted disruption of the I-
Apb chain. In addition, the H-2S and H-2b haplotypes have both lost functional expression
of I-E, allowing the effects of I-A and I-E to be isolated. In our initial studies, we
unexpectedly found that F, progeny between B6.SJL (H-2S) and B6 (H-2b) mice were
highly resistant to HgCl2 treatment and suggested that this might have been due to a
dominant negative gene linked to X-Ab. Such a possibility is not without precedent.

14
Recently, the clinical manifestation of arthritis in the autoimmune model of collagen-
induced arthritis was shown not to be exclusively limited to a particular MHC class II
antigen but may be due to linked genes which participate in intracellular loading and
selection of antigenic peptides (44). To distinguish between these two possibilities we
generated mice on the B6 background which expressed I-A8 and carried the genes linked
to the I-Ab region but did not express I-Ab. These mice were as susceptible to the
induction of ANoA as H-2S homozygous mice, despite carrying the genes linked to I-Ab.
We also examined the effects of the other class II molecule, I-E, in our system by
developing B6.SJL mice which carried the I-Ead transgene (45), allowing the functional
expression of I-E on the H-28 background. Contrary to a previously published report and
to our experience with I-A, the expression of I-E did not alter the ANoA response. Thus,
our experiments demonstrate that it is the expression of the resistant H-2 allele (I-Ab) itself
that down-regulated the ANoA response of Fj crosses between resistant and sensitive
haplotypes and strongly suggest that it is the H-2S molecule itself that is specifically
required. These results have important implications for understanding the mechanism by
which HgCl2 exposure causes a specific autoimmune syndrome.
Materials and Methods
Mice
C57BL/6J (B6; I-Ab, I-E', Ighb), C57BL/6J.SJL-//-2f (B6.SJL; I-A8, I-E', Ighb),
C57BL/6J-/g/?a ThyIa Gpia (B6.TC; I-Ab, I-E', Igha), A.BY/SnJ, A.SW/SnJ, SJL/J and
Balb/cJ mice were obtained directly or from breeders purchased from The Jackson

15
Laboratory (Bar Harbor, ME). Tg(H-2I:Ea)Bri39 (B6.I-Ead) mice were a gift from R. L.
Brinster (45). CD2â„¢ (MHC Class II deficient, B6 background) were purchased from
Taconic (Germantown, NY). These mice lack any cell surface expression of I-A due to
the insertion of a loss of function mutation into the I-Apb gene (46). B6.SJL.I-Ead mice
were developed by crossing B6.SJL with B6.I-Ead and then backcrossing to B6.SJL. The
B6.SJL.I-Ead mice were then selected using FACS analysis. These mice expressed I-E
molecules consisting of a P-chain of the b allele and an a-chain of the d allele (47,48).
B6.I-As/b and Bó.I-A*7' mice were developed by intercrossing CD2™( B6.I-A' -) with
B6 and the resultant F, bred to B6.SJL. The MHC class II expression was verified using
FACS analysis. To rule out an allotype effect on susceptibility B6.SJL.Igha mice were
developed by intercrossing B6.SJL and(B6.SJL x B6.TC)F1. FACS analysis was used to
select progeny which were I-As and Igha/b. These mice were then intercrossed and the
progeny selected which were B6.SJL.Igha using FACS and the strain was established
from a single breeding pair. All mice were housed in AAALAC approved facilities in
compliance with all applicable federal, state and local laws (Table 2-1).
HgCl-. Treatment
HgCl2 (Sigma Chemical Co., St. Louis, MO) was prepared in sterile, pyrogen-free
phosphate buffered saline (PBS). Mice were injected subcutaneously at a dose of 1.5
mg/kg three times weekly unless noted otherwise.

16
Indirect Immunofluorescence
Sera from mice was tested for the presence of ANoA by indirect
immunofluorescence using commercially prepared mouse frozen kidney slides (Sanofi,
Chaska, MN). The slides were incubated with sera diluted 1:50 in PBS for 30 minutes at
room temperature. For nonallotype-specific ANoA. the slides were incubated for 30 min
at room temperature with FITC-conjugated goat anti-mouse IgG (Fc fragment specific,
Jackson ImmunoResearch, West Grove, PA) diluted 1:50 in PBS. For the allotype-
specific ANoA determination slides, were incubated with rabbit anti-mouse IgG2aa or
IgG2ab (Nordic Labs, Capistrano Beach, CA) followed by FITC-conjugated goat anti¬
rabbit IgG (Jackson ImmunoResearch, West Grove, PA). These reagents had been pre¬
titrated by ELISA against an allotype-nonspecific rabbit anti-IgG2a antibody (Nordic
Labs) to produce equivalent sensitivities. Antinucleolar staining was evidenced by an
intense homogeneous staining of the nucleoli (Fig. 2-1).
Antibodies
MK-S4 (HB4; murine IgG2b anti-I-Aps,fiu (49)) and 14-4-4S (HB32; murine IgG2a k
anti-I-EaM (50)) were purchased from American Type Tissue Culture (Rockville, MD).
D3-137.5 (murine IgG2a anti-I-Apb antigen (51)) was originally obtained from Tonkonogy.
D77, originally developed against yeast fibrillarin but cross-reactive with fibrillarin of rat
and human origin (52), was a kind gift of Dr. John Aris (University of Florida, Gainesville,
FL). Aliquots of stock mAb were prepared from overgrown cell culture supernatant

17
which was affinity purified on a protein G column. Monoclonal antibodies were labeled
either with biotin hydrazide (53) or fluorescein isothiocyanate (FITC) (54), as needed.
Immunizations
Mice were immunized by an intraperitoneal injection of 50 pg of hen egg lysozyme
(HEL; Sigma Chemical Co., St. Louis, MO) in 100 pi of CFA (Difco Laboratories,
Detroit, MI) diluted 1:1 with sterile PBS. Fourteen days later the mice were boosted with
the same antigen in IF A. Mice were then bled via tail vein 8 weeks after the secondary
immunization and the sera stored at -20°C until analyzed by ELISA.
Anti-HEL ELISA
Determination of antigen-specific antibody levels were determined using an
indirect ELISA. Immulon® 2 microtiter plates (Dynatech Laboratories, Inc., Chantilly,
VA) were coated overnight at 4°C with HEL at 10pg/ml in bicarbonate buffer (0.1M, pH
8.2). Between all incubation steps microplates were washed three times with borate-
buffered saline (BBS: 25 mM Na2B407' 10 H20, 75 mM NaCl, 100 mM H3B03, pH 8.4)
containing 0.05% Tween 20. After blocking with BBS containing 0.5% BSA and 0.4%
Tween 20, serial dilutions of serum samples diluted in the blocking solution were
incubated overnight at 4°C. Either biotinylated goat anti-mouse IgG (Fc fragment
specific) or biotinylated goat anti-mouse IgM (mu chain specific)(Jackson
Immunoresearch, West Grove, PA) was added for one hour at room temperature. The
indicator system consisted of ExtrAvidin® alkaline phosphatase (Sigma Chemical Co., St.

18
Louis, MO) and the substrate Sigma 104® phosphatase substrate (p-nitrophenol
phosphate, disodium, hexahydrate). The substrate turnover was determined by the
difference between the OD405 and OD620 on a BioTek Instruments, Inc. (Winoski, VT)
Ceres 9000 microplate reader. The concentration of antigen-specific IgG and IgM is
reported in equivalent dilution factors of standardized reference B6.SJL sera. This is
defined by the formula: EDF = (dilution of a standard reference sera that gives the
equivalent OD of the test serum) X 104.
Flow Cytometry
Approximately 200 pi of tail vein blood was collected into heparinized tubes.
PBMCs were isolated using Lympholyte M (Cerdarlane, Hornby, Ontario, Canada)
density gradients. The cells were then collected into PBS, supplemented with 3% FCS
and 0.1% NaN3. For cell surface staining, saturating amounts of 14-4-4S , D3-137.5-
biotin or MK-S4-FITC were used as the first step. The second step consisted of
incubating with FITC-conjugated goat anti-mouse IgG (Fc fragment specific) or
phycoerythrin (PE)-conjugated streptavidin (Jackson Immunoresearch, West Grove, PA).
The cells were then washed three times in PBS and fixed with an equal volume of 2%
paraformaldehyde. List mode data were acquired on a FACScan flow cytometer (Becton
Dickinson, Palo Alto, CA) using PC Lysis software. Dead cells were excluded by forward
and side scatter gating. List mode files were then analyzed using Lysis II software.

19
Statistical Analysis
The difference between the number of mice showing an ANoA between B6.I-As/b
and B6.I-A87' was analyzed using Fischers exact test. A t-test was used to compare
baseline I-As MFI (mean fluorescent intensity) values between these two groups.
Differences between MFIs of the other groups was tested by an ANOVA (analysis of
variance) followed by Dunnett's multiple comparison procedure. Ap value of <0.05 was
considered to be significant.
Results
Fj Mice Between a Resistant and Sensitive Haplotvpe Were Resistant to ANoA Induction
In the course of our initial experiments with the HgCl2 model of induced
autoimmunity, we examined the ANoA response in an Fx cross between the susceptible
B6.SJL (H-2S; Ighb) and the resistant B6.TC (H-2b: Igha) strains. As expected, B6.SJL
(n=18) and B6.TC (n=13) were 100% and 0% susceptible, respectively. A representative
example is shown in Fig. 2-1, where a serum sample from a B6.SJL mouse treated for 5
weeks with HgCl2 shows a distinct nucleolar pattern. However, the vast majority (42/45)
of Fj intercrosses between these two strains were also resistant to HgCl2 treatment (Table
2-1). The resistance of the F, mice was surprising. Although the exact mechanism for
HgCl2-induced autoimmunity is unknown, the fact that the class II molecules are co-
dominantly expressed led us to expect that the expression of I-As would be sufficient to
confer susceptibility. To make sure this resistance was not due to the allotype difference,

20
B6.SJL.Igha (H-2S, Igha) mice were produced and found to be 100% susceptible to ANoA
formation. This showed that the Igh locus did not have an effect on susceptibility.
Finally, to rule out a dose effect in the Fj crosses, groups of 4-6 (B6.SJL x B6.TC)F1 were
dosed with either 3.0, 4.5, 7.5 or 10 mg/kg of HgCl2 for 3 weeks with no mice developing
an ANoA titer (data not shown).
Because it has been reported that the B6 background produces lower titer ANoA
responses than other H-2S strains, other crosses were tested. Initially, mice on the A
background were used. Consistent with previous studies, A.SW (H-2S) mice were 100%
(3/3) susceptible while A.B Y (H-2b) mice were resistant (0/4). Next, Fi crosses between
these two strains were generated. Despite the more susceptible A background, only
twenty percent (1/5) of (A.BY X A.SW)F1 mice showed ANoA positivity after treatment.
To rule out a maternal effect in Fx mice as previously suggested (9), the same strain was
produced except that the A.SW strain was used as the female parent. This cross, (A.SW x
A.BY)FX, resulted in only 11% (1/9) of the animals responding. In further attempts to find
an F, cross between resistant (H-2b) and sensitive (H-2S) mice in which the susceptibility
gene(s) was dominant, the following crosses were tested: (A.SW X B6)Fb (B6 X SJL/J)Fj
and (SJL/J X B6)Fb None of these animals developed an ANoA titer after mercury
treatment.
The specificity of the autoantibody response was verified by comparing an
immunoblot of sera from a B6.SJL mouse. This serum sample which reacted with a 34-
kDa nucleolar protein shown to be fibrillarin by mAb D77 (data not shown).

21
Figure 2-1 Serum from a B6.SJL mouse treated for 5 weeks with HgCl2 on mouse
frozen kidney slides and stained with FITC-conjugated goat anti-mouse
IgG as a second step showing intense staining of nucleoli.

22
Table 2-1 Congenie and Ft mice: serum ANoAa after 5 weeks of treatment with HgCl2
Strain Class II LA I-Ec IgGl nd %positivee
Haplotype Añ A„ Eñ E„ allotype
(b b) a 4 0
A.BY
b
b
b
A.SW
s
s
s
(A.BY x A. SW)FX
s/b
s/b
s/b
(A.SW x A.BY)F!
s/b
s/b
s/b
(A.SW xB6)Fj
s/b
s/b
s/b
(B6 x SJL/J)Fj
s/b
s/b
s/b
(SJL/J x B6)F1
s/b
s/b
s/b
B6.SJL
s
s
s
B6.TC
b
b
b
(BÓ.SJLxBÓ.TQF!
s/b
s/b
s/b
B6.SJL.Igha
s
s
s
(s
s)
a
3
100
(s/b
s/b)
a
5
20
(s/b
s/b)
a
9
11
(s/b
s/b)
a/b
8
0
(s/b
s/b)
b
6
0
(s/b
s/b)
b
6
0
(s
s)
b
18
100
(b
b)
a
13
0
(s/b
s/b)
a/b
45
7
(s
s)
a
9
100
flANoA, antinucleolar antibodies
A1.5 mg HgCl2/kg s.c. three times per week
calleles written in parenthesis denote nonexpression on the cell surface of the EpEa molecule
“'Number of mice
Traction of positive mice
The Ab Gene Product Conferred Resistance in Ft Mice
Because of the marked resistance of all (s/b) Fj mice tested and the possibility that
this effect was being mediated by non-class II molecules, the role of class II was tested
directly. Taking advantage of the availability of B6 mice with targeted deletions of the I-
Apb gene, mice heterozygous for this induced mutation were mated with B6.SJL mice to
generate two groups of mice within the littermates. One group possessed the genes linked
to I-Ab without expressing the molecule itself, while the other possessed these same genes
but also expressed I-Ab. Those mice that expressed I-Ab (I-As/b) were resistant (1/14) to
the development of ANoA, as expected from our earlier results, despite expressing the
susceptible I-As molecule. However, those that possessed both I-As and I-Ab-linked
genes, but not I-Ab expression itself (T-A87'), were as susceptible (17/20) to ANoA

23
formation as B6.SJL homozygous mice (Table 2-2). This clearly shows that it is the I-Ab
molecules themselves and not a linked gene which conferred resistance. As a negative
control, no mice (0/8) developed ANoA after treatment with PBS (Table 2-2).
Table 2-2 Effect of Anb deletion on ANoA° in HgCl,-treated(B6 SJL x B6.I-Ab/~) F, mice
Class II
Haplotype*
nc
Hgcy
NaCf
%positivei/
HgCl, NaCl
s/b
14
4
lg
0
s/-
20
4
85
0
“ANoA, antinucleolar antibodies measured after 5 weeks of treatment
mice either did (s/b) or did not (s/-) express I-Ab but were heterozygous at all the
MHC loci
^Number of mice
^Fraction of positive mice
e1.5 mg HgCl2/kg s.c. three times per week
f0.1 ml NaCl s.c. three times per week
^Significant difference between (s/b) and (s/-) mice (p < 0.05 using Fischer
exact test)
To rule out the possibility that the differential susceptibility between I-As/b and I-
A87' was due solely to increased expression of I-As and not a negative regulatory influence
by I-Ab, we examined I-A levels by haplotype-specific flow cytometry. Baseline I-As
expression was similar whether or not the mouse carried the disrupted gene (MFI = 45 vs.
49; p = 0.1185). MHC class II expression on PBMCs was similarly increased in both
strains when treated with HgCl2 (Fig. 2-2). I-As expression was increased by 44%
(¿?<0.Q5) in the mice which carried the disrupted allele, while it was increased by 48%
(¿K0.05) in those mice which expressed both haplotypes (s/b). The increase in I-Ab
expression in the mice which carried that gene was 48% (p<0.05). No significant increase
in class II expression was seen in those animals treated with PBS.

24
A
B
90
Uh
B6.I-A'
s/b
Figure 2-2 Effect of HgCl2 on class II expression in(B6.SJL x BóT-A^Fi mice. (A) Predose I-As
expression was similar in B6.I-As/b and B6.I-AS' mice. I-As expression was increased by 44%
in the mice which carried the disrupted allele, while it was increased by 48% in those mice
which expressed both haplotypes (s/b). (B) The increase in I-Ab expression in the mice which
carried that gene was 48%. No significant increase in class II expression was seen in those
animals treated with PBS. *p< 0.05

25
Expression of I-E Had No Effect on Susceptibility to HgCU
It has previously been reported in a number of autoimmune models that I-E
expression can down-regulate autoantibody titers. Intra-MHC recombinant strains have
given conflicting results in the HgCl2 model, but interpretation has been hampered by the
difficulty in isolating the effects of I-E expression. To test this more explicitly and to see if
competing expression by another class II molecule is a general property in the down-
regulation of the anti-nucleolar response to mercury, B6.SJL mice expressing the I-Ead
transgene were bred. Restoring the ability to express I-E in otherwise susceptible mice
had no effect on the ability to respond to HgCl2. All (5/5) B6.SJL.I-Ead mice retained the
ability to generate an ANoA titer upon HgCl2 treatment (Table 2-3). Mean I-E expression
in these mice (MFI = 280.7) was similar to Balb/c mice (MFI = 282.1) which normally
express the I-E gene product. I-As expression between B6.SJL and B6.SJL.I-Ead mice
was also similar (MFI = 267.9 vs. 253.2) (Fig. 2-3). The breeding scheme used to generate
the B6.SJL.I-Ead mice also produced (B6.SJL x B6.I-Ead )BCj mice which either did
(n=14) or did not (n=13) express the transgene. It has been suggested in the BxSB
autoimmune model that I-E expression could block autoantibody production by producing
an I-E-derived peptide that binds to I-Ab with such affinity that it competes with the
putative autoantigen-derived peptides for presentation (6). Starting with the same
transgenic parent strain as was used in the present experiments, we have also confirmed
this observation in the BxSB model (unpublished data). If this is a general property in
autoimmunity, it might be expected that I-E expression would convert an otherwise

26
resistant HgCi2-treated s/b heterozygous mouse into a sensitive one. Despite normal
levels of expression of I-E, however, all of these mice were resistant (Table 2-1).
iaoie z-j triect or l-u, on /vivoa
Strain Class II
Haplotype
respon
i
Afi
se
\
A«
I-E
Expression
nb
%positivec
B6.SJL.I-EKd s
s
s
+
5
100
(BÓ.SJLxBÓ.I-E^F! s/b
s/b
s/b
+
14
0
(B6.SJLxB6)F, ^ s/b
s/b
s/b
-
13
0
“ANoA, antinucleolar antibodies measured after 5 weeks of treatment with 1.5 mg
HgCl2/kg s.c. three times per week
^Number of mice
Traction of positive mice
^on-transgenic littermate of(B6.SJL x Bó.I-E^F!
Mean I-E Expression
B6.SJL.I-E
/V.
282.1
Tin ip tin
FL1-Height
~ib -1
B
Mean I-A Expression
B6.SJL
267.9
I2Y
1b “ TiT1 ilT5 Í51
FL 1-Height
To-
Balb/c
1
ib ' 1b 2
FL1-Height
lb 4
B6.SJL.I-E
253.2
ib ' ib 2 ib
FL1-Height
'lb 4
Figure 2-3 Class II expression in B6.SJL.I-Ead mice. (A) Mean I-E expression
was similar to Balb/c mice. (B) Mean I-As expression was not
affected by I-E expression. Single-parameter fluorescence histograms.
Increasing fluorescence intensity is plotted on the x-axis in log
fluorescent units versus cell number on the y-axis The numbers
represent the MFI (mean fluorescent intensity) of the cells expressing
the antigen of interest.

27
Responsiveness to Exogenous Antigen Was Inherited Co-dominantlv in F, Mice
The lack of ANoA's in our Ft mice led us to evaluate their responses to an
exogenous antigen (HEL). As expected, mice of the b haplotype (B6 .TC) responded
poorly while the B6.SJL mice mounted an intermediate titer. An intermediate response
was seen in the Ft cross (Figure 2-4). Balb/c mice responded with a titer approximately
10 fold higher than any of the mice on the B6 background (data not shown). Thus it
appears that the responsiveness in our Ft mice to HEL was inherited in a co-dominant
fashion.
Discussion
The mercury chloride response in rodents is a complicated one and classification of
responder and non-responder strains depends on the outcome being measured. For
example, by ANoA titer, the b haplotype is a non-responder. However, by measuring the
weights of draining popliteal lymph nodes of (DBA/2 X C57BL/10)F1 mice exposed to
HgCl2, b haplotype has been classified as intermediate responders.
The anti-fibrillarin response by susceptible mice, however, is very reproducible and
specific and therefore more amenable to mechanistic dissection. As a first step, it is crucial
to convincingly demonstrate that the anti-fibrillarin response is a property of the I-A
molecule itself and not a closely linked gene. Results in the literature thus far have been
suggestive but not conclusive. Robinson et al.(5) used two unrelated strains (A.SW and
C57BL/10) to generate the F/s. These mice were 57-100% susceptible to developing

28
ANoA, albeit of a lower titer. This led the authors to conclude that the response was
controlled by the H-2 region and one or more unlinked genes. Recently, Hultman et al.
(9) found that 75% of male and 11% of female mercury treated (SJL x B6)Fj mice showed
a clumpy IgG ANoA pattern with a low titer. These differences were attributed to the
non-H-2 genes. For example, the A genetic background seemed to enhance susceptibility
to the HgCl2 treatment, while the BIO background seems to strengthen the resistance.
They also suggested that the B6 non-H-2 background genes are more suppressive on
ANoA expression than equivalent genes from the B10 strain. A similar effect of non-H-2
genes was seen by Hultman et al. (12) when comparing the ANoA response in several H-
2s strains. B10.S mice showed a much lower titer than either A.SW or SJL mice after
HgCl2 treatment. This was again attributed to the non-H-2 genes.
A Anti-HEL IgG
B Anti-HEL IgM
1600
1200
b
800
w
400
0
B6.SJLB6.TC FI
Figure 2-4 Anti-HEL antibody levels 10 weeks after immunization. The response
to HEL was inherited co-dominantly in(B6.SJL X Bó.TQFj mice.
B6.SJL and (B6.SJL X Bó.TQFi mice respond to HEL while B6.TC do
not. (A) IgG (B) IgM; EDF +/- SEM

29
In the process of studying the cellular control of the mercury-induced anti-
fibrillarin response, we examined the susceptibility of a number of Ft mice in our own
laboratory. Contrary to previously published results summarized above, our F¡ mice
seldom produced a measurable ANoA titer most likely due to the strong suppressive
effects seen when using mice on the B6 background. This lead us to re-examine the role
of class II expression. Several genes linked to class II have been shown to participate in
antigen processing and presentation (44,55,56). The possibility existed that the genes
linked to the I-Ab molecule could be altering the processing and/or presentation of
fibrillarin resulting in a peptide which no longer is immunogenic. If this were the case then
mice which express I-As and carry the I-Ab linked genes without displaying the I-Ab
molecule itself would be expected to be resistant.
Since the ANoA response was not dominant in our F, mice we compared the
responses to HEL in which responsiveness should be inherited in a co-dominant fashion.
Mice of the b and s haplotype are classified as non-responders to HEL when measured by
PFC assays, although SJL (H-2S) mice do mount an intermediate serum titer (57). An
analysis of anti-HEL serum titers revealed that B6.TC mice responded poorly, as
expected, while both while B6.SJL and (B6.SJL x Bó.TQFj mice were intermediate
responders. It appears that there is a difference in the inheritability of responsiveness
between the HgCl2-induced ANoA response and that to a conventional exogenous
antigen.
To investigate whether the resistance of Fj mice is due to expression of the I-Ab
molecule or a linked gene, Ft mice on the B6 background were produced which lacked

30
expression of the I-Ab molecule due to disruption of the MHC class II I-Apb gene (46).
By using congenie strains which differed at only the H-2 region, we eliminated the effects
of non-MHC-linked genes. Because they are of a low-responder background, the B6
strains also allowed us to more carefully isolate the effects of the MHC. Finally, the B6
background permitted us to take advantage of the availability of class II transgenics and
targeted deletions. Our strategy was to cross the CD2â„¢ (B6.I-A'7") to a wild type B6.I-
Ab/b to obtain B6.I-Ab/" mice. These mice were then bred to a B6.SJL to obtain the two
strains used for analysis: B6.I-As/b and Bó.I-A87'. The incidence of ANoA in those mice
which expressed the I-Ab gene product was significantly decreased showing that the I-Ab
molecule itself and not a linked gene was responsible. To rule out the possibility that the
mice carrying the disrupted allele could be expressing a higher density of I-As molecules
thus leading to the increased susceptibility, we compared the predose mean I-As
expression between the 2 strains. The MFI of I-As expression for the mice carrying the
disrupted allele (I-A87-) was not significantly different from the I-As/b mice.
The possibility exists that the I-Aj* gene product could pair with the I-Aps
molecule resulting in an haplotype-mismatched class II molecule (Aab / Aps). Martin et
al. (5 8) showed that the inefficient assembly of haplotype-mismatched class II polypeptides
results from their inability to compete with the matched pairs. Different class II MHC aP
chain combinations vary greatly in their efficiency of cell-surface expression based on the
allelic origin of the a and P amino-terminal regions (59). If this pair was being expressed
we would have expected to see an increase in the I-As expression in the mice carrying the
disrupted allele (I-A^'), since the mAb MK-S4 recognizes the P chain of I-As, however no
difference was detected, and we believe this is an unlikely possibility.

31
Another possibility is that the HgCl2 had a differential effect on the two class II
molecules. If the HgCl2 had caused a greater increase in the expression of the I-Ab
molecule in those mice which expressed it, this could lead to a decreased presentation of
antigen by the susceptible allele (I-As), perhaps falling below the threshold required for T
cell activation. This is not likely since both class II molecules showed similar increases in
expression upon HgCl2 treatment.
To examine directly the effect of I-E expression in otherwise susceptible mice, we
generated mice on a B6 background expressing the I-As and the I-E molecules. Like other
H-2S mice, B6.SJL mice do not express I-E due to a deletion in the I-E„ gene (60). To
remedy this, we crossed B6.SJL mice with a B6 mouse carrying the I-E transgene (Tg(H-
2I:Ea)Bri39) (45). After backcrossing to B6.SJL mice, the progeny which expressed I-As
and I-E were selected. These mice remained 100% susceptible to ANoA formation
despite expressing the I-E molecule. Our results do not support the observation by
Mircheva et al. (6) that the expression of I-E suppresses the induction of ANoAs.
Hultman et al. (7) was also not able to confirm the suppression. Both of theses studies
examined the effect of I-E by using intra-H-2 recombinant strains, which meant that the
conclusions were based on inference from a number of comparisons of strains. We believe
that our system of using the introduction of a transgene to restore expression offers an
advantage over random recombinatorial events which can carry along other genes in the
process.
The possibility exists that these mice do not express I-E at physiological levels. If
I-E were being expressed at excessively high levels, antigen presentation by the I-As

32
molecule could be diminished. Conversely, low-level expression of I-E could result in
little physiologic effect. To rule out either of these artifacts, we compared mean I-E
expression between our mice and a strain which normally expresses I-E (Balb/c). There
was no difference in the mean expression between the two strains. The expression of I-A8
was also similar between B6.SJL and the B6.SJL.I-Ead mice. Moreover, expression of I-
E was also physiologic in that it was expressed on B but not T cells (data not shown).
Merino et al. (47) showed that the I-Ead transgene protected BxSB mice against
SLE. They proposed that overexpression the I-E a chain resulted in the generation of
excessive amounts of a peptide displaying a high affinity for the I-Ab molecule thereby
preventing autoantigen-derived peptides from being presented. This phenomenon has also
been seen in mice homozygous for the Ipr mutation (48). To investigate the possibility
that this might work in reverse to promote autoimmunity in some cases, we generated the
otherwise resistant (B6.SJL X Bó.I-E^Fj mice which either did or did not express I-E.
Theoretically, this could block I-Ab as a functional competitor and enhance an I-As-
mediated response. However, neither strain developed ANoA after treatment with HgCl2.
Thus, the expression of I-E does not promote the development of ANoA in a strain which
is otherwise resistant. The results are clear, although we cannot formally exclude the
possibility that a peptide derived from I-E was even more effective in blocking I-As.
However, if this were the case then we would have expected I-E to have decreased HgCl2-
induced susceptibility in the B6.SJL.I-Ead mice.
Taken together, our results conclusively demonstrate for the first time that HgCl2
responsiveness is regulated by the I-A haplotype itself and not a closely linked gene.

33
However, the ability to downregulate the immune response by coexpression in a resistant
haplotype in an otherwise sensitive strain is a complicated one. Antigen competetion,
proposed to be important in other autoimmune models (47,48), does not appear to have a
dominant role in the HgCl2-induced response.

CHAPTER 3
RESISTANCE TO HgCl2-INDUCED AUTOIMMUNITY IN HAPLOTYPE
HETEROZYGOUS MICE IS AN INTRINSIC PROPERTY OF B CELLS
Introduction
Systemic lupus erythematosus is a complicated systemic autoimmune disease with
a multigenic mode of inheritance interacting with a potentially complex array of generally
unknown environmental factors. Many of these features are also seen in the mercury
chloride model of murine autoimmunity, which offers a unique opportunity to study the
reproducible interaction between genetics and a simple, inorganic environmental agent.
Susceptibility to mercury chloride in mice is characterized by splenic hyperplasia,
hypergammaglobulinemia and autoantibody production (4,29), all of which appear as soon
as one week after the start of treatment (4). The principle autoantibody target is
fibrillarin, a U3 ribonucleoprotein found in the nucleolus and involved in the first step of
preribosomal processing (31,39). The antifibrillarin specificity is also seen in patients with
scleroderma and other connective tissue disorders (33,34). Recently, a comparison of the
specificity of the human and murine autoantibodies showed that they recognize similar, if
not identical, epitopes (40) and suggested an antigen-driven response in both.
Similar to human autoimmune disease, both MHC and non-MHC genes have been
shown to have a pronounced effect on susceptibility (12). As defined by the presence of
34

35
antifibrillarin antibody, mice of the H-2S haplotype are high responders, while those of the
H-2b and H-2d haplotypes are nonresponders (5). In contrast, non-MHC genes affect
primarily antibody titers but not specificity (12). Through the use of intra-H-2
recombinants, the presence of the HgCl2-induced antifibrillarin antibody response has been
mapped specifically to the I-A region (6). Previous studies in our laboratory have shown
that Fj animals between MHC-congenic susceptible H-2S and resistant H-2b mice to be
resistant to HgCl2-induced ANoAs (8). This is surprising, since the I-A molecule is co-
dominantly expressed in the Ft mice, and in other autoimmune models heterozygosity of
class II either enhances autoimmunity or modestly affects antibody titers (61,62).
Therefore, the profound resistance to HgCl2-induced ANoAs seen in all haplotype-
heterozygous mice tested raised the possibility that resistance was mediated by a
dominant gene linked to I-Ab. However, further experiments demonstrated that resistance
was, in fact, caused by co-expression of the I-Ab molecule itself, and that this outcome
was not merely the result of lower expression of the susceptible I-As haplotype on the
otherwise responsive B cells (8).
We have utilized adoptive transfer experiments to explore further the mechanisms
by which the resistant haplotype down regulated the HgCl2-induced antifibrillarin antibody
response. Our experiments demonstrated that resistance was due neither to I-Ab-mediated
alteration of T cell repertoire nor to inadequate I-As-restricted T cell help but was an
intrinsic property of the resistant haplotype-heterozygous B cells. These results suggest
the presence of a novel mechanism of regulation of a potentially autoreactive immune
response.

36
Mice
Materials and Methods
C57BL/6J (B6; I-Ab, Ighb), C57BL/6J.SJL (B6.SJL; I-As, Ighb), C57BL/6J-/g/z°
Thyla Gpf (B6.TC; I-Ab, Igha) mice were originally obtained from The Jackson
Laboratory (Bar Harbor, ME) and maintained in our breeding facility.d B6.SJL.Igha mice
were developed by intercrossing B6.SJL and(B6.SJL x B6.TC)F]. FACS analysis was
used to select progeny which were I-As and Igha/b. These mice were then intercrossed and
the progeny selected which were B6.SJL.Igha using FACS and the strain was established
from a single breeding pair. All mice were housed in AAALAC approved facilities in
compliance with all applicable federal, state and local laws.
Preparation of Chimeras
Preparation of chimeras was as previously described (63). Recipient mice were
provided with Septra® - (1% v/v) treated water the day prior to irradiation. On the day
prior to bone marrow transfer, the mice were treated with two doses of 525 rad of y-
radiation (Gammacell 40, Atomic Energy of Canada, Ltd., Ottawa, Canada) separated by
3-4 hours. The transfer involved i.v. reconstitution with a total of 107 bone marrow cells
from age- and sex-matched donors. Bone marrow cells were depleted of mature T cells
by incubation at 4° C for 30 min with a mixture of anti-mouse T-cell serum (Cedarlane
Laboratories, Hornby, Ontario, Canada), 172-4 (rat IgM anti-CD4 (64)), and 31 M (rat
IgM anti-CD8 (65)) followed by treatment with C at 37°C for 1 h (Low-Tox Guinea Pig

37
C, Cedarlane Laboratories, Hornby, Ontario, Canada). To prevent graft rejection the
B6.SJL mice receiving(B6.SJL x Bó.TQFj cells were given 0.1 mg i.p. ofMmTl (mouse
IgG2a anti-CD90.2 (66)) at the time of transfer. 172-4 and 31 M were obtained from Dr.
David Harris (University of Arizona, Tuscon, AZ) and prepared from overgrown cell
culture supernatant which was affinity purified on a protein G column.
In Vivo Treatments
HgCl2 (Sigma Chemical Co., St. Louis, MO) was prepared in sterile, pyrogen-free
phosphate buffered saline (PBS). Mice were injected subcutaneously at a dose of 1.5
mg/kg three times weekly after graft acceptance was verified using flow cytometry. Mice
were immunized intraperitoneally with 100 pg of human IgG (HGG; Sigma Chemical Co.,
St. Louis, MO ) emulsified 1:1 in CFA (Difco Laboratories, Detroit, MI). Booster
immunizations consisted of a second intraperitoneal injection of 100 pg of HGG in sterile,
pyrogen-free PBS.
Flow Cytometry
Approximately 5-6 weeks after bone marrow transfer 200 pi of tail vein blood was
collected into heparinized tubes. PBMCs were isolated using Lympholyte M (Cerdarlane
Laboratories, Hornby, Ontario, Canada) density gradients. The cells were then collected
into PBS, supplemented with 3% FCS and 0.1% NaN3. For cell surface staining,
saturating amounts of biotinylated D3-137.5 (mouse IgG2a anti-I-Ab)(51) and
fluoresceinated TIB 92 (mouse IgG2a anti-I-Ab)(67) or biotinylated AF6-78.25 (mouse

38
IgGl anti-IgMb)(68) and fluoresceinated DS-1 (mouse anti-IgMa)(69) or biotinylated
HIS51 (mouse IgG2a anti-CD90.1) and fluoresceinated 30-H12 (rat IgG2b anti-CD90.2)
were used as the first step. The second step consisted of incubating with phycoerythrin
(PE)-conjugated streptavidin (Jackson ImmunoResearch, West Grove, PA). The cells
were then washed three times in PBS and fixed with an equal volume of 2%
paraformaldehyde. List mode data was acquired on a FACScan flow cytometer (Becton
Dickinson, Palo Alto, CA) using PC Lysis software. Dead cells were excluded by forward
and side scatter gating. List mode files were then analyzed using Lysis II software.
Monoclonal antibodies were labeled either with biotin hydrazide (53) or fluorescein
isothiocyanate (FITC) (54), as needed. At the study termination the same procedures
were used with the exception that splenocytes were used for cell staining.
Indirect Immunofluorescence
Sera from mice collected 5 weeks after the initiation of HgCl2 was tested for the
presence of ANoA by indirect immunofluorescence using commercially prepared mouse
frozen kidney slides (Sanofi, Chaska, MN). The slides were incubated with sera diluted
1:50 in PBS for 30 minutes at room temperature. For nonallotype-specific ANoA the
slides were incubated for 30 min at room temperature with FITC-conjugated goat anti¬
mouse IgG (Fc fragment specific, Jackson ImmunoResearch, West Grove, PA) diluted
1:50 in PBS. For the allotype-specific ANoA determination slides, were incubated with
rabbit anti-mouse IgG2aa or IgG2ab (Nordic Labs, Capistrano Beach, CA) followed by
FITC-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA).

39
These reagents had been pre-titrated by ELISA against an allotype-nonspecific rabbit anti-
IgG2a antibody (Nordic Labs) to produce equivalent sensitivities. Antinucleolar staining
was evidenced by intense homogeneous staining of the nucleoli.
Allotype-Specific ELISA
The protocols used for measurement of allotype-specific serum total IgM and
IgG2a were minor modifications of previously described procedures (63). For serum total
IgM, samples were developed with affinity-purified donkey anti-mouse IgM (Jackson
ImmunoResearch, West Grove, PA). IgMa and IgMb were measured using DS-1 (69) and
AF6-78.25 (68) respectively. Serum total IgG2a was determined using rabbit anti-mouse
IgG2a (Nordic Labs, Capistrano Beach, CA) while allotype-specific IgG2a was measured
using either rabbit anti-mouse IgG2aa or IgG2ab (Nordic Labs, Capistrano Beach, CA).
These reagents had been pre-titrated by ELISA against the allotype-nonspecific rabbit
anti-IgG2a antibody (Nordic Labs) to produce equivalent sensitivities. The rabbit
antibodies were detected using alkaline phosphatase-conjugated donkey anti-rabbit IgG
(Jackson ImmunoResearch, West Grove, PA). Allotype-specific IgG2a anti-HGG was
measured using the same protocol except that HGG was used as the first step.
Results
MHC-Restricted T Cell Help Was Required to Produce ANoAs
Exposure to HgCl2 induces a wide range of physiologic effects in mice, including
the release of large amounts of cytokines, especially the Th2 cytokine IL-4. It was

40
therefore possible that the antifibrillarin response could result from non-MHC-restricted
interactions in genetically susceptible mice. To test this possibility, B6.TC mice were
reconstituted with T cell-depleted B6.SJL marrow. Because intrathymic positive selection
of T cells is mediated by the radioresistant thymic cortical epithelial cells (70), in these
mice all CD4+ T cells would be positively selected to interact with the host I-Ab and not
the donor I-As haplotype. Therefore, the Bó.SJL-derived B cells would not be expected
to receive MHC-restricted help. In contrast, central tolerance is mediated by bone
marrow-derived dendritic cells although thymic epithelial cells can make a significant
contribution (71), and therefore the animals would be tolerant to both haplotypes. Flow
cytometric analysis of peripheral blood lymphocytes showed that all of the B cells were of
donor B6.SJL origin. Surprisingly, despite the presence of these susceptible B cells and
the massive activation of cytokines following HgCl2 administration none of the animals
developed ANoAs (Table 3-1). The addition of B6.TC cells in the inoculum did not alter
the outcome. In marked contrast, syngeneically reconstituted B6.SJL mice responded
well to mercury chloride, demonstrating that lack of responsiveness was not an artifact of
radiation. We therefore conclude that the antifibrillarin specificity is mediated by I-As -
restricted T cell help.

41
Table 3-1 MHC-restricted T cell help was required to produce ANoAs*
Dpnor
Host
nb
ANoA Positive (%Y
B6.SJL & B6.TC
B6.TC
4
0
B6.SJL & B6.TC
B6.SJL
5
100
B6.SJL
B6.SJL
7
100
B6.SJL
B6.TC
5
0
“ANoA, antinucleolar antibodies
dumber of mice
Traction of positive mice
Absence of ANoA Production in Haplotype-Heterozvgous Mice Was Not Due to a
Difference in Thymic Education
Although the above experiments showed the importance of T-B cell collaboration
it is possible that I-Ab expression in the host and/or donor affected thymic education and
therefore eliminated a population of responsive T cells. This led us to investigate the
responses in (B6.SJL x Bó.TQFj hosts. When (B6.SJL x B6.TC)FX mice were used as
hosts and reconstituted with B6.SJL, they all (12/12) responded to HgCl2. When the same
hosts were used but the donor changed to B6.TC, none (0/6) of the mice produced
ANoAs (Table 3-2). Therefore, the donor haplotype determined whether or not
autoantibodies are produced when developing cells were positively selected on both
haplotypes.
Allogeneic Mixed Chimeras Developed ANoA Titers in Response to HgCU
In the rat model of mercury chloride-induced autoimmunity, allogeneic mixed
microchimerism has been able to induce a state of tolerance (72). In contrast, other
models have shown the potential for epitope spreading once tolerance is broken (73,74).

42
Therefore, it was of interest to determine whether the co-existence ofB cells sensitive and
resistant to the effects of HgCl2 could influence each other.
To test these possibilities, we lethally irradiated (B6.SJL x B6.TC)F, mice and
transferred T cell-depleted bone marrow from the following: 1) B6.SJL alone; 2) B6.TC
alone; 3) a combination of B6.TC & B6.SJL; and 4) (B6.SJL x Bó.TQFj alone. In all
four groups, T cells would be positively selected by either I-As or I-Ab. After
confirmation of mixed chimerism in group 3 (see below), the mice were treated with
HgCl2 and the results are shown in Table 3-2. As expected, (B6.SJL x B6.TC)F, mice
reconstituted with only B6.SJL bone marrow responded readily to HgCl2 while mice
given only resistant B6.TC bone marrow failed to develop an ANoA antibody titer. In
contrast, syngeneic reconstitution with resistant (B6.SJL x B6.TC)Fj bone marrow
resulted in a very poor ANoA response. These results essentially duplicated our
experience in analogous non-chimeric HgCl2-treated B6.SJL, B6.TC, and (B6.SJL x
B6.TC)F1 mice (8) and again demonstrated that radiation did not affect this outcome.
Strikingly, mice given a combination of resistant and susceptible bone marrow had
an intermediate result, with nearly 50% of the mice responding (Table 3-2). By allotype-
specific ANoA, all of the responders produced exclusively b allotype autoantibody and
therefore of donor B6.SJL origin (data not shown). This is not due to an intrinsic
property of the a allotype, for the a allotype B6.SJL-Igha strain responded equally well to
mercury (data not shown). Thus, the presence of ANoA in these mice resulted from the
loss of tolerance to fibrillarin in I-As B cells.

43
Table 3-2 ANoAa response in chimeric (B6.TC x Bó.SJLjF, host mice
Group
Source of Donor Bone Marrow
~
ANoA Positive (%)'
1
B6.SJL
12
100
2
B6.TC
6
0
3
B6.SJL & B6.TC
16
43.8
4
(B6.SJL x Bó.TCjF!
17
11.8
aANoA, antinucleolar antibodies
^Number of mice
Traction of positive mice
Lack of ANoA in Non-Responsive Mixed Chimeras Was Unrelated to B Cell Composition
Of interest, the response rate of 44% in the mixed chimeras (Group 3) was
significantly different than either the 100% seen in Group 1 (p<0.02 by z-test for
proportions) or the 11% seen in Group 4 (p<0.04) and suggested that the presence of
resistant class II could exert a negative influence even when not co-expressed on the cells
otherwise capable of responding. To evaluate for the possibility that this effect was
merely due to dilution of potentially susceptible I-As-bearing B cells by resistant I-Ab B
cells, we compared the B cell makeup of the positive and negative mice. The results are
shown in Figure 3-1. The median % of I-As expressing B cells was 58.5% in those mice
which failed to produce ANoAs while it was 60 % in those mice that did (Figure 3-1A).
Even mice which had as few as 38% of their B cells from the sensitive parent were able to
produce autoantibodies. Comparable results were seen when analyzed by B cell allotype
(Figure 3-1B). Therefore, these data suggest that the decreased responsiveness in the
presence of resistant B cells was not just a dilutional effect.

44
B Cells of Both B6.TC and B6.SJL Origin Were Functional in Group 3 Mixed Chimeras
It was possible that the presence of a and b allotype B cells in the mixed chimeras
influenced the development of antinucleolar antibodies through antigen non-specific
mechanisms. For example, in some combinations, allotype-specific suppression of immune
responses has been seen (75). This possibility was minimized by the use of allotype-
heterozygous host mice. However, to assess further the functionality of B cells of both
B6.SJL and B6.TC origin, allotype-specific total IgM and IgG2a ELIS As were performed.
In addition, the mice were also assayed for their response to immunization with a T cell
dependent antigen, HGG. As shown in Figure 3-2, total IgM and IgG2a of both allotypes
were present. Interestingly, despite the predominance of IgM of B6.SJL origin, IgG2a
was much better balanced between the two donors. Moreover, there was a good IgG2a
response to a T cell dependent antigen by B cells of both B6.SJL and B6.TC origin
(Figure 3-3 A), and both ANoA positive and negative mice responded equally well (data
not shown).
Lack of HgCk-Induced ANoA Response by Haplotype-Heterozygous Mice Was Not Due
to Absence of Anti-Fibrillarin-Specific I-As-Restricted T Cell Help
To determine whether or not the intermediate response seen in the mixed chimeras
and in HgCl2-resistant (B6.SJL x B6.TC)FX mice was due to a lack of I-As-restricted T cell
help, we reconstituted (B6.SJL x Bó.TQFj mice with a combination of syngeneic (B6.SJL
x Bó.TQFi and B6.SJL bone marrow. Fortyfive percent (5/11) of these mice produced

%I-A'
45
B6.TC + B6.SJL —> FL
A. Class II Expression g Surface IgM
Figure 3-1 Immunofluorescence on peripheral blood lymphocytes of parental
into Fj chimeras. Data are presented as percent of cells expressing
class II. A, Class II expression; B, surface IgM.
%IgM

46
B6.TC + B6.SJL -->Fj
A. Serum IgM
B. Serum IgG2a
- 100
- 90
- 80
- 70
- 60 .o
- 50
- 40
- 30
- 20
- 10
- 0
ANoA pos neg
X
0
i
0
X
1
X
o
Figure 3-2 Allotype-specific total IgM and IgG2a ELISA data of parental into Fj
chimeras. Data are presented as a percentage of total isotype. A, Serum
IgM; B, serum IgG2a.
%IgG2a

47
Anti-HGG titer
A.B6.TC+B6.SJL—>Fj B.BÓ.SJL+Fj ~>F
Figure 3-3 Allotype-specific anti-HGG ELISA data. Data are presented as a
percentage of total isotype. A, B6.TC + B6.SJL —> Fj mice;
B, Bó.SJL + Fj—>F1mice.
%IgG2a

48
ANoAs upon HgCl2 treatment. The fact that some of the mice responded meant that T
cells were available in which tolerance to fibrillarin was broken. Allotypic analysis
revealed that all of the autoantibodies were of the b allotype. It seems likely that they
arose from the B6.SJL donor since the B cells from the (B6.SJL x Bó.TQFj mice would
have produced autoantibodies of either a or b allotype and no a allotype autoantibodies
were detected even at a 1:10 dilution (data not shown). Thus, despite the presence of T
cells which provided help in an I-As -restricted fashion, those B cells which co-expressed
both haplotypes did not produce autoantibodies.
B Cell Composition Was Not a Determining Factor in Responsiveness
Figure 3-4 shows the B cell makeup of the B6.SJL + (B6.SJL x B6.TC)Fj —>
(B6.SJL x Bó.TQFj mice. Flow cytometric analysis of class II expression showed these
mice to be very well balanced (%I-As/b range: 33-54%). The predominance of IgM of the
b allotype (61-82%) is expected considering that approximately one-half of the B cells
from the (B6.SJL x B6.TC)F! donor would express IgM of the b allotype. Similar to the
parental into (B6.SJL x B6.TC)F1 mice no correlation was noted between the B cell
makeup and responsiveness (Figure 3-4).
Both Bonors Provided Functional B Cells
The apparent lack of response from the a allotype B cells led us to examine
whether these B cells were functionally equivalent. Figure 3-5 shows that IgM and IgG2a
of both allotypes were produced although the composition is somewhat skewed to the b
allotype. Again, this is an expected finding since the cells from the (B6.SJL x B6.TC)Fj

%I-A'
49
B6.SJL + Fj —>Fj
A. Class II Expression
B. Surface IgM
ANoA
Figure 3-4 Immunofluorescence on peripheral blood lymphocytes ofBó.SJL +
F, —> Fj chimeras. Data are presented as percent of cells
expressing class II. A, Class II expression; B, surface IgM.
%IgM

50
mice would produce either a or b allotype. Similarly, B cells of the a allotype responded
equally well when the response to a T cell dependent antigen (HGG) was analyzed in an
allotypic fashion (Figure 3-3B). A comparison between positive and negative mice
revealed no significant differences. Taken together, we conclude that the inability of
haplotype-heterozygous mice to respond to HgCl2 with an ANoA response is an intrinsic
property of the B cells.
Absence of ANoA Production in Haplotype-Heterozygous Mice Was Not Due to the
Presence of I-Ab-Restricted T Cells
It was possible that the non-responsiveness of the mixed chimeric as well as the
syngeneically reconstituted (B6.SJL x B6.TC)F1 mice was due to a negative regulatory
effect of I-Ab -restricted T cells. To test for this possibility we lethally irradiated B6.SJL
mice and transferred T-cell-depleted bone marrow from (B6.SJL x Bó.TQFj mice. To
ensure no carry over of T cells educated in an (B6.SJL x B6.TC)Fj mouse we treated the
recipients at the time of transfer with 0.1 mg i.p. of MmTl (mouse IgG2a anti-CD90.2
(66)). The T cells in these mice are unable to provide help through I-Ab since they
developed in an I-A 8 expressing host. Flow cytometric analysis confirmed that all the B
cells in these mice expressed both haplotypes. Interestingly, none (0/11) of the mice
produced ANoAs following HgCl2 treatment. The resistance of these mice could not have
been due to negative influences by I-Ab -restricted T cells.

51
B6.SJL + Fj —>
A. SemmlgM B. Serum IgG2a
Figure 3-5 Allotype-specific total IgM and IgG2a ELISA data of B6.SJL + Fj
—> F t chimeras. Data are presented as a percentage of total
isotype. A, Serum IgM; B, serum IgG2a.
%IgG2a

52
Discussion
HgCl2 treatment in mice induces a wide range of physiological responses including
marked increases in JL-4 production (20,25). The requirement for T cells in this model
was shown by the fact that nude mice on the H-2S background fail to develop
autoantibodies (15). Splenic CD4+ T cells of HgCl2-treated H-2S mice have been shown
to have a strong increase in IL-4 mRNA, whereas those of H-2d mice showed only a weak
increase (17). Interestingly, treatment of H-2S mice with anti-IL-4 monoclonal antibody
did not prevent ANoA production although it changed the isotype profile of the
autoantibodies (25). Despite the possibility for non-cognate help via cytokines, we have
shown that MHC-restricted T cell help was required to produce ANoAs. The B cells
were not merely responding to cytokines in the environment but required specific MHC-
restricted signals from T cells.
Cognate interactions are also required in several other models of autoimmune
disease. Double parental-into-F! chimeras were used in the graft-vs-host model to show
that autoantibodies were derived nearly entirely from B cells receiving direct alloreactive T
cell help (51). Bone-marrow chimeras were also used to show that autoantibody
production in Ipr mice was restricted to those B cells that received T cell help (76). Thus,
in three very different murine models of autoimmunity, in vivo autoantibody responses
were MHC-restricted.
When the difference in responsiveness between B6.SJL —> B6.TC and B6.SJL
—> B6.SJL mice was examined it became apparent that the possibility existed for
significant differences in the T cell repertoire. Using bone marrow transfers involving

53
class II deficient mice Glimcher et al (77) showed that it is the radioresistant host thymic
epithelial cells which mediate positive selection. In our model it was possible that lack of
expression of I-As on the radioresistant thymic population eliminated responsive T cells.
The use of (B6.SJL x B6.TC)FX mice as hosts allowed for positive selection on both I-As
and I-Ab and led us to conclude that responsiveness is dependent on the donor haplotype
when T cells for both haplotypes are positively selected.
Mixed chimerism has been shown to prevent autoimmunity in several systems.
Nonobese diabetic mice were protected from diabetes when they were made chimeric with
diabetes resistant bone marrow (78,79). It appeared that the autoimmune potential of the
NOD cells was restrained. However, when the amount of resistant cells was decreased, a
low incidence of insulinitis was seen. Using the rat model of mercury-induced
autoimmunity Delaney et al. (72) showed that the presence of resistant bone marrow cells
(microchimerism) converted an otherwise sensitive rat strain to a resistant one. If similar
to the rat model, it would be expected that the presence of immune competent cells of
resistant origin would prevent Bó.SJL-derived B cells from responding to HgCl2 with an
antifibrillarin response. Our experiments differed somewhat in that we used a resistant
(B6.SJL x B6.TC)FX mouse which had been coinfused with a combination of resistant and
sensitive bone marrow. However, B6.SJL alone as the bone marrow donor completely
restored the ability to produce ANoAs upon treatment with HgCl2. Despite containing a
significant amount of resistant cells we still obtained an intermediate response.
The presence of autoantibodies could play a role in spreading autoimmunity from a
dominant epitope to previously cryptic epitopes (80). Presumably, binding of antibody
can alter antigen processing, revealing new epitopes. Several other investigators have

54
shown that is it possible to break T cell tolerance to self antigens by coimmunizing mice
with self and foreign antigens which in turn generates cross-reactive B cells that can elicit
an autoimmune T cell response to previously cryptic self determinants on the autoantigen
(73,74,81,82). These forms of epitope spreading do not appear to be occurring in our
model. The production of autoantibodies by the B6.SJL B cells in the B6.SJL + (B6.SJL
x B6.TQF! —> (B6.SJL x Bó.TQFj and the B6.SJL + B6.TC —> (B6.SJL x Bó.TQFj
mice did not result in the loss of tolerance to fibrillarin by previously resistant B cells.
Several other differences exists between our model and that used by Delaney (72).
To induce chimerism without myeloablation the rats were transiently treated with the
immunosuppressive agent FK506. Some protection from the manifestations of HgCl2-
induced autoimmune disease was seen in the rats receiving FK506 alone. The authors
acknowledged that the results do suggest that transient immunosuppression was an
important component of the protection. We used a complete myeloablation procedure
which allowed us to forego any immunosuppressive treatment. Another important
difference lies in the fact that in the rat model regulatory T cells appear which render the
rats resistant to additional HgCl2 injections and can confer resistance to naive rats (83).
Regulatory T cells have not been identified in the mouse model and the animals do not
become resistant to further treatment.
The intermediate response seen in group 3 (Table 3-2) mice was a surprising
finding. It was possible that the ANoA negative mice failed to respond because the
number of potentially responsive B cells of I-As origin had been reduced by the presence
of resistant B cells of I-Ab origin. This mechanism appears unlikely in as much as the B

55
cell composition of the chimeras had no relationship to the development of an ANoA
response. Another antigen non-specific mechanism which could have accounted for the
lack of response involves allotype suppression. For example, in studying chronic graft-vs-
host disease Morris et al. found that in allotype-heterozygous recipients, the
autoantibodies were preferentially made by those host cells that expressed the donor
allotype, whereas those host B cells that expressed nondonor allotype were relatively
suppressed. In allotype-homozygous recipients, the donor cells frequently suppressed the
host allotype completely (75). To minimize the possibility that this phenomenon could be
occurring in our mice we used allotype-heterozygous mice as hosts. Allotypic analysis of
spontaneous as well as antigen specific antibodies demonstrated good participation by
both allotypes. Therefore, allotype suppression was not a factor in our mixed chimeras.
A more likely possibility for the poor responsiveness of haplotype-heterozyous
mice was lack of specific T cell help. By substituting (B6.SJL x Bó.TQFj for B6.TC as
the resistant donor in mixed chimeras also receiving susceptible B6.SJL bone marrow, we
provided a mechanism to verify the presence of I-As-restricted T cell help in individual
chimeric mice. In those mice with a positive ANoA, T cells capable of helping
antifibrillarin-expressing B cells of I-As origin must be present, and these activated T cells,
particularly with their reduced stimulation threshold, should also be capable of interacting
with antifibrillarin-expressing B cells of I-As/b origin. Moreover, by using haplotype-
heterozygous mice as co-donors, the overall expression of the resistant I-Ab haplotype was
reduced in cells of donor origin. Despite this, there was no increase in ANoA response
rate. More surprisingly, in the ANoA positive chimeras, this specificity was limited to the

56
b allotype, indicating that the haplotype-heterozygous B cells failed to participate in this
response. These results strongly suggest that lack of responsiveness is an intrinsic
regulatory property of B cells.
The presence of I-Ab restricted T cell help could have accounted for the
intermediate response seen in the mixed chimeras. The (B6.SJL x Bó.TQFj mice that
served as hosts provided an environment in which T cells would be selected to interact
with I-Ab expressed on the B cells from the B6.TC hosts. By transferring (B6.SJL x
B6.TC)Fj bone marrow into B6.SJL mice, there would be little potential for positive
selection of T cells on I-Ab. Despite the absence of these cells, there was no ANoA
response to HgCl2. This again points to an intrinsic defect in (B6.SJL x B6.TC)F1 B cells.
One possibility for this intrinsic property is the concept of MHC-guided processing
leading to determinant capture (84). This hypothesis states that when an antigen is taken
up by APCs and begins unfolding, different MHC molecules can compete for
determinants. Once it is bound by class II, the antigen is then trimmed down to its final
size while the remainder of the antigen including cryptic epitopes is discarded. An
example of this phenomenon was seen in the autoimmune disease insulin-dependent
diabetes mellitus (85). The response to the subdominant ANOD-restricted determinant of
HEL disappears when NOD mice were made transgenic by introduction of the Ead. The
responsivness was restored when scission of the HEL separated this determinant from its
adjoining, competitively dominant, Ed-restricted determinant. This suggested that the Ed
molecule bound and protected its dominant determinant on a long peptide while captured
neighboring determinants were lost during proteolysis. In our mice, I-Ab could effectively

57
be binding fibrillarin in the (B6.SJL x BóTC^ cells, thereby preventing I-A9 from
presenting fibrillarin and receiving T cell help.
Recently the molecular and antigenic properties of mercury-modified fibrillarin has
been examined. The exposure of fibrillarin both in vivo and in vitro caused a change in its
migration under non-reducing SDS-PAGE and resulted in a loss of reactivity to
autoantibodies. Mutation of the cysteines in fibrillarin resulted in a loss of mercury-
induced modification. The authors concluded that unmodified fibrillarin is the B cell
antigen while the T cell antigen appears to be mercury-modified fibrillarin (38).
Therefore, if our model is correct, the presence of I-Ab affected the processing of Hg-
modified fibrillarin in the context of I-A9 and provide a unique opportunity for testing the
role of antigen competition in an important environmental model of induced
autoimmunity.

CHAPTER 4
DISCUSSION/CONCLUSION
Humans are continually exposed to mercury via various routes resulting in high
levels in skin, nails, hair and the kidneys. Many exposures are a result of mercury being
widely distributed in the environment. The major natural source of mercury is the
degassing of the earths crust, and is estimated to produce between 2700 to 6000 tons per
year. Approximately 10,000 tons of mercury are mined each year. Even industrial
activities not directly employing mercury or mercury products give rise to substantial
amounts. Fossil fuels, for example, may contain up to 1 ppm resulting in 5000 tons per
year being emitted from burning coal, natural gas and from refining petroleum products
(86). Human are also exposed to mercury in the form of pharmaceuticals, cosmetics and
dental amalgams.
On the basis of toxicologic characteristics, there are three forms of mercury;
elemental, inorganic, and organic compounds. Both organic and inorganic forms of
mercury undergo environmental transformation. Metallic mercury may be oxidized to
inorganic divalent mercury, particularly in the presence of organic material. Divalent
mercury may, in turn, be reduced to metallic mercury when reducing conditions are
present. Anaerobic bacteria are capable of methylating the divalent form resulting in
dimethyl mercury. In tissues, methyl mercury undergoes biotransformation to divalent
mercury compounds (86).
58

59
Metallic or elemental mercury volatilizes to mercury vapor at ambient
temperatures, and most human exposure is by inhalation. Acute exposure can result in
corrosive bronchitis and interstitial pneumonia. Long term exposure to mercury vapor
targets the central nervous system producing a triad of signs including; increased
excitability, tremors and gingivitis. Renal effects are sometimes seen with chronic mercury
vapor exposure and may be similar to that which occurs following inorganic mercury
exposure. The brain is also a target organ for methyl mercury. Environmental exposures,
such as eating mercury containing fish can result in neurotoxic effects. These include
parathesia, ataxia, neurasthenia, vision and/or hearing loss, tremors and possibly coma
followed by death (86).
Inorganic mercury salts may be present as divalent (mercuric) or monovalent
(mercurous). Ingestion of high doses of this form of mercury causes corrosive ulceration
and necrosis of the gastrointestinal tract followed by circulatory collapse. If this initial
insult is survived, renal failure occurs within 24 hours due to necrosis of the proximal
tubular epithelium.
Chronic low-dose exposure to inorganic mercury can cause immunological
aberrations in both humans and rabbits (27). Immunosuppressive effects have been
reported in mice resulting in impaired host resistance (87). The immunoactivating
properties of inorganic mercury in mice have been widely studied and can be divided into
three major pathological sequelae: lymphoproliferation, hypergammaglobulinemia and the
development of autoimmunity (27). Lymphoproliferation is classified as an MHC-
independent effect while autoimmunity development is tightly restricted by haplotype.

60
Currently, the link between MHC haplotype and susceptibility to HgCl2 induced
ANoAs has remained unexplained. The experiments conducted for this dissertation were
begun with the idea of examining just how the MHC haplotype determines the response to
HgCl2. Through the use of H-2-recombinants it has been shown that responsiveness could
be mapped to the I-A region of the MHC (6). Mice of the H-2S haplotype are high
responders while those of the H-2b haplotype are resistant (5). In an Fj cross between two
different haplotypes, the products of both alleles are expressed on the same cell, so
expression is said to be co-dominant. Therefore, an Fj (s/b) mouse should be susceptible
to ANoA induction. This result has been reported in the literature although the response
was somewhat attenuated (5,9,43). When we generated s/b haplotype mice using H-2-
congenic founder strains, we surprisingly found that the mice were resistant. This led us
to examine how the resistant haplotype was exerting its effects.
It was possible that a gene linked to the resistant haplotype and not the I-A
molecule itself that could be preventing the response. To this end we generated Fx mice
between resistant and susceptible strains in which the resistant class II genes were
eliminated. These mice did develop ANoAs which showed that the downregulation was a
result of the resistant allele itself and not a linked gene.
We also directly examined what effects, if any, that the other class II molecule, I-
E, had on susceptibility. Previous studies in the murine mercury model have shown
expression to either suppress (6) or have no effect (88) on susceptibility. The protection
afforded by I-E is not without precedent. Expression of I-E in lupus prone B6.1pr mice
significantly reduced the production of autoantibodies as well as lowering the spleen and

61
lymph node weights (48). Using bone marrow chimeras in which I-E-positive and I-E-
negative B cells co-existed, it was determined that the amelioration of signs resulted from
a direct effect of I-E on the B cell This phenomenon may be a universal feature of
autoimmune disease on the H-2B background. In an unrelated autoimmune mouse model,
BXSB, which also bears the H2b haplotype, expression of I-E prevented the development
of SLE (47), and a potential mechanism was proposed. It was shown that the a chain of
the I-E molecule generates a peptide with high affinity to I-Ab, thereby competing with
the pathogenic autoantigen-derived peptides for presentation to B cells. Interestingly, if
this same mechanism were important in the HgCl2 model, we would predict just the
opposite results, in our (B6.SJL X Bó.I-E^ mice. Since the that lack of response in s/b
mice was due to expression of I-Ab, the I-E a-derived peptide could have preferentially
blocked I-Ab and enhanced the I-As-mediated ANoA response. However, we found no
enhancing effect on autoimmunity by co-expression of I-E on F: mice. One possibility for
this lack of effect was that an I-E-derived peptide might also have an equal or even greater
affinity for I-A8. Against this possibility was our results that co-expression of I-E on the
H-28 homozygous background did not attenuate the autoimmune response, as would be
expected based on the experience with the BXSB and lpr models of autoimmunity.
Therefore, the lack of autoantibody production in these Fj mice led us to investigate other
mechanisms whereby I-Ab could be down regulating the response.
For these studies, we turned to bone marrow chimeras. Adoptive bone marrow
transfer is an important tool that can be utlitized to dissect the cellular interactions in an
adaptive immune response by manipulating the environment in which T cell development
occurs thereby modifying the T cell repertoire. By taking advantage of the differences in

62
positive and negative selection seen between host and donor cells we showed that MHC
restricted T cell help was required to produce ANoAs. This result was not surprising but
was instrumental in interpreting future results. The next possibility we examined was
whether or not the differences in T cell repertoire could account for the differences in
susceptibility. By using an Fx cross between the two donor strains we eliminated the T cell
differences yet still saw differences in responsiveness.
The intermediate response seen when resistant and sensitive bone marrow was
coinfused into a haplotype heterozygous host was surprising. Several possibilities could
have accounted for this. The first possibility, dilution of responsive B cells, was ruled out
by noting a lack of any correlation between B cell composition and responsiveness.
Secondly, we needed to rule out that allotype suppression in which autoantibodies are
preferentially produced by those B cells expressing the same haplotype as the host was
occurring in these chimeras. We attempted to eliminate this possibility by using Fx hosts.
We also verified that allotype suppression was not an important mechanism in the
chimeras by observing a brisk response by both allotypes when immunized with an
exogenous antigen. The final possibility we examined that may have led to the
intermediate result involved the question of T cell help. By using Fx rather than B6 mice
as the resistant co-donor and observing an anti-nucleolar response by the B cells of donor
B6.SJL origin, we were assured that activated T cells capable of providing I-As-restricted
help to antifibrillarin B cells were present. Despite the presence of adequate T cell help
that should have been able to interact with (s/b)Fx B cells, no increase in susceptibility was
observed. The fact that these mice produced ANoA only of the b allotype combined with

63
the observation that no response was seen when I-Ab restricted T cells were eliminated led
us to conclude that the resistance of the Ft B cells was due to an intrinsic defect of the B
cells themselves.
Several theories have been proposed as to how mercury can lead to autoimmunity.
It has been suggested that mercury may alter self antigen by complexing with the antigen
itself. As mentioned above, divalent mercury is a highly reactive molecule with a
propensity to bind sulfhydryl groups of proteins and nonprotein thiols. Reactivity has also
been found towards hydroxyl, carboxyl and phosphoryl groups albeit at a lower affinity
(89). These modified self antigens may then be recognized by the immune system as
foreign and in turn induce an immune response capable of recognizing the modified self
antigen.
During T cell development, self tolerance is only induced to efficiently-presented
dominant epitopes but not to cryptic ones (90). Therefore, potentially autoreactive T cells
which recognize these cryptic epitopes that would arise by random rearrangement of the T
cell receptor would not be eliminated by negative selection. The dominant epitope
presented is influenced by a number of factors and is incompletely understood. Factors
that have been described include the haplotype, protein structure and/or conformation.
Changes in any of these factors could theoretically result in novel cleavage products. This
phenomenon has been observed for many of the autoantigens recognized by SLE patients.
Casciola-Rosen et al. (91) showed that during apoptosis the lupus autoantigens cluster
and are concentrated in the surface blebs of apoptotic cells. This could result in the
autoantigens becoming substrates for enzymes not present within the cell resulting in new

64
cleavage products. A follow-up study by the same group (90) showed that several of the
autoantigens targeted in scleroderma are susceptible to reactive oxygen species in a metal-
dependent manner. This cleavage generates unique epitopes resulting in autoantibody
formation. They propose that the repetitive episodes of ischemia-reperfusion, which
occurs in these patients as a result of vasomotor instability of their arterioles, are
responsible for the generation of the reactive oxygen species. In contrast, fibrillarin was
not fragmented under identical conditions, thereby suggesting an alternative mechanism
was responsible for development of this specificity in a subset of scleroderma patients.
Pollard et al. (38) proposed a mechanism whereby mercury can induce
antifibrillarin autoantibodies. They showed that mercury-induced cell death was
associated with a loss of fibrillarin antigenicity and modification of the molecular
properties of fibrillarin. By mutating the cysteines in fibrillarin to alanines they confirmed
that HgCl2 exposure in vitro leads to a disulfide bonded form of fibrillarin which is poorly
recognized by autoantibodies using either indirect immunofluorescence or
immunoprecipitation. This suggests that the mercury-induced change involves a
conformational structure change that is no longer detected by autoantibodies.
The above mechanisms do not take into account the tight link between
susceptibility to autoimmunity and the mouse MHC. We propose a model which
elaborates on both the modified self-antigen theory and takes into account the lack of
response seen in haplotype-heterozygous mice.
During T cell development, fibrillarin is processed and the immunodominant
peptide(s) is presented. T cells which recognize this peptide are deleted resulting in

65
tolerance to fibrillarin. On exposure to HgCl2, a disulfide bond forms which affects the
immunodominant peptide normally presented by I-As, resulting in the presentation of a
cryptic epitope. This cryptic peptide could potentially take a number of forms. The
production of a mercury-modified disulfide link could protect the area around that peptide
from being digested and the resultant cryptic peptide would be a normal sequence derived
from another portion of the fibrillarin molecule. This would be a remote effect. It could
also result from a conformational change that exposes a buried sequence to digestion.
Mercury-modified disulfide could provide a completely novel peptide sequence centered
around the cysteine-Hg-cysteine sequence. Arguing against this possibility is the fact that
no mercury was detected in the glomerular basement membrane in HgCl2-treated rabbits
using autoradiography (92). Another possibility is that the disulfide residue could be re¬
reduced during processing but one of the amino acids in the binding groove has Hg-linked
side chain that alters the immunogeneicity. The net result, then, would be the expression
of a self-epitope not previously encountered. The immune system would recognize this as
foreign, resulting in autoimmunity. In the case of a b haplotype mouse, the disulfide bond
doesn’t interfere with the processing and presentation of its immunodominant peptide. No
cryptic epitopes are revealed and hence no autoimmunity to fibrillarin develops.
Our experiments with haplotype heterozygous mice adds some interesting potential
insights into antigen processing and the expression of cryptic epitopes. We propose that
the cryptic epitope presented by I-As is either the same or overlaps with the peptide
normally presented by I-Ab. When an s/b haplotype mouse is exposed to mercury, a
competition for access to that part of fibrillarin that could be bound by either of the two

66
class II alleles will result. If the I-Ab molecule has a greater avidity for this peptide than I-
As, no new peptide is seen in the context of I-As and no autoimmunity results.
There is a precedent for the idea that different class II molecules actually compete
for the same or overlapping peptide. Sercarz et al. (84) showed that competition takes
place between I-A and I-E in the mouse. They coined the phrase “MHC guided
processing” to describe the process where the MHC class II molecule would encounter a
tightly folded protein antigen in one of the acidic compartments of the endosomal-
lysosomal system soon after preliminary processing had succeeded in partially unfolding it,
making one or a few determinants available. The dominant determinant would be the one
that initially won the competition for binding to the MHC and provided the binding was
stable, the residues bound within the MHC would be protected from further proteolysis or
binding by a different MHC molecule. Thus far, this has been described in in vitro models.
Our experiments in the mercury chloride-induced model would be the first demonstration
of its potential importance in vivo.
Our investigations of mercury-induced autoimmunity in mice have stimulated the
need for future investigations addressing the many questions spawned by our initial
research into this area. Several major questions need to be addressed to either support or
refute our model. Firstly, if in fact a competition is ongoing between MHC molecules in
haplotype heterozygous mice, will increasing the dose of mercury which in turn should
provide a greater pool of modified fibrillarin result in these animals now becoming
susceptible? This is a fairly straightforward question to address. The second and perhaps
more fundamental question to be addressed is the relationship between the T cell and B

67
cell epitopes. As previously discussed, the T cell epitope has been proposed to be
mercury-modified fibrillarin, while the B cell epitope is a conformational determinant lost
by exposure to mercury (38). This absence of an antibody response to mercury-modified
fibrillarin is a surprising one. One possible explanation is that binding of the mercury-
modified epitope by surface Ig protects that portion of the molecule from antigen
processing. The net result is that no mercury-induced cryptic determinants are presented
by those B cells with specificity for the mercury modification. While reasonable, this
raises yet another issue: how then is it possible for a specific immune response to develop
if the B cells that can bind to native fibrillarin do not have a mercury-modified T-cell
epitope to present? One possibility is that the mercury either directly or indirectly causes
an aggregation of fibrillarin in which some molecules are mercury-modified and some are
still in the native conformation. When a B cell internalizes this aggregate specifically via
its surface Ig receptor, which recognizes native fribrillarin, it processes the mercury
modified sites within the aggregate and presents it to T cells, which recognize this
modified form as foreign. While this hypothesis readily explains the potential for T-B
interactions in the generation of a Hg-induced response to native fibrillarin it re-opens the
question as to why there is no response to Hg-modified fibrillarin. It would seem unlikely
that surface Ig could bind every Hg-modified residue in an aggregate. Failure to do so
would allow the B cell specific for Hg-modified fibrillarin to present the cryptic epitope
and thereby receive T cell help .
We have shown that it is the I-Ab molecule that confers resistance to HgCl2-
induced autoimmunity in F! (s/b) mice. Through the use of bone marrow chimeric animals

68
we determined that this effect is not mediated by a change in the T-cell repertoire. It
appears that the MHC may be exerting its effects by manipulating the antigen
processing/presentation pathway. To address this, one would need to elute and sequence
the peptides that are actually being presented on the surface of the antigen presentation
cells. Similarly, to address the apparent paradox between the T cell and B cell epitopes
one would need to obtain the same information for both the T cell and the B cell
receptors.

LIST OF REFERENCES
1. Theofilopoulos, A. N. 1993. Molecular Pathology of Autoimmunity. In The Molecular
Pathology of Autoimmune Diseases. C.A. Bona, K.A. Siminovitch, M. Zanetti and A.N.
Theofilopoulos, eds. Harwood Academic Publishers, Chur, Switzerland, p. 1.
2. Condemi, J. J. 1992. The autoimmune diseases. JAMA 268:2882.
3. Tubbs, R. R., G. N. Gephardt, J. T. McMahon, M. C. Pohl, D. G. Vidt, S. A.
Barenberg, and R. Valenzuela. 1982. Membranous glomerulonephritis associated with
industrial mercury exposure. Study of pathogenetic mechanisms. Am. J. Clin. Pathol.
77:409.
4. Robinson, C. J., A. A. Abraham, and T. Balazs. 1984. Induction of anti-nuclear
antibodies by mercuric chloride in mice. Clin. Exp. Immunol. 55:300.
5. Robinson, C. J., T. Balazs, and I. K. Egorov. 1986. Mercuric chloride-, gold sodium
thiomalate-, and D-penicillamine-induced antinuclear antibodies in mice. Toxicol. Appl.
Pharmacol. 86:159.
6. Mirtcheva, J., C. Pfeiffer, J. A. De Bruijn, F. Jacquesmart, and E. Gleichmann. 1989.
Immunological alterations inducible by mercury compounds. III. H-2A acts as an immune
response and H-2E as an immune "suppression" locus for HgC12-induced antinucleolar
autoantibodies. Eur. J. Immunol. 19:2251.
7. Hultman, P., L. J. Bell, S. Enestrom, and K. M. Pollard. 1992. Murine susceptibility to
mercury. I. Autoantibody profiles and systemic immune deposits in inbred, congenie, and
intra-H-2 recombinant strains. Clin. Immunol. Immunopathol. 65:98.
8. Hanley, G. A., J. Schiffenbauer, and E. S. Sobel. 1997. Class II haplotype differentially
regulates immune response in HgC12-treated mice. Clin. Immunol. Immunopathol.
84:328.
9. Hultman, P., L. J. Bell, S. Enestrom, and K. M. Pollard. 1993. Murine susceptibility to
mercury. II. autoantibody profiles and renal immune deposits in hybrid, backcross, and
H-2d congenie mice. Clin. Immunol. Immunopathol. 68:9.
10. Hultman, P. and S. Enestrom. 1988. Mercury induced antinuclear antibodies in mice:
characterization and correlation with renal immune complex deposits. Clin. Exp. Immunol.
71:269.
69

70
11. Robinson, C. J., H. J. White, and N. R. Rose. 1997. Murine strain differences in
response to mercuric chloride: antinucleolar antibodies production does not correlate with
renal immune complex deposition. Clin. Immunol. Immunopathol. 53:127.
12. Hultman, P., S. J. Turley, S. Enestrom, U. Lindh, and K. M. Pollard. 1996. Murine
Genotype Influences the Specificity, Magnitude and Persistence of Murine
Mercury-Induced Autoimmunity. J. Autoimmunity 9:139.
13. Griem, P., E. Scholz, M. Turfeld, D. Zander, U. Wiesner, L. Dunemann, and E.
Gleichmann. 1997. Strain differences in tissue concentrations of mercury in inbred mice
treated with mercuric chloride. Toxicol. Appl. Pharmacol. 144:163.
14. Christensen, M. M. 1996. Histochemical localization of autometallographically
detectable mercury in tissues of the immune system from mice exposed to mercuric
chloride. Histochem. J. 25:217.
15. Hultman, P., U. Johansson, and F. Dagnaes-Hansen. 1995. Murine Mercury-Induced
Autoimmunity: the Role of T-Helper Cells. J. Autoimmunity 5:809.
16. Johansson, U., B. Sander, and P. Hultman. 1997. Effects of the murine genotype on T
cell activation and cytokine production in murine mercury-induced autoimmunity. J.
Autoimmun. 10:347.
17. van Vliet, E., M. Uhrberg, C. Stein, and E. Gleichmann. 1993. MHC control of
IL-4-dependent enhancement of B cell la expression and Ig class switching in mice treated
with mercuric chloride. Int. Arch. Allergy Immunol. 101:392.
18. Hu, H., M. Abedi Valugerdi, and G. Moller. 1997. Pretreatment of lymphocytes with
mercury in vitro induces a response in T cells from genetically determined low-responders
and a shift of the interleukin profile. Immunology 90:198.
19. Jiang, Y. and G. Moller. 1996. In vitro effects of HgC12 on murine lymphocytes. II.
Selective activation of T cells expressing certain V beta TCR. Int. Immunol. 5:1729.
20. Gillespie, K. M., F. J. Qasim, L. M. Tibbatts, S. Thiru, D. B. Oliveira, and P. W.
Mathieson. 1995. Interleukin-4 gene expression in mercury-induced autoimmunity. Scand.
J. Immunol. 41:268.
21. Prigent, P., A. Saoudi, C. Pannetier, P. Graber, J. Y. Bonnefoy, P. Druet, and F.
Hirsch. 1995. Mercuric chloride, a chemical responsible for T helper cell (Th)2-mediated
autoimmunity in brown Norway rats, directly triggers T cells to produce interleukin-4. J.
Clin. Invest. 96:1484.

71
22. Goldman, M., P. Druet, and E. Gleichmann. 1991. TH2 cells in systemic
autoimmunity: insights from allogeneic diseases and chemically-induced autoimmunity.
Immunol. Today 72:223.
23. Pietsch, P., H. W. Vohr, K. Degitz, and E. Gleichmann. 1989. Immunological
alterations inducible by mercury compounds. II. HgC12 and gold sodium thiomalate
enhance serum IgE and IgG concentrations in susceptible mouse strains. Int. Arch. Allergy
Appl. Immunol. 90 A7.
24. Lundgren, M., U. Persson, P. Larsson, C. Magnusson, C. I. Smith, L. Hammarstrom,
and E. Severinson. 1989. Interleukin 4 induces synthesis of IgE and IgG4 in human B
cells. Eur. J. Immunol. 79:1311.
25. Ochel, M., H. W. Vohr, C. Pfeiffer, and E. Gleichmann. 1991. IL-4 is required for the
IgE and IgGl increase and IgGl autoantibody formation in mice treated with mercuric
chloride. J. Immunol. 746:3006.
26. Doth, M., M. Fricke, F. Nicoletti, G. Garotta, M. L. Van Velthuysen, J. A. Bruijn, and
E. Gleichmann. 1997. Genetic differences in immune reactivity to mercuric chloride
(HgC12): immunosuppression of H-2d mice is mediated by interferon-gamma
(IFN-gamma). Clin. Exp. Immunol. 709:149.
27. Pollard, K. M. 1997. Induction of Autoimmunity by Mercuric Chloride. Fifth
International Workshop on the Molecular and Cell Biology of Autoantibodies and
Autoimmunity 24.(Abstract)
28. Hultman, P. and S. Enestrom. 1992. Dose-response studies in murine mercury-induced
autoimmunity and immune-complex disease. Toxicol. Appl. Pharmacol. 773:199.
29. Hultman, P. and S. Enestrom. 1989. Mercury induced B-cell activation and antinuclear
antibodies in mice. J. Clin. Lab. Immunol. 28:143.
30. Hu, H., G. Moller, M. Abedi Valugerdi, U. Johansson, B. Sander, and P. Hultman.
1997. Major histocompatibility complex class II antigens are required for both cytokine
production and proliferation induced by mercuric chloride in vitro Effects of the murine
genotype on T cell activation and cytokine production in murine mercury-induced
autoimmunity. J. Autoimmun. 70:347.
31. Reuter, R., G. Tessars, H. W. Vohr, E. Gleichmann, and R. Luhrmann. 1989.
Mercuric chloride induces autoantibodies against U3 small nuclear ribonucleoprotein in
susceptible mice. Proc. Natl. Acad. Sci. U. S. A. 86:237.

72
32. Hultman, P., S. Enestrom, K. M. Pollard, and E. M. Tan. 1989. Anti-fibrillarin
autoantibodies in mercury-treated mice. Clin. Exp. Immunol. 75:470.
33. Okano, Y., V. D. Steen, and T. A. Medsger. 1992. Autoantibody to U3 nucleolar
ribonucleoprotein (fibrillarin) in patients with systemic sclerosis. Arthritis Rheum. 35:95.
34. Kasturi, K. N., A. Hatakeyama, H. Spiera, and C. A. Bona. 1995. Antifibrillarin
autoantibodies present in systemic sclerosis and other connective tissue diseases interact
with similar epitopes. J. Exp. Med. 757:1027.
35. Takeuchi, K., S. J. Turley, E. M. Tan, and K. M. Pollard. 1995. Analysis of the
autoantibody response to fibrillarin in human disease and murine models of autoimmunity.
J. Immunol. 154:961.
36. Kubicka Muranyi, M., J. Kremer, N. Rottmann, B. Lubben, R. Albers, N. Bloksma, R.
Luhrmann, and E. Gleichmann. 1996. Murine systemic autoimmune disease induced by
mercuric chloride: T helper cells reacting to self proteins. Int. Arch. Allergy Immunol.
709:11.
37. Kubicka Muranyi, M., P. Griem, B. Lubben, N. Rottmann, R. Luhrmann, and E.
Gleichmann. 1995. Mercuric-chloride-induced autoimmunity in mice involves up-regulated
presentation by spleen cells of altered and unaltered nucleolar self antigen. Int. Arch.
Allergy Immunol. 705:1.
38. Pollard, K. M., D. K. Lee, C. A. Casiano, M. Bluthner, M. M. Johnston, and E. M.
Tan. 1997. The autoimmunity-inducing xenobiotic mercury interacts with the autoantigen
fibrillarin and modifies its molecular and antigenic properties. J. Immunol. 755:3521.
39. Kass, S., K. Tyc, J. A. Steitz, and B. Sollner Webb. 1990. The U3 small nucleolar
ribonucleoprotein functions in the first step of preribosomal RNA processing. Cell 60:897.
40. Lubben, B., N. Rottmann, M. Kubicka Muranyi, E. Gleichmann, and R. Luhrmann.
1994. The specificity of disease-associated anti-fibrillarin autoantibodies compared with
that of HgC12-induced autoantibodies. Mol. Biol. Rep. 20:63.
41. Bencko, V., V. Wagner, M. Wagnerova, and V. Ondrejcak. 1990. Immunological
profiles in workers occupationally exposed to inorganic mercury. J. Hyg. Epidemiol.
Microbiol. Immunol. 34:9.
42. Tubbs, R. R., G. N. Gephardt, J. T. McMahon, M. C. Pohl, D. G. Vidt, S. A.
Barenberg, and R. Valenzuela. 1982. Membranous glomerulonephritis associated with
industrial mercury exposure. Study of pathogenetic mechanisms. Am. J. Clin. Pathol.
77:409.

73
43. Stiller Winkler, R., T. Radaszkiewicz, and E. Gleichmann. 1988. Immunopathological
signs in mice treated with mercury compounds--! Identification by the popliteal lymph
node assay of responder and nonresponder strains. Int. J. Immunopharmacol. 70:475.
44. Walter, W., M. Loos, and M. J. Maeurer. 1996. H2-M polymorphism in mice
susceptible to collagen-induced arthritis involves the peptide binding groove.
Immunogenetics 44: 19.
45. Widera, G., L. C. Burkly, C. A. Pinkert, E. C. Bottger, C. Cowing, R. D. Palmiter, R.
L. Brinster, and R. A. Flavell. 1987. Transgenic mice selectively lacking MHC class II
(I-E) antigen expression on B cells: an in vivo approach to investigate la gene function.
Cell 57:175.
46. Grusby, M. J., R. S. Johnson, V. E. Papaioannou, and L. H. Glimcher. 1991.
Depletion of CD4+ T cells in major histocompatibility complex class II-deficient mice.
Science 253:1417.
47. Merino, R., M. Iwamoto, L. Fossati, P. Muniesa, K. Araki, S. Takahashi, J. Huarte, K.
Yamamura, J. D. Vassalli, and S. Izui. 1993. Prevention of systemic lupus erythematosus
in autoimmune BXSB mice by a transgene encoding I-E alpha chain. J. Exp. Med.
775:1189.
48. Creech, E. A., D. Nakul Aquaronne, E. A. Reap, R. L. Cheek, P. A. Wolthusen, P. L.
Cohen, and R. A. Eisenberg. 1996. MHC genes modify systemic autoimmune disease. The
role of the I-E locus. J. Immunol. 756:812.
49. Kappler, J. W., B. Skidmore, J. White, and P. Marrack. 1981. Antigen-inducible,
H-2-restricted, interleukin-2-producing T cell hybridomas. Lack of independent antigen
and H-2 recognition. J. Exp. Med. 153:1198.
50. Ozato, K., N. Mayer, and D. H. Sachs. 1980. Hybridoma cell lines secreting
monoclonal antibodies to mouse H-2 and la antigens. J. Immunol. 124:533.
51. Morris, S. C., R. L. Cheek, P. L. Cohen, and R. A. Eisenberg. 1990. Autoantibodies in
chronic graft versus host result from cognate T-B interactions. J. Exp. Med. 777:503.
52. Aris, J. P. and G. Blobel. 1988. Identification and characterization of a yeast nucleolar
protein that is similar to a rat liver nucleolar protein. J. Cell Biol. 107:11.
53. O'Shannessy, D. J., M. J. Dobersen, and R. H. Quarles. 1984. A novel procedure for
labeling immunoglobulins by conjugation to oligosaccharide moieties. Immunol. Lett.
8:213.

74
54. Goding, J. W. 1976. Conjugation of antibodies with fluorochromes: modifications to
the standard methods. J. Immunol. Methods 73:215.
55. Ceman, S., R. A. Rudersdorf, J. M. Petersen, and R. DeMars. 1995. DMA and DMB
are the only genes in the class II region of the human MHC needed for class II-associated
antigen processing. J. Immunol. 154:2545.
56. Riberdy, J. M. and P. Cresswell. 1992. The antigen-processing mutant T2 suggests a
role for MHC-linked genes in class II antigen presentation [published erratum appears in J
Immunol 1992 Aug 1;149(3): 1113]. J. Immunol. 745:2586.
57. Hill, S. W., D. E. Kipp, I. Melchers, J. A. Frelinger, and E. E. Sercarz. 1980. Multiple
H-2 and non-H-2 genes controlling the antilysozyme response: alternative gene
constellations can lead to responsiveness. Eur. J. Immunol. 70:384.
58. Martin, J., B. Y. Wei, and C. S. David. 1993. Expression of MHC class II A alpha k
transgene. Pairing with endogenous beta-chain is dependent upon affinity and competition.
Transplantation 56:1233.
59. Layet, C. and R. N. Germain. 1991. Invariant chain promotes egress of poorly
expressed, haplotype-mismatched class II major histocompatibility complex A alpha A
beta dimers from the endoplasmic reticulum/cis-Golgi compartment. Proc. Natl. Acad.
Sci. U. S. A. 88:2346.
60. Mathis, D. J., C. Benoist, V. E. Williams, M. Kanter, and H. O. McDevitt. 1983.
Several mechanisms can account for defective E alpha gene expression in different mouse
haplotypes. Proc. Natl. Acad. Sci. U. S. A. 80:273.
61. Nygard, N. R., D. M. McCarthy, J. Schiffenbauer, and B. D. Schwartz. 1993. Mixed
haplotypes and autoimmunity. Immunol. Today 14:53.
62. Hirose, S., G. Ueda, K. Noguchi, T. Okada, I. Sekigawa, H. Sato, and T. Shirai. 1986.
Requirement of H-2 heterozygosity for autoimmunity in (NZB XNZW)F1 hybrid mice.
Eur. J. Immunol. 76:1631.
63. Sobel, E. S., T. Katagiri, K. Katagiri, S. C. Morris, P. L. Cohen, and R. A. Eisenberg.
1991. An intrinsic B cell defect is required for the production of autoantibodies in the lpr
model of murine systemic autoimmunity. J. Exp. Med. 773:1441.
64. Ceredig, R., J. W. Lowenthal, M. Nabholz, and H. R. MacDonald. 1985. Expression
of interleukin-2 receptors as a differentiation marker on intrathymic stem cells. Nature
374:98.

75
65. Sarmiento, M., A. L. Glasebrook, and F. W. Fitch. 1980. IgG or IgM monoclonal
antibodies reactive with different determinants on the molecular complex bearing Lyt 2
antigen block T cell-mediated cytolysis in the absence of complement. J. Immunol.
125-2665.
66. Kremmer, E., S. Thierfelder, U. Kummer, R. Lederer, and J. Mysliwietz. 1989.
Neutralization of immunosuppression by antibodies against variable as well as constant
regions of monoclonal anti-Thy-1 xenoantibodies and their ability to be suppressed by
initial T cell depletion. Transplantation 47:641.
67. Oi, V. T., P. P. Jones, J. W. Goding, and L. A. Herzenberg. 1978. Properties of
monoclonal antibodies to mouse Ig allotypes, H-2, and la antigens. Curr. Top. Microbiol.
Immunol. 57:115.
68. Stall, A. M. and M. R. Loken. 1984. Allotypic specificities of murine IgD and IgM
recognized by monoclonal antibodies. J. Immunol. 132:187.
69. Sieckmann, D. G, A. M. Stall, and B. Subbarao. 1991. A mouse monoclonal antibody
specific for an allotypic determinant of the Igha allele of murine IgM: genetic and
functional analysis of Igh-6a epitopes using anti-IgM monoclonal antibodies. Hybridoma
70:121.
70. Cosgrove, D., S. H. Chan, C. Waltzinger, C. Benoist, and D. Mathis. 1992. The
thymic compartment responsible for positive selection of CD4+ T cells. Int. Immunol.
4:101.
71. Sprent, J., H. Kosaka, and E. K. Gao. 1992. T cell tolerance after bone marrow
transplantation in mice. Bone Marrow Transplant. 10 Suppl 7:5.
72. Delaney, C. P., N. Murase, M. Chen Woan, J. J. Fung, T. E. Starzl, and A. J.
Demetris. 1996. Allogeneic hematolymphoid microchimerism and prevention of
autoimmune disease in the rat. A relationship between alio- and autoimmunity. J. Clin.
Invest. 97:217.
73. Mamula, M. J., S. Fatenejad, and J. Craft. 1994. B cells process and present lupus
autoantigens that initiate autoimmune T cell responses. J. Immunol. 752:1453.
74. Mamula, M. J., R. H. Lin, C. A. Janeway,Jr., and J. A. Hardin. 1992. Breaking T cell
tolerance with foreign and self co-immunogens. A study of autoimmune B and T cell
epitopes of cytochrome c. J. Immunol. 749:789.
75. Morris, S. C., R. L. Cheek, P. L. Cohen, and R. A. Eisenberg. 1990. Allotype-specific
immunoregulation of autoantibody production by host B cells in chronic graft-versus host
disease. J. Immunol. 144:916.

76
76. Sobel, E. S., V. N. Kakkanaiah, M. Kakkanaiah, R. L. Cheek, P. L. Cohen, and R. A.
Eisenberg. 1994. T-B collaboration for autoantibody production in lpr mice is cognate and
MHC-restricted. J. Immunol. 152:6Oil.
77. Markowitz, J. S., H. Auchincloss,Jr., M. J. Grusby, and L. H. Glimcher. 1993. Class
II-positive hematopoietic cells cannot mediate positive selection of CD4+ T lymphocytes
in class II-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 90:2119.
78. Li, H., C. L. Kaufman, S. S. Boggs, P. C. Johnson, K. D. Patrene, and S. T. Ildstad.
1996. Mixed allogeneic chimerism induced by a sublethal approach prevents autoimmune
diabetes and reverses insulitis in nonobese diabetic (NOD) mice. J. Immunol. 756:380.
79. Mathieu, C., K. Casteels, R. Bouillon, and M. Waer. 1997. Protection against
autoimmune diabetes in mixed bone marrow chimeras: mechanisms involved. J. Immunol.
755:1453.
80. Satoh, M., A. K. Ajmani, L. Stojanov, J. J. Langdon, T. Ogasawara, J. Wang, M. A.
Dooley, H. B. Richards, J. B. Winfield, T. H. Carter, and W. H. Reeves. 1996.
Autoantibodies that stabilize the molecular interaction of Ku antigen with DNA-dependent
protein kinase catalytic subunit. Clin. Exp. Immunol. 705:460.
81. Lin, R. H., M. J. Mamula, J. A. Hardin, and C. A. Janeway,Jr. 1991. Induction of
autoreactive B cells allows priming of autoreactive T cells. J. Exp. Med. 773:1433.
82. Fatenejad, S., M. J. Mamula, and J. Craft. 1993. Role of intermolecular/intrastructural
B- and T-cell determinants in the diversification of autoantibodies to ribonucleoprotein
particles. Proc. Natl. Acad. Sci. U. S. A. 90:12010.
83. Bowman, C., D. W. Mason, C. D. Pusey, and C. M. Lockwood. 1984. Autoregulation
of autoantibody synthesis in mercuric chloride nephritis in the Brown Norway rat. I. A
role for T suppressor cells. Eur. J. Immunol. 77:464.
84. Sercarz, E. E., P. V. Lehmann, A. Ametani, G. Benichou, A. Miller, and K. Moudgil.
1993. Dominance and crypticity of T cell antigenic determinants. Annu. Rev. Immunol.
77:729.
85. Deng, H., R. Apple, M. Clare Salzler, S. Trembleau, D. Mathis, L. Adorini, and E.
Sercarz. 1993. Determinant capture as a possible mechanism of protection afforded by
major histocompatibility complex class II molecules in autoimmune disease. J. Exp. Med.
775:1675.
86. Goyer, R. A. 1996. Toxic Effects of Metals. In Casarett and Doull's Toxicology. Fifth
ed. C.D. Klaassen, ed. McGraw-Hill, New York, p. 691.

77
87. Ellermann Eriksen, S., M. M. Christensen, and S. C. Mogensen. 1994. Effect of
mercuric chloride on macrophage-mediated resistance mechanisms against infection with
herpes simplex virus type 2. Toxicology 93.269.
88. Vladutiu, A. O. 1991. Treatment of autoimmune diseases with antibodies to class II
major histocompatibility complex antigens. Clin. Immunol. Immunopathol. 6T.\.
89. Utschig, L. M., J. G. Wright, and T. V. O'Halloran. 1993. Biochemical and
spectroscopic probes of mercury(II) coordination environments in proteins. Methods
Enzymol. 226:71.
90. Casciola Rosen, L., F. Wigley, and A. Rosen. 1997. Scleroderma autoantigens are
uniquely fragmented by metal-catalyzed oxidation reactions: implications for pathogenesis.
J. Exp. Med. 185:71.
91. Casciola Rosen, L. A., G. Anhalt, and A. Rosen. 1994. Autoantigens targeted in
systemic lupus erythematosus are clustered in two populations of surface structures on
apoptotic keratinocytes [see comments], J. Exp. Med. 779:1317.
92. Druet, P., A. Bernard, F. Hirsch, J. J. Weening, P. Gengoux, P. Mahieu, and S.
Birkeland. 1982. Immunologically mediated glomerulonephritis induced by heavy metals.
Arch. Toxicol. 50:187.

BIOGRAPHICAL SKETCH
Gregory Alan Hanley was bom in Batavia, New York, November 24, 1965. He
received a Bachelor of Science degree from Geneseo State, a branch of the State
University of New York. He then attended the University of Florida, College of
Veterinary Medicine and was awarded the Doctor of Veterinary Medicine in 1993. After
graduation Dr. Hanley received an NIH postdoctoral fellowship to pursue advanced
training in the veterinary speciality of laboratory animal medicine and training in a basic
science leading to a Ph.D. Dr. Hanley’s focus during his studies concentrated on the
induction of autoimmunity in mice using mercuric chloride, the results of which are
included in this dissertation.
78

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
kJJ
Stephen M. Roberts, Chair
Associate Professor of Veterinary
Medicine
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Eric S. Sobel, Cochair
Assistant Professor of Medicine
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
)avis
of Veterinary Medicine
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Joel Schiffenbauer s â– 
Associate Professor of Medicine
This dissertation was submitted to the Graduate Faculty of the College of
Veterinary Medicine and to the Graduate School and was accepted as partial fulfillment of
the requirements for the degree of Doctor of Philosophy.
May, 1998
■W-Í2A
Dean, College of Veterinary Medicine
Dean, Graduate School