Elucidation and characterization of a potential cis-acting regulatory element in a murine Mhc class II gene


Material Information

Elucidation and characterization of a potential cis-acting regulatory element in a murine Mhc class II gene
Physical Description:
viii, 128 leaves : ill. ; 29 cm.
McIndoe, Richard Alan, 1962-
Publication Date:


Subjects / Keywords:
Research   ( mesh )
Genes, MHC Class II   ( mesh )
Alleles   ( mesh )
Gene Expression Regulation   ( mesh )
Muridae   ( mesh )
Chromatin -- ultrastructure   ( mesh )
Enhancer Elements (Genetics)   ( mesh )
Molecular Sequence Data   ( mesh )
Base Sequence   ( mesh )
Department of Pathology and Laboratory Medicine thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Pathology and Laboratory Medicine -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1991.
Includes bibliographical references (leaves 115-127).
Statement of Responsibility:
by Richard A. McIndoe.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001664803
oclc - 25668294
notis - AHX6577
System ID:

Full Text







This dissertation is dedicated to my wife Mary and son

Robert for the joy they have brought to my life.


I would like to express my thanks and appreciation to all

the people who have affected my life over the last few years.

First and foremost, I would like to mention my mentor Dr.

Edward K. Wakeland. Without his help and guidance this

dissertation would not have been possible. He has taught me

the value of critical thinking and the scientific method. I

hope our friendship will last into the future.

My appreciation is also extended to Drs. Ammon Peck and

Art Kimura for interesting conversations and a hearty laugh

now and then. My thanks are also extended to Dr. William

Winter for his friendship and thought-provoking questions.

Over the years I have seen quite a few students leave the

laboratory of Dr. Wakeland and hope that I will see them in

the future. My thanks go out to Tom McConnell, Vickie Henson,

and Marge Price-LaFace who made lab work considerably more

interesting during my first two years as a technician. This

dedication would be lacking if I did not mention Roy Tarnuzzer

and Stefen Boehme for their fun camping trips, lunch outings,

and general lab humor (I think the gels are still holding

strong in the darkroom). Of more recent vintage, I would like


to thank C.C. Lu, Ivan Cheng, Ying Ye, Mary Yui and Karen

Wright for their kindness and understanding during those

trying moments. My thanks also go to Wayne Potts, Kasinathan

Muralidharan and Jo Manning for forcing me to look at things

in a different perspective.

Finnaly, I would also like to thank those people in

other laboratories who have been helpful and kind. My thanks

to Lena Dingler and Jane Strandberg for some fun barbecues and

trips. My thanks go to Lee Grimes for his friendship and

unique perspective on the situation.







Major Histocompatibility Complex . .
Genomic Organization of the Mouse Mhc .
Genetic Organization of the I Region .
Genomic Organization of Class II Genes .
Biochemical Properties of Class II
Polypetides . .
Structure of Class II Genes . .
Regulation of Class II Gene Expression .
Biology of Class II Gene Expression .
Analysis of Class II Gene Expression Using
Mutant Cell Lines . .
Motifs and Regions Contributing to Class II
Gene Expression . .
Trans-acting Factors . .
How are Class II Genes Regulated? .


Inbred Mice . .
RNA Isolation . .
Northern Analysis .
Tissues and Cell Lines .
B cell Isolation . .
DNAse I Hypersensitivity Assay .
Probes Used In DH Assay .
DH Assay . .

Band Retardation Assay .. .. .
Primers Used For Synthesis of Probes .
Preparation of Nuclear and Whole Cell
Extracts . .
Preparation of Radiolabeled Probes .
Isolation of Cold Competitors .



. 48
. 48

. 39

. 39
. 39
. 40
. 40
. 41
. 44
. 44
. 44

* *

Band Retardation . 55
DNAse I Footprinting Assay . 55
Preparation of Double Stranded DNA Probes 55
DNAse I Footprinting . 56

4 RESULTS . . 58

Ab Expression Varies Among Different Tissues 58
Expressing and Non-expressing Tissues have
Different Chromatin Structures 61
Polymorphism of DH Site Position Among Different
Haplotypes . . 62
DNAse I Footprinting in Intron 2 . 95

5 DISCUSSION . . 101

Chromatin Structure of Ab . .. 101
Potential cis-Acting Regulatory Element
In the Second Intron of Ab . 107



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



Richard A. McIndoe

August, 1991

Chairperson: Dr. Edward K. Wakeland, Ph.D.
Major Department: Pathology and Laboratory Medicine

The murine Mhc class II genes encode molecules that are

involved in the recognition and presentation of foreign

antigen to regulatory T lymphocytes. A DNAse I

hypersensitivity (DH) assay was used to study the chromatin

structure of a murine Mhc class II gene (Ab) in high (B

cells), low (kidney) and non-expressing (liver and brain)

tissues. Six DH sites were detected in the high and low

expressing tissues, their DH site designation and positions

relative to the +1 codon are as follows: Dl, -3200; D2, -

1500; D3, -400; D4, +790; D5, +1700; and D6, +2200. Non-

expressing tissues had a marked lower sensitivity to DNAse I

and were missing sites D5 and D6. The lack of sensitivity and

absence of DH sites is indicative of inactive chromatin and

may contribute to the non-expressing phenotype observed. The

locations of DH sites in several alleles of Ab were determined


to assess the evolutionary stability of these DH sites in this

highly polymorphic gene. This analysis divided Ab alleles

into two groups with chromatin structures that could be

distinguished on the basis of two polymorphic DH sites unique

to transcriptionally active tissues. These additional DH

sites mapped into a polymorphic retroposon insertion that

distinguished the two groups of alleles. Using B cell nuclear

extracts, a band retardation analysis of the first 600 bp of

intron 2 containing the DH sites (D5,D6) correlating with Ab

transcription revealed two 50 bp segments with unique protein-

DNA interactions. A DNAse I footprinting assay was performed

on these two segments and the protected nucleotides were

determined. Sequence motifs with identity to immunoglobulin

heavy chain enhancer (IgH) elements (gE2 and ME3) were found

within the DNAse I protected regions. These elements have

been shown to be sufficient to confer lymphoid-specific

expression of the IgH enhancer. These data suggest the

presence of an enhancer in the second intron of Ab.



The major histocompatibility complex (Mhc) class II genes

are a cluster of tightly-linked and highly polymorphic loci

that encode molecules involved in the recognition and

presentation of foreign antigen to regulatory T lymphocytes

during the initiation of an immune response (Uhr et al. 1979;

Benacerraf, 1981; Klein 1979). Serologic studies have defined

two isotypic forms of class II molecules in the mouse, the A

and E molecules (Cullen et al. 1976). Both are cell surface

glycoproteins composed of an a and P chain that non-covalently

associate to form a heterodimeric molecule. Class II

expression is primarily found in B cells (constitutively) and

antigen presenting cells induciblee by y-interferon,IFN-y).

The molecular basis for the tissue specific expression

has been under investigation for quite some time. In vitro

studies have defined cis-acting sequences and trans-acting

factors that have an influence on the transcriptional activity

in both human and mouse Mhc class II genes (Celada and Maki

1989a; Didier et al. 1988; Celada et al. 1988; Accolla et al.

1985b; Miwa et al. 1987; Dellabona et al. 1989; Reith et al.

1989; Reith et al. 1988; Dorn et al. 1989; Sugita et al.

1987). Using transient transfection assays, restriction

fragments from both mouse and human class II genes linked to

heterologous promoters have defined cis-acting elements that

can control tissue specific as well as IFN-y inducible

expression (Sullivan and Peterlin 1987; Wang et al. 1987; Dorn

et al. 1987a; Sherman et al. 1987; Basta et al. 1987; Gillies

et al. 1984; Tsang et al. 1988; Sloan and Boss, 1988; Koch et

al. 1989; Calman and Peterlin 1988). However, with the

exception of motifs associated with the promoters of class II

genes, the regulatory regions that have been defined by these

experiments are not always consistent from gene to gene with

respect to positionss, size, and influence. For example,

Tonegawa and coworkers mapped an enhancer in an area 2.7 Kb 5'

of the leader exon in the Ebd gene (Gillies et al. 1984). The

area that contains this elements) is 2.0 Kb in length and

requires the entire 2.0 Kb fragment to enhance expression.

Mathis and coworkers describe enhancing ability associated

with a restriction fragment that extended 2.1 Kb 5' of the

leader exon of the Eak gene and would retain activity when

broken into two halves (Koch et al. 1989). Further

characterization demonstrated the enhancing activity could be

localized to discrete sequences on both halves of the

fragment. These results are in contrast to what was observed

by Tonegawa for the Eb gene and also differ from results with

human Mhc class II genes.


It is clear from these data that the molecular mechanisms

responsible for Mhc class II gene expression are not

completely understood and that each of these genes may have

distinct regulatory features. Differences in binding

specificities and affinities of cloned X and Y box nuclear

proteins are evidence for the latter (Kouskoff et al. 1991;

Celada and Maki 1989a; Kobr et al. 1990; Boothby et al. 1989;

Reith et al. 1989). This dissertation focuses on studying

Mhc class II gene regulation on the expression of Ab. In

particular, I have analyzed the chromatin structure of Ab in

tissues with differing levels of transcription. Here the

distribution of Ab DNAse I hypersensitivity (DH) sites in

these tissues as well as protein-DNA interactions associated

with selected DH sites. The results indicate a change in Ab

chromatin conformation in expressing versus non-expressing

tissues in vivo, predominantly at specific sites in intron 2.

Further characterization of the protein-DNA interactions at

these sites, leads me to postulate the presence of an

intragenic enhancer in the Ab gene.


It has been almost four decades since the discovery of

the major histocompatibility complex (Mhc), also known as the

H-2 complex, in inbred lines of mice. Gorer and his coworkers

(Gorer 1938) first characterized the serology of tissue

transplantation using inbred lines of mice as a model system.

He postulated that the genes involved in the rejection of

tissues were linked to an erythrocyte antigen (denoted antigen

II) (Gorer 1936) and should be called histocompatibility

genes, thus the nomenclature H-2 (Gorer et al. 1948). Since

the discovery of the H-2 complex, a number of biological

phenomena, mostly immunological, has been attributed to the

Mhc (Klein et al. 1983).

Major Histocompatibility Complex

The Mhc contains a cluster of genes located on

chromosome 17 (in the mouse) that encode polymorphic molecules

essential for the presentation of foreign antigens (both

endogenous and exogenous) to regulatory T lymphocytes. It has

been divided into six major regions (K, I, S, p, Qg, and Tla)


containing genes to encode three classes of immunologically

related molecules (I ,II and III) (Klein 1975). The class I

genes can be divided into two major subclasses: the classical

transplantation antigens (K, D, L) (Stephan et al. 1986), and

those genes expressed on nucelated blood cells (0a) and

certain thymocytes and luekemias (fTl) (Stanton et al. 1978;

Winoto et al. 1983). The former genes are expressed on all

nucleated cells and are responsible for the presentation of

viral antigens to cytotoxic T lymphocytes. The function of

the Qa and Tla antigens is not known. Class II genes are

found in the I region and are expressed primarily on B

lymphocytes and monocytes (Klein 1975) and are critical in the

presentation of foreign antigen to regulatory T lymphocytes

(Zinkernagel et al. 1976; Marrack and Kappler 1988). The

class III genes are found in the S region and encode some of

the complement components (C2, Bf, C4) (Muller et al. 1987b).

Genomic Organization of the Mouse Mhc

The genomic organization of the loci in the Mhc of two

laboratory inbred mice, C57BL/10 (Flavell et al. 1985; Weiss

et al. 1984) and BALB/c (Steinmetz et al. 1982a; Steinmetz et

al. 1986), has been determined (Figure 2-1). Steinmetz and

coworkers (1986) used a technique called chromosomal walking

to clone forty nine overlapping cosmid clones defining 600 Kb

of DNA containing two class I (K and K2) and seven class II

0) r-4

0 4)




0 "



.,4 -H


I -

*3 0
Ol r
Cr 4)
P4 1


= 0


g- ,,,
w --.





- 0








genes (Ab3 to Ea) from BALB/c mice (Figure 2-1). This was the

first time in which the K and 1 regions were physically linked

by overlapping clones. A second gene cluster of 330 Kb has

been cloned from the S region and contains the genes coding

for complement components (C2, Bf, Sp, C4) and two homologous

genes (21-OHA, 21-OHB) coding for steroid 21-hydroxylase

(Muller et al. 1987b). Telomeric to the I region, a third

gene cluster has been isolated containing 13 class I genes

spanning from the D to Qa regions (Stephan et al. 1986). This

500 Kb segment of DNA contains the D, Qal-10, TNF-a and TNF-B

genes (Muller et al. 1987a). Winoto and coworkers (Winoto et

al. 1983) have isolated 54 cosmids containing 36 class I

genes, 31 of which are in the Tla region (Figure 2-1). In

summary, the mouse H-2 and Tla complex, from BALB/c mice,

spans 2.0 centimorgans on chromosome 17 (Stephan et al. 1986)

and contain 36 class I genes, 7 class II genes, and 6 class

III genes.

Genetic Organization of the I Region

The existence of the I region came about from immunizing

guinea pigs and inbred mice with synthetic polypeptides

(Schwartz 1982a). Using the appropriate crosses, McDevitt and

Sela (1965) demonstrated that the immune response differences

between CBA and C57 mice to the synthetic peptide

poly(tyr,glu)-poly-DL-ala-poly-lys ((T,G)-A-L) was due to the

influence of a single genetic locus (immune response (Ir)


genes). It was later established that the Ir genes were

linked to the mouse Mhc (McDevitt and Tyan 1968). This

discovery led to the utilization of H-2 recombinant inbred

lines to create a fine structure genetic map of the I region

(Schwartz 1986). Using synthetic polypetides (e.g. (T,G)-A-L)

and complex antigens (e.g. MOPC173, Staphylococcal nuclease)

many investigators mapped several Ir genes to the region

between the K and S locus in the H-2 complex (Lieberman et al.

1972; Lozner et al. 1974; Murphy et al. 1976). As each

laboratory mapped the Ir gene for their particular antigen, it

became clear that there were several Ir genes. A genetic map

of the I region soon emerged in which there were five

subregions: A, B, B, f, C (centromeric to telomeric).

The B locus. The B subregion was originally mapped by

Lieberman and coworkers (Lieberman et al. 1972) to explain the

genetic control of an antibody response to a myeloma protein,

MOPC 173. Several other antigens have also been mapped to the

B locus (Lozner et al. 1974; Klein et al. 1983). However,

using a T cell proliferation assay, Baxevanis and coworkers

(1981) observed that they could convert nonresponders into

responders by the addition of anti-I-A monoclonal antibodies

or by removal of Lyt-2+ lymphocytes (T cells). The authors

concluded the T lymphocyte population plays a large role in

determining the responder phenotype, making it unnecessary to

postulate a separate B subregion. The molecular cloning of


the I region had confirmed the nonexistence of the B region

(Steinmetz et al. 1982b).

The J locus. The J subregion was originally defined as

an I region determinant expressed on T suppressor (Ts) cells

(Murphy et al. 1976). The existence of the I-J determinant

came about from reciprocal anti-I region antisera between

inbred congenic recombinant mouse strains B10.HTT, B10.S(9R),

B10.A, B10.A(5R) and B10.A(3R) (Murphy 1978; Murphy et al.

1976). These mice were also used to genetically map the

position of the new locus, I-J (Ia-4). The presence of the J

determinant on allotypic Ts cells was detected by killing this

population with Ia alloantisera plus compliment (Murphy et Al.

1976). However, the biochemical data supporting the existence

of this Ts specific determinant is lacking.

Within the last decade three lines of evidence have

emerged that dispute the existence of the J subregion within

the I region. First, the systematic molecular

characterization of the I region failed to substantiate the

existence of the J subregion (Steinmetz et al. 1982b). Using

restriction fragment length polymorphism (RFLP) analysis to

map the intra I region recombinant inbred mouse strains, the

authors were able to map the suspected position of J to a 3.4

Kb stretch of DNA within the second intron of the Eb gene

(Steinmetz et al. 1982b) This intron is the site of a

recombinational hotspot within the I region (Steinmetz et al.

1986; Steinmetz et al. 1987). A closer examination of the 3.4


Kb stretch of DNA in six I region recombinant inbred strains

(including strains originally used to define the I-J

subregion) narrowed the possible location of J to a 2.0 Kb

fragment (Kobori et al. 1984) mapping it entirely within the

Eb gene. Finally, the most convincing piece of evidence comes

from Kronenberg and coworkers (1983), in which they failed to

hybridize DNA from the suspected coding region of J to

poly(A)* RNA isolated from cloned Ts hybridomas.

The C locus. The C subregion was originally discovered

with an H-2'h2 anti H-2h4 antiserum (David and Shreffler 1974)

and was later reproduced (Sandrin and McKenzie 1981). Further

support was given by Rich and coworkers (1979), when they

reported that an antisera containing C-specific antibodies

reacts with a suppressor factor produced in an allogeneic

mixed lymphocyte reaction (MLR). Okuda and David (1978) also

reported an MLR that occurs in congenic strain combinations

that differ at the C locus and can be inhibited by anti-C

sera. Recombinant H-2 inbred strains were then used to map

the location of the C locus to a position telomeric to Ea and

centromeric to C4. Although this chromosomal segment has not

been completely cloned and characterized, the molecular data

available does not support the existence of a C locus near the

Ea locus. Furthermore, other investigators have not been able

to detect any activity in the C-defining H-2h2 anti-H-Z1h4

combination using a variety of serological methods (Klein et

al. 1983).

Genomic Organization of Class II Genes

The only two regions to stand the test of both

serological and biochemical assays are A and E. These two

loci contain seven class II genes and span approximately 500

Kb of DNA starting telomeric to K and centromeric to the C4

gene (Steinmetz et al. 1982b; Hood et al. 1983; Hood et al.

1982; Wake and Flavell 1985). Four of these class II genes

(Ob(Ag2), Aa, Ab, Eb) are contained within A and two others
(Eb2, Ea) within E (Figure 2-2). The seventh class II gene

(Pb (A.3)) is located 90 Kb distal to the K gene and is known

to be a psuedogene. Only four of the seven class II genes

have been shown to encode gene products. The Aa and Ab

polypeptides noncovalently associate to form the A molecule

and the Ea and Eb polypeptides join to form the E molecule

(Jones et al. 1978; Uhr et al. 1979). Although there is

evidence that the Pb and Ob genes are transcribed, there is

does not seem to be a protein product produced and their

function is not known (Wake and Flavell 1985).

Biochemical Properties of Class II Polypetides

The gene products of A and E are heterodimeric

glycoproteins made of one light (P) and one heavy (a) chain

that noncovalently associate on the cell surface. The P chain

has a molecular weight of approximately 28,000 with a chain

length of about 220 amino acids. In contrast, the a chain has

a molecular weight of approximately 34,000 with a chain length

0 *d

H *




--) 0









-4 -D
c 1

.r4 or
44 r




0 0c

wD U


of about 230 amino acids (reviewed by Klein 1975; Klein et al.

1983). Each class II peptide has a carbohydrate side chain

and it is postulated to contribute to the difference in

molecular weight (Klein et al. 1983). Each chain has two

extracellular domains (al-a2 or 81-B2) of approximately 90

amino acids each, a transmembrane region of about 30 residues

and a short cytoplasmic tail of about 10-15 residues. Three

of the four extracellular domains (81, B2, a2) have a

disulfide bridge, while the fourth (al) does not. The peptide

binding properties of class II molecules is associated with

the two polymorphic membrane distal domains (al and 31) (Klein

et al. 1983). Brown and coworkers (1988) have developed a

hypothetical model for the three dimensional structure of the

class II molecule antigen binding site (ABS) based on a

comparison of the patterns of conserved and polymorphic

residues of twenty-six class I and fifty-four class II amino

acid sequences. The model predicts the al and B1 domains

associate in a symmetrical fashion to form the ABS, and it is

made up of eight B-pleated sheets (four from each chain) and

two a-helices.

Structure of Class II Genes

The last ten years has seen a surge of nucleotide

sequence data coming from the scientific community. A number

of laboratories have sequenced cDNA as well as genomic clones

of both human and mouse class II genes (Hyldig Nielsen et al.


1983; Benoist et al. 1983; Mathis et al. 1983; Larhammar et

al. 1983; Malissen et al. 1983; Estess et al. 1986; Saito et

al. 1983). There is a striking similarity between the domain

organization of the protein product and the exon-intron

organization of the genes. Both a and B genes contain an exon

coding for a signal peptide as the first exon followed by two

exons coding the two extracellular domains (al and a2 or Bl

and 32) (Figure 2-3). The 8 genes have three exons coding for

the transmembrane (exon 4), intracytoplasmic (exon 5) and 3'

untranslated (exon 6) regions. In contrast, the a genes have

the transmembrane, intracytoplasmic and part of the 3'

untranslated domains encoded in exon 4, while the rest of the

3' untranslated domain is in exon 6.

Regulation of Class II Gene Expression

As stated previously, the primary role of class II

genes is the presentation of foreign antigen peptides (via the

ABS) to regulatory T lymphocytes. In addition, they are also

critical during positive and negative selection of the T cell

repertoire in the thymus (MacDonald et al. 1988; Teh et al.

1988). Both of these functions play key roles in restricting

and regulating the immune response. It is not surprising, that

the gene products of class II genes are found on a specific

tissue distribution, mainly cells of the lymphocyte and

monocyte lineages.

0 rw




1 0

a ..u

r-4 ra




O *

6 -


- Cl)
- 0)
Co C
Co 0)
Cu 0)


Biology of Class II Gene Expression

One unique aspect of class II gene expression is the

variety of expressing patterns observed: constitutive,

inducible and non-expression (reviewed by Benoist and Mathis

1990). The major cell types that are found to express class

II genes are B cells, T cells (in the human) and macrophages

(Benoist and Mathis 1990; Paulnock King et al. 1985; Pullen

and Schook 1986; King and Jones 1983). Distinct mechanisms

have been proposed to govern the class II gene expression in

B cells and macrophages (Maffei et al. 1987). While some

other limited cell types have also been shown to express class

II genes, most require the help of y-interferon (Collins et

al. 1984; Wuthrich et al. 1989). It seems clear that the

regulation of class II gene expression is at the level of

transcription (Salter et al. 1985), although some have

postulated a contribution in post translational modifications

(Buerstedde et al. 1988).

Class II gene expression in B cells varies during the

differentiation from pre-B cells to antibody secreting plasma

cells. There are no detectable levels of class II mRNA or

protein in murine pre-B cells (Polla et al. 1986). As the

cells differentiate into mature resting B cells, basal levels

of class II molecules are observed on the cell surface in a

constituative manner (Mond et al. 1980; Robbins et al. 1988).


Additionally, class II expression can be increased by using a

variety of inducers: IFN-y, IL-4, LPS, phorbol esters, calcium

ionophores and cross linking of surface Ig receptors with

anti-Ig sera (Mond et al. 1981; Monroe et al. 1984). Finally,

when the cells reach the dead end plasma cell stage, class II

gene expression stops completely (Latron St al. 1988). Some

investigators have argued the existence of a trans-acting

factors) in the plasma cell that actively suppresses class II

expression, even across species barriers (Dellabona et al.

1989; Latron et al. 1988).

With some exceptions, murine T cells do not normally

express class II molecules on their cell surface (Benoist and

Mathis 1990). However, upon activation by mitogens or

alloantigens human T cells have been shown to express class II

molecules (Charron et al. 1980; Triebel et al. 1986; Robbins

et al. 1988). The reason why there is a difference in

expressing phenotypes is not known, but some have postulated

a defect in murine regulatory factors (reviewed by Benoist

and Mathis 1990).

Cells of the monocyte lineage, namely macrophages, do not

constituatively express class II molecules. They can,

however, be induced to express high levels on their cell

surface using IFN-y (King and Jones 1983; Rosa and Fellous

1984) or TNF-a (Chang and Lee 1986). The effects of

supernatants from mitogen stimulated spleen cells on Ia

expression lead to the discovery of IFN-y as an inducing


factor. Since then, it has been shown by a number of

laboratories to be a potent inducer of class II genes in

macrophages as well as other cell types (Collins et al. 1984).

TNF-a has also been shown to act synergistically with IFN-y

(Arenzana-Seisdedos et al. 1988) and both are thought to use

a Ca++ dependent pathway during the induction.

Analysis of Class II Gene Expression Using Mutant Cell Lines

The analysis of class II gene regulation has been

assisted by the existence of mutant cell lines. The two

sources of these cell lines are from either bare lymphocyte

syndrome (BLS) patients or chemically induced from B

lymphomas. The BLS patients suffer from a genetic disease

characterized by a deficiency in Mhc gene expression. Both

class I (Touraine et al. 1978) and class II (Rijkers et al.

1987) negative expressing patients can be found. The defects

causing loss of gene expression are thought to lie in the

absense of transcription factors. However, recent studies by

Kara and Glimcher (1991) have implicated changes in

accessibility of the trans-acting factors to their cis-acting

elements as contributory to the la' phenotype of BLS cell


In contrast to BLS derived cell lines, many investigators

have produced chemically induced la" mutants from existing Ia*

lymphomas. Some of the mutants were derived from a Burkitt

lymphoma cell line, Raji (RJ 2.2.5, RM2, RM3), and another


from a human B-lymphoblastoid cell line (6.1.6) (Accolla et

al. 1985b; Calman and Peterlin 1987; Salter et al. 1985).

These mutant cell lines have been shown to be deficient at the

level of transcription and have no mRNA or protein products.

Fusion between human RJ 2.2.5 and mouse la* B cell lymphomas

recovered class II gene expression (Accolla et al. 1985b;

Accolla et al. 1985a), demonstrating the defect was in the

absense of a trans-acting factor. This phenomena was found

only to work with B cells and not macrophages (Maffei et al.

1987), implying that distinct molecular mechanisms were

operating in the two cell types. Accolla and coworkers (1986)

were able to demonstrate the factor was unlinked to Mhc and

located on chromosome 16 in the mouse.

Motifs and Regions Contributing to Class II Gene Expression

To elucidate the crucial cis-acting regulatory elements

responsible for the different phenotypic expression patterns

observed, investigators have manipulated genes and

reintroduced them into biological expression systems. This

approach has not only yielded some important DNA intervals for

inducible and contitutive expression, but also some of the

trans-acting factors themselves. The biological systems that

have been employed range from cell-free transcription systems

to transfected cell lines to transgenic animal models. The

large amount of work to date involves the placement of class

II promoter fragments (of varying length and sequence) in


front of a reporter gene (e.g. chrloramphenicol-acetyl-

transferase) and transfection of them into various cell lines

(both Ia" and Ia). The activity of the fragments are

indirectly measured by quantification of the protein

translated from the reporter gene. Enhancer activity

associated with various class II gene segments are also

evaluated in a similar fashion, with the exception that the

reporter gene has a promoter. The transgenic mouse

experiments have yielded some very interesting results for the

Ea gene (Van Ewijk et al. 1988; Dorn et al. 1988; Widera et

al. 1987). This is the only gene that has been systematically

deleted and introduced into mouse embryos to assay for various

regulatory regions. These experiments give the investigator

a look at elements that can influence regulatory events

throughout the normal differentiation of these cell types in


The B cell control region

The functional significance of the B cell control region

was uncovered in the Eaak and Ead transgenic animal models

(Widera et al. 1987; Dorn et al. 1988; Van Ewijk et al. 1988).

These animals were produced using constructs of the Ea gene

with different deletions in the 5' region (-2000 bp to -1300

bp). The transgenic animals containing at least 2000 bp of

DNA sequence 5' of the leader exhibited a normal tissue

distribution of expression for the transgene. However, those

transgenic animals containing only -1300 bp of 5' sequence


expressed the transgene in the thymus and IFN-y induced

macrophages, but failed to express in the B cell population.

The authors concluded that the region from -2000 bp to -1300

bp contains a B cell specific control region (Figure 2-4).

Curiously, this region appears to exert much less of an

influence when transfected into B-cell lymphomas (Van Ewijk et

al. 1988; Dorn et al. 1988). When the 5' deletion extends to

only -1180 bp a few B cells are observed expressing the

transgene (Koch et al. 1989), leading some to postulate the

presence of a repressor within -1300 to -1180.

Dorn and coworkers (1988) cloned and sequenced the B cell

control region looking for sequence motifs that could explain

the activity observed. The sequence data (Dorn et al. 1988)

coupled with functional data (Koch et al. 1989) has revealed

some interesting parallels between this region and the

promoter proximal motifs (Figure 2-4). First, a copy of the

X and Y box motifs, termed X' and Y', are found between

positions -1402 and -1336 in the 5' region in the reverse

orientation. The binding specificities of the X'and Y' motifs

were shown to be similar to their counterparts using band

retardation and methylation interference assays (Dorn et al.

1988). Second, the X' and Y' boxes exhibit similar enhancing

activity when tested in a transient transfection assay using

these segments linked to a reporter gene (Koch et Al. 1989).

Second, a copy of the B motif is present between -1716 and -

1674 (Figure 2-4). This motif has been shown to be

0 ) .-.
> o

a) *-I 0

i -4 *
* I 04)
t4) a4

0 0,

0-4 0 4
M0 .1


pM 0
o #

0.4 4)

1) 4 oI

(* .a) .0
S4) -0

U a)
0 0
x 0O
x CM

rz(0 a

N tU "
. to m

0 0
.4 Id


C1 -


I r~1'
Is C>

- co
E-] C





a control element in a number of enhancers (e.g. SV40 and HIV)

and serves as a binding site for a family of factors (e.g. NF-

KB). Third, a W box is located within -1322 to -1180 (Dorn et

al. 1989). This motif was originally discovered in a search

for B cell specific binding proteins and is important

component of the Ea enhancer. Finally, there are sequences

with homolgies to both the Ephrussi and Pu box motifs within -

1824 to -1216.

The promoter region

A considerable amount of research has gone into

analyzing the promoter regions of both human and mouse class

II genes. The region -200 to -10 bp has had strong promoter

and enhancer activity associated with it. Investigators have

used transient transfection, band retardation, DNAse I foot

printing, site directed mutagenesis and linker-scanning

mutagenesis experiments (Finn et al. 1991; Tsang et al. 1988;

Sullivan and Peterlin 1987; Thanos et al. 1988; Sherman et al.

1987; Gillies et al. 1984; Kobr et al. 1990; Celada et al.

1988; Peterlin et al. 1987; Kara and Glimcher 1991; Celada and

Maki 1989; Miwa et al. 1987; Boothby et al. 1989) to elucidate

the promoter proximal elements essential for class II

expression. Furthermore, there is quite a bit of consistency

observed between the different promoter regions assayed in

both mouse and human class II genes. The following is a

breakdown of the elements that have been found to have a

functional impact on class II expression.


The X and Y box motifs. These elements were the first

sequence motifs to be identified and characterized. They were

originally discovered by doing a sequence alignment among the

promoters of human and mouse class II genes. The promoter

element consists of a 15 bp X box (CCTAGAGATGATG) and a 14 bp

Y box ( TCrGATrGG ) separated by 18 to 20 bases that are

conserved in length but not in sequence. Overall, the X box

sequences are much less conserved than that of the Y box

seqences (Boothby et al. 1989). The functional data on the

significance of these motifs illustrates several important

features. First, deleting the X-Y box, either clean (Dorn et

al. 1987a) or by 5' deletions (Thanos et al. 1988; Sloan and

Boss 1988; Tsang et al. 1988), from the promoter abolishes

both promoter and enhancer activity in B lymphomas and IFN-y

induced cells. Second, the X and Y box appear to be critical

for the accurate and efficient initiation of transcription of

an Eak transgene in transgenic animals (Dorn et al. 1987a).

However, the Eak protein was detected on the surface of B

cells and IFN-y induced macrophages in these animals.

Illustrating the aberrantly initiated transcripts can be

translated into Eak protein. Interestingly, only about half

of the B cells from the mice with the Y box deleted were Eak
+ staining cells. Finally, fragments containing the X-Y box

have enhancing activity in B cells but not in plasmacytomas,


fibroblasts or T cells (Sloan and Boss, 1988; Koch et al.

1989; Tsang et al. 1988).

The Z/W region (S box). The W region, sometimes called

the Z region, was first characterized by 5' deletion

experiments (Basta et al. 1987; Thanos et al. 1988) and has

been implicated to assist in both IFN-y induced and

constituative B cell expression. Linker scanning experiments

establish a role for a 7 bp sequence called the S box.

Mutations within this segment have a strong down regulating

effect. The exact interplay of the S box with the X-Y box

motif is not understood.

The octamer motif. The octamer motif has been found only

in the human DRa gene a few base pairs 3' of the Y box

(Sherman et al. 1987). This sequence is termed the "octamer"

motif because of its identity to the octamer sequence found in

immunoglobulin genes. This motif seems to contribute to DRa

expression in B lymphomas (Sherman et al. 1987). Why the DRa

gene has this additional transcriptional element is still a

mystery, but some have postulated it may account for the high

activity of this gene in human cells (Benoist and Mathis


Intragenic enhancers

The majority of research looking at the regulation of

class II genes has been concentrated on the 5' region of these

genes. Very few studies have addressed the possibility of

regulatory elements downstream from the transcriptional start


site. Some investigators have described intragenic enhancers

in the DRa, DQa and =Db genes (Sullivan and Peterlin 1987;

Wang et al. 1987), but the exact interplay of these elements

with the 5' regulatory regions is not understood. Peterlin

and coworkers have described enhancers in the first intron in

both the DRa (Wang It al. 1987) and DQa (Sullivan and Peterlin

1987) genes. In contrast, the second intron of the =Db gene

contained the enhancer element (Sullivan and Peterlin 1987).

These elements fulfilled the requirements of an enhancer by

being position and orientation independent. In both cases,

the exact nucleotide sequence exerting the enhancing influence

is not known. However, there are sequences with identity to

the Ephrussi and enhancer motifs located within the first

intron of the DRa gene.

Chromatin Structure

Chromatin structure has been postulated to play a

significant role in the transcriptional activation of a

variety of genes in many different systems (Fritton et al.

1984; Weintraub 1983). Over the last fifteen years nucleases

have been found to be effective tools for the dissection of

the structure of chromatin. In particular, active chromatin

is preferentially digested by the nuclease DNAse I (Gross and

Garrard 1988). It was soon realized the areas in active

chromatin hypersensitive to digestion by DNAse I were regions

of the genes that influenced transcription (reviewed by Elgin

1981). A number of investigators have demonstrated the


appearance of DNAse I hypersensitivity (DH) sites upon

transcriptional activation of a gene (Fritton et al. 1984;

Larsen and Weintraub 1982). The exact biochemical nature of

DH sites is not completely understood.

There have only been a handful of studies dealing with

the chromatin structure of class II genes (Peterlin et al.

1987; Liou et al. 1988a; Gonczy et al. 1989). Nonetheless,

changes in chromatin conformation (or sensitivity to DNAse I)

have been correlated with transcriptional activity of class II

genes. Liou and coworkers (Liou et al. 1988a) detected three

DH sites in the Aa gene, two in the 5' region and one in the

first intron. A DH site next to the promoter was specific for

those tissues expressing class II genes. Peterlin and

coworkers (1987) also detected three DH sites in the DRa gene.

However, one DH site was in the promoter region and two in the

first intron. In contrast to Aa, the two DH sites in the

first intron were specific for actively transcribing tissues.

The authors later show this region to contain enhancing

activity (Wang et al. 1987). Finally, Gonczy and coworkers

(1989) demonstrate the loss of DRa expression in EBV

transformed cell lines from SCID patients was due to

inaccesibility of trans-acting factors for their cis-acting

sequences. They detected the loss of a DH site in the

promoter region that was critical for the expression of the

class II gene. It appears from these studies that chromatin


conformation may play a large role in regulating class II gene


Trans-acting Factors

When the functional relevance of the promoter region

motifs became apparent, many investigators set out to

characterize the protein-DNA interactions associated with

them. Furthermore, the genes encoding some of the proteins

that bound to those sequences were also cloned. Using band

retardation, methylation interference and DNAse I protection

experiments, investigators identified protein or protein

complexes that bind to specific sequences. The methods used

to isolate the genes were to make cDNA libraries cloned into

a Agtll expression system and the protein products were

screened with a 32P-labeled double stranded oligo specific for

the motif of interest.

W box binding proteins

These proteins were characterized by Dorn and coworkers

(1989) while searching for protein-DNA interactions that bound

to the W motif in the Ea B cell control region. Using a

series of biochemical assays, they were able to demonstrate

that two separate proteins, NF-W1 and NF-W2, bound to this 38

bp sequence. Although both proteins recognize the core motif

GTTGCATC, they have different binding properties as determined

by band retardation, methylation interference and

oligonucleotide mutagenesis experiments (Dorn et al. 1989).


Furthermore, NF-W2 was found to be present in all tissues

examined, while NF-W1 was found in B lymphomas that expressed

class II genes as well as a pre-B lymphoma (70Z/3) induced

with LPS. The two proteins differ in molecular mass by 20,000

daltons, but are indistinguishable by proteolysis experiments.

The cDNA for these proteins has not been cloned.

X and X2 box binding proteins

There has been a considerable amount of work done

characterizing the protein-DNA interactions associated with

the X box motif. There have been two proteins cloned that

bind specifically to the X box core sequence and two proteins

that recognize a sequence that overlaps into the 3' end of the

X box (termed X2 box; Figure 2-4).

X box binding proteins. RF-X was one of the first

factors cloned that specifically binds to the conserved X box

(Celada et al. 1988). This protein was first described by

Reith and coworkers (1988) as a factor that binds to the DRa

X box and is missing in B cells from SCID patients. Using

methylation interference assays the contact points have been

determined to be in the center of the motif and it requires a

fairly large DNA target (Reith et al. 1989). Furthermore,

this protein exhibits a gradient of affinities for other X

boxes (DRa > DPa > DOa).

The second protein found to bind to the X box comes from

the mouse system. NF-X was initially identified as binding to

the X box of Ea and DRa promoters in band retardation


experiments as a doublet or triplet (Dorn et al. 1987a; Miwa

et al. 1987; Kouskoff et al. 1991). Nuclear extracts prepared

from a variety of cell types were tested for NF-X activity

using a band retardation assay (Kouskoff et al. 1991). NF-X

specific bands were identified in all extracts tested,

including la* B lymphomas (WEHI-231, TA3) as well as la" pre-B

cells (70Z/3) and fibroblasts (LMTK). Unlike RF-X, this X box

binding protein was found in B lymphoblastoid lines from SCID

patients. Interestingly, the protein contact points, deduced

by methylation interference, were very similar to that found

for RF-X. Kouskoff and coworkers have determined by Ferguson

analysis the size of the NF-X complexes to be several hundred

kDa and a core DNA binding domain of 10 kDa (Kouskoff et al.

1991). Futhermore, they demonstrated that NF-X is probably

identical to the EF-C factor found to interact with the

polyoma and hepatitis virus enhancers.

X2 box binding proteins. The X2 motif was initially

identified in band retardation assays in which cross

competition experiments demonstrated multiple proteins binding

X box motifs (Boothby et al. 1989). A closer examination of

the binding properties revealed a protein-DNA complex

overlapping the 3' nucleotides of the X box. Two cDNA clones

have been isolated from the mouse that code for proteins that

bind to this motif, mXBP and mXBP-2 (Liou et al. 1988b). The

human homolog, hXBP-1, has also been cloned (Liou &et Al.

1990). The possible importance of this sequence motif comes


from the observation that the X2 region contains a cAMP

responsive element (CRE) or TPA responsive element (TRE),

implying that X2 might belong to the fos/jun/CREB family of

regulatory elements. Confirming this interpretation, the

hXBP-1 cDNA clone codes for a 250 amino acid protein which

contains a central leucine zipper with an adjacent basic

region (Liou et al. 1990). This protein domain is

characteristic of several DNA-binding transcription factors

(e.g. myc, fos, jun, cEBP) (Liou et al. 1990). In fact, Ono

and coworkers have recently shown the human X2 binding

protein, hXBP-1, is required for transcription and forms a

heterodimer with c-fos (Ono et al. 1991a; Ono et al. 1991b).

Y box binding proteins

The 14 bp transcriptional control element known as the Y

motif contains a CCAAT box in reverse orientation. The CCAAT

sequence is known to be a crucial component of several

promoters and can operate in reverse orientation (Santoro et

al. 1988; Dorn et al. 1987b). More then one Y box binding

protein has been cloned and the binding is dependent on the

core CCAAT motif (Dorn et al. 1987a; Didier et al. 1988).

The first Y box protein to be characterized was NF-Y

(Dorn et al. 1987a). Dorn and coworkers (Dorn et al. 1987a)

used a band retardation assay to scan the promoter region of

the Eak gene. Using deletion analysis, the authors described

protein-DNA interactions specifically associated with the Y

box motif. A careful examination of the properties of NF-Y


has established it to be a unique CCAAT box binding protein

(Dorn et al. 1987b). Furthermore, NF-Y is a metalloprotein

composed of two subunits of approximately 32 kDa and 42 kDa,

both of which are required for binding.

The second Y box binding protein, YB1, was identified by

screening a Agtll library with a double stranded 32P-labeled

oligo containing the Y box (Didier et al. 1988). YB-1 binding

has an absolute requirement for the CCAAT motif and relative

specificity for the Y box. The molecular weight of this

protein is approximately 39,000 Da, contains 18% basic

residues and putative nuclear localization signals (Didier et

al. 1988). Interestingly, YB-1 mRNA levels inversely correlate

with DRb chain mRNA levels. This conclusion lead the authors

to postulate that YB-1 may be a negative regulatory factor.

How are Class II Genes Regulated?

The regulated expression of class II genes requires the

interaction of promoter elements, upstream regions and

intragenic enhancers. Even though there is a considerable

amount of information known about the cis-acting elements,

trans-acting factors, 5' regulatory regions and intragenic

enhancers that individually contribute to the Ia phenotype,

investigators still do not know how this complex system

interacts to give a cell the potential for constituative as

well as inducible expression. It seems clear that the

promoter region motifs are an essential core unit of class II


expression, while the upstream region is a control element for

a specific tissue. If accessibility of these trans-acting

factors for their cis-acting sequences is vital, then

chromatin structure may play a larger role then previously

thought. The possible cooperative role that intragenic

enhancers play is still a mystery. Furthermore, it is quite

possible that each class II gene may have slight differences

in which to regulate that expression. This is evidenced by

the difference in binding affinities (Kouskoff et al. 1991)

and specificities (Celada and Maki 1989; Kobr et al. 1990;

Boothby et al. 1989; Reith et al. 1989) of factors for their

sequences as well as the appearance of unique elements

associated with specific class II genes (Sherman et al. 1987).

This dissertation takes a closer look at the regulation of the

murine Ab gene. Using DNAse I hypersensitivity, band

retardation and DNAse I protection assays, evidence for a

possible intragenic enhancer in the second intron of Ab will

be presented.


Inbred Mice

All mice used in this study were maintained in our mouse

colony. The inbred mouse strains are maintained by full sib

mating with a single line of descent. Their properties and

origins have been described previously (Wakeland and Klein

1983; Wakeland and Darby 1983).

RNA Isolation

Total cellular RNA was isolated using an acid guanidinium

thiocyanate-phenol-chloroform extraction procedure described

by Chomczynski and Sacchi (1987) with slight modifications.

Briefly, tissues were homogenized with a motorized dounce in

4M guanidinium thiocyanate, 25mM sodium citrate pH 7.0, 0.5%

sarcosyl and 0.1M 3-mercaptoethanol (GTC solution). Equal

volumes of water saturated phenol, sodium acetate (pH 4.0),

and chloroform were added to the GTC-tissue mixture, incubated

on ice for 10 min and centrifuged at 10,000X g for 25 minutes

at 4C. The aqueous phase was removed and the RNA

precipitated with an equal volume of ice cold isopropanol. The

precipitate was resuspended in 0.3 ml GTC solution and

reprecipitated with 0.3M sodium acetate (pH 5.2) and



isopropanol. The RNA was lyophilized and resuspended in

diethylpyrocarbonate (DEPC) treated water. Finally, the RNA

was reprecipitated with 0.3M sodium acetate (pH 7.0) and 2X

volume 100% ethanol, centrifuged, washed with 70% ethanol, and

dissolved in 0.3 ml DEPC treated water. RNA concentrations

were calculate from absorbance at 260 nm.

Northern Analysis

RNA (20 Mg) was denatured at 650C for 10 min in a 12.3M

formaldehyde, 50% formamide loading buffer and electrophoresed

through a 1.0% agarose gel containing 40mM MOPS, 10mM sodium

acetate, 1mM EDTA, and 2.2M formaldehyde. The RNA was

visualized by ethidium bromide (EtBr) staining and UV light.

The gel was denatured in 50mM NaOH, 10mM NaCL, neutralized in

0.1M Tris HCL, pH 7.5, equilibrated in 40mM TBE and

electrobloted to a nylon membrane (Zetabind, AMF Cuno). The

resulting blot was hybridized with an [a-32P]ATP radiolabeled

5.5 Kb Abd genomic clone described previously (McConnell et

al. 1988). The blot was washed at 65*C in 0.1X SSC,0.1% SDS

and exposed to X-ray film (Kodak, XAR-5).

Tissues and Cell Lines

The tissues used in the DNAse I hypersensitivity and band

retardation assays (liver, brain and kidney) were extracted

from the animals under general anesthesia. The A20 BALB/c

derived B cell line was kind gift from Dr. Howard Johnson and


was maintained in RPMI with 10% fetal calf serum, 0.05mM P-

mercaptoethanol, 50 U/ml penicillin and 50 jug/ml streptomycin.

B cell Isolation

B cells were isolated from splenocytes using negative

immunoselection. Briefly, total splenocytes were harvested

from each animal by perfusion of the spleen with 10 ml of

phosphate buffered saline (PBS). The red cells were lysed and

the remaining splenocytes were washed once in PBS. The cells

were incubated with a monoclonal anti Thy 1.2 (HO-13-4) or

anti Thy 1 (HB 23) and a rabbit anti-mouse brain polysera

(preabsorbed with mouse red cells) for 40 minutes on ice. The

splenocytes were washed and treated with low-tox rabbit

complement (Cedar Lane, 1:10 dilution) for 40 minutes at 37

C. The live cells were collected using a 50% Percoll

gradient. The isolation was monitored by staining the cells

before and after complement lysis using indirect

immunofluorescence (against the class II A molecule) and

analyzed on a fluorescence activated cell sorter (FACS)

subsequent to this treatment (Figure 3-1); > 85% of the

lymphocytes expressed A molecules on their cell surface. A

reduction of at least 50% of the original splenocyte

population was observed repeatedly for all strains and

consequentially this criterion was used to monitor the

Figure 3-1. FACS analysis of B cell isolations. All
histograms give mean channel number on the X axis and
increasing fluorescence on the Y axis. A) BALB/c spleen cell
negative control. The primary antibody was omitted from the
assay. B) T cell and B cell populations before complement
lysis. C) T cell and B cell populations after complement


(T Cell)



(B Cell)

isolation once the B cell enrichment was established by FACS


DNAse I HvDersensitivity Assay

Probes Used In DH Assay

The 3' probes used in the analysis were isolated from a

5.5 Kb EcoRI Abd genomic clone (McConnell et al. 1988). A 2.1

Kb BamHI-PvuII fragment was used to detect DH site positions

from the 3' side of the Ab gene in BALB/c, B10.SAA48, and NOD

B cells (probe 1, Figure 3-2). Alternatively, a 1.7 Kb

HindIII fragment was used as a 3' probe for the Ab gene in

C57BL/6 and B10.WB B cells (probe 2, Figure 3-2). The 5'

probe is a 500 bp BamHI-SacI fragment isolated from the cosmid

41.1 (Steinmetz et al. 1984) containing the Abd gene (probe 3,

Figure 3-2). An EcoRI-HindIII 400 bp probe (probe 4, Figure 3-

2) and double digests were used to confirm the positions of

the DH sites. The relative positions of the probes to the Ab

gene are shown in Figure 3-2.

DH Assay

Nuclei from each haplotype and tissue type were isolated

by the same procedure. All isolations were done on ice. The

cells were suspended in a nuclear isolation buffer (NIB, 60mM

Tris, pH 8.2, 60mM KCL, 15mM NaCL, 0.1mM EGTA, 5mM MgCL,

0.25mM Sucrose, 1mM DTT, and 0.5mM PMSF) with Triton X-100

(0.1%). The cell membrane was shed by mechanical shearing

U0 *>iC (
- ) > 4 .Q )
4 0) 0
O t 4 04Q
,- m4 1 0
0 0 -a r.1
o 0 .# *
0 0 0
4-4 ) 44 H
$4O P

4J 0 0O H
., 0 0
SM 4 ) 4 P

i M 4 4 I


$4 VpwOi
4.4 0) 0H
n4 *H r-
1 0 -I4 0
m c *H o U
it 0 0 4 0
*H U) 4.) W :

*> p 0 .) Q H

*H i0nO0 $4H
S o a0

*) O1 Q 0 H H
$4 *H O4
4*JM k 4 0 HI

0 ) 'I 0 k H
U) =- o **

to 0MU) -)H M H

,P 0 010 0

HM 5. Er ,

< 0 t0
0) 0) HO

0) V 0-
O$4 H

3() ( 4)
to 4 I OH 0

0 ) 0 A
Q)I .M O *4

0 0 0 l
C 4 -H 4 0CO
.0 0OM

V n 41 mOB

UN 0 OC .
W o f *HO*-iuo


c-- -2


co ~CD c


using a motorized dounce for 10-20 strokes. Nuclei were

pelleted by centrifugation, washed in ice cold NIB (to

eliminate excess Triton X-100), and resuspended in 4.0 ml NIB.

Quantitative and qualitative assessment of the nuclei was

accomplished by staining with Trypan blue and direct

visualization under light microscopy.

The DNAse I hypersensitivity assay consisted of treating

the nuclei with a fixed concentration of DNAse I (BRL) and

incubating at 37C. Aliqouts were taken at varying time

intervals and lysed in a 1% SDS, 100mM EDTA, 0.4 mg/ml

Proteinase K solution and incubated at 65*C overnight. DNA

was isolated on a nucleic acid extractor (ABI, model 340A),

restricted with the appropriate enzyme (see Results),

electrophoresed through a 0.7% agarose gel, transferred to a

nylon membrane (Zetabind, AMF Cuno) and hybridized with the

relevant 32P labeled probe as described previously (McConnell

et al. 1986). DH sites were deduced using the size of the

subfragments generated from DNAse I cleavage of the native


Band Retardation Assay

Primers Used For Synthesis of Probes

A polymerase chain reaction (PCR) based protocol was used

to amplify and isolate [a-32P] or [y-32P] labeled probes to be

used in the band retardation assays. The primers used in this

assay are given in Table 3-1 and their relative position

Table 3-1. Oligonucleotides used for PCR reactions in band
retardatation and DNAse I protection assays.

Location Primer Sequence"


Intron 2



All oligonucleotides given in the 5' to 3' direction. Odd
numbered primers have a Sal I restriction site on 5' end and
even number primers have a BamHI restriction site on the 5'
* These oligonucleotides are used in the DNAse I protection
assay. All of these primers have a BamHI restriction site on
the 5' end.

H *0) C4
4)C Q) 0
0 V4'>
M- 0 0

g *H

0 0

E1 0) -

1 .,.4 ^ 0
*.o4 0 00

*, -.-

4) 9 :
* 4 to 0
* > O C

I 000M 4-
14 N 4
4j #0
a) 0a 4)

44 >dzt 0 0

- 01



<8 ,



within the Ab gene is shown in Figure 3-3. All primers were

synthesized at the University of Florida core DNA synthesis


Preparation of Nuclear and Whole Cell Extracts

Nuclear extracts. The nuclear protein extracts were

isolated from A20 B lymphoma cells using the method of Dignam

et al. (Dignam et al. 1983) with slight modification.

Briefly, the nuclei from approximately 109 A20 B cells was

isolated as described previously (see DH Assay above). Once

the nuclei were washed in NIB they were resuspended in 2.0 ml

Buffer C (420mM NaCl, 1.5mM MgCl2, 20mM HEPES, pH 7.9, 0.2mM

EDTA, 0.5mM PMSF, 25% (v/v) glycerol, 0.5mM DTT), the membrane

shed with a motorized dounce for 15 strokes and the mixture

was stirred gently with a magnetic stirring bar on ice for

30.0 mins. The mixture was then centrifuged at 15,000 rpm for

30.0 min and the clear supernatant was dialyzed against 100X

Buffer D (100mM KC1, 20mM HEPES, pH 7.9, 0.2mM EDTA, 0.5mM

PMSF, 0.5mM DTT, 20% (v/v) glycerol) for 2 hours on ice,

changing the buffer every hour. The dialysate was centrifuged

in a microfuge at 4C for 15.0 mins and the precipitate was

discarded. The supernatant was aliqouted and frozen at -700C.

Whole cell extracts. The whole cell extracts were

prepared as described previously (Manley et al. 1980). The

procedure is done on ice and the reagents and hardware are

prechilled before starting. Briefly, the tissues or cell line

were frozen in liquid N2 in a mortar, ground to a fine powder

and transferred to a glass mortar containing 25 ml. of NLB

(110mM KC1, 15mM HEPES, pH 7.6, 5mM MgC12, 1mM DTT, 0.1mM

PMSF). The mixture was thawed for 2.0 min, subjected to a

motorized dounce for 10-20 strokes and transferred to a 50 ml

screw cap centrifuge tube. Three ml of 4.0M AmSO4 was added

and the extract gently rocked on ice for 30 min. The protein

mixture was centrifuged at 19,000 rpm for 30.0 min and the

supernatant removed to another screw cap centrifuge tube. A

fine ground powder of AmSO4 (0.3g/ml) was added to the

supernatant, rocked gently on ice for another 30.0 min and

centrifuged at 15,000 rpm for 20 min. The supernatant was

discarded and the pellet resuspended in 1.0 ml NLB. The NLB

mixture was dialyzed against 100X NEB (40mM KC1, 20mM HEPES,

pH 7.6, 0.1mM EDTA, 10% glycerol, 1mM DTT, 0.5mM PMSF) for two

hours on ice, changing the buffer every hour. The dialysate

was centrifuged for 1.0 min in a microcentrifuge, the

supernatant removed and frozen at -70*C in 100 Al aliqouts.

The concentration of protein in each extract was

determined using a BioRad Protein Assay system.

Preparation of Radiolabeled Probes

The probes used in the band retardation assay were

radiolabelled using a PCR based methodology. The primers used

to amplify the probes are given in Table 3-1. The template

from which the amplification products were derived is a

plasmid (Bluescript M13%, Promega) containing the 5.5 Kb

genomic clone of the Ad gene. The PCR reactions were as

follows: long of plasmid containing the Abd gene was added to

the PCR buffer (500mM KC1, 100mM Tris, pH 7.6, 15mM MgCl2,

0.1% gelatin (w/v), 5gM of each dNTP) containing 10 pmoles of

each primer, 5.0 Al [a-32P]dATP (3000 Ci/mmol, Amersham) and

5U of Taq polymerase (Perkin-Elmer Cetus). The reaction

mixture was subjected to 20 cycles of denaturation at 94*C for

1.0 min, annealing at 60C for 2.0 min and extension at 72*C

for 2.5 min in a Perkin-Elmer Cetus Thermocycler (model N801-

0150). Following PCR amplification, the amplified products

were electrophoresed through a 1.5% agarose gel and visualized

by EtBr staining and UV light. The probe was excised from the

gel and isolated in an IBI electroelution chamber. Briefly,

the DNA fragment was submerged in a low salt buffer (2.0mM

Tris, 0.5mM NaCl, 0.02mM EDTA) and electrophoresed into a salt

bridge (10M NH4 Acetate). The DNA was recovered by

precipitation from the salt bridge by adding 2X volume

ethanol. The probe was washed with 70% ethanol, lyophilized

to dryness and resuspended in 20 Ml sterile H20.

Isolation of Cold Competitors

Nonradioactive double stranded DNA competitors used in

the band retardation assay were prepared in the same manner as

the probes. The only difference in the protocol was the dNTPs

were increased to 20AM and the [a-32P]dATP was omitted in the

PCR reaction mixture. The number of cycles, reaction times,

temperatures and PCR fragment isolations were all done as


Band Retardation

The band retardation assays were performed in a 20li

reaction volume containing 10mM Tris, pH 7.6, 50mM NaCI, 1.OmM

DTT, 1.0mM EDTA, pH 8.0, 5% glycerol using 20,000 cpm of [a-

32P] labelled DNA, 2-3Ag poly(dI-dC) :poly(dI-dC) and 2-311 (4-

6gg) of nuclear extract or 1.5-2.0il (10Ag) whole cell


Radiolabeled DNA probes were added to the reaction

mixture after a 5.0 min prebinding (on ice) of the cold

competitor and protein extract. Binding was allowed to

proceed for 15.0 min on ice, 2.0l bromphenol blue tracking

dye (50% glycerol, H20) was added and the samples

electrophoresed for one hour through a 5% nondenaturing

polyacrylamide gel containing 10mM Tris, 1mM EDTA, 200mM

glycine at 4C. The gels were allowed to pre-electrophorese

for 20 min before loading the samples. After electrophoresis,

the gels were fixed in 10% methanol, 20% ethanol and 5% acetic

acid, dried and exposed to X-ray film (Kodak XAR5) overnight.

DNAse I Footprinting Assay

Preparation of Double Stranded DNA Probes

The probes used in the DNAse I footprinting assay were

isolated in the same manner as the probes in the band


retardation assay with a few exceptions. Briefly, 100 pmoles

of primer (Table 3-1) were radiolabeled in 50mM Tris, 10mM

MgCl2, 5mM DTT, 50lig/ml BSA using 5.0 Al [y-32P] dATP (7000

Ci/mmol, ICN Biochemicals) and 2U T4 kinase. The reaction was

incubated at 37"C for one hour, brought up to 500~l with

sterile H20 and centrifuged at 6500x g in a Centricon-3

(Amicon) for 1.75 hours. The retentant containing the labeled

primer was then collected as per the manufacturers

instructions. The probes were synthesized using one

radiolabeled primer and one cold primer in the same PCR

protocol as described previously. The PCR products were

electrophoresed through a 1.5% agarose gel, excised out and

isolated on an IBI electroelution chamber as described


DNAse I Footprinting

The binding reactions (20Al) were carried out as

described previously with the exception of doubling the

protein extract and using 300,000 cpm of end labeled probe.

After an incubation of 12 min on ice, the binding reactions

were placed at room temperature for 3.0 min. MgC12 was then

added (final concentration, 2.5mM), followed immediately by

DNAse I (BRL). The concentrations of DNAse I were 0.5pg/ml

for the nascent DNA control and 3.0pg/ml for mixtures

containing nuclear extracts. The reaction times were 45.0 sec

for the former and 4.0 min for the latter and stopped by the

addition of EDTA (final concentration, 20mM). The mixture was


extracted once with phenol:chloroform (1:1) and precipitated

using 0.3M sodium acetate and 2X volume ethanol at -70*C. The

precipitate was isolated by centrifugation, washed once with

70% ethanol, lyophilized to dryness and resuspended in 4.041

formamide with bromphenol blue. The samples (2.041, 150,000

cpm) were loaded onto a 20% denaturing polyacrylamide gel

containing 7M urea and lX TBE. The gels were pre-

electrophoresed for 1 hour and NH4Acetate (final

concentration, 1M) was added to the bottom chamber to create

a salt gradient during the run. The samples were

electrophoresed for approximately 6 hours and the wet gel was

exposed to X-ray film overnight. Sequences were aligned with

reference to the G and G+A Maxam-Gilbert ladders of the end-

labeled DNA.


Ab Expression Varies Among Different Tissues

Northern analyses were used to quantitate Ab gene

expression in liver, brain, kidney, and spleen tissues from

BALB/c (Abd) and C57BL/6 (AbW) mice. Approximately 20 Ag of

total RNA from each tissue was electrophoresed through a 1.0%

formaldehyde agarose gel and hybridized with a 5.5 Kb genomic

clone of the Abd gene. Figure 4-1 illustrates the variable

expression found in these tissues from BALB/c mice. The 1.3

Kb Ab transcript was present in both kidney and spleen tissues

(lanes marked K, KP, and S) and undetectable in either liver

or brain tissues (lanes marked L and B). The level of

expression in spleen tissue (containing B cells) was at least

25 fold higher then that of the kidney tissue as assessed by

densitometric analysis (BioRad, Model 260). Similar results

were obtained with tissues from C57BL/6 mice (data not shown).

The kidney was perfused with 3.0 ml. PBS before isolation

of the RNA to verify that the expression found in the kidney

is specific for that tissue and not due to passenger

leukocytes. As shown in Figure 4-1, the level of expression

was the same with or without perfusion (compare lanes K and

Figure 4-1. Representative Northern blot of different
tissues from BALB/c mice. RNA (20 pg) was electrophoresed
through a 1.0% agarose gel, transferred to a nylon membrane
and hybridized with a 5.5 Kb EcoRI-EcoRI Ab genomic clone.
The 1.3 Kb Ad transcript is labeled. L, liver; B, brain; K,
kidney; KP, kidney perfused (3 ml. PBS); S, spleen.




KP), indicating that at least some kidney cells transcribe Ab.

These results are consistent with previous serological data

showing class II expression in several tissues of the kidney

(Halloran et al. 1988; Barrett et al. 1987; Halloran et al.


Expressing and Non-expressing Tissues have Different
Chromatin Structures

The northern analyses demonstrate that there are three

types of expression in vivo. The chromatin structure around

the Abd gene in each of these tissues was assessed using a

DNAse I hypersensitivity (DH) assay. All DH sites shown were

confirmed by assaying from the 3' and 5' direction using

multiple probes (probes 1-4, Figure 3-2) and double digests.

Representative results used to identify DH site positions in

the Abd gene in splenic B cells is shown in Figure 4-2a. The

DNA samples were digested with BamHI, electrophoresed through

a 0.7% agarose gel, and hybridized with probe 1 (Figure 3-2).

Six DH sites were detected in the Abd allele, their DH site

designation and positions relative to the +1 codon are as

follows: Dl, -3200; D2, -1500; D3, -400; D4, +790; D5,

+1700; and D6, +2200. Figure 4-3 summarizes these findings

and presents the genomic positions of the DH sites in Abd.

Liver and brain tissues were assayed to determine the

chromatin structure of the Ab gene in non-expressing tissues.

Figure 4-2c shows a representative DH assay comparing liver

and kidney tissues (non-expressing vs. low). Sites D1-D4 were


found in both expressing and non-expressing tissues, while

sites D5 and D6 were present only in the expressing tissue.

The DH site distribution and sensitivity found in brain tissue

is identical to that of liver (data not shown). These results

indicate that Ab expression correlates with the acquisition of

2 new DH sites (D5 and D6) located in intron 2 and increased

sensitivity at all other sites, with the exception of site D2

(compare the 3 min data points, Figure 4-2c). This indicates

that the chromatin structure of Ab is in a tighter

conformation in the non-expressing tissues than in the

expressing tissues.

The DH assay was performed on kidney tissue from BALB/c

mice to evaluate potential differences in the chromatin

structure between a high and low expressing tissue. Figure 4-

2b shows a representative Southern blot used to localize DH

site in the Abd gene in low expressing kidney tissue. All six

DH sites found in B cells were present in kidney tissue.

These results indicate that transcriptionally active Ab genes

have a characteristic chromatin structure and that the rate of

transcription may be independent of the chromatin


Polymorphism of DH Site Position Among Different Haplotypes

We have previously demonstrated that two evolutionary

lineages of Ab alleles diverged a minimum of 5 million years

Figure 4-2. Representative DNAse I hypersensitivity assays
of various tissues from BALB/c mice. DH site designations are
given on the right of each blot. Marker sizes are shown on
the left. All panels were digested with BamHI and hybridized
with probe 1 (Figure 3-2). A) High expressing B cells
isolated from total splenocytes. Above each lane is the
concentration of DNAse I (/g/ml) added to each sample. The
samples were incubated for 3.0 min at 37*C and the reaction
was terminated with 100 mM EDTA, 1% SDS, proteinase K (0.4
mg/ml). B) DH assay of low expressing kidney tissue.
Samples were incubated with 30 ig/ml DNAse I and aliqouts were
taken at varying time intervals. The control sample (lane 1)
was incubated at 37C without DNAse I for the duration of the
assay. Above each lane is the minutes of digestion at 37*C.
C) Comparison of non-expressing and low expressing tissues.
Lanes 1-9 are liver nuclei, while lanes 11-13 are kidney
nuclei. Above each lane is the minutes of digestion with 30
j/g/ml DNAse I at 37'C. The control sample was treated the same
as in B above.

0 .1 .4 .5 .6 .7 .8

A l

-l V ,:

~ j

0 1 3










0 1 2 3 4 5 6 9 12

.. 4.





0 1 3

MtAi .

*HI 9 *









* "0 to
r-4 0 -H

04.) 4J

g4) fl
-M H
>.0 )
S o d

0 > wo 00

SM .q -

$ 0H

*4.) H3

40 0 --
-M H

> 0 01'

-4.) to

O- M
S*HH 0
4rI .> ) 0

g H 0 M

OS 0

V r (>0

0 .


04 & 0 H
Q ) IB ,Q *


I- I- I-

CO- Co- C1-

\- \- 2- \-E

-4g ,-48g M-

48 48 48

o co- c- <

1 co C c
!| ? 1
I i ^0


ago (McConnell et al. 1988). These lineages differ in > 10%

of their intron sequence and can be distinguished by the

presence or absence of an 861 bp retroposon insertion in the

second intron separating the P1 and P2 exons (McConnell et al.

1988). Two prototypic alleles from lineages 1 and 2 are Abd

(BALB/c) and Abb (C57BL/6), respectively. To determine

whether the positions of these DH sites are conserved in

highly divergent allelic lineages of Ab, the DH assay was

performed on the same tissue distribution in C57BL/6 mice. A

representative DH assay of C57BL/6 B cells and the DH sites

generated are given in Figures 4-4 and 4-5, respectively. The

DNAs were digested with HindIII and hybridized with probe 2

(Figure 3-2). In addition to the six DH sites present in the

Abd allele (D1-D6), the Abb allele contains two unique DH sites

located in the retroposon insertion in the second intron

(D7,D8). Site D1 was determined by a double digest using

BamHI and HindIII and hybridizing with the 5' probe (data not

shown). The DH site distribution and sensitivity observed in

the low and non-expressing tissues of C57BL/6 was similar to

that observed in BALB/c (data not shown). For example, the

kidney tissue exhibited an identical chromatin structure to

that of the B cells, while the liver and brain tissues had a

decreased sensitivity to DNAse I and lack sites D5-D8 (Figure


The appearance of two polymorphic DH sites (D7,D8) in the

retroposon of C57BL/6 mice suggests this particular sequence


is responsible for the generation of these sites. Therefore,

the chromatin structure of the Ab gene was assessed in other

representatives from each lineage (McConnell et al. 1988).

Figure 4-6 presents a summary of the results obtained with B

cells from each strain. Sites D7 and D8 are only found in

those lineages that contain the retroposon insertion.

Unique Protein-DNA Interactions Associated
With DH Sites D5 and D6

A wide variety of functional sequences are associated

with DNAse I hypersensitive sites. For example, DH sites

occur around enhancers, silencers, upstream activation

sequences, promoter elements, terminators, topoisomerase sites

and recombination loci (reviewed by Gross and Garrard 1988).

Thus, it seems clear that these sites are nearly always

associated with cis-acting DNA sequences. Furthermore,

several classes of nuclear proteins have been found associated

with a subset of these sites, including RNA polymerase II,

topoisomerase I and II and transcription factors.

Given these facts, I wanted to assess the protein-DNA

interactions associated with the DH sites correlating with

tissues that are actively transcribing the Ab gene. A band

retardation assay was performed to scan 600 bp in intron 2

(containing these sites) looking for unique protein-DNA

interactions. A PCR based protocol was used to isolate probes

amplified from a genomic clone of Abd. Ten primers (Table 3-

1, Figure 3-3) were used to generate five probes of

Figure 4-4. DNAse I hypersensitivity assay of C57BL/6 B
cells. Marker sizes and DH site designations are given on the
side. Nuclei were isolated and incubated with 15 pg/ml DNAse
I for varying time intervals at 37*C. The minutes of
digestion are given above each lane. The control sample (lane
1) was treated the same as in Figure 3b,c. The DNAs were
digested with HindIII and hybridized with the 1.7 Kb HindIII-
HindIII 3' probe (probe 2, Figure 3-2).

0 1 369

6.7- D2

Ii7 D3

4.3- D4


.., D7
2.3- 4D8

a) 4) -r-4 a) tO
OH Ha)
* 40 M *q
() 4 a) Q 4
*r QH r-1

r* -I B M

0tO A 4 >4
eCI, a)
p to4)~
44 10 4-)

0 ri^"

0 giU

QM u m'
a *Mg
4 l 4 )

0 r) 0)



0 >
~a) *-I5
a) aP (

-lH *^

*.) .4
tLa),-4t Hn3

<1 4J (0

-4 V S

# <8 C M
S* M &> *i- "
4 0 0) ) H

H kf 'U-H1

*M 4 oU)4)
j r xa m
0. 4 ) rj





4) to0 N >

0 P4 14 H 0 0
-,1- 4 14 1 -
V- ,M 0 4)
OHr-l4 0
r-4r t<| 0 0)

to.I 0

04 0

,. to4 0
S0 4 0
.4 v

4) 0

m0 E 0

5 "- a o
i 01 p 4 ,-

IA0g '- *4

4) .d 0t
10 -0 $-.-

4) 0E oJ 0

0 H 0 -H

V *V H 0 *4- H
in ,n 0 p 4 r i

V*fd | -j
Id 0 N 04 03

44 .Q *r4 IV p r-4 (fl




co Cm

co CD
co-/ c -
S8 8









u. .


I ~


"r -

w -48


I -

I -






approximately 150 bp each and overlapping each other by 50 bp.

The radiolabeled probes were allowed to react with A20 B

lymphoma protein extract for 15.0 min on ice and

electrophoresed through a 5% non-denaturing polyacrylamide

gel. The gel was fixed, dried and exposed to X-ray. Figure

4-7 illustrates the protein binding patterns observed for all

five probes spanning the 600 bp in intron 2. As can be seen,

incubation of the first four probes with the A20 nuclear

extract generated protein-DNA complexes that migrated slower

than the free DNA, however, the fifth probe does not. This

result would imply that there are four unique protein-DNA

interactions within this region of intron 2. However,

competition experiments for each probe are required to deduce

the specificity of the protein-DNA interactions found within

this region.

Using a series of competition experiments in which cold

specific and non-specific DNA competitors are added to the

reaction mixture prior to the addition of radiolabeled probe,

the specificity of the protein-DNA interaction was deduced.

Figure 4-8 is an example of a competition experiment using the

second probe. The first two lanes illustrate the binding

reactions without the presence of competitors, while lanes 3-9

demonstrate the competition of cold competitors with the

labeled probe. Of the five fragments and BRL 1Kb DNA ladder

used as cold competitors, only the fragment corresponding to

the probe (lane 3+4) completely competed for the binding

activity. This demonstrates the binding observed is specific

for the second probe. Furthermore, the protein-DNA

interaction associated with this fragment was localized to the

50 bp segment between primers 2 and 5 (Figure 4-8), since the

first and third fragments could not compete for the


The competition experiment for the third probe is

illustrated in Figure 4-9. The lanes were set up in the same

manner as in the previous figure, where lanes 3-9 represent

the addition of cold competitors. Unlike the second probe,

two fragments were able to compete for the binding activity.

The first is the equivalent sequence (lane 5+6) and the second

is the adjacent fragment (lane 7+8). Because the BRL 1Kb DNA

ladder failed to compete for this fragment, one can conclude

the protein-DNA interaction was specific for this probe.

Furthermore, the region of binding in the third probe can be

localized to the 50 bp segment shared between them because of

the cross competition between the two adjacent fragments (5+6

and 7+8). This was confirmed by using the 50 bp region (lane

7+6, Figure 4-9) to successfully compete for the binding.

To deduce the binding location and specificity of the

fourth probe, a competition experiment was performed (Figure

4-10). Similar to the third probe, the fourth probe was

successfully competed away from binding by two fragments, the

identical sequence (lane 7+8) and the adjacent fragment (lane

5+6, Figure 4-10). Again, the activity observed in the fourth


probe appeared to be the 50 bp segment shared between the same

two fragments described above. Likewise, the 50 bp region was

able to effectively compete for the binding (lane 7+6).

To confirm that the protein-DNA interaction observed in

the two probes was actually the region shared between them,

the 50 bp segment was radiolabeled and a competition

experiment was performed. Figure 4-11 demonstrates that the

identical sequence (lane 7+6), third (lane 5+6) and fourth

(lane 7+8) fragments were able to effectively compete with the

labeled probe. These results confirm that the protein-DNA

interactions observed in the third and fourth probes was

actually in the region shared between them.

The competition experiment performed on the first probe

is illustrated in Figure 4-12. The protein-DNA interaction

associated with this region appeared to be non-specific

because the BRL control competed for binding as well as the

identical sequence (lanes BRL and 1+2, respectively). This

result indicates the protein(s) is not uniquely interacting

with sequences in the first fragment. Additionally, a

titration of competing fragments (50-100ng) confirmed the non-

specific control is just as effective a competitor as the

identical sequence (Figure 4-13).

The band retardation assay was performed with the second

and third probes using whole cell extracts from A20, liver,

brain and kidney tissues to determine if the protein(s)

binding within these regions are present. Figure 4-14

Figure 4-7. Band retardation experiments of the five probes
spanning the 600 bp in intron 2 containing DH sites D5 and D6.
The thin line represents the intron and the black box denotes
the end of exon3. The five black bars represent the probes
used in the assays with the experiment below. The numbers
represent the primers used to generate the probes (Table 3-1,
Figure 3-3). In each experiment 20,000 cpm of probe was run
without (first lane) or with (second lane) A20 nuclear
extract. Each experiment used 341 (=7Ag) of nuclear extract
and 3Ag of poly(dI-dC):poly(dI-dC).

D5 Intron 2 D6
I 3 4 7 8
1 2 5 6 9 10

M al

Figure 4-8. Competition experiment with the second probe.
A diagram showing the relative positions of the probe and
competitors within the second intron is given. The probe used
in this assay is denoted by the black bar, while competitors
are dotted. The first lane is the free probe (20,000 cpm)
without the addition of A20 nuclear extract. The remaining
lanes contain 3pl (z7ig) of nuclear extract and 3ig of
poly(dI-dC):poly(dI-dC). Each competitor (30ng) is given
above the lane and is indicated by the two primers used to
generate the fragment. BRL: non-specific control, 1Kb DNA

Intron 2

3 4 7 8
1 2 5 6 9 10



+ +
- -I

+ + +
- in I+-

+ + +

IL k

o _1
+ n-

+ +

Figure 4-9. Competition experiment with the third probe.
A diagram showing the relative positions of the probe and
competitors within the second intron is given. The probe used
in this assay is denoted by the black bar, while competitors
are dotted. The first lane is the free probe (20,000 cpm)
without the addition of A20 nuclear extract. The remaining
lanes contain 341 (=7Ag) of nuclear extract and 3jug of
poly(dI-dC):poly(dI-dC). Each competitor (30ng) is given
above the lane and is indicated by the two primers used to
generate the fragment. BRL: non-specific control, 1Kb DNA

Intron 2

3 4 7 8
2 5 6 9 10


+D CM 4t O 0 (
+ + + + + +
) c r3 ,- 0) IN CO

- + + + + + +

+ +

Li Li

w^B ito U

Figure 4-10. Competition experiment with the fourth probe.
A diagram showing the relative positions of the probe and
competitors within the second intron is given. The probe used
in this assay is denoted by the black bar, while competitors
are dotted. The first lane is the free probe (20,000 cpm)
without the addition of A20 nuclear extract. The remaining
lanes contain 34l (=7/g) of nuclear extract and 3Ag of
poly(dI-dC):poly(dI-dC). Each competitor (30ng) is given
above the lane and is indicated by the two primers used to
generate the fragment. BRL: non-specific control, 1Kb DNA

Intron 2

1 2

4 7
5 6

9 10


N wt <0 T- CD O j
+ + + + + + cr
-- -- -O- c I- o
- + + + + + + + +


Figure 4-11. Competition experiment using the region shared
between the third and fourth fragments as a probe. A diagram
showing the relative positions of the probe and competitors
within the second intron is given. The probe used in this
assay is denoted by the black bar, while competitors are
dotted. The first lane is the free probe (20,000 cpm) without
the addition of A20 nuclear extract. The remaining lanes
contain 3A1 (=7Ag) of nuclear extract and 3ig of poly(dI-
dC):poly(dI-dC). Each competitor (30ng) is given above the
lane and is indicated by the two primers used to generate the
fragment. BRL: non-specific control, 1Kb DNA ladder.

Intron 2

5 6



(D (D 0
+ + + i:
.- LO P-. i

- +

+ + + +


Figure 4-12. Competition experiment with the first probe.
A diagram showing the relative positions of the probe and
competitors within the second intron is given. The probe used
in this assay is denoted by the black bar, while competitors
are dotted. The first lane is the free probe (20,000 cpm)
without the addition of A20 nuclear extract. The remaining
lanes contain 341 (z7Ag) of nuclear extract and 3jig of
poly(dI-dC):poly(dI-dC). Each competitor (30ng) is given
above the lane and is indicated by the two primers used to
generate the fragment. BRL: non-specific control, 1Kb DNA

Intron 2

3 4 7 8
1 2 5 6 9 10



- P- +

- + + + + + + +

i L L4 U i

C 0)
0 r
.4 4J


4-) Vi-

No -
A u



0 t( 0
(- -H1

(u Wv '-


04 *r4
V 4)

to ,
9 -H


44 4J
* H 0

4 -p 4

0 0
& .
M $4

5I >i4-
P4 to

4 -H

-H (A ~
*E-4 4) j

g 01
>) 4*-I
tv > tp IA V-
o'i W1 V
44 to to

0 t

H **
.a ..
4 u

H U.

0 0
tp "


to 4) 4)
4 > >
4 0 0


M +


I +

o W

1 40 C
HHl 01

4 0

QO iMi
0 4

I.m .

01 4c0
o)* o

r. 0 .

,' H

m o) u 0

o q

o r- 4

.,4 4o 0

01 tr m

S*- -

rj .. t

ON > > V








presents the autoradiographs produced from these experiments.

Protein-DNA complexes within the second probe were found using

the brain and liver whole cell extracts, while none were

detected using kidney whole cell extract (Figure 40-14b). It

is not clear whether the kidney tissue does not contain the

protein(s) or that there are very low levels in the extract.

However, the brain, liver and kidney tissue clearly contained

protein(s) that can bind to the third probe (5+6, Figure 4-


Conclusions that can be inferred from the previous band

retardation assays are as follows: There are two unique

protein-DNA interactions located in the first 600 bp of intron

2. These interactions are localized to two 50 bp segments.

Their positions relative to the +1 codon start at +1850 and

+2050. These regions are well within the estimation for the

positions of the two DH sites (D5, +1700 and D6, +2200) found

in expressing tissues. The presence of protein-DNA

interactions within these DH sites leads us to postulate the

presence of cis-acting regulatory elements within this segment

of intron 2.

DNAse I FootprintinQ in Intron 2

The band retardation experiments demonstrated specific

protein-DNA interactions within the two DH sites in the second

intron. To delineate the factor binding site further, an in

vitro DNAse I footprinting analysis was carried out using two