Inhibition of human complement by extracellular lipoteichoic acid from Streptococcus Mutans BHT


Material Information

Inhibition of human complement by extracellular lipoteichoic acid from Streptococcus Mutans BHT
Physical Description:
x, 132 leaves : ill. ; 28 cm.
Silvestri, Louis Joseph, 1952-
Publication Date:


Subjects / Keywords:
Streptococcus   ( lcsh )
Lipoteichoic acids   ( lcsh )
Pathogenic bacteria   ( lcsh )
Microbiology and Cell Science thesis Ph. D
Dissertations, Academic -- Microbiology and Cell Science -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis--University of Florida.
Bibliography: leaves 120-131.
Statement of Responsibility:
by Louis Joseph Silvestri.
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 - 026220234
oclc - 03928986
System ID:

Full Text







In all sincerity, no occasion or project thus far undertaken has had a more humbling effect on my life than the completion of this Ph.D.dissertation. It is unfortunate that only now in retrospect can I clearly see the tremendous debt I owe for the help, patience, understanding, and knowledge so generously contributed by my mentor, my friends and wife.

If I were to formally thank everyone who contributed in someway to the successful completion of my Ph.D. dissertation, it is likely that my acknowledgements would read like the listings of a telephone book. I do however

feel compelled to thank a Few very special people.

First of all I would like to thank my mentor and friend, Dr. E.M.

Hoffmann. Ed possesses the rare ability to earn respect rather than having to demand it. His tolerance for my idiosyncrasies was unlimited (almost). He gave me direction yet lef. me with alternatives; hb challenged my intellect, vet never made me feel ignorant; he provided the foundation on which I am still building my scientific character. Most of all, he was (and is) a friend.

I would also like to thank the members of our laboratory familyy" (Suzanne, Jean, Bert and Tom) for their help, patience and tolerance during these difficult last days. One cannot help but reflect upon the many events that shape the complex web of friendships within the laboratory. All of you will always he regarded as the closes; t of friends.

I wish to express my gratitude to Ron Craig not only for his friendship, but also for the professional technical assistance that he afforded.

I would also like to expres my appreciat ion to the employees of 'eachin; Resources, for dedication above and beyond the call of duty. I would

especially like to acknowledge the professional artistic assistance of Margie Summers, Margie Nibl.ack and John Knaub.

I would like to thank Steve Hurst for being there when I needed a friend and for helping with some of the last minute photography.

The typist, Joanne Hall, deserves a partL cularlyv pecial mention.

If not for her personal concern and dedication the deadlines would never have been met. She worked on this dissertation under conditions for which no degree of monetary reinbursement could possibly compensate. I thank you Joanne and I sincerely hope you never have to go through that again!

Finally, and most importantly, I wish to thank my wife Lvn for her infinite patience and encouragement, Her attitudes, her ideals, her "being" is so much a part of me that it would be hopelessly futile to list all the things for which I am indebted to her. She is a friend and lover, a typist and an occasional laborato ry techniiLnn. S he is ': driving force in life and rightfully so, I dedicate this dissertation to her.

And t ith onie ou oCot i ao to est t he jumped at Athe' orc '

-- have I I'

e(t o d Ato he "T Ond' ag at K O e ba AaA. Scrud
struggle got &L head ito dau igh nl adsi cetul



ACKNOWLEDGMENTS ........................................ i

LIST OF TABLES ......................................... v

LIST OF FIGURES......................................... v

GLOSSARY OF ABBREVIATIONS.............................. viii

ABSTRACT............................ ................... ix

INTRODUCTION ........................ ................... I

MATERIALS AND METHODS................................... 17

RESULTS. .............. .................................. 11

DISCUSSION. ............................................. 1

LITERATURE CITED .................................. .....- n

BIOGRAPHICAL SKETCH.... ........................ .......... 1




1. Partial Puriffication of LTA byv AS-M
Gel Filtratio n ................................... 57

2. I. Results from Partial Purification
of LTA .......................................... 62

3. TI. Results from Partial Purification
of LTA ...... ....................... ............... 63

4. Percent Recovery of LTA During Octyl
Sepharose Purification .......................... 70
1 4
5. I) distribution of C-Phosphatidvl Choline
During PCV Purification of LTA .................. 72

6. Percent Reucoverv (of ,ITA from ;ario us,
Steps of PCV Puriffication........................ 73

7. Summairized Chtumicni Comnposition ,f Vairiouis
LTA Containing Sources ................ .... ........ 7

8. Specific Activitv Determinations of Puri filed
LTA............................................. 70

9. Effect of LTApcx on th Abiliti of Cls tc
Consume (' s ;rc (: 2 A ctivitv ....................... 07

10. C(ompa rison cf thb Re I t. ive Numbeors o f Ef fect ie C1 Mol I ,cul ,s Capable o f Tran; i 'r
from EACl TI 'r acted with L T'Apc. ........... ......... 1(1

11. he Inhilbition on f FA I ysi s bvy Iip(t( ,ic hol ic A\c ids from S,,vo ril Bictori i l Snurco, ...... ..... 7...7




1. Titration of whole human complement
after incubation with crude extracellular lipoteichoic acid (LTAcx).............. 33

2. Dose response inhibition of whole
human complement after iticuhat on
with varying concentrtions of LTAcx ............ 35

3. Titration of (C3 in whole human scrum
after treatment with LTAcx....................... 18

4. Complement cmpl)onent titration of
whole human srra after treatment
w ith LTAcx ...................................... n0

5. Inhihition of complemen t med it d
lysis of FA treated with varvi tng
concentrations of T,TAcx ......................... .3

6. Passive heminllgg t nation (PlA\) of EA
t reated wi th varying concon t rat i ons;
of LTAcx .................................. ...... 45

7. Effects of ,TAcx treatment on the
SYsis of var ous comp element component
intermediates ............ ...................... 48

8. PIIA of vnriouis LITArx tr heated complement
romponul nt ii te rm ed ilt i s ......................... 0)

9. EIffrct of 'lTAcx t r atm'nt o1n 1 it, 0 ve i

10. Effect of LAcx on hemolvt ic ant ibodv t it rat ion . ........ ..... ..................... .

11. Partial p; i rif cait ion o f TA hv A5-M gel filtr.ition....... ........................... 59

12. Partial. purificition of ITA by AS-M pI, 1 iltr ti o it i1 TA I I nri h( d
start in g materi I al............................... 1

13. Purification of ITA by ('ty. Sopharoso
hydrophobic gel chromatography ..................

14. Simultaneous removal of ,salt and propano l
from LTAosx by LI1-20 g.l, chromatography ......... 69

15. Carbohydrate analysis of TA containing
preparations by gas liquid chromatography ....... 76

16. Passive hemagglut i 'tin tton (PHIA) t itrati ion
and inhibition of complement mediated lysis
of EA treated with varying concentrations
of LTApcx........................................ 81

17. Effect of LTApcx on the complement
mediated lysis of various cellular
complement component intermediates ............... 83

18. Effect of LTAcx and LTApcx on functionally purified human C1 ........................... Q

19. Immunodiffusion and precipitation analysis
of various steps in the purification oFi
human Clq.........................................

20. Disc gel electrophoresis of purified
human Clq .......................................

21. DEAE elution profile of human C(ls............... 03

22. Immunoelec trophoresis of human Cls and
C ls ............................................... 95

23. Effect of LTApcx on the ability of C1:
to hydrolize TAMe ............................... 99

24. Difference in complement medilated yvtic
susceptability of ITApx treated EAC4 versus
EAC( 4........................................... 103

vi i


A: Antibody C: Complement Cl, C2---C9:a Complement components. Horizontal bars
above the component des ignation denotes
a biologically active state. CVF: Cobra venom factor E: Erythrocyte EDTA: (Disodium) Ethylenediamine tetraactic acid ECTA: Ethyleneglycol-bis (f Amino Ethyl Ether) N,N tetra
acetic acid

LTAcx: Crude extracelular lipoteichoic acid LTApcx: Extracellular LTA purified via phosphatidyl-choline
vesicle adsorbtion

LTAppx: Partially purified extracellular LTA LTAosx: Extracellular LTA purified via Octyl Sepharose gol.
column adsorbtton

LPS: L ipopolysncc hna ride PHA: Passive hemagg 1 utination PHAg: Passive humagglutination (modlifted method) TA: Teichoic acid TANE: p-Tosyl-1-arginine methvlester

All complement nom~,n I;ltur, follows the WHOfI recommendations

(Bull. Wld. Hlth. Org. 39:939, 1968).

Vi i i.

Abstract of Dissertation Presented to ithe
Graduate Counci of the UlniversiLv of Florida
in Partial Fulfit lment of the Requirements for the
Degree of Doctor of Philosophv


Louis Joseph Silvestri

December, 1977

Chairman: Edward M. Hoffmann
Major Department: Microbio logy and Ce ll Science

A number of biological and chemical similarities exist between the lipopolysaccharides (LPS) of gram negative microorganisms and the lipoteichoc acids (LTA) of gram positive organisms. The potent affects of LPS on the complement system are well documented: however, the effect of LTA on this host defense system has not been adequately studied. Furthermore, all studies thus far conducted have ben limited to the interaction of LTA with whole fluid phase compcloent. In this investigation it was demonstrated that extracellulr LTA from the cariogen ic mi croorgan ism Sutreptococcu as mrt;ns BIT was not only capable 1 of spontaneously binding to sheep erythrocvte target cells but was also capable of rendering them refractory to complriment mediated Iv i:s. Purification of the I,'A to homogenoittv was achieved by a combination of go[ filtration and adsorhtion to phospholipid chol ine vesicles (artificint membranes). By utilizing v;ritous cellular complement component intermediate complexes and firnctiona llv purif ied complement components. experiments were conducted to define the site and mechanism of Inhibition


by LTA. The site of inhiiliton "as determined to ocrur between the formation of the SAC1 and SAC142 Compllex. Bcaus C Cl is no 1 longer necessary after formation of the C3 convertase (SAC42), lack of inhibition after this step implies a direct effect on Cl activity. Although experimental data derived from utilizing CI, Clq, Cls, and C1s were suggestive, data did not unequivocally establish this as the precise mechanism of inhibition. No evidence for fluid phase consumption of hemolysin Ab, CI, C4, or C2 by ,TA could be demonstrated. Evidence for the inhibitory activity of LTA from severe unrelated genera is presented and the possible role of LTA in periodontal disease is discu.I



As reviewed by Wlicken and Knox ( L,2), a number of chemi c;1 and

biological similarities exis:t between the l ipopolysa chcarides (LPS) of gram negative bacteria and the lipoteichoic acids (ILTA) of gram positive organisms. Because of these similitudes ou.r laboratory began to investigate whether LTA possessed anti.complementary activity analogous to that associated with the LPS endotoxin (3-8). Although there have been concentrated efforts to define the site and mechanism of LPS inhibition of complement, very few investigators have reported data on the possible effects of LTA on the complement system (9). This is somewhat surprising since the interaction of LTA, LPS, and complement almost certainly plai a significant role in the etiology of periodontal dis:eses~5. Bacterial products and sernm components in the gingival crevices of the oral cnvitv have been shown to act ivae complement by both the classical (10, 11) and the alternative pathways (12,13). In fact recent evidence suggests that bone loss (a major clinical manifestation of acute periodontal disease) may occur via osteoclast activation due to the interaction of complement and prostaglandin E (14). Prostagla dins are naturally occurring cyclized derivatives of unsaturated long chain fatty acids (15) and their conc:ntrations are dramatically ly elevated in inflamed gingival tissues (16). It is of interest to note that both LTA and LPS are also capable of initiating osteoclast mediated boue resorbh t inn (17) and this activity prrc,,ed.s without the contribution of complement t (or prostgland ni The potential fo syvnergism cannot be overlooked, and indeed t.'S endotoxins hayv lIong hoon implicated ;a partiriplnrs in tlho development of peri dental lo sions (,-R). Analogous LTA activity could be of significant clinical, import especially


in light of the fact that gram positive bacteria represent the majinr cellular constituent of dutntal plaque nat the early stages of plaque formation (18). Most of the gram positive organisms found in dental plaque have been isolated, cultured, and identified. The production of copious amounts of extracellular LTA by several of thpce organisms has been well established (19,20). In fact, growing urender conditions estimated to reflect the growth rate in the oral cavity, Wicken and Knox have shown that the cariogenic bacterium Streptococcus m-tans BHT produlc s some eleven fold greater amount of extracellular ITA in tlhe culture fluid than that contained within the cells themselves (1,2). Therefore, if an effect on complement by LTA can be demonstrated in vitro, an in vive model can be readily envisioned. Preliminary experimentation with a crude LTA containing extract from S. mutans RBT did indeed indicate that complement activity was consumed However, consumption or alteration of complement activity can he due to a number of specific or non-specific factors. Because of the complexity of thi; system, n thorough understanding of the possible interactions is necessary before any mod 1 attempting to define a site and mechanism -f inhibition can he elucidated.

The complement system of vertebrates is comprised of at least

eighteen discrete plasma proteins capable of interacting in a specific and sequential fashion. There are two pathways by whtch this hiochemical cascade may b( initited and they a rc eferrvd to as the ci -a sical and the altern;ltive pathwa;vs ol complement activat ion. low'ver, regardless of how the activation scheme is init I tc td, the h io ogii consoquerlnces of act ivat ion are the same for both pathwayvs:

Silvestri et al. 1 q76 Ahst. Ann. Meet!in ASM, p77.

1). philogogenic activity modinted via complomant reaction b'y-prodr,:tc 2). increased opsonic susr cptibility of fnreigri s l .tmliu es 3). irreversible physinchemial membrane damn ge--a d ultimately, c 'ytol.vsis--of susceptahle target cells. A,]thoiugh the importance f coimplement as a component of the hIost defense system has been suspected for quite some time, only recently has its biomedical significance been firmly established. Indeed, the participation of complement in host resistance to infections and in several disease mechanisms is a topic which has generated considerable research interest in recent years (21,22).

The classical pathway of complement contains elven discrete glv;aproteins representing nine distinct components referred to sequentially as Cl through C(9. Cl is actually a multimolecular complex of three distinct proteins (Clq, Clr, and (Ils) and the aggregate is held together h the divalent calcium ions (23). Removal of caliilm inns hv chlelating agents such as ethylenediaminetetrnacetic naid (EDITA) results in the disassociation oE Cl into its subcomponents with concomi tnnt loss of acti ity

(24). Activation of the classical pathway is character ized by a dependence on IgG or igM antibodies complexed with antigenss. The classical pathway also specifically requires; the components Cl, (2, and C7 ans well. as the divalent cations ca;lciumi and magnesium. Although the componentCl3. and C5 through C9 are us iylv considered part of the classira syvsterm, they are :;shirrod by tlh' alt'runti\v' patlhwav ;ind li,: .thu re n ti cllon idr' d .As unique components of the clawsicnl1 system per so.

Thie rong'm nition .unl init iat ion f unct ion within ros 't to i mu oglobul) ins r s ides with the sui omr ponent (25,2 ). Clrq itself i; rather peoculir protein consisting of three different polypeptide -,h ins

(17). Chemically, Clq Contains npproximatel 10' I cirb h'dr"ite, 3K,

hdroxyproline, 27 hlvdromxvlvsin and 18 W yvrine. Tihi unusml colagenlike composition makes it unlike any plasan protein yet described (28,29).

When complement is activated by ant:ii ody-antigen complexes such as exists on the surface of an antibody sensitized erythrocyte (EA), it undergoes a self assembly process sequentially depositing the entire fluid phase cascade onto the surface of the target. Specifically, Clq recognizes a previously sequestered binding site located in the Fc fragment of IgG and IgM (30,31). The three polypeptidc chains of Clq are physically arranged in a manner perhaps analagous to a six headed mace or bola with each "head" representing a binding site for an igG molecule

(32). Thus each Clq molecule has six bind ing sites for IgG (and presumably the same number for gM). Internal activation of Cl probably is the result of a conformational change in Clq w:ich in turn induces a change in the pronzvme Clr (33). Once CI.r is act ivated to Clr it is endowed with enzymnti.c activity through which the pranzvm, ('1Cs i' converted to CL esternso. (Cls)(34,35, 3,) C(s is a sericn esternse and is inhibited by diisapropylphosphofl luoridate (DFP) (371. This estrs act ivitv call he used to hydrolvyze the svnthetiic subscrates p-Tosyl-l rginilne methvlester (TAMe) and N-actyl-l---tyrosnine ethv!oster (ATiC) (38). ecrentlyv Toos and Raepple have demonst rated thit rn, polyanion were cahpahle of inhibiLtLng the activity of Cl either hv ip rF ering with Clq hiding to the antLinbody-antigen complex x, or 1by pVrvent iIng int.era'ctL iO ol i itl with C; ( q,40(). Altho gh hindin, of '1 .iia ll lI OAdIs to .ACt iC t iOn, the two processes are noLt integral--tl" with modific'd trvptophl n ('1) and the human immunogloblul in s-;ubcltas 1qGk (A2)--ho th behind (Cliq b t d, not activate Cl.

After active at ion, Cls enzymatical y cleaves (CA into a large (CMbl

and small frament (C4a) (63). The cleava; of (4 expose s a rembrne attachment site on the CAh molecule and it will attach to the antihbodvantigen complex at a site juxtaposed to the C1-antibodv complex (44,45). Cls then cleaves C2 into C2a and C2h (46) with C2a attaching to the C4b site and C2 being released into the fluid phase. Thus, the molecular complex C.b2a is formed and is referred to as C3 convertase because it is capable of splitting and activating C3 (47,48). (C3 convertase is also an esterase, and although C3 is its natural substrate, it nlso hydrolyzes the ester bond of acetyl-glycl-lysine methyl ester (49). The catalytic site of C3 convertase is believed to reside in the C2a subunit and even after release from the CA4 complex, cvtol yticaly inactive C2a retains esterase activity, but is no ]clnger capable of cleaving C3

(49). The enzymatic half-life of CAb2a is quite ephemeral--nly 10 minutes at 37'. However, if the C2 is first oxidized by troltmernt With iodine (applicable to human but not guinea pig C2) not only is the binding of C2a to C4b enhanced, but the half life of the birolecular complex is increased 20) fold (50). No doUt the tLins int association of C2a with the C2 and (A3 complex pla s a vital role in controlling the comple-ent react ion by tomporarilv l in g tiH the functional assocr iation of these complex enzymes.

Once C3 is cleavd into C3l and (C3, the small (:a fram:nn t is

released into tihe fluid phase and C l becomes as--s int tedi wi th the C4b2a complex anj with other non-h mnol\t i r itr; on th .irg,'t membrane ( '7). The associ tion of (3b with he (, convertasc" its activity so that now M, becomes the natural substratc of this triinoiculr ormpleox. The C423b complex is ref.rrd to as (*5 conlvertase (l1) and like C '2, is a hihIv s"ecialimed protease. lust as (l is the onylv knon pret in

sul)strat for C42, C5 is the only known substrate for C1423.

Once CS is cleaved into C5a and (C51, C n in; r e l-,r sed inii tht: fl iid phase and C5h transiently acquires the abit it v to I id one O Inl cii each of C(6 and C7 (52,53). With this, a self-assemblv process is initiated and results, without any further enzymatic activity, in the formation of the stable C5b-9 comp].ex (54). It should be noted that the small by-product fragments C3a and C3a are endowed with marked philogogenic activity (55,56,57). Some of these activities include release of histamine from mast cells, contraction of smooth muscle tissue, directed chemotaxis of polymorphonuclear Icukocytes,and vasoldiation both in conjunction and independent of histamine activity (58). Such potent pharmacological activities obviously play a major role in the normal course of the inflammatory response.

Once the C5b67 complex is formed, it too can hind inonspecificallv to areas on the membrane other than at the location iF the CS convertoise

(52). The trirmolecular association of C567 provide s tire molecular rran emernt for the adsorptivle binding of one mol cutlc of C(:R which in tLurn provides a binding region for up to six molecules of () (54). A Low grade lesion of the target membrane occurs :.ith cnly t:he, additionn of C(:8 L ti t complex (+9); ,ut with the binding of :9, a ten component macroi olcr!.lr complex is formed which greatliv 'enhlli; ces the rate of tll):c 'et 1 "tlvsis (54). It should h" noted that the CIh 7 compl x ,r even the C b complex r;irl attach to lOln-se5ll ti z e'd "i 111m nno lit Iby-st Illid'" r l Is rnd t uIs promot )t ( 1 feormirnal vI tol vt i event. 'hi phi om,, n hen termed "reactve ivysis" (60) and is controlledd iby the rnp id dcnv of thie 11111unbund omple recis (1t m62)

The precise merch:ni:m by which complement mediaites cvtol,sis of

susceptible target cel is is not clearly understood. (One h 'octis, in light of the newly disrcvred tributvrini n' activitv of C7. is tOat the lytic event is caiised by an enzymatic attack on the lmlrhrao ne (631. Ilowever, no enzymatic degradation products have ever been disrnvere>d in either lysed cell membrnnus or in ruptured svnthet:ic lipid hil;aers (64). The two most favored models taree the "doughnuilt" insertion hvoerihis (05) and the C8 insertion model (29). The former model purports that the C5b-9 complex inserts inself into the membrane as a "prefabricated h e'le" allowing the exchange of inLra- and extracellular material vi rn an internal hydrophilic channel (63). HIowever, the model fails to explain how: the hydrophilic complement components enter the hydrop obic expanses of the membrane. In addition, :lthlrogh electron microscpv has revealed apparent ultrastructure doughnut shaped "les ons"' n the surface of i: s lysed by complement (66), freeze etchlin g tLchniques have shown that the ultrastructure alterations are confined to the neteorleaflot of the mmlmbrane, i.e. the lesion does not penetrate the mTmbrane (67). The (' insertion model embraces most of the sali nt features of the2 do..ughnut ::odeli but in addition postulates that thie a and y chai ins of Ci ;ire inserted into the channel formed by the surface macromolecular complex. h c -, d y chains thus extend into the membrane hilaver causing disruption if orderlv structure.

In add ition to the rentraints placed on the ncompl lent icasnl diTin to the rapid decay of several of the i.termdiatos, there are two naturally o rcurring inhibi tors of complement present in the icr of main and. probably in all vrtebr t s. The first inhibitor is r ief rre'd Ito is Cis inhibitor and, as the name implies, it directly abrogates the her,lvtic and sternlvti .:at ivity of CL (8,fM ). The sn cond inhibitor i

referred to as C3b inactive tor and cleaves both soluble and rill bound C3b into two antigentally distinct fragments, C3c and C3d (70). As a result, C423 loses C5 convertase activity, and C3h activation of both the alternative pathway and the immune adherence phenonenon is abolished (71,72,73). This latter activity can be visualized by the clustering of cells bearing C3b on their surface around other cells displaying C3b receptors. Such receptors have been shown to be present on human erythrmcytes, polymorphonuclear leukocytes, platelets, macrophages, and on lymphocytes (74,75). The attachment of C3b not only plays a direct roie in the increased opsonization of target cells (76), but C3b binding to B lymphocytes has been postulated to play a role in B-cell activation as well (77).

The second pathway by which complement may be activated is referred to as the alternative or properdin pathway. llistoricallyv, the existence of this pathway had been suggested as enrlv as 1954. At th at time, Pillemer and his associates reported the discovery of a; new protein in normal human sera (78). Properdin, as it was called, was c capable of reacting non-specifically ith diverse nnaturally occurring polysn'charides and l ipopo I ysaccharides ultimately resulting in the activation of complcnimnt. This process ostens bihlv occurred without the interaction of antibody and was proposed as a m:ajo pathwayv hv which s lsceprtibl hauci;itora and viruses werp destroyed. However, th s provocative hypothesis was discarded as opocryphaLn and the described octtviiLs were attributed to the presence ofI atral annti Lbodie (70). lThe controversy remained utnresolved until recent vears whIen igorous immunochmical techniques wore employed in the isolation, puriFication, and determi nation nf function of many of these components.. The mn~inticip ted compl'xitv of the properdin

system has spawned a multiplicity of modolq attempting to eliucidate its precise mode of initiation and function. Clearly, a plethrn of diverse stimuli are capable of activating this pathway, and this fact alone imposes a formidable constraint on any molecular model. Some of the more common naturally occurring activators of the alternative pathway include bacterial and fungal cell wall 1 constituent s such as ILpopolschiaride, zymosan, and inulin (a polyfructose) (71,80-83). In addition, aggrega tes of some immunoglobulin classes (84,85), some types of animal cell membrane constituents (86,87), and antibody-coated budding virus infected cells (88,89) also stimulate this pathway. The alternative pathway can even be activated by substances of relatively defined chemical nature such as benzyl-B-D-frc topyranoside (90), rolyglu coso with rapititious a 1-3 and branched a 1-6 linkages (91), din it rophnylatod albu- in (92), and many polyanionic substances. Cobra venom factor (a non-lipolytic, non-hemolytic glycoprotein isolated from tihe venom of the cohra Naga naja) is also a potent activator of complement cvtolvtic potential, but it appears to act as a (3b analog and is thus unique in its mode of alternative pathway activation (93,94,95). Potentiat ion of this system requires devalent magnesium ions and the interact ion of at Least five novel serum proteins. By convention, the names of these proteins are TF (or initiating factor), P or F (properdin). Factor B (Cl pr:n tivator), Factor B (C3 activator), and Factor ID or I ((:3 proctivator convert:-se). To date, all of the above components have been isolated, puri f ied, and character: ,d as to molecular weight, cl oct rophoretic mobility, and sodimentation coefficients (83,96-8S). C3b (t the classical pathwav) plays an intregal role in the a lternative pathway (71,96,99), and thus it in essence forms the luncrtion point of the two systems. Because all terminal

components (C3, C5-9) are shared, the biological consequences o ativation encompass all the processes previous ly described (immune adherence, opsonic activity, annphylatoxin production, membrane attack complexes, etc.).

There are similarities between some of the more salient features

of the classical pathway compared with those of the alternative pathway. Analogous to Cl1q, IF seems to function as the recognition unit for the properdin pathway, but its relationship to another factor (referred to as a C3 nephritic factor from the sera of patients with membranoproliferative glomerulonephritis (100) and its mode of activation is poorly understood (96). Factor D is capable of enzymnaLtcally cleaving Factor B into Ba and l-b (29,94). u the presence of C3h, a bmolecular complex CIhBb is formed (29) which is endowed with C3 splitting activity similar to the C3 convertase (C4hb2a) of the classical pathway. Furthermore, just as CAb anchors the classical convertasc to the memhr:ine allowing Cn to exert its enzvmatin activity, so too cytophilic C3b niiichors: tlhe (ibBb complex to the membrane allowing the enzyma tic activity ,of Factor Bb to he expressed (83). Both complexes merely gain addit ional C3( to modulate C5 cleaving activity (99). Thus, the prescec of (C3 not onlv prevents aln "abort" due to rapid decay of either convtrtase, but because C(b is utilized as part of the alternative pathway convertasc, it participates in a type of amplification loop. Tn other words, the more C3b that i formed from either pathway, the more C3 c;leaving potential is endowed upon the proprdin C3 conlvertaso. Properdin (P) se'ms to st ilize rthe fragile C3hBlb complex hut its possible role in stahil izinlg the classical C3 convertase has not been investigated (on). Netcworthv, however, is the potent effect pro prdin exertS on the C3b inhibitor (9). By

modulating the action of this enzyme, propordin at least inlrirectly plays a role in stabilizing the classical pathway sequence.

The recognition of foreign substance; by a host usually leads to

the neutralization and eradication of these substances by immune lymphocytes, phagocytic cells, specific antibodies, complement, or an amalgamation of these factors. However, in instances where antigenic substances interact directly with host tissue, the reactions of the host's immunological defense system could sometimes result in a considerable amount of autodestruction. LTA represents a class of antigens that are capable of spontaneous cytophilic binding to mammalian tissue (101,102,103). As a result, host tissue acquires a new "anitgenic face" and may now react with natural or induced antibodies to the LTA. Furthermore, antibodies directed primarily at LTA determinants may cross react with similar determinants of the host's tissue. Such a mechanism has been proposed for the high incidence of rheumatic fever and glomerulonephr itis in patients recovering from post streptococcal infections (l(.1(13). Recently, acyl.ated heteropnlysaccharides (LA) isolated from the cell membranes of several lactobacillus species were shown to replace pigeon excreta antigens in complement consumption tests d iagnostic for pigeon breeders disease (9,1(16). Thus, precedence may already be established for LTA's role in the manifestation of several clinical maladies. In addition, the chemical and biological similarities between LTA and .PS (1,2) plus the ability of LTA to stimulaLe bone resorbtion (17) make LTA a likely candidate for a role i.n periodontal disease. n tho other hand, LTA lacks some of the biological nctiviti is 5 ssociatod with ILPS such as pyrogenicity in rabbhits (2,107) and a mitgenic effect on B-cclls

(2). Since theso activities have been slown to reside with the compl x

Lipid A of LPS (108,109) and since the unique sugars and hydroxvacy esters of Lipid A are absent in LTA, it i s not surprLsing that :asscla.:t activities are absent as well. As a class. tcichic and lipoteichoic acids are wall and membrane components of gramn positive bacteria (107,108)i LTA is typically membrane associated and consists of a glycolipid covalently linked to a polyg lycero]phosphate halckhone which may carry cnrbn-hydrate and D-alanine substi.tuents (2). Teichoic acids (TA), however, are never associated with cell membranes; they lack the terminal glvcolipid coupling, and they may have a backbone nf either poly\tlcerolphosphate or polyribitol phosphate (2). LTA may be converted functionally to polyglyceroal TA by spontaneous deacylation in an aqueous environ1
ment, or mil.d alkaline, or acidic hydrolysis (107). The molecular weight of LTA (93) is probably between 3000-12000 but be cause of its tendency to form micelles in an aqueous environment, the apparent: molecular weight as determined by gel. filtration is approximately four million (110).2 Because LTA possess the glyly ipid moiet v, they are amphipathic molecules exhibitirng a propensity to spontaneously :ssociate with proteins and biological membranes (103). Mammalian red blood cells can be "coated" by spontaneous adsorbtion with an 1.TA contaniing extract and the cells can subsequently be agg:luinnted with an anti-TA serum. Passive henagglutination (P'HA) performed in this manner with sheep red blood cells has previously been reported by many invest i-.or4 s who discovered

SPersonal comn inicat ions from R. Crnic, Qpt. of rVS, 'niv. of rl.; K. Knox and A. J. Wickun, Institute for l)Dent l R research, Sydney, Australia; and personal Iunpublished data.

SD.ta supported hv prsonal oexp rience (soe Figuinron 11 and 12), and personal communication from R. Craig.

ervthrocyte--sensit izing antiens in cell free sa line wnshings or spent culture fluid from several gram positive organimn (I101.02). l. These so called "Rantz antigens" were recently shown to possess properties associated with LTA (11). Because only acylated LTA wil hind to erythrocytes, PHA provides a means of quantitating the amount of LTA in a preparation without having to contend with deacylated TA contamination.

The biological role of TA and LTA to the microorganism has been a subject of considerable disputation by several investigators in recent years. Thus far, at least three roles have been tentatively assigned: i). TA and LTA seem to fEunction as "carrier" molecules for membrane and cell wall. components, i.e. amphipathic LTA may be used by the cell to transport needed hydrophobic molecules through hydrophilic zones which would otherwise pose an almost impenetrable harrier. Fielder and Glaser have established that incracellular LTA serves as a lipid carrier for the biosynthesis of cell '.all ribitol tLichoic acid in Staphylococcus aureus (112,113). Chat rj ce and Won g (114) have demonstrated that LTA may serve as the acceptor in which nascent peptidoglyran polymers are synthesized. 2). LTA seems to be involved in cell wall division and regulation. Holtje and Tomasz have reported that ITA exhibits an inhibitory effect on the function of ,nutolvtic cmnzymos during the division cycle of pneumococcus (115). It is intoresttilg to note that !;imilar functions have been describe d vby Cleveland, et al. working w ith a strain of Streptooccus faccalis (11,!.17,118). In these sy;vtems. 17TA is dencylated and released iito the enviroricnt as 'TA. Once the c.'ncmontration of LTA is sufficient lv lwered, or the concentration of auitolvtic: enzymes is sufficiently e lcvatd I, cel wall auteol\ysis begins at the division zone. This autolvtic .ctivitv theni allows for insertion of additional

cell wall material. 3). [,TA or TA may contribute to the nvorrall electrostatic charge of gram positive nrganismn. Al thth membra ne localized, the long polar tails of many ['LTA penetrnt:e the thick peptidoglycan layer and become externalized (107). These, together with ,he TA which are covalentlv linked to the ce.ll wall (Ir8) present a myriad of antigenic faces to the external environment (1i10,12). This antigenic presentation is of serological import since these antigens are often genus, species, group, or type specific (103,120). In addition, these polar tails generate a net negative charge by exposing the phosphate groups of the polyglycerol or polyribitol backbone. This net negative charge has been teleologicaIly assigned the functinn of maintaining electrostatic repulsion and dispersion of the bacterial cell (121). Since LTA has been shown to sequester certain cations such as magnesium (122), an additional function as a site of divalent caionic c on \ver r nce has also been postulated. The association with ma gnesium ions appears to be more than casual since protopL.asts of Lactobacil us casei placed in a magnesium ion free or chelated medium rapidly lose their 1,TA from the cell membrane.

Anti-LTA titers (of both the Igi and Ig, 'asses) have been reguLarl v reported in mice, rabbits, and man (2,123). Several clinical studies have reported increases in anti-LTA titer-- incl uding r: t ibhodi es of the class [gA--after acute gram positive infections (121.125). P gs, guinea pigs, and r;at exhibit a low level of natural inmuonitv t, 1,TA and rerontly', there have been reports of silivary [:A production as a result of gastric intuhation of monkeys with Stre pt,oc..cus mnutalns h715 s;ro t \po C. There is no doubt that TA and ITA of a 1I grnm positive ganera thus f;r investigated contain antige nic mniorit:is and that nmdcr i'rtii n circumst-ances,


LTA can be immunogenic (2). of particular interest is the fact that the attachment of streptococc-al LTA to erythrorvtes could he revc rsWiblv transferred from the Crvthrocvtrs to other t[ssu: cells (10(.l ,126). The possible significance of this "transfernhilitv" in relation to rheun tci fever and glomerulonephritis and pigeon breeders disease has been previously discussed (9,104-106). However. despite this precedence the significance of the binding of LTA to oral epithelial cells in gingival pockets has not yet been investigated. Not only os ITA mediate bone resorbtion as previously indicated, but spontaneous hybrid micells of LPS and LTA are known to occur, thus compounding the possibility of in situ immunological modulat ion. There is little doubt of the availabitit of extracellular LTA in this environment--Streptococcus mutans BUT alone has been reported to produce excess of 50 u~ of l'TA/ml in culture media

(21)). Recently, Wicken and Knox have studied the cxcrtion ofr xtracellular LTA from this organism in a chemostat under steady state logarithmic growth conditions. Results indicated that a genera tin time of 10-14 hours (estimated to reflect that actual in vive growth rate of this organism in the oral cavity) i)rodiced the maximal amount of 'xtranellular LTA (I). Considering its ud iquity and the cariogenic nature of Strepto-C coccus mu tans BIlT (127-130), the secretion of cop ious amounts of hioloj irally active LTA into the oral rnvitv has the potcnial of considerable influence on the host-parnsite relationship.

Th'L objectives of the project were than defined ans follow:

(L) To establish if an LTA-containing extraccelular extract of

Pers ona l communication o f A. 1 icken.


Streptococcus mutans BuT was capable of inhibiting complement mediated cytolysis of target sheep erythrocytes.

(2) To purify the extracellutar LTA of S. mrtans BilT to homogenity.

(3) To describe tihe nature of any anti-complementary activity that purified extracellular LTA may exhibit.

(4) To determine the site of action of any such inhibition.

(5) To determine the mechanism by which purified extracellular LTA may enhibit anti-complementary activity.

(6) To determine if the I,TA from other gram positive genera and species can be shown to demonstrate anti-complementary activity.


Crude extracellular LTA (LTAcx). The initial. studies were carried out utilizing LTAcx prepared in Australia by the method of Wicken and Knox (110). Streptococcus mutans BHIT was grown to late stationary phase in a New Brunswick Microfirm fermenter at 370C, under anerobic conditions (95' N2 and 5 CO ) in a complex medium.

Later experiments utilized LTAcx prepared at Cainesville,

Florida. The original method was modified as follows. A Pellicon
Cassette system (Millipore Corp., Bedford, MA) equipped with 5.0 ft of PTGC filter materiaJ was used to dinlvzc Todd-Hlw itt broth (Di fo Laboratories, Detroit, MN). A 100 ml culture of early log pha:e S. mutans BHT was inoculated into 10 liters of dialyzed medium and incubated at 370 for 24 hours. The cells were harvested using a Delaval Cyrotester (Po'ighkeepsie, NY). The superna:t was pa ss through the Pellicon Cassette system (loaded with 1.0 ft of 0.45 1 microporous membrane) to remove remaining c lls and debris. The cellFree spent fluid was then frac tionnted and concentrated by passage through 5.n ft PTCC membrane (nominal molecular weight exclusion limit of 10,0(0). The filter retentatc was washed in situ with syvcral liters of water, collpIcted and Ivophilized. The freeze-dried retenrate, designated as LTAcx. w~as stored in a de si'ator at -20"C.

Solutions for mpl-mnt ,j;i ypvs. Isot nic Ver.nal huffured sodium chloride (V.;S), dextrose celatin Veronal bu"ffr with added

CaC12 and NgC12 (DGVB), EDTA containing Veronal buffer (O.0'W : EDTA-GVB) and gelatin Ve rnal I huf fer with added CaiUl, and IMgC( (GVB) were prepared as previously described by Hoffmann (131).

Human complement (HuC). Fresh human blood samples were obtained from the Gainesville Plasma Corp., Gainesville, F L. The blood was allowed to clot at room temperature for about 60 minutes, and the serum was separated by centrifugation at 500 X at O'C. The serum was collected and stored at -700C.

Guinea pig complement_ (GC). Fresh frozen uine pig complement was purchased from Pel Freeze Laboratories (Rogers, AR). The serum was shipped in dry ice and it was stored at -70'C after arrival in the laboratory.

_Comjlement components. Purified guinea npi5 Cl and C2 were prepared according to Nelson et al. (132) and Ruddv and Austin (133,134). Functionally purified guinea pig C3, CS and C9 and human C1, C5, C6 and C7 were purchased from Cordis Laboratories (Miami, FL).

Erythjrocytes (E) Sheep blood was taken by vCni ncture from a single animal maintained at the animal research laboratory of the J. Hillis miller Health Center ((;;G inesville, FL). One hundred milliliter volumes of blood were collected in cqual volumes of sterile Al:aever's solution (135) and the blood was tored at /4" for up to three weeks.

Antibody sensit imod ;he ._rythrocytes (E.A). Rabbit anti

sheep E stromata was obtained from Cordis laboratories (Miami, FL). Sensitization of washed sheep E was performed as rcccmmended by the supplier.

Complement component intermediate complexes. Sheep E in various stages of complement fixation were used in this study. EAC1, EAC14 and EACi42 were prepared by methods described by Borsos and Rapp (136). EAC1423567 were prepared by the procedure described by Hoffmann (137). Unless otherwise indicated, guinea pig Cl, C8 and C9 were used in all instances, and the remaining C components

were from human serum.

Treatment of cells and cellular intermediates with LTAcx.

Unless otherwise indicated, cells were washed and suspended in VBS at a concentration of 10'/mi. Equal volumes of these cells and LTAcx were mixed and incubated at 37' for 20 minutes with continuous shaking. The mixture was then placed in an ice bath for 10 minutes. At the end of incubation DGVB was added to the mixture and it was centrifuged at 500 g for five minutes. The supernate ,was discarded and the cells were suspended and washed thrice with DGVB (00 for 10 minutes at 500 g) to remove any unlound material. The cells were then resuspended in DGVB at a concentration of 10i/mI. A sample of the cells were tested for cell-bound ILTA using passive hemagglutinn tion with rabihit anti-lTA. The rema ining, cells were used in experiments to detct acqu ired rosist:nce to homolvsis.

Passive hemaggutinat ion (PlA). Passive hemalgl ut inat ion was carried out using a micrt. itration system. F[ft-y i. of a VBS dilution of anti-TA readded to the first row of wells of a round bottom microtiter plate (Cook Engineering Co., Alexandria. VA) and 2D iii (one lron From the calibrated pipetes supplied with the system) of VIS were added to the other wells on the plate. The anti-serum was serinllv dit.-ed in situ and one drop of ILTAcx


treated cells was added to each well. Controls frr spontaneouss or nonspecific agglutination consisted of wells that contined .anriserum and sheep E which had never been exposed to lTAcx. Treated sheep E plus VBS constituted another control. The microtiter plate was incubated at 370C on a Cordis Micromixer (Cordis Laboratories, Miami, FL) for 15 minutes. The plates were removed from the mixer and the cells were allowed to settle for two hours at 370C, followed by three hours at room temperature.

Modified Lassive hemagglutination (PHAg). A modification of the above technique was used to semi-quantLtate the amounts of LTA present in various preparations. The same apparati were used, but instead of antibody, LTA-containing extracts were added to the bottom wells and serially diluted in situ as described. After each LTA source was diluted, one drop of sheep ervthrocvtes (1in/ml in VBS) was added to each well and the plate was then incubated at 370C for 20 minutes and at 00C for 10 minutes. The cells were kept in suspension by vibrating the plate on a Cordis Micromixer during both incubation periods. One drop of CVB was then added to each well and the plate was centrifuged at 200 g for 5 minutes. The entire plate was then abruptly inverted over absorbant paper towels and allowed to drain for approximately one minute. One drop of (CVB was again aldded o each well and the plate was \'ibr ted at O0C for 5 minutes to resu spend the pellet. An additional drop of GVB was added per well and the plate was again centrifuged at 200 g for 5 minutes. This washing procedure was repeated thrae times and the cells were then finally resuspended in one drop of cGV. One cop of anti-LA (diluted 1:1000 in VBS) was then

added to each well and the plate was incubated at 370C for 15 minutes on a Cordis M!icromixer. The plate was removed from the mixer and the cells were allowed to settle for two hours at 370C, followed by three hours at room temperature.

Inhibition of complement mediated lv__s. EA coated with

LTA (EALTA) were tested by mixing 0.1 ml of EAITA (10 rells/ml) in DGVB and 0.4 ml of DGVB diluted HuC. The HuC was diluted so that a maximum of 80 percent lysis was produced in EA which had not been treated with LTAcx. The mixture was incubated at 370C with continuous shaking for 60 minutes. One milliliter of ice cold EDTA-GVB was added, the mixture was centrifuged for 5 minutes at 500 g at OC and the supernatent fluid was recovered. The optical density of the supernntent fluid was determined at a wave length of 414 nm. Inhib ition of hemolysis was calculated for each concentration of LTAcx used by comparing the extent of lysis in each assay with a control reaction mixture which cont lined EA that had not been treated with lTAcx.

Effect of LTAcx on the titer of antibodies specific for sheet

ervthrocvte stromata. 3ecaue ITA associate with some proteins (138S it was necessary to perform a hemolytic nntihody titration to determine if the ability of the immunog lobul ins to fix complement at the cell surface wa heing affected by lTAcx treatment. The possibility of similar it igens in LTAcx and sheep ervthrocvte stromltna was al;o considered. Equal volumes of LTAcx (00 o g/ml in VBS) and rabbit anti-sheep E stromata were inciuhated together at 37C for 20 minutes. A control consisted of incubating an

equal volume mixture of VRS and anti-sheep erythror:te stromata for the same time at the same temperature. The ant ibodies were then titrated using limiting amounts of complement (135).

C1 fixation and transfer. The number of Ci molecules bound to an antigen-antibody complex can be measured by the CL fixation and transfer test described by Borsos and Rapp (130). In a modification of this procedure, an attempt was made to quantitate the number of C1 molecules fixed to EA which had previously been treated with LTApcx. Buffer controls and EALTA were prepared as previously described, and after washing were resuspended at 108 cells/ml in DGVB. Equal volumes of EA and EA were inLTA VBS
cubated with Cl at 30C for 15 minutes. The cell mixtures were washed twice with DGVB, and resuspended in CVB at a cell concentration of 1 X 10 8/ml, 5 X 10 7/m, and 1 X 10 /ml. One volume of each cell concentration was added to one volume of EAC4 (at

1 X 10 cells/ml) to permit transfer of Cl from EA C1 to EAC4. The cells were incubated at 300C for 15 minutes. and then C2 and C-EDTA were added in relative excess as described previously.

A variation of the C( transfer assay was performed by treating preformed EACI with LTA or buffer control as described. The resulting EACI were resuspended to 1 X 10 cells/ml in GVB and
the amount of Cl capable of transfer was measured as described above.

Gel Filtration. LTrAcx was fractionated on a 2.5 m X 100.0 cm column of Bio-Gel A-5M, 200-40/ mesh (Biorid Laboratories, Richmond,

In this instance, "x" represent s LTA or the appropriate b,"ffer treated control.

CA) using a modification of the method described by Wic:ken and Knox (110). The column was equilibrated and eluted using 0.01 'I Tris carbonate (Sigma Chemical Co., St. Louis, MO), pl 6.8.

Hydrophobic Affinity Column chromatography. Because of the hydrophobic nature of the fatty acid moieties of lipoteichoic acid, adsorbtion to a stationary phase of a chromatographic column was used in an attempt to further purify the LTA. LTAppx in buffer A (0.01 M Tris carbonate pH 6.8, 1.0 M Nacl was loaded on a 25.0 X 2.25 cm column packed with Octyl Sepharose (Pharmacia Fine Chemicals, Piscataway, NJ) and equilibrated in the same buffer. After eluting with 150 m! of starting buffer A, the reservoir was then changed to buffer B (0.01 M Tris carbonate pH 6.8) and another 100 ml were eluted. Buffer C consisted of 250 ml of a gradient ranging from 10-70 % propanol (by volume) in 0.01 M Tris carbonate, pH 6.8.

Octyl Sepharose is a derivative of the cross linked agarose Sepharose CL-4B. The terminal n-octyl groups of this agarose gel. confer a hydrophobicity to the matrix. By exploiting this property it was hoped that polar or neutral non-interacting components would be removed by elution with solutions of high ionic strength. The lipoteichoic acid would then be eluted from the matrix with an organic solvent such as propanol. (It is imperative that all tubing, connections and gaskets used throughout the column he constructed of a material that is resistant to organic solvents).

1This method represents a nodific:ltion of a procedure described by A... Wicken and K. Knox (Svdney, Australia) via personal communication.

Removal of salt and ronl From LTA conta in ig extracts.

Removal of salts and/or propanol from various preparations was rapidly and quantitatively accomplished by gel filtration utilizing 1,11:20 (Pharmacia Fine Chemicals, Piscatawav, NJ) as the solid nhase support matrix. The most commonly employed column was 50.0 cm X 2.5 cm but a larger 65.0 cm X 3.0 cm column was sometimes utilized. The column was packed and equilibrated with deionized water. Sample preparations usually involved rotary flash--evaporation (Buchler Instruments. Fort Lee, NJ) in order to reduce the volume of sample to 15-20 ml. Elution of product was carried out at a pressure head of aproximatel 50 cm water and approximately 4.0 ml effluent were collected per tube.

Phosphatidyl choline vesicle (PCV) _urification of LTA -(a) Preparation of PCV. Although reported as the method of choice by other invest igators, in our hands POctyl Sepharose u purificat io, of LTA from Streptococcus mutans BHT resulted in a product still highly contaminated with polysaccharides In an attempt to achieve homogeneous purification of LTA, a modification of the above ment-ioned hydrophobic adsorbtion principle wa;s employed. In this procedure, artificial membrane vesicles were prepared with 1),nhosphatidyl choline dipalmitoyl (1PC) (Sigma Chemical Co.) as the sole constituent via a modified method of Hli.. (1.40). In brief, '0. 0 m of PC was placed in each of several 10 ml high speed glas.- Corex (cntriffuge tubes (Corning Glass iJWorks, Corniiig, NY') and (1dLssolved with one mti chloroform. The solvent was gently evaporated in a 500C water bath while rotation ng the tubes (so as t.o coat the bottom 5 or 6 cm of the tube with PC. Once dry, rhe tube were pL.iced in a Ivophiliat ion flask and any residiln solvTnt was removed in \la.cuo.. ()ne milliliter Wichen, A.J.., and Knox, K.--Personal communic.-tion.

of 0.01 M Tris carbonate pH 6.8 was then added to each tube and they were placed in a 500C water bath. Once warmed, the tubes were vigorously vortexed (Vortex Genie Mixer, Scientific Industries Inc., Bohemia, NY) and the cycle of warming and vortexing was continued until a milky emulsion was formed. Fifteen milliliters of 0.01 M Tris carbonate were then added to each tube and the tubes were centrifuged at 27,000 g for 30 minutes. The supernatent fluids were then decanted, the pellets were resuspended in 1.0 ml Tris carbonate buffer and warmed to 500C in a water bath. The tubes were gently swirled (butnot aggitated) to dissolve and resuspend the pellet The resulting phosphatidyl choline vesicles (PCV), devoid of very small vesicles, were then used to adsorb LTA from LTAppx.

(b) Prearation of PCV-LTA. Two milliliters of TAppx at a concentration of 1.5 mg/ml in 0.01 M Tris carbonate, pH 6.8 were added to each centrifuge tube containing 1.0 ml of PCV. The tubes were covered with parafilm (American Can Co., Neehaw, iS) and incubated for 90 minutes in a 370 shaker water bath. Thirteen milliliters of

0.01 M Tris carbonate were then added to each test tube and they were centrifuged at 27,000 g for 45 minutes. The supernates were discarded and the pellets were gently resuspended in 1.0 ml of Tris carbonate buffer at 500C as previously described.

Fifteen milliliters of buffer were tlhn added to anch pellet, the tubes were gentlv swirled and then centrifuged as described. The pellets were washed three times in this manner. The final pellet was drained and then dissolved in 5.0 ml of chloroform/methanol (3 + I v/v). The tubes were then covered with aluminum foil and allowed to sit at room temperature for 60 minutes.

A Millipore 15 m analytical filter holder (Millipore Corp.,

Bedford, MA) was loaded with a 3.0 w fluoropore membrane (Millipore Corp.) and washed with several volumes of the chloroform/methnnol solvent. The test tubes were all sequentially decanted into the apparatus and the contents were allowed to filter by gravity through the membrane. Each test tube was washed with several volumes of warmed chloroform and decanted into the filtering apparatus. Finally, the barrel and filter were washed in situ with warm chloroform. The filter was removed after air drying in situ and placed in 10.0 ml. of deionized water warmed to approximately 40'C. All centrifuge tubes and the barrel of the filtering apparatus were washed with warm deionized water and all products were combined. The resulting product was passed through a 25 mm Swinne> filter (Millipore (Corp.) loaded with a 5 o microporous membrane (Millinore (Corp.) to remove particulate debris. The membrane was washed in situ with several volumes of warm deionized water. The filtrate was collected directly into a lyophilization flask and was then shell frozen and lyophilized. The final product was stored in a dessicator at -20(C.
C Phosphatidyl choline analysis. In order to detect any phospholipid contamination of the LTA throughout the previously described PCV purification, radioactive PC was used to label the phospholipids in the vesi.les. Approximatelyv 2. 3 Ci. (3 X 10 6 D M) of C labeled phosphatidyL choline (Amerslham Searle Corp, Arlington Heights, TI,) were added to 4(0 mg of phosphntidvl choline dipnlmitoyl in a 30 ml Corex centrifuge tube. Phosphatldvl cRholine vehicles were prepared from this and the non-labeled contents of three other tubes bv the methods previo uyiiv described. Fifty microliter


samples from the C containing test tube were taken at each ' PPO (2,5 diphenyloxazole), and 0.01% POPOP (1,4-di (2-(5-phenyvoxazolyl)benzene) were added to each vial. The degree of 14C-PCV contamination of the final product was determined by placing the entire LTA-containing-fluoropore filter in a scintillation vial with 5.0 ml scintillation fluid. The possible influence of quenching by the fluoropore filter was investigated by adding equal aliquots of 14
C-PC to two scintillation vials one of which contained a fluoropore filter in addition to scintillation fluid. No npprociabtle difference in CPM was observed. Disintegrations per minutee (1)PM) values were calculated from a standard quench curve constructed for use with chloroform. Standard ratios were determined for each sample and percent efficiencies were extrapolated from the standard quench curve. This volume was then used to correct counts per minute (CPM) to PPM. Unless otherwise indicated, the samples were counted for 10 minutes in a Beckman LS-733 liquid scintillation counter (Bockman Instruments, Fullerton, (CA).

Colormetric assays. Phosphorous was determined by the method of Lowry et al. (141) with absorbancies measured at R20 nm. .t al carbohydrate was measured hy the phenol sulfuric acid assay as described by Duboijs et al. (142). Total protein was performed on o;amples using the Bio-Rad Protein Assayv (Bio-Rad Laboratories, Rockville Center, NY). Samples and the standard curve were prepared ftlltowing the manufacturer's recommndat ions.


Gas Iiquid Chromatography. Carbohydrate analvsis was performed after treatment of the samples with 1.0 N P ,s in sealed ampules for 8 hours at 105C. Upon cooling, the sent was broken and exactly

0.2 ml of mannitol (either at 5.0 mg/mi or 1.0 mg/ml depending on carbohydrate concentration of the sample) was added as an internal standard. The contents of each vial were quantitatively transferred to 15 ml centrifuge tubes (Corning Glass Works) containing 0.3 g BaCO3. Each centrifuge tube was heated in a boiling water bath and alternately vortexed until the pH approached neutrality as indicated by full-range pH paper (Micro Essential Laboratory, Brooklyn, NY). All tubes were centriufged at 500 g for 5 minutes and the supernates were removed and collected in appropriately labeled 13 mm screw cap tubes fitted with teflnn lined lids. The centrifuge tubes containing BaCO3 were washed once with one ml of deionizced water and the supernates were appropriately pooled.

After lyophilization, the hydrolyzed carbohydrates were converted to trimethylsilyl ester (TMS) derivatives by the addition of 0.2 or 1.0 ml (depending upon carbohydrate concentration) of TRI SIL Z (Pierce Chemical Co.). Samples were warmed to approximately 600C in a water bath for 15-3() minutes before use and assayed using a Packard 800 series gas chiromatograph equipped with a flame ioniza t ion detector. 'The gas chromant graplic column (153 cm X

4 cm) was packed with SE-40 1UTRAPHIASE 37: on Chromosorh W (1P) 80/[00 mesh matrix (Pierce Chemical Co., Rockford, IL). Column and detccror temperatures were set at 16(0'C and 1S C, repr tively. The N. carrier gas was set at approximately 30 cc/minute.

Amino acid analysis. Amino acids and amino sugi:rs were measured on a JEOL model JLC-6AH automated amino ac Ld ana lyser (JIEOl, Inc.. Cranford, NJ). Sample hydrolysates were prepared as described by Grabar and Burtin (143).

Clq Cs, and Cls purification. Highly purified human Clq

was prepared from whole human sera by the method of Yonemasu nnd Stroud (144). Highly purified human Cis and Cls were prepared by a minor modification of the method described by Sakai and Stroud (35). For the final resolution step, Bio-Rad Cellex-D DEAE with binding capacity of 1.07 meq/g (Cellex-D, Bio-Rad Laboratories, Rockville Center, NY) was substituted for fibrous DEAE cellulose Whatman DE-23. The DEAE was washed and prepared according to the manufacturer's specifications. Final elution of the product was accomplished with the use of the same elut ing buffer is described, but instead of a stepwise elution of the column, an ionic gradient from

0.2 0.4 RSC (relative sodium chloride concentration) was utilized.

Disc acrylamide gel electrophoresis of Cfl, Cis and Cs s. This was

carried out essentially as described by Yonemasu and Stroud (i.4 ) but without the use of sodium dodecyl sulfate (SDS).

CIs inhibition assays. The ability of Cls to consume C2 activity was assayed by a modification of the method described by Sakai and Stroud (35). Briefly, 0.1 ml of CLs (approximately 8.0 X 107 site forming unit SFU/ml) plus 0.1 ml LTApcx (100 ig/ml in D)VB) were ircubated at 30" For L5 minutes. One tenth millii liter of (2 was then added at a concept ration of appreximately 9.0 X 10 effective moleculc;a/ml andt inclhatd at 37" for 10 minutes. At the end of the incubation, 9.7 ml cold I)(OVR were added to the mixture resulting in a 1:1.)0 dilution of the (C. The (2 was then serially diluted and .! ml aliquots from each dilution were added to 0.1 ml of


EAC14 (10 cells/ml in DGVB). The mixture was incubated at 130 for .Io minutes and cooled to O0C in an ice bath for 2.0 minutes. Three tenths of a milliliter of C-EDTA (1:37.5 in 0.04 M EDTA-CVB ) were then added to each test tube and the mixtures were incubated at 370C for 60 minutes. At the end of the incubation period, 1.0 ml of cold EDTA-GVB was added, the tubes were centrifuged, and the supernntes read for release of nxvhemaglobin at a wave length of 414 nm. External controls consisted of C2 with no Cis nor LTApcx, C2 with Cis but not LTApcx, and C2 with LTApcx but no Cis. The usual internal controls (spontaneous lysis, color correction, no C2, and total ]ysis) were included at all times. Results were expressed as percent inhibition of C2 consuming ability compared with a control containing only Cls and C2.

The ability of Cls to hydrolize the synthetic substrate p-Tosyl-larginine methylester (TAMe) was determined as described by Nagaki and Stroud (38). Inhibition assays were performed by incubating equal volume,7
of CIs (approximately 8.0 X 10 SFU/ml) and LTApcx (npproximtelv 101 ,a.i at 370 for 10 minutes. Residual Cls activity was then determined as described (38-40).

C1q inhibition assays. The effect of LTApcx on the ability of purified Clq to bind to antibody sensitized sheep erythrocytes was determined by methods described by Loos et al (39) and Raepp e et al. (40). Fqual volumes of (Alq (approxim;leclv 1.5 X 1()0 SFU/mi.) and 'Apex (10) g/ml) were incubated at 37 for 1() minutes. Residual Clq activity was then determined as indicated nbove.

Inhibition of whole human complemen-t by crude

lipoteichoic acid (LTAcx). To determine whether L,TAcx had any effect on whole human complement, equal volumes of LTAcx nnd whole human complement were preincubated at 370/30 minutes. After preincuahtion, the complement source was serially diluted in DGVB and the residual hemolytic activity was titrated. As shown in Figure 1, approximately 50'O of the whol-e complement hemolytic activity (measured in CH50 units) was consumed. Furthermore, as seen in Figure 2, this consumption was dependent on the concentration of the LTAcx used.

Titration of complement components in whole human sera after treatment with LTAcx. One mechanism for fluid phase consumption of whole complement could have been the interaction of natural antibodies in the human sera with LTA or some other antigenic substance in the crude extract. The result would be the fixation of Cl and subsequent activation of (4 and C2 via classical pathway. Another explanation for decreased hemolytic activity could have been the activation of the alternative pathway in a manner analogous to LPS. To differentiate between these two modes of activation, individual component titrat ions were performed on huma;in scra incihatted with lTAcx. Tn addition, (: titrations were carried out in the presence of oetyleneglycol-his (i Amino E:thyl Etlher) N,N totra;iectic acid (EC.\A) and Mg ions. This c l, t inlg agent preferentiallv hinds Ca ins (1 I,1 f and by reinforcing the E(I'A buffer with Mg ions one can effectively deplete the available Ca ions yet inninta in relatively high levels of Mg ions. Thus, the Ca ion dependent classical pathway is blocked, hut the alternat ive pathway can function relat ivel-v unimpai red (145, 157).

Figure 1. Titration of whole human complement after incubat.iln
with crude extracellular lipoteichoic acid (LTAcx).
Symbols: (o) Non-treated control: (e) Secrum triatcd
with LTAcx at 500 ig/ml.




O WC vo
o o


LO It -O- O
0 0

O 0 u u co



Figure 2. iose response inhilbition of whole humin compl ement
n after incubh;tion with varving concentrate ions oF
I'TAcx. The non-treated control is ibh reviaLoted
,s NTC.

LL I oD aO to o z

0 O



OOOO /n0 090 V0
lwsi~n09H -vnisi

A typical component Litration in serum trrotead with L.TA'x is depicted in Figure 3. In this example, the LTAcx treaonted sirulm was serially diluted in DGVB. Next EAC142, C5, 6, 7, and CS-9 were added sequentially to the dilutions. Since all components were added in excess, C3 became the limiting factor in contributing to the hemolvsis of the target cells. Percent lysis in each test tube was mathematically converted to Z (the average number of SAC1423 sites per cell) and this was plotted against the reciprocal of the serum dilution. Percent inhibiticn of site forming units (SFU) was then calculated from 7=1 values or percent inhibition of CH150 units was determined from value es associated with Z= 09. Figure 4 represents a composite of multiple component titrations from whole human sera treated with LTAcx. As can he seen in this figure, CL and C4 activities were consumed to omie degree, however, more than 501 inhibition of C2 activity was observed. As indicated, C3 activity was also consumed during preincuibation of complement with LTAcs, but incubation with purified [iHU prordu d no inhibiticn of C3 hemolytic potential. No C3 consumption occurred if the incubation was performed in the presence of the chelator ethvlenediamine tetra acetic acid ( EDTA) and Less than 7" if incubaited in the presence of EcTA-Mg ions. The above results indicated the necessity for divalent cations as cofactors mediating the consumplhtion of C3 in the presence of LTAcx. In addition, there appeared to he a requ irpment for other setum factors (possibly natural AB and/or components of the alternative pathwavt since purified C3 activity remain d unaffected ,v inieiition wirh LTAcx.

Figure 3. Titration of C3 in whole human serum :EFter trentnont
with LTAcx. Symhols: (a) Non-treated control; (0) Serum trreatd w ith lAcrx it a concentrate n of 2 T0
lg/ml. After incubation, sera were t itrnted for
residual C3 activity according to procedures
described in Material.s and Methods.

0 O

xx <

00 0


Z .J -o

(Iw/S.1.S 0I IS JO dI3VtJfNN U9Vd3AV)Z
< F- -r) 1

Ex I

owin/ss 2II IRs jo 1R1HVnN 39V3/W) Z

Figure 4. Complement component titration of whole human sra
after treatment with LTAcx. The sern were incubated
with the LTAcx (50(0 ug/ml) then titrated for residual activity of the components indic ated as described
in Materials and Methods.


/ 7


/ //>' <>,k'/ 7 ,// CK / ,/ /7k 77 C/ .,

II Ill/

SIo 0 0 0 0 0 S'I IND 09HD :JONOIIHlHNI %

Inhibition of complement Lysis of LTAcx treated EA. During an experiment in which EA treated with LTAcx were testeded for reactive lysis, it was discovered that the LTAcx treated cells exhibited Lesq hemolysis than even the buffer treated controls. This serendipitous observation led to the discovery that LTA treated EA were refractory to complement mediated lysis. To confirm these results, various concentrations of LTAcx were used to treat EA. After tlhe treated cells were extensively washed they were tested for their suseptibility to lysis by complement. The same cells were also tested for the presence of cell-bound LTA using the passive hemagglutination technique (PH.A) w th anti-LTA. The results shown in Figures 5 and 6 indicated that both the extent of inhibition of hemolysis and PHA titers were LTAcx dose dependent. There was a decline in both activities n nlv after the iTAc:. had been diluted to a concentration of 62. 5 ig/ml. The decrease in titer below this concentration indicated that the test cells were no longer saturated with LTA. There was a conco('mitannt drop in inhibition of lysis at 62.5 pg/ml. EA which were treated with uninoculated culture medium (dialyzed Todd-llewitt broth) were unanI fec ted when compleemnt was added.

Effect of LTAcx on lysis of sheeP E and sheep E: cellul-ar

intermediates. The tr(atm(nt of EA with LTAcx caused the cells to become relatively resistant to complement mcdi ated Isis. This could have been dUo to an cI c ct ion te ant ;libody nlo ls al' : In effect on on or more of the complimonut componuits., or an al. rat iion of the c'll membrane.

To further invest-i ~tc the nature of the comp lemnt inhhiit inn

associated with LTAc, hep E. sheep E., and various ; sheep F rnmc, o lent

Figure 5. Inhibition of complement mneditd lysis of EA
treated with varving concentrations of LITArx.







0 i


S (D U)

Figure 6. Passive hcmagglutination (PlIA) of LA treated with
varying concentrations of LTA\x.


cN x
z 0

0 0

rn N r


WOo 0 0 o 0O0 o o 0
- q 3 Cn t a %t N W cN 0 O ro 0 I Od

component intermediatesn were treated wi. th LTAcx and analyzed for suIsceptibil[ty to complement mediated Ivsis. The LTAcx treated cells were also tested for bound LTA using PRIA with ant-'LTA. Results indicated that E, EA, and EACI4 were al.l refractory to complement medinted lvsis and that LTA was detectable on the surfaces of the cells (Figures 7 and 8). However, EACL423567 which had been treated with LTAcx were not resistant to lysis despite the fact that LTA was detectable on t:he cells (Figure 8). Thus, the inhibitor appeared to affect a complement component required for lysis os EACI4, but which was unnecessary for Lysis of EAC1423567.

In an attempt of focus on the site of inhibition, the ability of LTAcx to affect the hemolytic susceptibility of EACL42 was examined. This intermediate possesses C3 convertase activity (C42) which is involved in the generation of SAC423 and SAC,\4235. However, C( is not required for lysis of the intermediate once SAC142 have been formed (148). Failure of LTAcx to inhibit this intermediate would indicate that C3 convertase was not the step in thie complement sequence affected by the LTAcx.

Sheep EACi42 were treated with LTAex according to the protocol.

that has been described. For this experiment, tire different amounts of C2 were used to generate EAC1.42 from EAC.4. The results clearly indicated that there was no inhibition of the intermediate comply I'x EAC14'2 (Figure 9). testing by PHA with antibodies specific for ILTA confirmed the presence of LTA on the surfaces of the cells at the samo relative conlcen trations found when the other intermediate complexes were tested.

Effect of LTAcx on anti-sheep ervthrocvte antibodies. Some

Figure 7. Effect of LTAcx treatment on the lysis of various
complement component intermediates. Each cellu Nr
intermidLate was prepared and then treated with
LTAcx (125 pg/mi). Lysis was developed using procedures described in Materials and Methods.
Percent inhibition was calculated by comparison
against buffer treated controls.




ui <
I W .' W

W --

CT o




Figure 8. PHA of various LTAcx treated complement component

QQQ, < 000 uJ

< L C',

C~j LLJ )
rO~ Lc
cr Lu FLU _

0 Ld ui

ZOI~~L 'X '-Ui -LIIN


Figure 9. Effect of LTAcx treatment on the lysis of EAC142.
Various limiting concentrations of C2 were used to
prepare EAC142 cellular intermediates. The cells were then treated with LTAcx (250 Vg/ml) and lysis
was developed using procedures described in Materials
and Methods. Symbols: (o) EAC142 incubated with
LTAcx; (e) EAC142 incubated with buffer.




0 O0


-1/-1 .__ZI VS JO 838VinN 39V83AV
a 0..


substances in LTAcx might be capable of interacting with the ant ibodies used to sensitize sheep E. This internctinn could then lead to an impairment of C1 activation and result in reduced lysis. Such a mechanism might be the reason why E and EA become resistant to lysis after treatment with LTAcx. Therefore, antibodies to sheep erythrocyte stromata were incubated with LTAcx. The mixture was then diluted to the point where the LTAcx:-related inhibition could not be detected and the antibodies in the mixture were titrated (135). It was found that antibodies that had been preincubated with LTAcx had the same titer as antibodies that were incubated for the same time and temperature with VBS (Figure 10).

Partial purification of LTA. Partial purificat Lon of LTA and

the complement inhibitor was acc mplished by gel lil trntion n the LTA: through an A-5M Biogel column. The results of a typical experiment arc, shown in Figure 11. Arens of antigenicity were resolved by immnodiFfusion in an agarose gel utilizing ann nti-sernm specific for the IAl backbone. Fractions were pooled as indicated (A-F), and each pool was dialyzed against water and subsequently lvophlized. Note that po ls B, C, and E contained high levels of phosphorus and tit the zones of antigenicity were also located in these areas. Utilizing extra ca e I cl(ar extracts from S. mutans and other microorgan isnms, similar frActi onat ion profiles under comparable lc con ditions were obtained by Wicken and Knox (110) and ilewiuis and Craig. Analysis by these workers revealed that the second phosphorouns contain ing pvak (peak 11) conta ined LA\ whereas the trailing phosphorus peak contained deyiclated TA. and wall teichoic

Personal commun ica t ion.

Figure I). Effect of ITArx on homovltic antihody titration.
Antibodies to shoep red hlood cell stromaeta (Ah) were
incuhbated with ITAcx (95OO ug/ml) and residual hiemnal-svin
activity was titrated y procedures described in
Materials and Methods. LUsis nf cells was developed with whole guinea pig complement. Symbols: (o) Ab incubated with LTAcx; (e) Ab incubated with bnffer.







0 U

cr a


to (D

S0 0 0 0 0 0 0 0 0Sco a- CO4 iN 3 HJ

SIS 9O01J3H IN3313d

acids. Because peak 11 represented partinl ly puri [i' d ext ran:ciliLliar lipoteichoic acid, the recovered material was designated LTAppx.

A sample of each pool was rehydrated to 50 ye/ml and reacted with EA, according to standard procedures (Materii!s and i:thods). Fach FA preparation was analyzed using the complement inhibition nssav and tested for bound LTA by PHA. Onlv the pools contaiini; n LTA (as demonstrated by PHA) caused inhibition of complement mediated 1vsis (Table 1).

Despite the excellent separation of LITA from most of the material that absorbed light at a wave length of 260() nm and p resumably from all deacylated LTA or TA, two persistent problems arose with this purification procedure:

1). Polysacchar le contamina tion accounted for a major portion Vf the mass recovered in peak I1I, and

2). The total mass of lTAppx under peak II was anlmos t immcasrai small.

In an attempt to at least increase the yield of peak 11 material, a Millipore Cassette system was emp Loyed to both concentrate and frtnctionate the spent culture supernate (Materials nnd Methods). This method of LTA enrichment proved highLlv successful as evidenced bv lhe results in Figure 12. Even after va;llues are corrected for the greater mass of crude extract applied on the latter column the mans vield ,f I ,\Apx was some fifteen fold grestut r than thnt obtained with previously employed procedures (F igur.e 11).

An analysis of result tracing th' partial puri ication of LTA is summarized in Tables 2 and 3. It ;houtl he noted that the total amount of Pi, mass, protein, and A 260 absorbing mat oral decreased several thousand fold in the purificnt s n process, whereas the total

j 0 C 0 0 c; -A
-0 4- 0 C C 00 -S0 -u

(o rr

71 u;

0 -'-4

E --4 ,--4
Li Z) ;) k.42

C-C o -e
4.-.. 1I -0 HO H
O. O


O G0 0 01
O . RJ .-i 0 -.0 H 22HIn'


A -:>a,--H- C 0 0CC)

4J J Q)" -0I It- '- 42

.0 4.= M C

0 0
O. c O 42 42 0 3C
0 U. o -.. U'0 a 0

I 0. L1 4 0

O- - 0J
C1 f4 4 -- -( SC4C' "0*0
42 r0 422 42 4

0 U. 0 0.42 .0 - 34
-HI I cJ UC .,-i 02) C

.-. 42 U 4 U ..
0. 0 ,0 0 o vi 0.

a0 0 A *H-d.
C i4
U 4u)) 440 ,,u 4 4Li) 0 0 d 00 0 0 0 .- 4 U "1 'iU 414 > 4-'~~~r '4 "22

0 n 42 0

0.4. >1 (4 .4' 2 "4 42i 0n 0 LU U.. 0, 4 ~ 0 J =
04 Ld .)'4 4243~
4iz 42 .

Figure 11. Partial purification of LTA by A5-M gel filtr-ition.
Symbols: (e) A 26I0 aorbance (m'xmql1 :Ibsorh;lnce
wavelength for nuclcic acids): (*:) A,. absorhbnce (maximal absorhance wavelength for carbohydrates ns
determined by the Pheno S lIfuri c A:id assayv ); (A)
Pi concentration in n-moles As dctermind by the Lowry PL nssay; (+) Antigenicity as dctcrmined by
Ouchterlony gel di ffusios n 1sinC a nntiscra
directed against LTA backhone.

o O (salo -U) !d Ln -6 0





N~l r3- c O. 0

-.- .

0 V 0 0
I9 I

cso dz

0 6 qFt~v 6

6; 0 0nV c

Figure 12. Partial puWrification of LTA b\ A5-M gel filtration
with LTA enriched starting -mterial. Svmbol.s: (e)
A absorhance: (o) Pi concentration in u-mmlu's/mli
as determined by the Lowry i assay: (+) Antig&"nicity
as determined by PHA using ;ntiqern directed against
LTA backbone.

Iw /!d saloW 71
O o 0 0 0 o LD ,q N 0 cO 0 n o c a N





.0/ 0


0 Ll o --


4+ 0

O il

I- b



09 CJ HJ "(NJ('4


o 0 0 L
g- C 00 0 0 0

-., E-4

O a a-*-a --u

-- 1

c ::

*, a am ,
O Li

LL "

5 4 0 0 0 0 Li < Li
LiLiLi c : cl I:

.1 -4

It -4 0
0n C- It L U

lu*h 3

O E i LA, (" H 0 4
O 03 O 2. Co
0 0 0 i
.. 3 X 0 X C-H bc m U e o

C L * C
-- L

UJ'H '

C Cr a~
0 C

t 4.4

4.J I 0 0 4- 0 -0 cCO O

m~ -cr
CLe O J O N 0 04
o E * "H H 03 ( C~I 4 4J 4cC

u u 1 -4 .-i
Cd UH: ~ i

ci1 Cr:
t O

04 Un4"O*
C- -C CC3 (
O 0 -. 0 Ct

CE--4 O H U
0 1-4 C3
Urf -I 0 C l O r

-1 x W -4 u V) C

U-'w -4 4 C 4-I C 4CO


3~e en7 N CO >

vj Li AJ w) c ..0 O 0C -C w~~ M r X C T'

O LO( 0 0. Li
-H et

1-1 --4 '
E~ 'L: H

CL 3 .0
o M- *-4

U m l ru (P
'c Cr U O > 0 4 0

e U - -1 -4 N x X en e-i
< r.. e N ru t F -4d O C L 4

wn cm a.1 .-4 al c= i
0~. -C 0 0 0

0t O t c.
I [ a-4 (I-i -H r~ ..0 ..CO

(j rH .-tCff) W (nU
C:" T- ft U -4C

U) CL t u-- t4 H tD -4

Sa C 0 + 0d -- at a- X C 0 :-s 1J L4 000 0 .0 0 u

e-4 0 aU *r4
Cr U O 3 *C

O O 0 O CL W,- 0: 0u < Cr r 0 *< 3 C OZ CO U vi (--4 3 > cO
I. Ej 0I 0 -d (' ... to
-4 1 0 0 0 > *H 0

u 0 0 f 0 0 .4 C C 0 C O 00 0 < J
cil ~ ~ ~ ~ ~ et -O 0i )'.(O ~ r

O3 'I' a O 0 CL v 0r 1 01 0 O I.-.) ~ ii Ci >u

0J I J' NY c ed ru H 4 0 3 0
1 -0 al a -U -- -e 0
(1 --4 1 0 -3 4c C 4- C J 4 5 0 i-* 3.r 0 O o 0 t~- 0- .. "O -1
CO( 3 H- 0j
CO -n O 1U C C

amount of LTA in the sample, PHA titer, and percent lytic inhibition of EA remained relatively unchanged or increased in value.
Purification of LTA by hVdrophobic interaction gel chromatogra hy. A 25.0 X 2.25 cm column packed with Octyl Sepharose ;ind equilhrated in buffer A was prepared as described in Naterials and Methods. Approximately 6.0 mg of LTAppx dissolved in 10.0 ml of buffer A were applied to the column. As can be seen in Figure 13, a small amount of phosphate containing material passed unimpeded through the column. A slightly greater mass of polysarcharide was also excluded without binding No additional material eluted from the column with buffer B. Point C on the graph marks the location where a 10-70% propnnol gradient was begun. Point D represents the point where a signficant volume decrease per test tube was observed. Since fractions were collc:ct.d n a "Idrops per tube" basis, the presence of propanol in che effluent causes a change in surface tension of the drop resuLting in decreased volume per drop. The ultimate result is a decrease in the volume per tube. This, test tube volume provided a convenient means o, monitoring the progress of the propnnol gradient.

It should he noted that despite the use of a grncient (the original procedure called for a single step-wise clution with 50" propnlol) significant amounts ,f carbnWhydrnte eiteld with the LIA. As indicated on the the graph, all areas containing phosphates a lso contained L'TA as detetced using PHA. The Iact thAt a smAill ount of LTA pa ssd inhound through the column sugp pests- that either the column binding capacitv was exceeded, r perhaps the ITA was onlv part inllv n'viated and not capable of tenacious hvdrophhir hindin .

Figure 13. Purification of LTA by Octyl Sepharose hydrophobic gel
chromatography. Symbols: (o) A260 :ibsorbance; (A)
concentration of carbohydrate (n-mole /mt) as dprermined by the Phenol-Sulfuric Acid assay. (Concentrati,,ns were
determined using glucose as a standard carbohydrate.
(o) Concentration of Pi (n-moles/mi) as determined by
the Lowry Pi assay; (+) zones of antigencity as
determined by PHA using antisera directed against LTA backbone; (A) elution with buffer A (1.OM NaCI, 0.01 M Tris-carbonate pH 6.8); (R) elution with buffer B (0.01 M Tris-carbonate, pH 6.8); (C) elution with a 10-70' gradient of a propanol-buffer B mixture; (D)
elution volume at which significant reductions of volume/tube were observed, indicating elution of propanol.

IW/!j saloW u

. . 4-....


-- O

C "




0n )4. 0

cCc co o a:)co co ae j 0 N 0 rt) < m -c

Iw/(3on1) dv3 salow u


All test tubes containing greater than 25.0 n-moles Pi/ml were pooled. The entire peak (approximately 22 ml) was loaded on to a 65.0 cm X 3.0 cm column packed with LII-20 equilibrated with deionized water. Four and two tenths milliliter of effluent were collected per test tube a a flow rate of approximately 30.0 ml/hour. The results of this procedure, which simultaneously removed salt and propanol, nre shown in Figure 14. The column effluent was monitored at a wave length of 220 nm and was also screened for LTA by PHA (++++) using a single dilution sample. In addition column fractions were tested for the presence of chloride ions by placing one drop of a saturated AgNO3 solution on a coverslip containing one drop from each test tube. Any resulting precipitation was evaluated on a +1 to +5 basis and plotted accordingly. It was empirically determined that not only Cl reacted with the AgNO3 resulting in insoluble AgC1, but the NaN3 and tries carbonate in the buffers reacted as well. The presence of propanol was monitored indirectly by changes in test tube volume. Since LTA. azide, and tris carbonate all absorb at a wave length of 220 nm the combination of ultraviolet light screening, the AgNO3 precipitation test, and visual inspection of volume changes per test tube proved to be invaluable for rapidly discerning the location and separation of LTA from contaminating salts and solvents. The entire contents of peak I were pooled, frozen, and lyophilized. The Final product was referred to as LTAosx (extracellolar l ipoLeichoic acid purified by OcLvl. Spharose hydrophobic a ffinityv gel chromatography). The typicAl mass yield from s!(h a procedure was about 60-700. Prcent recovery of LTA at var Ins points in the procedure is summarized in Table 4.

Figure 14. Simultaneous removal of salt and propnot free LTAosx
by LH-20 gel. chromatography. Symbols: (o) A220 :Ibsorbance; (o) Volume/test tube: (+) Antigenicity as determined by PItA; (Shaded Area) relative degree of precipitation of salt and other low molecutiar weight materials
as determined by AgNO3 test.


39NI .LS31 /3iNN1OA

0 0 0 0 0 In ro J -+ + + + + O

o131 ONbV

\ '0 0

0 0
0 0

1 0 a



O 0






0 0 0 t 0 0 ozz v


go 4 4 N O f- )o

UO 0 N r-



li a.n 0 u to
O 4 0

>-J cl cf 00 "
O rd <

41 cn m o J_- L~


I- 0
i to ON Nio

4 o

O 0

ct vi a
H w 0 C
3 EE

4-1 w
o a r- Q -4 COC
o 0i 0 <1 '-0, 04

UHO O) (0 C! 0~ 1'- 1-4 C)a o N

0 F-1 4u 0

QH (-4

4-4 4

-.J u

0 0
CO _r

o C

uN 4a -4a
t4-4 r-u

C: m 0H- 0j V40 0 IJ .14 rd i
0 t 0 0

3 ~ t~ Q. Ol O
-~Q -4U~ 0 0 0
4-1 -4 aa~

0d c'o C

O3C 0 ) I a 3 C
'H fl 0
U ~~~ aa(-t
o~ Uu

I- u 0 N N1 0 0 C CO-A 0 0 0 '0 U 0) (-C '1 r-- -af 04

04I U-H 4U

tnC '-- C) E O 040


c~o to CO CO.- 04 O 0 '-I~~e e
u-i)C i -JO u Z WN .-u I-- ---4

lc OOC C COP C)0

CO 0:n ) QI C ) P .
HUd HO-d 0 3 04

U U- C ) -1
a --0 4-C, -'0 04uiu a
o3 X 4' '0( -4to 'HM '0 0) U 1!) 0. -C,) Ud CC U r 0.: -1: -i1C Om CO 04 (0r r o
-r '~ ( O0'0 ~m Cofl
.2 r 0. 0. 0C4-C C'. C' 2

Phosphatidyl choline vesicle (PVC) purification of LTA using C labelled phosphatidyv choline. Approximately 5 X 10 DPM of C labelled phosphatidyl choline were added to 40 mg of phosphatidyl choline dipalmitoyl. Phosphatidyl choline vesicles (PVC) were prepared as described in Materials and Methods. Three test tubes containing identical volumes and concentrations of non-labelled PCV were prepared simultaneously and 2.0 ml of LTAppx (1.5 mg/ml) were added to each test tube. Fifty microliter samples from the C containing test tube were removed at various steps during the purification process and analyzed as described (Materials and Methods). A standard chloroform quench curve was constructed and all reported counts represent corrected DPM values. Table 5 depicts the distribution of 14C counts at various steps in the purification procedure. Utilizing this procedure as described, essentially no contaminating phospholipid could be detected in the final product. The typical mass yield of product via PVC purification was about 10-15%. Percent recovery of LTA at various steps in the procedure is summarized in Table 6.

Comparison and summary of LTApcx versus LTAosx. As indicated in Table 7, both methods of LTA purification removed the majority of protein as compared to the total amount abailablo in the LTAppx. Both methods ostensibly recovered > 80% of the original ITA. However, the major difference between the two products is reflected in the percent total, mass recovery and the concomitant increase in percent carbohydrate in the final LTAosx product. This latter difference can he most readily discerned by observing the composite gas chiromatograph tracings in Figure 15. The carbohydrate standard (CHO-STD) depicts the typical chromatograph of glucnse and ;glactoe after preparing trimthylsillvl


c o O O" r-. 'a -o C 7 -i O C >
-4 ao l- a < o 0 C C CD 0 m O O C -O c i O

0 S W 4 C 00

<- -0

Sa a c

4-1O 4- N 4o t o 4 3
O 0 4 O e-1 C "O C C1 0i O 00 01 m try ,-4 4 i a ,-a I

C) :)L 0 CL -41 c-0 co c _r Ln CD 0 C o a oa CD c cc o c)-a ca -- -- a a-> O a C) 0.0 0000 000C o a u L c. V1 *I

O .M
U Ga

--4C) 3, C o -. -T C) C) -4 Lc- O] C .;

- a c O #O O- O CO --. H- . r 4 r 0 C) C- c-i fq ? ) 0 N c CD O i .r4 m

-W 0 r. I) Hn

O t C) C : i - c- ) O O -4 C O

",Inc a c ..- o O .0 0 -) Oc C2c --a


L) Li

S c C) C C) C) c- C C) c --4
I0 C Cc > CC 1 0 co c- w Li r r. r L i C
a w- c.:. ca )x Li) Li

w a


0 0

w o D C C: .

o co O

o a on 00 0 0 0

c o-' o oo 00

Ooic4 >0 w o 0 0 c .. $o uc

SHO 0 o

- on -0 0 "T L. CO C) 0

O cd c o -4 0 a 0 0 c

C) 0

-4 T- 7 1

.2 D .-4 N (l 2 m .

c 0- r- .c Oc 0 r. C Z C



f O CL
ID .-.

N C>': -- .-ac L

,,, ... 41 ,
CS 0

CD, Li 3 C i <- .. U ,/

C 0 1
Lr n


0 n

4 C: .,q .-,4- 4 0- C '2 Q rU C 0 Li C C C C Li ,- L .i

o.0 .-,0- a

< C'-, ,nLi 0C Li CC S i 0 C .: l) u C- C C) C-, *H-) C) C C

-'- .H U -u 0Ti Li--* n a

-"C 2 4 4 LC i C CC
C- 0 c L i Cr C 1(. C -4

3) I Li C- 3 C.a

Figure 15. Carbohydrnte analysis of ITA containing preparalt-o,ns
by gas liquid chromatography. Abbreviations: (MAN) Mannitol; (GLC) Glucose; (CAL) Galactose. Mannitol
was incorporated as an internl standard with all







LIA osx LTA p~x



ester (TMS) derivatives as described in Miterial and Methods. An internal mannito standard is included with all samples. The LIAppx chrnmatograph represents the typical carbohvdrnte profile achieved with partially purified LTA. The tracings for LTAosx and LTAppx contrast the qualitative and quantitative differences in carbohydrate content. The second two chromatograms,dencylated cardiolip in ((3 P2) and cardiolipin, were included as a comparison of how a naked polyglycerol phosphate backbone might be expected to react under the described conditions. The base line instability of the G3 P2 looks remarkably similar to the profile of the purified LTApcx. The procedure for purifying deacylated cardiolipin requires passage through Sephadex columns. It is quite conceivable that the minute quantities of unidentified carbohydrates which are indicated may be due to dextran contamination from the column. However, it would be difficult t to account for the same source of contamination for the LTApcx since gel. chromatography was not used in the final purification. On the other hand, the similarity of the indicated chromatogram tracings may be more than more coincidence and may ref lect a:ctun! reactions of the derivatizing agent with the pol.:yglycerol phosphate backbone. This latter hypothesiis s suipprted by the fact that an unidentified trailing "carbohyvdrato" peak of sign i i.cnnt mass appears in both the cardiotipin and C3 2 chromatographs. The RE value of this peak is similar (huitv t s,,sp iciously d isparate) to the rcat: tion t ine normally observed for N-glucIse. However, if inded this peak does represent B-glucose, one is hard pressed to, ration alize why a corresponding n-glucose peak does not occur as wel. In either case, it is

Dencyl:ited cardioL ipin was prepared by the method of Wi lkinson (14 q) and was kindly provided by R. Craig, University of Florida.

apparent that the LTAosx st ill contains if icant nmounts of carbohydrate contamination in contrast to the i noer vilid, but high!: prified LTApcx.

A summarv of specific activities relative to PH;A activist is pr.sented in Table 3.

Inhibition of complement mediated l':sis of LTApcx treated EA\.

Once a highly purified preparation of LTA was obtained, it was necessary to confirm the results that had been previously established with LTAcx. As can be seen in Figure 16, not only were LTApcx treated EA refractory to complement mediated lysis, but in addition the general profile was remarkably similar to LTAcx treated EA. As indicated, LTApcx was used in concentrations five-fold to ten-Fold less than those used with LTAcx to achieve comparable degrees of inhibition. PHA titers typically indicated LTA saturation of the cells.

Effect of LTApcx on_ the .vsis of various cel Lular complement

component intermediates. Sheep E, EA and various complement intermediates were treated with LTApcx (100 c /ml in DVR) and washed extensively in DGVR. Percent lysis and inhibition of CH5 units were determined as described in Miaterials and Metihods. The results of these experiments are summarized in Figure 17. As with LTAcx treated cells. E, EA, and EAC1 were al 1 ref ractorv to lvsis 1by n mln ment. Li kewise, EACI42 and 'ACI-7 inritWrmediates were total llyv iif:iIfe. etd by the presence of LTApex on their cell s rfaceq. The on lv observahl, diffr'nce in the activity of LTApcx on ce:tlllalr intermedi teos versus LIArx was that E.\:li" cells were inhibited to a l esser degree with LTAp'x than TArcx.

Effect of LTAcx and LIApcx on fluid_ phase Cl. As previo us v

discussed, CA is not ronuired for yIvsis once SAC42 are formed but it is

t"1 sq 4m 1 1 O I n
r co cs n
-0 cn I

<"1 C LO N- M

N0 .


Otf a- l

oo : 0 co

,4 CL : t

-siv e-1 e
Cs ..o r- CO Cs .I * O 2.CO 0 -Z n- Cl C
N-tf- 0 0
c I0


JO C O D C U 4O4 C r' .) 0 C) CI -4l C '] C 0 "O
*,- ..- .. -. cC O tt O

ca~O 0 0
Co 0 \ O~ O '0 0
0 4 0 ,l -- 0

a4 .. ro a 7*
o C CO

m~ 0) MOi
41n -> 'H

&-4 & H -O U ** U

-n o rt a C -'

GO 4HOC H H HH-GO 0410 H ) .? f r 1 : U U 0 3-4 04 --t .- C
09) .0 Cs. '-0

U 54

C. 4. C

09 U 0 CC .

410 0 *- ~ ~ 3 iV
0 Ls 04 0 0-v (
W I L-. -. 0 jOC U ~ ~ 4-O -r CO 'CU 0i U3 Cs C CCU
'-t C:. 400 a


c C C *l U

C4. 04 cl Co Cl r C
'TCs .0 C: 0.)r:c,

-]~~- C) oSi" -3 C

Figure 16. Passive hemagglutination (PH1A) titration and inhibhit ion
of complement mediated lysis of EA treated with varying
concentrations of L'TApcx.

] TITER 1600 3200 3200 3200 3200 C 6400 S3200

S1600 Z 800
uLL 400
1 200

CL 6.25 12.5 25 50 I00

:: 0 % PHA
50 % PHA S100 % PHA

> 70

- 40

z 10

w 6 25 12.5 25 50 100


Figure 17. Effect of ITApcx on the complement mediated 1vsi~ of
various cellular complement components intermediate.



I o


w~lU O O-- O rcr
< <0 w- 32


n O O N O


essential until that point is reached (148). In addition, mnny polyanionic substances are known to directly affect Cl by interfering with Clq binding or Cl esterase (CIs) activity (39,40). Because LTA is polyanionic due to the polyglycerol phosphate backbone and because cellular intermediates beyond the EACL42 step were no longer inhibited, it seemed reasonable to hypothesize that LTA was behaving like a polyanion and directly affecting Cl.

To test this hypothesis, LTAcx and LTApcx were preincubated with functionally purified human Cl at 300/15 minutes. Residual Cl activity was titrated as described by Rapp and Borsos (139) and activity was compared against buffer treated controls. The results shown in Figure 18 indicate that although LTAcx consumed C1 activity, purified LTApcx did not.

Purification of human C1l, Cls and C1s. In an attempt to further eludicate a possible site and mechanism of C inhibition, human C1 subcomponents were purified by the methods and modifications previously described (Materials and Methods). Although homogeneity beyond functional purity was not essential, the methods employed yielded highly purified products. Figure 19 demonstrates Clq homogeneity by immunodiffusion against several monospecific antisera. Precipitation bands of identity were observed in adjacent wells containing the whole human serum starting material, the purified Clq final product, and a hi ghlyl enriched C1lq prior to final precipitation (Figure L9, plate 5, we11 numbers A. (, and E). As can be observed in Figure 20, disc gel electrophorests of the final product revealed a single dark staining band which barely migrated into the separation gel. These results are consistant with the observations of other investigators (144).

Figure 18. Effect of LTAcx and LTApcx on functionally purified
human Cl. The upper graph represents a residual Cl
titration after incubation with LTAcx (500 vg/mi).
The lower graph represents the restIlts from an analogous experiment using LTApcx (500 ug/ml) instead of LTApcx in the incubation mixture. Symbols: (0) C
incubated with biiffer; (o) Cl incubNaited w ith the
appropriate LTA containing extract.



1.8 1.4

. I .0
_J 00.6 U0.2 to L. I 6,000 8,000 4,000 2,000


Z2.2 w2


06 02

16,000 8,000 4,000 2,000

Figure 19. Immunodiffusion and precipitation analysis of various
steps in the purification of human Clq. Purification was achieved by repeated fractional precipitations of whole human sera in buffers varying in ionic strength,
pH, and concentrations EGTA or EDTA (144).
Well designations:
(A) Whole human sera (starting material);
(B) Supernate from first precipitation;
(C) Supernate from second precipitation;
(D) Supernate from third precipitation;
(E) Material from resuspended pellet prior to final precipitation.
(F) Final product (purified Clq).
Plate designation: (1) Center well contains anti-IG(;
(2) Center well contains Anti-IgA; (3) Center well
contains anti-whole human scra: (4) Center well
contains anti-IgM: (5) Center well contains anti-Clq.

'. ia


~LB~lg lldQQ ~J

Figure 20. Disc gel electrophoresis of piirified human C1q.
Cathode was at the top.

9 ) E.