A Study on Cross-Reactivity of Anti-DNA Antibody with Glycosaminoglycans

~aoki Kashihara *, Shuzo Hirakawa, Yasuaki 1\Iino, Hirofumi l\lakino and Zensukc Ota

Third Department of Internal ,'J1edicine, Okayama University .lledical School, Okayama 700, Japan

.\ eta :\led Okayama 47 4 255 259 (1993 )

A Study on Cross-Reactivity of Anti-DNA Antibody with Glycosaminoglycans

~aoki J{ashihara *, Shuzo Hirakawa, Yasuaki 1\lino, Hirofumi 1\Iakino and Zensuke Ota

Third Department of Internal Medicine, Okayama L'niversity J!edical School, Okayama 700, Japan

To study the pathogenesis of lupus nephritis, the cross reactivity between anti-DNA antibody and glycosaminoglycans (GAGs) was investigated. :Monoclonal anti-DNA antibodies were obtained from hybridomas by the fusion of l\1RL/ Ipr/lpr splenocytes with murine myeloma cells. Some of these monoclonal anti-DNA antibodies showed cross reactivity with GAGs, such as hyaluronic acid, chondroitin sulfate and heparan sulfate. To elucidate the mechanism of cross reactivity, inhibition assays with propanol and polyethylenimine (PEl), a cationic agent, were carried out. Increase of the concentration of PEl (0.6 2.0 % vol/vol) resulted in a dose dependent decrease in the binding ability of anti-Dl\A antibody to GAGs. Propanol, an organic reagent which disrupts the van der Waals bonds between epitopes and paratopes, showed little inhibitory effect on the binding activity of monoclonal anti-DNA antibody to GAGs. These results indicate that the binding of anti-DNA antibody to GAGs is due to a charge interaction rather than van der Waals forces. Anti-DNA antibody which can react with GAGs in the glomerular basement membrane seems to play an important role in the pathogenesis of lupus nephritis.

Key words : anti-DNA antibody, cross-reactivity, glycosaminoglycan, lupus nephritis

High titer of anti-DNA antibody, especiall y anti­double stranded DNA antibody activity in the sera is specifi c for the patients of systemic lupus erythematosus (SLE) and correlates with clinical fmdings and severity of

renal involvement (1). However, the pathogenic signifi­cance of anti-DNA antibody in lupus nephritis remains uncertain.

There are two proposed mechanisms by which anti­D A antibody could cause lupus nephriti . One is the circulating immune complex (CIC) mechanism: circulat­ing DNA/ anti-D A antibody immune complex is present in the sera of patients with lupus nephritis. However, the direct demonstration of D~A both in CIC and glomeruli has been d:ifti.cult, and reported only in a few studies (2, 3). Another mechanism is the in situ IC formation mechanism: Izui demonstrated that DNA had a high affinity for collagen molecule in GBM in vitro (4). They proposed that circulating anti-D:\A antibodies would bind

*To whom correspondence should be addressed.

255

to D:-.JA bound to GBM e:md form the immune complex in situ. However, there is a lack of du·ect and concrete evidence which supports either of the e two hypotheses.

Recently, Faaber el al. showed the cross reactivity of anti-DNA antibody with GAGs (5, 6), and proposed a new hypothe is with re pect to the pathogene is of lupus nephritis. They suggest that anti-DNA antibody binds directly to the GAGs to form the immune complex in ·itu

and causes lupus nephritis. We cultivated the monoclonal <:mti-D A antibodies

and studied the cross reactiyjty with GAGs to investigate

the preci e mechanism of thi · cross reactivity.

Materials and Methods

,\1ice. l\lRL-lpr/ lpr mice W<'n' purcha~Pd from Ki\\"a labora-

tory, Japan and Bi\LB/ C mice were obtained from the animal

colony of our ins titute.

Cell fusion and cloning. Somatic cell hybridization was carried

256 Ka~hihara d a/.

out as discribcd by Kohler and :YrilstC'in (7). Spleen cells from MRL-Ipr/ lpr mice were fused with the non-immunoglobulin secret­ing aminopterin sensitive murine myeloma cell line. ~S 1. Anti­DNA antibody-producing hybridomas were cloned by repeating the procedure of limiting dilutions.

DNA and glycosamirwf5lycan~. Calf thymus D 'A, pur-chased from Worthington Biochemical Corporation, ew Jersey, CSA, was dissolved in 0.01 :vi Tris-ITCI buffer, pll8.0containing 0.001 M I ~ DT/\. Chondroitin sulfate and hyalulonic acid were obtainncd from Sigma Chemical Company, St. Louis, Missouri, USA lleparan sulfall' was obtained from Scikagaku Kogyo, Tokyo, .Japan. Tlwsc GAGs were dissol ved in phosphate buffered saline (PBS), pl l 7.4.

l~nzymc- link('d immwwsorbcnl assay. Anti- D A and anti-c; AG antibody assays wer<' perform(•d by rnzyme-linked im­munosorbent assay (ELISA). Each well of polystyrene microtiter plate ISumitomo, Tokyo, .Japan) was precoatrd with lOOJ.ll of 0.75 mg/ ml of prolamine sulfat<' and incubated overnight at 4 C. After washing with PBS, lOOJ.l l/wcll of DNA (200,ug/ ml) or GAG (500,ug/ ml ) were adckd to each well and incubated overnight at 4 C. Thl' plat(•s wcrc washed three times with PBS to remove excess <mtigcn. Unbound sites on the plates were subsequcnLiy hlockcd for 1 h at room tl'm prature with 1 % ( wt/ vol) bovine surum albumin (13Si\ ) in PBS cmd then washed three times with PHS. One hundrC'd ,u I of a monoclonal antibody to be tested was added to each well and incubated for 1 h at room temprrature. After washing three t[mes with Pf3S, 100 ,u I of peroxidase­conjugal ·d goat <mti-mousl' !gG (Cappel, Pcnn~ylv;mia, USA), diluted 1: 3,000 in PBS containing ] % BSA werC' added and incubated again for 1 h at room tempcratur·e. Then 100,u l of ~ubstrate solution, 1 rng/ ml o-phcnylencdicullinc (W ako Pure Clwmic;u Industries, Tokyo, Japan) dissolved in 0.1 M citrate buffer. pH 5.0. containing 1 ,u 1/ ml of 35 % ( vol/ vol) ll20 2 was added to each well. After 10 min , the reaction was stopped by adding 50 ,u I of 3 I ICI. Thr optical absorbance was measured at 492 nm by <m ETA RE1\DER 810-RAD MODEL 2550.

Crithidia assay. Crithidia assay was carried out to verify anti-double ,;trandcd DNA activity. Crithid ia luci llae was plated on glass slick . .\lonoclomu antibodies were placed on the slides c:md F lTC-cnnjugated goat anti-mouse lgG antibodies 1\Wl' applied. F luorescence of kinetoplasts was obst>rVl'cl h~· immunofluorescent microscop~.

ELIS1\ u•ith dij{erent salt concentration. To irn·estigate the influence of ionic strength on the interaction of monoclonal anti­bodies and D A or GAGs, ELISAs ll'ere performed with \'ru·iou · concentrations of aCI. l\lonodonal w1ti-D lJ\ antibodies were appliecl to th(• antigC'n-coatt•d wells in the presence nf different 0JaCI wncentrations (0.03 1!\I).

Inhibition studies. To elucidate the mechanism of interactions between cmti-D TA antibodies and ])~ ,\ or G.\ Gs. El JSA wC're performed in the presence of a cationic agent. PEI or prop<mol. 1onoclonal antibodies 11·ere appied to ;mtigen-coated \\'ells in tllC'

presence of increasing concentration of PEl \0.03 1 % \'OI 1·ol) or propcmol (0.03 4.00% \'OI \on. The following procedw-es 1\'CI'e

performed under standard conditions. As a control study, the scm1e inhibition studies were performed using tll)Toid microsomal antigen and monoclonal anti-microsomal antibody, which were produced in our laboratory (8).

Results

Production of monoclonal anti-DNA antibodies. The cell-fusion procedure was carried out twice and yielded 98 cell lines producing anti-DNA antibody.

Cross-react·ivity of monoclonal anti-DNA antibodies with GAGs. Seven clones of anti-DNA antibody produc­ing hybridomas were able to cross-react with GAGs (Table 1). Anti-DNA antibody activity of these clones was also conftrmed with Crithidia assay. There was a correlation between anti-DNA activity and anti-GAG activity in these monoclonal antibodies (Figs. 1, 2)

Irifluence of the salt concentration on the binding of monoclonal anti-DNA antibodies to DNA and GAG. Decrease of binding activity to DNA and GAG in the presence of increasing concentration of NaCl in a dose­dependent manner was demonstrated (Fig. 3). These data indicate that bindings of these monoclonal antibodies to D A and GAGs were not mediated \-vith non-specific charge interation because binding activities were not signiflc;mtly reduced with 0.3 M NaCI in which concentra­tion, non-speciftc charge interaction was inhibited (9). It

was also demonstrated that aCl at the concentrations used in this study did not dissociate the DNA or GAG from the plate (Data was not shown).

Effect of PEl on the binding to DNA and GAG. The presence of increasing amount of PEI caused dose-

Table 1 Binding ac tivrty of monoclonal <mti-Dr\r\ antibodies to

glycosaminogl; ran'

Antigen llybr·icloma

r· lorw' llcpanm Chondroitin IJ1·alulonic D~ . ..\

sulfate sulfate acid

lG3 0.599 0.413 0.609 0.409 L\9 1.273 1.820 1.523 1.667 3G2 0.380 0.3~8 0.351 0.647 .3Bl 1.033 1.1 -14 1.096 1.015 389 1.778 1.092 1.351 1.746 9C-l 0..128 0.905 0. 9~ 1.989 7C1 0.759 1.664 0.326 1.345

.\II ,-a] ups represent optical ab,orbance at 492 nm.

Cros~-Rea('tiYrty of ,\nu-D:'\A :\ntihody 257

r~o. 95 N

"' P<O .O Ol ..,. 0 0 •

•

" ~ "' 0. C.J • "i' 0.5 .lJ • " <

• 0.5 1.0

Anti-DNA OD 492

Fig. 1 Con·elation between anti-Dt\A activity <md mlti-hcp<mm sulaft.c· acti1·ity of monoclonal anti-Dl\ ,\ tmtibodics.

~ '0 <=

_o

-<I)

u <I)

0..

OL---~----~----~----~--~----~----~ 0 0.25 0.38 0.5 0.75 1.5

NaCl concentration ( M

Fig. 3 lnOuence of inonic strength on the binding actil' itie. of monoclonal anti-Dt-.;A m1tibody (li\9 clone) to D\'A <md glyco <uni noglycan . Binding of monoclonal <mtibocly to D:\ A ( ~ ) and hcparan sulfate ( • • ).

dependent reduction in anti-D A and anti-GAG activity (Fig. 4). Five % (vol/ vol ) of PEI reduced anti-DNA activity and anti-GAG activity by 50 % and 65 %, respectively. But PEI did not show an inhibitory effect on

r ~ o _ 90 P <0.025

Ant i-DAN

•

•

0 0 492

F'ig-. 2 Corr·datio n between Wlli -Di\ ,\ a('lil ity and <Ulti-dHJndruilin ~ulfatl•

acti1·ity of monoclonal an ti-D:\ i\ <mtibodies.

"' " 'tl

" 3 -'

~ 50

0

"' ~ a.

Polyethyleni mine concentration ( % )

Fig. 4 Lnnucncc of polyPUl) lenimine on tlw binding of monoclonal

anti hodiP!-. to D~/\. gi)Co~aminoglycan cUld Ul)Toid rnirro,omal <mtigen . Binding of monoclonal anti-0:'\/\ antibody (l /\9 dum• to]);'\,\ r ) <mel heparan sui fall' ( • • ). Bmcling of monoclon<J anti- mitTU!-.O tmJ antibody tu tl1yroirl micro!-.Olll<J ;mtigl>n • • ).

the other antigen-antibody interaction; microsomal anti­gen and monoclonal anti-microsomal antibody.

Effect o.f propanol on the binding to DNA and GAG. Propanol is an organic agent which inhibit van der Waals

258 Ka;,hihara el rd.

= 50 > ~

0 ~------~----~~------~------~ 0 0.625 1.25 2.5 Propanol concentrat>on ( % )

Fig-. 5 lnflu l'nCP of propanol on thl' hmdin ~ of nwnorlon;tl an tihodil's to 1>\' :\, gly('();,am in ogi)C<Ul ;u1d Lh) roul mirro;,omal anligl'n. 13inding of mono­clonal an li-D :i\ <tntihody (1 ;\ 9 clont· ) to Dl\ :\ ) and lwparan sulfatl' t • • I. Hind in~ of monoclonal anti-microsomal anti body to Ul) mid mi-cro;,orna l antigl'n • • I.

force between antigen and antibody (10, 11). As shown in Fig. 5, increase of prop<mol concentration had little effect on the binding activity of monoclonal antibodies to D A and GAG_ On the other hand, the binding activity of monoclonal anti-microsomal antibody to microsomal antigen was inhibited by propanol.

Discus ion

Anti-D A antibody has been believed to play an important role in the pathogenesis of systemic lupus erythematosus ( I .. E)_ But the mechanism that involves anti-D A antibody that cause, tissue injurie has not been elu cidated. Development of monoclonal <mti-Dl A antibodic - by somatic cell hybridization techniques has facilitated the analysis of the character of ;mti-D_ A

· antibody and its target antigen. Studies us ing monoclonal anti-D A antibody re,·ealed

that anti-D A antibody could react to a broad spectrum of antigens including those with molecular structures that were apparently different from D A. Cross-reacti,·ity with the synthetic polynuclcotides of different compo,- ilion ha. been reported. The cross-reactive <:mtigcnic determi­nant seemed to be in the ugar-phosphate backbone (12). Anti-DNA antibody were able to react with cardiolipin

and phospholipids. Phosphodiester linked pho phate groups were identified as possible cross-reactive moieties

(13 15). Furthermore, cross-reactivity of anti-D A anti­body extended to a variety of antigens, such as IgG, platelet, Raji cell and endogenous bacteria (16 18). F aaber el al. also . reported the cross-reactivity with

GAGs (5, 6). The wide variety of cross-reactions is diffJcult to explain with the existing data. This situation seems rich in possibilities for further investigations.

The repulsive and attractive forces that constitute the non-covalent interaction between antigens and antibodies (10) are charactrerized by a) Dispersion (van der Waals) force and b) E lectrostatic (Coulombic) interaction. The organic reagents, propanol and acetic acid, decrease the surface tension of the liquid medium and disrupt the van der Waals bond to dissociate the antigen-antibody com­plex. 1icrosomal antigen and anti-microsomal antibody complex is eff1ciently dissociated by propanol. However, the bindings of monoclonal anti-DNA antibodies to DNA and GAG were not inhibited by propanol at the concentra­tion in which microsomal antigen and anti-microsomal antibody was dissociated. From these date, it may be concluded that van der Waals force does not contribute

the interaction between anti-Dl A antibody and DNA or GAG.

Both D\TA and GAG are negatively charged. Ebling ;mel Halm reported that anti-D. A antiboilies eluted from NZB/ W F1 mouse kidneys had higher isoclectric points than those in the sera. They speculated that cationic anti-D A antibodies were pathogenic in lupus neplu·itis (19). In the present study, the cationic agent, PEl, inhibited the binding of anti-DNA antibody to DKA and GAG. It is pos ible that the binding of anti-D A anti­body to GAG is mediated by simple charge-charge inter­action. Because the 0.3 M ionic strength, which can prevent non-specif1c charge-charge interaction ilid not affect the interaction between anti-DNA antibody and GAG. this interaction does not seem to be mediated with a simple force between opposite charges. From these re, ult , it appear that this charge-charge interaction exists between epitope and paratope.

Van der Waals force decreases at a rate inverse to the 7th power of the intermolecular distance. The distance between two molecules should be very close for van der \Vaals force to work efficiently. On tl1e other hand, electrostatic force is inversely proportional to the second power of the intermolecular distance and could work at distances where van der Waals force could not work.

Cro,.:;- fkacti\it) of .\llli -D\'. \ .\ntiholh 259

There wa a significant correlation between anti-D1 t\ activity and anti-GAG activity in these monoclonal <:mti­bodies. However, D A and GAG do not hm·c similar molecular structures other than repeating negative charged residues. It seems to be reasonable that vcu1 der Waals force does not work effectively between such molecules.

In this study, the cro s-reactivity of monoclonal CU1ti­D A antibody with GAG was demonstrated under vcu·ious conditions. Hepcu·an sulfate proteoglycan, which is one of the GAGs in glomeruli, is an important compo­nent of anionic site in the GBM and necesscu·y to maintain both charge and size selective bcuTiers (20, 21). The cross-reaction betvveen anti-D A antibody with GAG may play two possible roles in the pathogenesis of lupus nephritis. Direct binding of anti-D A cu1tibody to GAG may lead to the formation of the immune complex in situ and cause the activation of complements and tis ue damage. Another possibility is that anti-D A cu1tibody may neutralize the negative chcu·ge of GBM to disturb the chcu·ge selective barrier and result in proteinuria.

References

1. Tan E:\'l. ch ur PI l, Carr HI <llld Kun kel II G: Droxyrihonucleic ar id (1)\'A) in Ll1c serum of patienL~ wiL/1 systemi(' lupus pr·yllwmato~ u ~ . .J Clin In vest (1966) 45, 1732 1740.

2. ll crbcck Rl, Bcu·dana E.\1, Kohlc·r PF and Carr HI: D A: Anti­f) ' A rom pl t>xPs : Tlwir rk·LL·ction in sy"lemi(' lupu~ erylhl•malo;,u;,

sera. J Clin lnvcsl (1973) 52. 780 79:5. 3. Sana II and :\1orimoto C: Isolation of D:\:\ from D\'.\ CUlti- IJ\' .\

<Ultiboclr immune complexes in ")'Slemic lupu~ rrythrmatosu, . . ] lm­munol ( 1981) 126, 538 539.

4. lzu i S. Lambert PI! and :\lir,cher PA: In ritro demonstration of a particulcu· affmi ty of glomerulm· basPment membrane ;md ('O II a~en for D\' :\ . .J Ex p :\ led (1976 ) 144. <128 443.

5. Faabcr P, Capel P.JA, Hijkl' GP:VI. Vierwindf!n G, van dP PullP LBA and Kol'nc !tAP: Cross-rPartivity· of anli -IJ J\ A <mtibodie~ with proteo~lycan;,. Clin Exp lmmunol 0984 ) 55. 502 50 .

6. Faabcr P, Rijke TP:\1, \·an de Putte LBA, Capel P.L\ and BerdPn .JIL\1 : Croso.;-rL•acli\·tt) of hurmm <mel murine anti- IJ\' ,\ illllibodics \\iL/1 hcparan ,u lfall' .. J Clin lm c;,L 1986 ) 77. 182 1 1830.

7. Kiihler (; and \1ilstein C: Continuous culture;. of fused cells 't'tTcUn~

antthnth of prwltfmed sJl!'riftcrl~ . \'aturv 19751 256 . 195 197. 8. I lir;tkawa S. :\litushima K. Shtmitu I. Sunada \I. Sutuki S and Ow

Z: Prndurtion of monodonal ;mtihodit·,.. to human U1)TOid mirrchonw antigens. Fmnt Th) roidnl \ 1986 > 2, I 115 1 119.

9. Ttjss!'n 1': l'rac li n• anrl tht•n r) of t•nt.) lilt' irnrnuno;ts;,a) ,..; in l .ahora­

to ry TPchniqucs in 13iorlwrnislry and :\1nlerular Biology. Burdon <Uld Knippt'nhe r~ t•cl:-. bi Ed. \ 'ol l5. El~l'\icr . . \m,..l t•rdam 1985 pp:l7:l

:w. l 0. Ttjs:-,t•n 1': I 'rarltce and Llwor~ of en1~ nw irnmunt>as-,ay s: in I .altora-

tnr) Terhniqup;, in 13iodwmisU') and :\lolt•rular Binlng). Burdon and 1\nippcnhl'rg l'tk 1,1 hi. \ ·ol1 5. Ebt'\ it•r . . \nbtl'rrlam 1985 pp123

126. 11 . i)p (;root 1':1\. l .;mwr;, \I C, :\ardt•n 1 .. \ . Sntt•enk lt.JT iU1d \CUl (),.,

C.J: Di,.,,..ociation of D\' .\ anli-DI\ \ romplt•xt·~ at h i ~h pi I. lmmunol Commun tl980 ) 9. 515 528.

12. Ballard 1)\\. ;Uld \ ' os;, F\\· : :\ lonol'ional murim• antt-nuciPw ami

anlilmd) \\ ilh doultlt·- ~tr;utlk·d ~pt•rtfrril~. \lol lmmunol 1982 19. 793 799.

13. l .aft'r E:\1 , l\aurh .J . . \ndrlt'it•w,ki C, \ludd D, Furit• ll. FunP B. Schwartz HS and Stolbr Il l l: I 'o l) spt•nfw rnontwlonal lupus autoantihodit·;, n'acli\ t' wil11 both pol) nudt•oLidl'" <Uld pho,plmltpirl,. ,) l·:xp !\It'd (1981 ) 153. 897 909.

14. 1\.oikl' T, Tomioka I I and Kuma~;ti , \ : t\ntihod it•s rros;,-r·l'<H'Lt \·t· \\'tUl Di\l\ and rardtolip in in pati t• nls with sy;,ll'lllic luplh Pr) tlwtnalo~u~ .

Clin l ~xp lmrnunol !1982) 50, 298 302. 15. lchikawa 1\.. Shimada 1..;, \'awata Y. l ~hii T. Tomioka II. Yoshida S

and Koike T: :\lonodonal h\ hndoma anli -cardioltprn ant1hodH•s from

SLE mict•. Clin Exp lrnmunol ( 1988 ) 7 4, 110 Ill.

16. .laroh I ., Tron F, fkt rh .J F and I ,ou\ard 1): .\ nitlnorlona l <U1 t i-D~ ,\

<llllihody also hinds to rd!-,.,urfan• protl'in (s ). Proc i\'ati J\ cad Sn l !:1 .\ (198 1) 81. 38<13 3845.

17. T ro n 1: . .Jacob I , and Bach .J F: l{inding of a munnl' monoclonal anti -D\' t\ antiltod ~ to l{;tj i n·lk Implica tion' for Lilt' intPrprl'tation of tlw i{aji ct• ll a;,"a) for immum• compkxPs. 1·\rr .J lmmunol (198·1 14.

283 286. 18. C<u-ro ll 1', Swfford IJ , Schwartz I{S and Stollar BD: :\1unm• mono­

clonal anli-D\'1\ auloantiltodtt•s htrrd t11 Pnrlo~t'llllll., harlt'n;t. .J lm ­tnunol (1985 ) 135. 1086 1090.

19. llahn BI I: Charactt•ri;,tic~ of pallww·nic suhpopu lation~ of antihodiP~ to 1),\ ,\ . An.hritis ltheum 11982 J 25 , 71 7 752.

20. K;mww· YS: Hiophysiolo~) of 1-':lonwnrl ar· frltration and protl'inut·ia. Lth lrm><;l (198~ ) 51 , 7 21.

21. K<mww· YS and lto;,enzwci~ I ,j : C l o~~in~ of tlw glonwru lar bast•menl nwmhram' . . J C(:'ll Bioi r1982 93 , 189 19 1.

HcceiH·d .J<mu<u·) 7, 1993 ; itt-c-PptPd FPioruar:-- 10. 1993.