koji · mikio okamura, yoshiharu kanayama, koji tadanao takeda, takatoshi inoue abstract the...

Annals ofthe Rheumatic Diseases 1993; 52: 14-20

Significance of enzyme linked immunosorbentassay (ELISA) for antibodies to double stranded andsingle stranded DNA in patients with lupusnephritis: correlation with severity of renal histology

Mikio Okamura, Yoshiharu Kanayama, KojiTadanao Takeda, Takatoshi Inoue

AbstractThe correlation between renal histology andclass specific (IgG and IgM) antibodies todouble stranded DNA (dsDNA) and singlestranded DNA (ssDNA) was studied byenzyme linked immunosorbent assay (ELISA)in 40 untreated patients with systemic lupuserythematosus (SLE). The levels of IgGantibodies to dsDNA were significantly higherin patients with World Health Organisationclass IV nephritis than in those with class I,class II, or class III nephritis. IgG anti-bodies to ssDNA were higher in patients withclass IV than in those with class II nephritis.IgG antibodies to dsDNA showed a close cor-relation with the histological activity scoreand the amount of electron dense deposit.IgG antibodies to ssDNA showed only a weakcorrelation with the renal histological activityscore. IgM antibodies to dsDNA and IgMantibodies to ssDNA were not correlated withrenal histological features. Patients withmoderate to severe nephritis had a lower ratioofIgM antibodies to dsDNA to IgG antibodiesto dsDNA than those with mild nephritis.These results indicate that the measurementof IgG antibodies to dsDNA is predictive inevaluating renal histological activity in patientswith SLE.

(Ann Rheum Dis 1993; 52: 14-20)

The First Departmentof Internal Medicine,Osaka City UniversityMedical School,1-5-7, Asahimachi,Abeno-ku, Osaka 545,JapanM OkamuraY KanayamaK AmastuN NegoroS KohdaT TakedaT Inoue

Correspondence to:Dr Okamura.Accepted for publication19 August 1992

Antibodies to double stranded DNA (dsDNA)are characteristic of systemic lupus erythema-tosus (SLE),' 2 and these antibodies have beenconsidered to be the principal factor in thepathogenesis of lupus nephritis.3 ' Manyworkers have reported a significant correlationbetween serum titres of antibodies to dsDNAand the severity of the disease, particularlyassociated with lupus nephritis.' 57 Relativelyfew studies, however, have examined the cor-relation between antibodies to dsDNA and thehistological severity of lupus nephritis."'0Although Steinman et al, using a syntheticdouble stranded copolymer of deoxyadenosineand deoxythymidine to detect antibodies todsDNA, showed a significant correlationbetween serum titres of antibodies to dsDNAand renal histological severity,'0 other workersusing a conventional dsDNA preparation todetect antibodies to dsDNA, could not confirmthis.8 9

An enzyme linked immunosorbent assay(ELISA) for antibodies to dsDNA is a sensitiveand specific method for their detection and

Amastu, Nobuo Negoro, Shinichi Kohda,

measurement," 12 and also for the measure-ment of antibodies to single stranded DNA(ssDNA). " 13 An ELISA offers the opportunityto measure the different antibody classes and toavoid the measurement of non-specific DNAbinding proteins. Using a chromatographicallypurified and SI nuclease treated dsDNA pre-paration as the solid phase antigen, we used anELISA for the measurement of antibodies todsDNA of the IgG and IgM classes and assessedthe correlation between serum levels of anti-bodies to dsDNA and histological severity in40 untreated patients with SLE. Althoughantibodies to ssDNA have been thought to beassociated with disease activity in patients withSLE,'4 no study has correlated them with renalhistological severity. We have examined thisand appraised the significance of each parameterin evaluating renal histological severity inpatients with SLE.

Patients and methodsPATIENTSSerum samples were obtained from 40 patientswith a clinical diagnosis of SLE at the FirstDepartment of Internal Medicine, Osaka CityUniversity Hospital. All patients met the 1982revised criteria for the classification of SLE.'5None of the patients had received steroids orimmunosuppresssive drugs at the time of col-lecting the serum samples. The serum sampleswere stored at - 80°C until tested. A renalbiopsy sample was taken either at the pretreat-ment stage (30 patients) or within two weeks ofthe start of treatment (10 patients). Patientswere judged to have clinically active nephritis ifthey showed the presence of either activeurinary sediment, proteinuria (over 500 mg/day),or an increased serum creatinine concentrationgreater than 124 ,umol/l.

DNA PREPARATIONSDouble stranded DNA prepared from calfthymus DNA (Sigma) was purified by methy-lated albumin kieselguhr chromatography16 andS1 nuclease treatment.'7 Single stranded DNAwas prepared by heating a 0-1 mg/ml solution ofdsDNA in phosphate buffered saline (PBS), pH7-2, for 15 minutes at 100°C followed by rapidcooling in an ice bath.

DOUBLE STRANDED DNA COATED MICROTITREPLATESMethylated bovine serum albumin (BSA) (10

14

on 25 April 2019 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.52.1.14 on 1 January 1993. Dow

nloaded from

http://ard.bmj.com/

DNA antibody and lupus nephritis

jig/ml) was incubated in Immulon type I flatbottomed microtitre plates (Dynatech Labora-tories, Alexandria, VA, USA) and, after rinsingwith PBS, 0-2 ml dsDNA solution (2-5 jig/ml)in PBS was added, and was incubated for 18hours at 5°C. After incubation the dsDNAsolution was decanted and 0 3 ml gelatinsolution (1 mg/ml) was incubated in the wellsfor at least 18 hours. Just before use, theimmobilised DNA was redigested with S1nuclease solution at 4 U/ml in 0 03 M sodiumacetate buffer, pH 4 4, containing 0 1 M NaCl,1 mM ZnCl2, and 100 ,ig/ml BSA (Cohn'sfraction V, Sigma). After four hours shaking atroom temperature the wells were rinsed twicewith PBS containing 0 05% Tween 20 (PBS-Tween).

SINGLE STRANDED DNA COATED MICROTITREPLATESIn Immuron type I plates 0-2 ml ssDNAsolution (10 pi/ml in PBS, pH 7 2) was incubatedfor 18 hours. After incubation the wells werecoated with gelatin solution as described earlier.

STANDARD ANTIBODIES TO DOUBLE STRANDEDAND SINGLE STRANDED DNAPlasma was obtained from a patient with typicalclinical and serological features of SLE who hadreceived plasmapheresis. Onehundred millilitresof plasma was applied to dsDNA-cellulosechromatography plates (Sigma), washed withPBS until the absorbance of the wash at 280 nmreturned to baseline, and subsequently elutedby increasing the salt concentration in thePBS.'8 The peaks at 280 nm were pooled anddialysed against PBS. The concentrations ofIgG and IgM in the eluates were determined bythe radial immunodiffusion method. Antibodiesto ssDNA were isolated as described earlierusing ssDNA-cellulose chromatography(Sigma). 18 These antibodies to dsDNA andssDNA were used as standard antibodies for theELISA.

ENZYME LINKED IMMUNOSORBENT ASSAYThe assays for antibodies to dsDNA wereperformed on dsDNA coated microplates, andfor the determination of antibodies to ssDNA,ssDNA coated plates were used. The two assayswere performed according to the methodsdescribed by Rubin et al.'2 Briefly, wells werewashed twice with PBS-Tween and 0-2 mlserum, diluted 1:200 in PBS-Tween containing5 mg/ml bovine gammaglobulin (BGG, Cohn'sfraction II, Sigma), 1 mg/ml BSA, and 1 mg/mlgelatin, were added. After two hours of incuba-tion with agitation, the contents were removedby aspiration and the wells were washed threetimes with PBS-Tween. Peroxidase conjugatedrabbit antihuman IgG, or IgM (Dakopatts,Copenhagen, Denmark) was diluted 1:400 inPBS-Tween containing 5 mg/ml BSA and1 mg/ml BGG, and 0-2 ml was added to eachwell. After agitation at room temperature fortwo hours the wells were rinsed once with PBS-Tween and twice with PBS. A 0-2 ml aliquot

of substrate solution, containing 1 mg/ml2,2-azino-di-(3-ethylbenzthiazoline-,B-sulphonicacid) (Sigma) and 0 005% H202 in Mcllvaine'sbuffer, pH 4-6, was added. The absorbance wasdetermined after 30 minutes using a Titertekmultiscan spectrophotometer at 414 nm. Foreach assay a standard curve was establishedusing a purified known amount of antibodies todsDNA and ssDNA as a standard antibody.

CRITHIDIA LUCILIAE IMMUNOFLUORESCENCETESTA standard indirect immunofluorescence tech-nique was employed,'9 using slides preparedwith Crithidia luciliae (Kainos, Tokyo, Japan).FITC conjugated antihuman IgG (Dakopatts)was used for staining. Serum specimens werescreened with an initial serum dilution of 1:10in twofold dilutions.

RENAL HISTOLOGYThe specimens for light microscopy were fixedin buffered formalin solution and embedded inparaffin. The following stains were used:haematoxylin-eosin, periodic acid-Schiff, andMasson's trichrome. World Health Organisation(WHO) criteria were used for the light micro-scopy classification of the major forms of lupusnephritis20 as class I (minor abnormality), classII (mesangial alteration), class III (focalproliferative), class IV (diffuse proliferative),and class V (membranous). Histologicalfeatures were scored according to renal activityand chronicity indices.2' Individual pathologicalchanges were scored on a semiquantitative basis(0-3+). The activity index was defined as thesum of individual scores of the following items:glomerular cell proliferation, leucocyte exu-dation, karyorrhexis/fibrinoid necrosis (x 2),cellular crescents (x 2), hyaline deposits, andinterstitial inflammation. The chronicity indexconsisted of the sum of individual scores of thefollowing items: glomerular sclerosis, fibrouscrescents, tubular atrophy, and interstitialfibrosis.

For electron microscopy the specimens werefixed in 2-5% glutaraldehyde, postfixed in 2%osmium tetroxide, dehydrated in gradedalcohol, and embedded in epoxy resin. Ultra-thin sections were stained with uranyl acetate

2-0

c 0-5

a,II0

0c

1-

0

01 1 5 10 25 50IgG anti-dsDNA (,Lg/ml)

Figure 1 Standard curve established in an assayfor IgGanti-dsDNA by enzyme linked immunosorbent assay(ELISA).

15



j.com/

Ann R

heum D


nloaded from

http://ard.bmj.com/

Okamura, Kanayama, Amastu, et al

Tabk I Incidence of abnormal laboratory data according to histological classiftcation in patients zwith lupus nephritis

Laboratory parameter Class*

I II III IV V Total(n=S) (n= 12) (n=S) (n= iS) (n=3) (n=40)

Haematuria/leucocyturia 0 3 2 7 1 13Proteinuria 0 4 3 12 3 22Nephrotic syndrome 0 0 0 8 2 10Renal insufficiency 0 0 0 3 0 3IgG antibodies to dsDNA positive by CLIFTt 1 5 3 14 1 24IgG antibodies to dsDNA positive by ELISAt 4 12 5 15 3 39IgM antibodies to dsDNA positive by ELISAt 3 10 5 14 3 35IgG antibodies to ssDNA positive by ELISAt 4 10 5 14 3 36IgM antibodies to ssDNA positive by ELISAt 5 7 3 10 3 28

'According to WHO morphological classification, class I=minor abnormality, class Il=mesangial alteration, class III=focal pro-liferative nephritis, class IV=diffuse proliferative nephritis, and class V=membranous nephritis.tCL-IFT=Crithidia luciliae immunofluorescence test; ELISA=enzyme linked immunosorbent assay.

and lead citrate. Based on the examination ofelectron micrographs of multiple areas, theamount of overall electron dense deposit wassemiquantitated on a scale of 0 to 3+ bymodification of the method of Grishman et al.22

NORMAL CONTROLSSerum samples from 40 healthy laboratory staffwere collected and assayed in the same way asthe patients' serum samples.

STATISTICSThe statistical significance of the results wasevaluated by an analysis of variance (ANOVA)and if significant differences were found by theanalysis p values were determined by Duncan'smultiple range comparison test. An unpairedStudent's t test was also used where appropriate.To determine correlation coefficient values thenon-parametric rank order correlation methodofSpearman was used. All values were expressedas the mean (SD).

IgM anti-dsDNA

S

0

0

0

0

000 0

0 0 * .

I 11 III IV V

I

* 0

8 0 8 I

0

..._................. . . . . . . . . . . . . . . . . . ..*. X . X . . . v . v . v . - . . . .... ...............

............. *****

. . . . . . . . . . . . . . .*************.......***************...***************.*.........**....*...*. . *. * . . . . . . .

. **.. . . . . . . . . . . *.**.*. ......*****.****.*..

*. . . . . . . .* * * * * * * * * ******.**........ . . . . . . . . . . . **... . . .

. . . . . . . . . . . . . . . . . ......*..**.*. . . . . . . . . . . . . . . . . . ..... ....

I 11 IIIIV V

WHO class

Figure 2 Correlations ofIgG and IgM anti-dsDNA with histological classes oflupusnephritis. S.ignificant differences were noted between class I and class IV (p<O0Ol), class IIand class IV (p<OOOI), and class III and class IV (p<OOl)forIgG anti-dsDNA.ForIgM anti-dsDNA no difference was shown between histological classes. (0) indicatespatients with clinically active nephritis. (0) indicates patients without clinically activenephritis. Dotted zones mean normal rangesfor IgG and IgM to anti-dsDNA.

ResultsFigure 1 shows the linearity of a typicalstandard graph with the ELISA IgG antibody todsDNA. The IgM antibodies to dsDNA, IgGantibodies to ssDNA, and IgM antibodies tossDNA standard curves were equally satis-factory. An upper limit of normal was setarbitrarily at two standard deviations above themean of the values for 40 healthy controls.

Table I lists the laboratory features at thetime of the collection of serum samples inwhich the renal histology was classified accord-ing to the WHO classification. Of the 40untreated patients 22 (55%) had clinicalsymptoms of active nephritis, whereas 38patients (95%) presented histological abnor-malities (three patients with class I nephritisshowed abnormalities by electron microscopy).Although most of the patients in classes I and IIshowed little or no haematuria or proteinuria, orboth, four patients showed substantial pro-teinuria. Three patients who had class IV lupusnephritis with diffuse proliferative lesions andprominent electron dense deposits showed nohaematuria or proteinuria.By the ELISA, 39 (98%) patients were

positive for IgG antibodies to dsDNA, 35 (88%)were positive for IgM antibodies to dsDNA, 36(90%) were positive for IgG antibodies tossDNA, and 28 (70%/o) were positive for IgMantibodies to ssDNA. Although the levels ofIgG antibodies to dsDNA were highly correlatedwith the titres of the Crithidia luciliae immuno-fluorescence test (r,=0'960; p<0-001), thesensitivity is much better for the ELISA(97 5%) than for the immunofluorescence test(67-5%).

Figure 2 gives the concentrations of IgG andIgM antibodies to dsDNA with their histologicalclasses. The levels of antibodies to dsDNA inclass IV were higher than those in class I

IgG anti-dsDNA50 r

25 F

10 1

5

1

0-5 1

0-1 1

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

16

0

....................................................................................................................................................................................................................................



j.com/

Ann R

heum D


nloaded from

http://ard.bmj.com/


Table 2 Comparison of antibodies to DNA between renal histological classes. Values given as mean (SD) (pg/ml)

Renal No of IgG antibodies IgM antibodies IgM antibodies IgG antibodies IgM antibodies IgM antibodieshistology patients to dsDNA to dsDNA to dsDNA to ssDNA to ssDNA to ssDNA

IgG antibodies IgG antibodiesto dsDNA to ssDNA

Class I 5 0-19 (0-13) 0-99 (0 86) 7-12 (7 88) 2-44 (1 53) 0 74 (0 47) 0-81 (0 90)Class II 12 0 52 (0 50) 1-56 (1-17) 4-68 (4 68) 2-70 (1-85) 0 79 (0-83) 1-26 (3-12)Class III 5 1-33 (1-16) 1-55 (068) 1-94 (1-40) 3-80 (3-01) 0-54 (0 32) 0-25 (009)Class IV 15 11-59 (9.9)*t 2-03 (1-50) 0-26 (0 29) 11-85 (18-69)§ 0-73 (0-65) 0-14 (1-40)Class V 3 2-09 (2-15) 1-46 (0 72) 0-98 (0 69) 2-16 (0 85) 1-44 (0-84) 0-62 (1-28)Classes I and II 17 0-42 (0 46) 1-39 (1-12) 5 40 (5-91) 2-62 (1-77) 0-78 (0-76) 1-12 (2 68)Classes III and IV 20 9-02 (9 94)t 1-88 (135) 0-71 (1-03)f 9-84 (16-62) 0-69 (0-59) 0-17 (0-16)

*p<0-01 v class I, class III, by Duncan's test.tp<0 001 v class II, by Duncan's test.fp<0-01 v classes I and II, by Student's t test.§p<005 v class II, by Duncan's test.

IgG anti-ssDNA IgM anti-ssDNA

0

I§

0 00

0

* 8*

0

0

0

0

.S

0

..........***..***..*...................

.. . . . . .....

.. . . . . . . . . . . .*.. . . . . . . . . . . . . . . . . . .............*. . . . . . . . . . . . . . . . * * ***. . . . . . . . . . .*... . . . . . . . . . . . . . . . .**

*.........*. . . . ... . . . . .

. . . .**.. . . . . . . . . . . . . . . **... ..... . . . . . . . . . . . . . . . . ...********.. . . . . . . . . . . . . . . * * ********.. . . . . . . . . . . . . . . . . . .... *.....*** **.

. . .

......***.* ......

.......... **** .***

...***************.

I 11 III IV V

0 0

0

8:0

S.

I0

0

. . . . . . . . . . . . . . . . . .

.......... .........

..@. . . . . . . . . . . . . . . . . . ......... .....

. . . . . . . . . . . . . ........**. . . . . . . . . . . . . . . .***.. . . . . . . . . . . . . . . . . . ....

.. *... *..

. . . .

*. . . . . * . . . . . ....... ............,

. . . . . . . . . , . . . .

................***

..........*..*..*..-.**.*................I 11 Ill IV V

WHO class

Figure 3 Correlations ofIgG andIgM anti-ssDNA with histological classes oflupusnephritis. A significant difference was noted between class II and class IV (p<005)forIgGanti-ssDNA . No difference was shown between histological classes forIgM anti-ssDNA.(0) indicates patients with clinically active nephritis. (0) indicates patients without clinicallyactive nephritis. Dotted zones mean normal rangesfor IgG and IgM to anti-ssDNA.

Table 3 Rank order correlations between renal histology and antibodies to DNA

Activity Chronicity 7otal Electronscore score pathological dense

score deposits

IgG antibodies to dsDNA 0-741** 0407' ' 0 710'' 0 810**IgM antibodies to dsDNA 0-268 0-113 0-249 0 308IgG antibodies to ssDNA 0-434** 0 368" 0-448:'' 0-446'IgM antibodies to ssDNA 0-168 -0079 0073 0212

*p<0.05; *p<OOl.

(p<OOl), class II (p<OOOl), and class III(p<OOl), though there was a dispersion oflevels among the classes. The patients withclinical symptoms of nephritis had a highermean (SD) level of IgG antibodies to dsDNA(7-80 (9-98) tig/ml) than in those withoutclinical symptoms of nephritis (1-12 (2 47)[tg/ml) (p<001). No difference was foundamong patients with different histologicalclasses in the levels of IgM antibodies todsDNA. Figure 3 gives the correlations of IgGand IgM antibodies to ssDNA with the histo-logical classes. The levels of IgG antibodies tossDNA in patients with class IV nephritis werehigher than those in patients with class IInephritis (p<005). The patients with clinicalsymptoms of nephritis had higher levels of IgGantibodies to ssDNA than those without clinicalsymptoms of nephritis (8179 (16-43) ptg/ml v3 04 (2 69) ig/ml; p<005). No difference wasfound among patients with different histologicalclasses in the levels of IgM antibodies tossDNA. Table 2 gives all the data for comparisonof antibodies to DNA among renal histologicalclasses. The levels of IgG antibodies to dsDNAwere significantly higher in patients withmoderate to severe nephritis (class III and classIV) than those with mild nephritis (class I andclass II), whereas there was no differencebetween the two groups in the levels of IgMantibodies to dsDNA. The geometric meanratio of IgM antibodies to dsDNA/IgG anti-bodies to dsDNA was significantly lower inpatients with moderate to severe nephritis thanin those with mild nephritis (p<001).

Table 3 gives results of the Spearman rankorder correlation of antibodies to dsDNA andssDNA with the renal activity score, chronicityscore, and total pathological score by lightmicroscopy and with the overall electron densedeposits. Correlation for IgG antibodies todsDNA with each of the histological scores wasgood, with that for electron dense depositsbeing the best and activity score being thesecond best (rs=0810; p<001; r,=0741;p<001 respectively) (figs 4 and 5). Correlationsfor IgG antibodies to ssDNA with each of theparameters were weak. IgM antibodies todsDNA and IgM antibodies to ssDNA were notcorrelated with those parameters. There was aweak correlation between the level of IgGantibodies to dsDNA and IgG antibodies tossDNA (r=0-394; p<001). There were three

70

50

25

10

5000

o.5

O-t .

--------------------

17

1



j.com/

Ann R

heum D


nloaded from

http://ard.bmj.com/


40 r

30

r, = 0810p < 0-01

20 F

.

10 [

nh -

0

.

0S

I

mJ6

0

0

.

1 2Electron dense dep(

Figure 4 Close correlation ofIgG anti-dsLhistological activity score. Correlation coeffiSpearman rank order analysis is indicated bl

40 -

r[ = 0741 0

p < 001

30

20

10 .

0 5 10

Activity score

FigureS Close correlation ofIgG anti-ds]amount ofoverall electron dense deposits. Ccoefficient by Spearman rank order analysisr, value.

patients with class IV nephritis v

low IgG antibodies to dsDNA andIgG antibodies to ssDNA.

DiscussionWe measured the levels of IgGdsDNA, IgM antibodies to dsD1}bodies to ssDNA, and IgM;ssDNA by ELISA in the seruruntreated patients with SLE who-

logical severities were assessed. Measurementsfor antibodies to DNA have been consideredas useful in the management of patients withSLE because of their diagnostic specificityand their probable correlation with diseaseactivity. It has been reported that the levelsof antibodies to dsDNA correlate with renaldisease in patients with SLE,' -71' butnot all studies are in agreement with thesereports.8 9 These results should be interpretedwith some caution. First, the antibodies todsDNA were often measured with the use ofconventional dsDNA antigen which is thoughtto be nearly always contaminated with ssDNA.23Chromatographically purified and S1 nucleasetreated dsDNA preparations are consideredoperationally native." 1 This view wassupported by the results in this study that thelevels of IgG antibodies to dsDNA were highlycorrelated with the titres of Crithidia luciliaeimmunofluorescence test (rs=096; p<0-00l).As it has been shown that the antigenicity ofdT-dA is different from that of dsDNA,24 the

3 purified dsDNA preparation used in this studyis thought to be more suitable to detect anti-

osit bodies to dsDNA. Second, interpretation of the)NA with renal data on antibodies to dsDNA may be misleading,cient by without renal biopsy data. In this study they rS value. incidence of clinically active nephritis was 55%,

whereas renal histological abnormalities werefound in 95% of patients. There were threepatients with class IV lupus nephritis whoshowed no clinical symptoms of renal disease.Discrepancies between clinical data and renalhistological features have been reported byother workers.2127 Thirdly, as it is likelythat adequate immunosuppressive treatmentinduces rapid changes in serum levels of anti-bodies to DNA and changes in renal histologicalfeatures, data obtained from the untreatedpatients with SLE are particularly important.

In the present study increased levels of IgGantibodies to dsDNA were found in 98% ofuntreated patients with SLE, and a similar highincidence has been reported by other workersusing ELISAs." We found that the levels ofIgG antibodies to dsDNA in patients with classIV nephritis were significantly higher thanthose of patients with classes I, II, and III,nephritis. Although other workers have reportedthat antibodies to dsDNA, detected by the

15 20 conventional dsDNA preparation, in patientswith diffuse proliferative nephritis were higher

DNA with than those in patients with focal and segmentalorrelation nephritis, the difference was not clear and notis indicated by statistically significant.28 Cloughand Valenzuella

showed that patients with diffuse proliferativenephritis had a higher ratio of IgM to IgGantibodies to DNA than those with focal neph-

vith relatively ritis, even though the total amount of antibodyhigh levels of to DNA was the same in the two groups.29

Others have reported that the ratio of IgG toIgM antibodies to dsDNA did not significantlydiffer in patients with active nephritis and inpatients without clinical evidence of renal

antibodies to disease.30 31 We tried to compare the ratio of4A, IgG anti- IgM to IgG antibodies to dsDNA betweenantibodies to patients with histologically mild nephritis (classn samples of I and class II) and those with moderate to severese renal histo- nephritis (class III and class IV). The ratio of

E1-

z0

VC

0,

ECD

4

:._zaU,'0

0,

18



j.com/

Ann R

heum D


nloaded from

http://ard.bmj.com/


IgM to IgG antibodies to dsDNA was signi-ficantly lower in patients with moderate tosevere nephritis than in those with mild nephritis.These results may be explained by the fact thatthe levels of IgG antibodies to dsDNA weresignificantly higher in patients with moderate tosevere nephritis than in those with mild neph-ritis, whereas the levels of IgM antibodies todsDNA were the same in the two groups. Ourresults are consistent with the results ofPennebaker et al who reported that patientswith predominantly IgG antibodies to dsDNAshowed more severe histological changes.3'These findings are also compatible with thereport that immunoglobulin deposition in lupusnephritis, detected by immunofluorescence,and antibodies to DNA, eluted from the kidneysof patients with severe lupus nephritis, wereprimarily IgG.'4Owing to the marked variability of lupus

nephritis, we used a semiquantitative scoringsystem and assessed the correlation with anti-bodies to dsDNA. We found a significantcorrelation between the levels of IgG antibodiesto dsDNA and each of the histological para-meters. A previous study, using a conventionaldsDNA preparation, showed no correlationbetween antibodies to dsDNA and renal histo-logical activity.8 As the levels of IgG antibodiesto dsDNA closely reflect the electron densedeposit as well as the activity score by lightmicroscopy, they may provide useful informationto assess the severity of lupus nephritis non-invasively.Our results support the general view that

antibodies to dsDNA play a crucial part in thepathogenesis of lupus nephritis. 4 We identifiedthree patients who showed no clinical signs ofrenal disease and were proved to have class IVlesions. The presence of such patients withclinically occult diffuse proliferative lupusnephritis has also been reported by otherworkers.26 27 As these patients had high levelsof IgG antibodies to dsDNA, we suggest thatthis assay may be helpful in deciding whether arenal biopsy is necessary or not.The prevalence of antibodies to ssDNA in

patients with SLE is high and these antibodieshave been thought to be closely correlated withdisease activity,'4 data supporting the presenceof ssDNA antibodies to ssDNA immunecomplexes in the affected glomeruli have beenreported.3 No previous studies have attemptedto correlate antibodies to ssDNA with renalhistological severity in patients with lupusnephritis. Although we found a weak correlationbetween IgG antibodies to ssDNA and renalhistological severity, the correlation was lessclear than that of IgG antibodies to dsDNA.These results are consistent with the generalview that ssDNA antibodies to ssDNA immunecomplexes are not important in the pathogenesisof lupus nephritis.32 As we found threepatients with class IV lesions who had highlevels of IgG antibodies to ssDNA and relativelylow levels ofIgG antibodies to dsDNA, however,antibodies to ssDNA may play a part in thepathogenesis of nephritis in some patients withlupus.

It has been previously shown that patients

whose antibodies precipitate and bind to DNAhave a greater incidence of nephritis,33 and lowavidity antibodies34 or complement fixing anti-bodies35 are claimed to be more nephritogenic.Our study, however, suggests that quantitativedifferences of antibodies to DNA play a majorpart in determining the histological features oflupus nephritis. Therefore whether qualitativedifferences in antibodies to DNA criticallyaffects the pathological features of lupus neph-ritis needs further study.

1 Pincus T, Schur P H, Rose J A, et al. Measurement of DNA-binding activity in systemic lupus erythematosus. N EnglJMed 1%9; 281: 701-5.

2 Hughes G R V, Cohen S A, Christian C L. Anti-DNA activityin systemic lupus erythematosus. Ann Rheum Dis 1971; 30:259-64.

3 Andres G A, Accinni L, Beiser S M, et al. Localization offluorescein labeled anti-nucleoside antibodies in glomeruliof patients with systemic lupus erythematosus. J Clin Invest1971; 49: 2106-18.

4 Koffler D, Agnello V, Kunkel H G. Polynucleotide immunecomplexes in serum and glomeruli of patients with systemiclupus erythematosus. AmJt Pathol 1974; 74: 109-24.

5 Schur P H, Sandson J. Immunologic factors and clinicalactivity in lupus erythematosus. N EnglJ7 Med 1%8; 278:533-8.

6 Ballou S P, Kunshner I. Immunochemical characteristics ofantibodies to DNA in patients with active systemic lupuserythematosus. Clin Exp Immunol 1979; 37: 58-67.

7 Chubick A, Sontheimer R D, Gilliam J N, et al. An appraisalof tests for native DNA antibodies in connective tissuedisease: clinical usefulness of Crithidia luciliae assay. AnnIntern Med 1978; 89: 186-92.

8 Hill G S, Hinglais N, Tron P, et al. Systemic lupuserythematosus. Morphologic correlation with immunologicand clinical data at the time of biopsy. AmJf Med 1978; 64:61-79.

9 Appel A E, Sablay L B, Golden R A, et al. The effect ofnormalization of serum complement and anti-DNA anti-body on the course of lupus nephritis. AmJ Med 1978; 64:274-83.

10 Steinman C R, Grishman E, Spiera H, et al. Binding ofsynthetic double stranded DNA by serum from patientswith systemic lupus erythematosus. Correlation with renalhistology. Am J Med 1977; 62: 319-23.

11 Miller T E, Lahita R G, Zarro V J, et al. Clinical significanceof anti-double stranded DNA antibodies detected by a solidphase enzyme immunoassay. Arthritis Rheum 1981; 24:602-10.

12 Rubin R L, Joslin F G, Tan E M. An improved ELISA foranti-native DNA by elimination of interference by anti-histone antibodies. J Immunol Methods 1983; 63: 359-66.

13 Gripenberg M, Linder E, Kurki P, et al. A solid phaseenzyme-linked immunosorbent assay (ELISA) for thedemonstration of antibodies against denatured, singlestranded DNA in patient sera. Scand J Immunol 1978; 7:151-7.

14 Koffler D, Schur P H, Kunkel H G. Immunological studiesconcerning the nephritis of systemic lupus erythematosus.J Exp Med 1%7; 126: 607-23.

15 Tan E M, Cohen A S, Fries J F, et al. The 1982 revisedcriteriaforthe classificationofsystemic lupus erythematosus.Arthritis Rheum 1982; 25: 1271-7.

16 Winfield J B, Faiferman I, Koffler D. Avidity of anti-DNAantibodies in serum and IgG glomerular eluates frompatients with systemic lupus erythematosus. J7 Clin Invest1977; 59: 90-6.

17 Kloz J L, Minami R M, Teplitz R L. An enzyme-linkedimmunosorbent assay for antibodies to native and denaturedDNA. J Immunol Methods 1979; 29: 155-65.

18 Albert B, Herrick G. DNA-cellulose chromatography. In:Grossman L, Moldave K, eds. Nucleic acids part D,Methods in Enzymology. Vol. XXI. New York: AcademicPress, 1971, 198-217.

19 Sontheimer R D, Gilliam J N. An immunofluorescence assayfor double-stranded DNA antibodies using the Crithidialuciliae kinetoplast as a double-stranded DNA substrate.J Lab Clin Med 1978; 91: 550-9.

20 Churg J. Renal disease. Classification and atlas of glomerulardisease. Tokyo: Igaku-Shoin, 1982, 12749.

21 Austin H A, Muenz L R, Joyce K M, et al. Diffuseproliferative lupus nephritis: identification of specificpathologic features affecting renal outcome. Kidney Int1984; 25: 689-95.

22 Grishman E, Porush J G, Churg J, et al. Renal biopsies inlupus nephritis: correlation of electron microscopicfindings with clinical course. Nephron 1973; 10: 25-36.

23 Steinman C R, Deesomchok K U, Spiera H, et al. Detectionof anti-DNA antibody using synthetic antigens. J7 ClinInvest 1976; 57: 1330-41.

24 Gilliam A C, Lang D, Lospalluto J J, et al. Antibodies todouble-stranded DNA: purification and characterization ofbinding specificities. J Immunol 1980; 125: 874-85.

25 Rothdield N J, McLuskey R J, Baldwin D S, et al. Renaldisease in systemic lupus erythematosus. N Engl 7 Med1%3; 269: 537-44.

19



j.com/

Ann R

heum D


nloaded from

http://ard.bmj.com/


26 Mahajan S K, Ordonez N G, Feitelson P J, et al. Lupusnephropathy without clinical renal involvement. Medicine(Baltimore) 1977; 56: 493-501.

27 Eiser A R, Katz S M, Schwartz C, et al. Clinical occultdiffuse proliferative lupus nephritis. Arch Intern Med 1979;139: 1022-5.

28 Tron F, Bach J F. Relationships between antibodies to nativeDNA and glomerulonephritis in systemic lupus erythema-tosus. Clin Exp Immunol 1977; 28: 426-32.

29 Clough J D, Valenzuella R. Relationship of renal histo-pathology in SLE nephritis to immunoglobulin class ofanti-DNA. Am Med 1980; 68: 80-5.

30 Feldman M D, Huston D P, Karsh J, et al. Correlation ofserum IgG, IgM and anti-native DNA antibodies with renalclinical indexes of activity in systemic lupus erythematosus.Rheumatol 1982; 9: 52-8.

31 Pennebaker J B, Gillian J N, Ziff M, et al. Immunoglobulinclasses of DNA binding activity in serum and skin in

systemic lupus erythematosus. Clin Invest 1977; 60:1331-8.

32 Koffler D, Agnello V, Winchester R V, et al. The occurrenceof single-stranded DNA in the serum of patients withsystemic lupus erythematosus and other disease. J ClinInvest 1973; 52: 198-204.

33 Edmonds J P, Johnson G D, Ansell B M, et al. The value oftests for antibodies to DNA in monitoring the clinicalcourse of SLE: a long-term study using the Farr test andthe DNA counter immunoelectrophoretic method. ClinExp Immunol 1975; 24: 9-15.

34 Leon S A, Green A, Ehrlich G E, et al. Avidity of antibodiesin SLE. Relations to severity of renal involvement. ArthritisRheum 1977; 20: 23-9.

35 Beaulieu A, Quismorio F P Jr. Kitridou R C, et al.Complement fixing antibodies to dsDNA in systemic lupuserythematosus: a study using the immunofluorescentCrithidia luciliae method. Rheumatol 1979; 6: 389-%.

20



j.com/

Ann R

heum D


nloaded from

http://ard.bmj.com/

koji · mikio okamura, yoshiharu kanayama, koji tadanao takeda, takatoshi inoue abstract the...

Documents