Scand. J. Immunoi 40. 383-388, 1994

Non-Coordinated Biosynthesis of Early ComplementComponents in a Deficiency of ComplementProteins Clr and Cls

A. CHEVAILLER, C. DROUET*t, D. PONARD*, C. ALIBEU*, S. SURANITI| ' .F. CARRERE, G. RENIER, D. HUREZ & M. G. COLOMB*tLaboratoire d'Immunopathologie, CHU, 49033 Angers, Cedex. France*Laboratoire d'Immunologie. CHRU Grenoble, Hopital Sud, 38130 Echirolles. FranceiUnite INSERM 238, DBMS-ICH. CENG 85X. 38041 Grenoble Cedex, FranceXService de Medecine B. CHU 49033 Angers Cedex, France

Chevailler A, Drouet C, Ponard D, Alibeu C, Suraniti S, Carrere F, Renier G, Hurez D, Colomb MG.Non-Coordinated Biosynthesis of Early Complement Components in a Deficiency of ComplementProteins Clr and Cls. Scand J Immuno] ]994;40;383-8

We report on a 60-year-old woman with systemic lupus erythematosus and a total (95%) Clr and apartial (36%) Cls deficiency. The patient complained about cutaneous lesions on forearms and legswithout other systemic involvement. Elevated anti-nuclear, anti-native DNA and anti-SSA antibodieswere present. The finding of persistently depressed levels of haemolytic complement activity (CH50) onboth serum and plasma, associated with normal levels of C3, C4 and C2 components, and normalalternative pathway haemolytic activity showed a deficiency of an early component of the classicalpathway. IndeedClrcomponent was below the limits of detection whereas Cls component was lowered(36%). The depressed CH50 was only corrected by purified Clr. Biosynthesis of Clr and Cls bypatient's monocytes was spontaneously norma] but not up-regulated by interferon-7 for Clr alone,whereas the biosynthesis of Cls, but also of interleukin-6, was increased, indicating a specificdisregulation of Clr. The deficiency was associated with a lupus syndrome and a fatal assumed septicshock. This is in agreement with other reported cases.

A. Chevailler, Laboratoire d'Immunopathologie, CHU, 49033 Angers Cedex 01, France

INTRODUCTION

Complement plays an important role in non-specific hostdefence against a wide variety of micro-organisms by gen-erating a normal inflammatory response leading to theirneutralization, phagocytosis and killing. Deficiencies foreach component except factor B and C4 binding protein,have been described previously [1,2] and the overall incidenceof these disorders have been estimated between 0.03 and

This work was in pan presenied as a poster al the 4th European Meeting onComplement in Human Diseases. May 1992, Leiden. The Netherlands.

Abbreviations: ANA, antinuclear antibodies; Cl-inh, Cl-inhibitor; IFN-o,interferon-o; lFN-^, inlerferon-/^; IFN-u2, interferon-a2; IFN-7, inter-feron--/; IC, immune complexes; lL-6, inlerleukin-6; RID, radial immuno-difTusion; RIA, radioimmunoassay; RF, rheumatoid factor; SLE, systemiclupus erythematosus.

' Present address: Service de Medecine, CHR, 27, rue Seigneur, 49043Saumur Cedex, France.

0.09% [3]. Most of these deficiencies are inherited althoughsome acquired deficiencies exist. Hereditary complementdeficiencies result either from a markedly diminished orabsent synthesis ofa component or result from the synthesisof a functionally inactive molecule [1. 2]. Homozygousdeficiencies are often associated with diseases as a conse-quence ofthe failure of activation ofthe complement system,namely immune complexes, associated diseases or pyogenicinfections [1-3].

Among the whole complement deficiency group, Clr/Clsdeficiencies, recently and extensively reviewed [4-6], are veryrare disorders. Twelve individuals in eight families havebeen reported to have these disorders, without biosyntheticstudies. They have a combined deficiency with total Clrdeficiency usually associated with decreased Cls level. Thisfeature might be a consequence ofthe close genetic linkage ofClr and Cls genes on chromosome 12 [7, 8]. The lack of afunctional Cl molecule is associated with diseases reflecting

383

384 A. Chevailler et al.

the inability to process immune complexes (IC) in the absenceof a functional classical pathway, namely systemic lupuserythematosus (SLE), SLE-like disorders, vasculitis, glo-merulonephritis. In some cases, recurrent pyogenic infec-tions have been observed 14, 6].

We report on a 60-year-oid woman with a SLE-like syn-drome associated with total Clr and partial Cls deficiency. Astudy of the biosynthesis of Clr and Cls proteins reveals aspecific disregulation of Clr synthesis by the patient's mono-cytes under cytokine stimulation.

In August 1989, the woman complained of lupus-likecutaneous lesions on forearms and legs without any othersystemic involvement. There was a several year history ofelevated blood pressure, There was no previous history ofrash, joint pain, cutaneous lesions or abnormal urinarysediment. The sedimentation rate was 132 mm/hour. Adiagnosis of lupus-like syndrome was made on the presenceof antinuclear antibodies (ANA), positive at 1:1000 with anhomogeneous pattern of fluorescence, of anti-dsDNA asshown by positive Farr radioimmunoassay and Crithidialuciliae test, of anti-SSA antibody and of a positive skinbiopsy only with the anti-/i antiserum. Tests for immunecomplexes, rheumatoid factor (RF) and cryoglobulinaemiawere all negative. The total haemolytic complement activitywas persistently undetectable despite normal C3 and C4serum levels, and prompted further investigations. Thecreatinine level and the urine were normal. A temporalarterial biopsy showed a vasculitis without giant cells. Corti-costeroid treatment was prescribed (80mg daily) and thepatient improved,

During the follow-up she was seen twice, when the corti-costeroid treatment was tapered for new lupus cutaneouslesions without other systemic involvement again. She wasalso hospitalized once for a varicose ulcer.

She died suddenly in February 1992 from an assumedseptic shock with intracerebral haemorrhage as assessed bybloody cerebrospinal fluid at the lumbar puncture. Nomicroorganism could be isolated from blood, urine orcerebrospinal fluid.

MATERIALS AND METHODS

Complement component levels. Serum and plasma samples fromIhe patient and from her children were stored at - 8 0 X within 6 h ofcollection and were thawed once before assay.

Complement component levels were determined either byimmunonepheiometry (BNA. Behring Rueil-Malmaison, France),radial immunodiffusion (RID) [9] or radioimmunoassay (RIA) [10]using anti-complement component monospecific antisera (Clq, C3,C4, Cl-inhibitor [Cl-inh], factor B (Behring); Cls, C2, C5. factor H(Atlantic Antibodies. Scarborough, ME, USA); factor I (ICNBiomedicais, Orsay, France); Clr (a rabbit polyclonal antibodyraised in ihe laboratory).

Haemolyiic activities. The functional activity of the classicalpathway in serum was measured either by the CH50 assay [II]testing the ability of serum to lyse antibody-coated sheep erythro-

cytes or by a kinetic haemolytic assay [12]. Purified proenzyme Clrwas a generous gift from Dr G. Arlaud (Grenoble, France).

The functional activity of the alternative pathway was measuredby the alternative pathway-complement haemolytic assay (AP-CH50) [13] testing the ability of serum to lyse rabbit erythrocytes.

In vitro Clr and Cls secretion by cultured monocytes. Ficoll-Paquewas purchased from Pharmacia, St Quentin-Yvelines, France.RPMI-I640 medium and fetal calf serum were from Gibco, Cergy-Pontoise, France. Human interferon-ft2 (IFN-o2) (l()''units/mg)was from Shering-Plough. UNICET. Dardilly, France, and affinitypurified human interferon-7 (IFN-7) (!.2 x 10*units/mg) was fromCellular Products Inc., NJ, USA. All other chemicals and reagentswere of analytical grade.

Blood samples from the patient were collected in the presence ofheparin (Vacutainer, Becton-Dickinson, Pont-de-Claix, France).Monolayers of adherent monocytes were established from Ficoll-separated peripheral blood mononuclear cells and maintained aspreviously described [10]. Secreted Clr and Cls proteins werequantified in culture supernatants by a radioimmunoassay as pre-viously described [10] using '"i-radiolabelled purified Clr and Clsas references, and the relevant antibody at half-saturating concen-tration. Secretion of proteins was related to DNA contents of lysatesfrom culture wells estimated according to West et at. [14].

Other immunologic studies. HLA-A, -B and -DR typing wasperformed by the microlymphocytotoxic assay [15]. InterIeukin-6(IL-6) levels in the supernatant ofthe patient's cultured monocyteswereassessed by ELISA (Medgenix. Fleurus, Belgium) following theprocedure indicated by the manufacturer. They were related to DNAcontents of lysates as for Clr and Cls secretion. RF and IC weredetermined by immunonepheiometry using IgG-coated and Clq-coated latex beads, respectively (BNA, Behring).

RESULTS

Standard measurement of complement

A markedly depressed level of CH50 activity in serum orplasma, normal serum levels of C3 and C2, increased C4 andCl-inh serum levels (Table I) and normal AP-CH50 activityindicated an early component deficiency in the classicalpathway. A null CH50 value was consistently observedthroughout the follow-up, whatever the clinical status was.

Identification ofthe ClrjCls deficiency

Serum Clr level was below the detection limit by immuno-nepheiometry or RID (Table 1 and Fig. 1) and markedlydepressed as assessed by RIA (Table 1). Cls serum level waslowered (36% of normal values) whereas Clq level wasnormal (Table 1). The CH50 activity was partially restoredby adding purified proenzyme Clr (Table 2).

Familial study

The patient had two children; both were healthy. Sera of bothchildren had no complement abnormalities as shown in Figs Iand 2. Their monocytes have not been available for biosyn-thetic studies.

Clr/Cls Deficiency with Abnormal Biosynthesis Regulation 385

Table 1. Levels of complement components and functional activity in patient's serum and plasma

Immunonepheiometryserum

RIDserum

RIAserum

2a

21,6

FunctionalActivity

serum < 5%plasma < 5%

100%

NormalRange

72- 123%72-123%40-48/ig/l72- 124%40-48/ig/l

206-348mg/l127-331 mg/172- 124%

451-861 mg/1I16-128mg/l98-177 mg/185- 105%85- 105%80-120%80-120%80-120%

ClqClr

Cls

97%

Cl-inhC4C2C3BC5HICH 50

601 mg/1646 mg/I

1130 mg/1279 mg/1207 mg/1

125%< 12%

37%

100%

150%160%

AP50

RID: radial immunodiffusion; RIA: radio-immunoassay; CH 50: functional activity ofthe classical pathway; AP 50: functional activity ofthe alternative pathway.

Clr/Cls secretion by patient's monocytes

Secretion of Clr and Cls by the patietit's resting tnonocytes(Fig. 3) were low but considered normal, when comparedwith values obtained in the case of healthy donors previouslyreported [16], with a I to 1 stoechiometry between the twoproteins. In the same culture, Cls secretion was stitnulated byIFN-7 as previously shown for nonnal individuals [16],whereas Clr secretion was not. A similar stimulation wasobserved by IFN-fi2 (data not shown). This defect ofthe Clrsecretion appeared selective, as, for instance, IL-6 secretionwas stimulated by IFN-7 (Fig. 3), as expected from theprevious report of Navarro et al. [17].

Table 2. Haemolytic functional assay. This was performed withdeficient patient's serum before and after reconstitution usingproenzyme Clr to 50mg/l. Absorbance at 660 nm of an antibodysensitized sheep erythrocyte (EA) suspension (3 ml at 10** cells/ml)is measured after addition of 25/il of serum. The time required tolyse 50% of the EA suspension is determined

Serum Time (s)

NHS*PatientPatient + Clr**

144522240

* NHS: normiil human serum; ** purified proenzyme Clr.

DISCUSSION

In this report we describe a 60-year-old woman with a lupus-like syndrome, a total (95%) Clr and a partial (36%) Clsdeficiency.

Up to now only 12 individuals in eight families have beenreported and recently reviewed [4-6]. Most of these studieswere performed before the 80s and before Clr and Cls genelocation, in a close tail-to-tail arrangement on chromosome12, was described [7, 8]. All but two patients with isolated Clsdeficiency [18, 19]. suffered, as our patient, from a total Clrdeficiency associated with a partial Cls deficiency leading theauthors to postulate, at that time, a possible linkage betweenboth genes. In an extensively studied Puerto Rican family [20,21] with two clinically affected subjects, De Bracco et al. [22]concluded an autosomal recessive inheritance of this de-ficiency and Mittal et al. [23] demonstrated the absence ofgenetic linkage with the HLA system of this deficiency, incontrast with C2 deficiency. Similar conclusions were drawnfrom another Puerto Rican family with two affected and twohealthy Clr/Cls deficient subjects [24, 25],

From a biological point of view, the definitive proof of ourpatient's Clr deficiency was given by reconstitution experi-ment where only the addition of purified proenzyme Clr toour patient's serum has normalized the haemolytic activity, asreported previously [20, 26, 27]. We cannot exclude thatpurified Clr has already acquired the configuration changenecessary for Cl activation; this could be reflected by thepartial restoration of the haemolytic activity of serum afterreconstitution (Table II). This observation is in agreement

386 A, Chevailler et al.

Fig. 1. Clr determination by radial immunodiffusion. RID wasdone in an agarose gel buffered with 10mm EDTA, 150mm NaCland 25 mm Na2HPO4, pH 7,4. Individuals 1-2, II-l and II-2 wereas described in the familial study (see Fig. 2). a: Normal HumanSerum (NHS) = 200% (two doses of undiluted serum);b: NHS= 100%; c: NHS = 50%; d: 1-2 = not detectable; e: l-2 = notdetectable; f: II-1 =75%; g: 11-2 = 85%.

with previously reported data in a reconstitution assay usingpurified Cl [28].

In the absence of extensive fatnilial studies we cannotconclude that this deficiency is hereditary. The late onset ofthe patient's illness does not argue against this hypothesis assome complement deficiencies are well known to give clinicalsymptoms only in the second or third decade or even later [3].We exclude selective activation or consumption of comple-ment as a possible mechanism of the deficiency for at leastthree reasons. Firstly, we cannot explain an activation of theclassical pathway with an imbalanced Clr and Cls consump-tion. Secondly, as already stated by Pickering et ai [26, 29],increased serum C4 concentration, a substrate of Cls, arguesagainst a Clr/Cls deficiency owing to activation. And lastly,the same is true for the high Cl-inh levels observed that mightresult from a reduced utilization under conditions wherereduced serum Clr/Cls levels preclude normal Cl acti-vation. The persistently increased C4 and Cl-inh levels inthis patient's serum is consistent with previous observations[20, 24, 25, 27]; these findings are in agreement with theassumed relationships of Clr/Cls with C4 and Cl-inh invivo [30]. We may also bear in mind the possible efifects of the

A l

A l

B8

BI4

1927

DR6

nR6

o1929

Al B37

A2 BI5

C3

C4

ClrCIS

CH50

DRl-

1130

646

<12

360

mg/1mg/I

%

%

1955 1

AI B8 DR6

Al B37 DRl

C3 1370 mg/1

C4 330 mg/I

Clr 85 %

Cls ND

CH50 100 %

.6 1963

A2 B37

Al BM

C3

C4ClrCliCH50

DRlDR6

1210 mg/I

348 mg/I

75 %

ND

100 %

Fig. 2. Familial study. The HLA typing and the complement leveldetermination were carried out as described in the Materials andMethods section. For the nonnal values see Table I. ND: not done.

inflammatory syndrome that explain, in part, the high levelsofthe other components (Table 1).

We observed low levels of Clr and Cls secretion by thepatient's resting monocytes; these low values are in agreementwith other previous studies [16, 31], IFN-Q, -3 and -7 havebeen shown to enhance monocyte synthesis of several com-ponents including Cl subcomponents [16, 32]. In the presentstudy, we therefore determined the modulation of thepatient's monocyte synthesis of Clr and of Cls. Cls mono-cyte secretion was up-regulated by IFN-7 in the same order ofmagnitude as previously reported with monocytes fromhealthy donors [10]. Taking into account this last result andby comparing this with the normal cell secretion of IL-6, tbepatient's monocytes can be considered nonnal. In contrast,the absence of up-regulation of Clr secretion from thepatient's monocytes suggests a selective defect in Clr geneexpression. These results might illustrate a non-coordinatedexpression of C1 r and C1 s genes, and independent regulationof these tail-to-tail arranged genes [7, 33].

As stated by Atkinson [34], deficiency of an early comple-ment component of the classical pathway may be one of themajor mechanisms, among several other reasons, leading toinappropriate deposition of IC. Such deficient sera are unableto generate C3b via the classical pathway leading to increasedIC precipitation and loss of immune adherence. Incidentally,the previously reported cases of Clr/Cls deficiencies pre-sented with isolated glomerulonephritis, or SLE-like disorder[18, 19], or isolated infectious diseases [35, 36], or both, likeour patient [20-22, 24, 25, 27].

Clr/Cls Deficiency with Abnormal Biosynthesis Regulation 387

I1B0 a

I

time (days)

Fig. 3. Clr and Cls biosynthesis by cultured patient's monocytes. Patient's monocytes were grown as described in the Material andMethods section. They were or not (resting monocytes) activated by lOOOU/ml IFN-7. At sequential 48h intervals, the I ml supernatantvolume from each well was replaced by the same volume of culture medium supplemented or not with the monocyte activation factor.Each point on the figure corresponds to a duplicate assay of Clr (A resting or A activated cells), Cls (O resting or • activated cells) andIL-6 (G resting or • activated cells) in supernatants of two culture wells.

The biological profile of our patient does not meet thewhole set of criteria for the diagnosis of SLE as proposed bythe Atnerican Rheumatism Association [37] as only threecriteria were present, namely the cutaneous lesions, theANA and the anti-dsDNA antibodies. The main peculiari-ties of our patient are the late onset (at the age of 59 years) ofher lupus-like sytidrome and the absence of renal involvementin contrast to the previously reported cases (age at thediagnosis: 6 to 31 years). Nevertheless she had a SLE-likedisorder and her death may be owing to a fulminant pyogenicinfection, both diseases potentially associated with Clr/Clsdeficiency.

From data of our patient's serum we observed an imbal-ance in the Clr/Cls ratio, what is usually 1:1. The finding ofa1:10 ratio is peculiar and may reflect the specific abnormalregulation of Clr biosynthesis observed in vitro. Furtherstudies on the Clr gene expression should explain thisbiosynthesis disregulation.

ACKNOWLEDGMENTS

We are grateful to Dr C. Leclech (Dermatologie, CHU.Angers, France) and to Dr P. Bourrier (Reanimation Medi-cale, CHU, Angers, France) who took care ofthe patient andallowed access to her file. We are indebted to Dr G. Arlaud(Institut de Biologie Structurale, Grenoble, France) for thekind gift of proenzyme Clr. We thank F. Czopaki (CHU,Grenoble, France) for expert technical assistance.

REFERENCES

I Morgan BP. Complement deficiencies and diseases. In: Morgan

BP, (ed.) Complement. Clinical aspects and relevance to disease.London: Academic Press Inc., 1990:78-96.

2 Colten HR, Rosen FS. Complement deficiencies. Annu RevImmunol 1992; 10:809-34.

3 Ross SC, Densen P. Complement deficiency states and infection:epidemiology, pathogenesis and consequences of Neisserial andother infections in an immune deficiency. Medicine 1984;63:243-73.

4 Loos M, Heinz H. Complement deficiencies: the first componentClq, Clr, Cls. Prog Allergy 1986;39:212-3I.

5 Reid KBM. Deficiency ofthe first component of human comple-ment. Immunodeficiency Rev 1989;l:247-60.

6 Kolbe K, Reid KBM. Genetic deficiencies of the complementsystem and association with diseases-early components. InternRev Immunol 1993;10:17 36.

7 Tosi M, Duponchel C, Meo T, Couture-Tosi E, Complementgenes Clr and Cls feature an intronless serine protease domainclosely related to haptoglobin. J Mol Biol 1989:208:709-14.

8 Tosi M, Duponchel C, Meo T. Julier C. Complement Clssequence and linkage to Clr. Biochemistry !987;26:8516-24.

9 Mancini G, Carbonara AL, Heremans JF. Immunochemicalquantitation of antigens by single radial immunodiffusion.Immunochemistry 1965;2:235-54.

10 Drouet C. Reboul A. Colomb M. Identification of a humannon-interferon lymphokine activating monocyte complementbiosynthesis. Biochem J 1989;263:!57-64.

11 Mayer MM. Complement and complement fixation. In: KabatEA, Mayer MM, (eds) Experimental Immunochemistry, 2ndedn. Springfield: Thomas, 1961:133-240.

12 Boackle RJ, Pruitt KM, Mestecky J. The interactions of humancomplement with interfacially aggregated preparations of humansecretory TgA. Immunochemistry 1974;lI:543-8.

13 Platts-Mills TAE, Ishizaka K. Activation of the altemativepathway of human complement by rabbit cells. J Immunol1974;113:348-58.

388 A. Chevailler et al.

14 West DC, Sattar A. Kuman S. A simplified in .situ solubilizationprocedure for the determination of DNA and cell number in tissuecultured mammalian cells. Anal Biochem 1985;147;289-95.

15 Terasaki PI, McClelland JD. Microdroplet assay of humancytotoxins. Nature 1964:204:998.

16 Drouet C, Reboul A. Biosynthesis of Clr and Cls subcompo-nents. Behring Inst Mitt !989;84:80-8.

17 Navarro S, Bebili N, Bernaudin JF, Vainchenker W, Doly J.Regulation of the expression of IL-6 in human monocytes. JImmunol l989;142:4339-45.

18 Pondman KW. Sloop JW, Cormane RH, Hannema AJ. Abnor-mal C'l in a patient with systemic lupus erythaematosus. JImmunol 1968;101:811.

19 Suzuki Y, Ogura Y. Otsubo O, Akagi K, Fujita T. Selectivedeficiency of Cls associated with a systemic lupus erythaema-tosus-like syndrome: report ofa case. Arthritis Rheum 1992;35:576-9.

20 Day NK, Geiger H, Stroud RM et al. Clr deficiency: an inbornerror associated with cutaneous and renal disease. J Clin Invest1972;51;1102-8.

21 Moncada B, Day NK, Good RA, Whindhorst DB. Lupuserythematosus like syndrome with a familial defect of comple-ment. N Engi J Med 1972:286:689 93.

22 De Bracco MME, Windhorst D. Stroud RM, Moncada B. Theautosomal recessive mode of inheritance of Ctr deficiency in alarge Puerto Rican family. Clin exp Immunol l974;16:183-8.

23 Mittal KK. Wolski KP, Lim D, Gewurz A, Gewurz H, SchmidFR. Genetic independence between the HLA system and deficitsin the first and sixth components of complement. Tissue Antigens1976;7:97-IO4.

24 Lee SL, Wallace SL. Barone R, Blum L, Chase PH. Familialdeficiency of two subunits ofthe first component of complement,Clr and Cls, associated with a lupus erythematosus-like disease.Arthritis Rheum 1978;21:958-67.

25 Chase PH, Barone R, Blum L, Wallace SL. Lupus-like syndromeassociated with deficiency of Cls: family studies. Ann RoyalCollege Phys Surgeons Canada 1976:9:93.

26 Pickering RJ, Naff GB. Stroud RM, Good RA, Gewurz H.

Deficiency of Clr in human serum. Effects on the structure andfunction of macromolecular C L J Exp Med 1970; 141:803 -15.

27 Rich KC, Hurley J, Gewurz H. Inborn Clr deficiency with a mildlupus-like syndrome. Clin Immunol Immunopathol 1979;I3:77-84.

28 BorsosT. Cl activation. Immune complex mediated activation ofthe classical component pathway. Behring Inst Mitt 1989;84:93-101.

29 Pickering RJ, Michael AF, Herdman RC, Good RA. Gewurz H,The complement system in chronic glomerulonephritis: threenewly associated aberrations. J Pediat 1971;78:30-43.

30 Sim RB, Reid KBM. Cl: molecular interactions with activatingsystems. Immunoi Today 1991;12:307-l 1.

31 Bensa JC, Reboul A, Colomb MG. Biosynthesis in vitro ofcomplement subcomponents Clq, Cls and Cl-inhibitor byresting and stimulated monocytes. Biochem J l983;216:385-92.

32 Hamilton AO. Jones L. Morrison L, Whaley K. Modulation ofmonocyte complement synthesis by interferons. Biochem J1987;242:8O9-I5.

33 Kusumoto H, Hirosawa S, Salier JP. Hagen FS, Kurachi K.Human genes for complement components Clr and Cls are ina close tail-to-tail arrangement. Proc Natl Acad Sci USA1988:85:7307-11.

34 Atkinson JP. Complement deficiency: predisposing factor toauto-immune syndromes. Clin Exp Rheumatol 1989:7(Suppl.3):95-101.

35 Garty BZ, Conley ME, Douglas SD. Kolski GB. Recurrentinfections and staphylococcal liver abcess in a child with Clrdeficiency. J Allerg Clin Immunol 1987;80:63l-5.

36 Ellison RT, Curd JG, Kohier PF, Reller LB, Judson FN.Underlying complement deficiency in a patient with dissemi-nated gonococcal infection. Sex Transm Dis 1987;14:201 4.

37 Tan EM, Cohen AS. Fries JF et al. The 1982 revised criteria forthe classification of systemic lupus erythematosus. ArthritisRheum 1982;25:127I-7.

Received 16 March 1994

Accepted in revised form 17 May 1994