further characterisation of occurring antibasement

Journal of Clinical Pathology, 1978, 31, 639-644

Further characterisation of a spontaneously occurringantibasement membrane antibodyARTHUR J. AMMANN AND JOSEPH R. GOODMANFrom the Department ofPediatrics, Immunology Section, University of California, San Francisco, andVeterans Administration, San Francisco, California, USA

SUMMARY A spontaneously occurring antibasement membrane antibody has been furthercharacterised according to antigenic specificity, immunoglobulin class, and tissue localisation usingimmunofluorescent and immunoelectron microscopic techniques. The autoantibody reacted withthe basement membrane of kidney tubules, Bowman's capsule, and the epithelial portion of ileumbut not with the basement membrane of skin, cornea, glomerulus, or oesophagus. It also reactedwith bile canaliculi, sarcolemmal sheath, and salivary duct. On electron microscopy the antibodywas distributed along the basement membrane of Bowman's capsule and renal tubules. Somereactivity against collagen was observed. Antibody activity was found in both IgG and 1gM fractions.In immunodeficiency disorders, the autoantibody was found only in patients with selective IgAdeficiency.

A spontaneously occurring autoantibody reactingwith selected membranes and distinct from otherbasement membranes was described by us in 1971(Ammann and Hong, 1971a). This antibody wasfound in patients with coeliac disease and IgAdeficiency. Seah et al. (1971) had previously describedan autoantibody with similar tissue localisation,which they termed reticulin antibody and which theyfound primarily in patients with coeliac disease anddermatitis herpetiformis. Subsequent studies by Alpand Wright (1971) and von Essen et aL (1972) con-firmed both the tissue localisation of the autoanti-body and the occurrence in coeliac disease. Rizzettoand Doniach (1973) reported on the occurrence ofreticulin antibodies in an extensive evaluation of3628 sera from patients with a wide variety of dis-orders. They defined the tissue localisation of theautoantibody, the specificity for certain diseasestates, and the immunoglobulin class distribution.They further suggested that five types of autoanti-body could be demonstrated and termed these:reticulin R1, reticulin R2, Kupffer cell, sinusoidaladherent cell cytoplasmic, and reticulin sinusoidalRs.

In this report we describe detailed studies on thetissue localisation and distribution of antibody

Received for publication 14 July 1977Amended version accepted 12 December 1977

reactive with basement membrane using immuno-fluorescent and immunoelectron microscopic tech-niques. This autoantibody appears to be identicalwith that described by Seah et al. (1971) and thereticulin R1 antibody described by Rizzetto andDoniach (1973).

Material and methods

Blood samples were obtained from patients,immediately separated, and stored frozen at - 20°C.These samples were used to perform indirectimmunofluorescence at 1-10 serum dilutions onvarious tissues obtained from rat and humansources, as previously described (Ammann andHong, 1971b).

Electron microscopy was performed as follows:Normal rat kidney was cut and placed in ice-cold 1%glutaraldehyde in 0-1 M cacodylate buffer for 45minutes. Blocks were washed in buffer and 35 ,uslices were cut on a tissue chopper. This wasfollowed by incubation with either the patient'spurified IgG or normal human serum for 45minutes at room temperature and the sections werewashed three times with 0-1 M cacodylate buffer.These, plus a third control preparation not in-cubated with serum, were all incubated in ferritin-labelled goat anti-human IgG for 30 minutes, washedthree times with 0-1 M cacodylate buffer, fixed in

639

on April 25, 2022 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.31.7.639 on 1 July 1978. Dow

nloaded from

http://jcp.bmj.com/

A. J. Ammann and J. R. Goodman

buffered 1 % glutaraldehyde, post-fixed in 1%osmium, and prepared for electron microscopy.

Collagen was prepared from the skin of newborninfants and was supplied to us by Rothbard andWatson (1972). This preparation was used forabsorption of antibody in a concentration of1 mg/ml with varying amounts of serum.The autoantibody was separated into IgG and

IgM fractions, using sodium sulphate precipitationand Sephadex G200 column chromatography. Thepatient from whom this autoantibody was obtainedlacked serum IgA.

Results

Separation of the serum into IgG and IgM fractionsand subsequent indirect immunofluorescence withanti-IgG and anti-IgM antibody indicated that auto-antibody activity was present in both immuno-globulin fractions. No autoantibody activity waspresent in IgA as the patient lacked serum IgA.

After absorption of the autoantibody with highlypurified collagen, no appreciable change in the dis-tribution or intensity of autoantibody reactivitycould be demonstrated by immunofluorescent tech-niques. Table 1 summarises the distribution ofreactivity of the autoantibody against various tissues.Figure 1 demonstrates a typical pattern on rat renaltubules.

Table 1 Tissue reactivity of antibasement membraneactivity

Structure Anti-IgGfluorescent antibodyreactivity

Bowman's capsule +Glomerular basement membraneKidney tubule basement membrane +Liver portal tracts +Sarcolemmal sheath +Epithelial cell basement membrane (ileum) -Oesophagus basement membraneSalivary duct (striated duct) basementmembrane of serous cells +

Cornea basement membrane

Electron microscopy using ferritin-labelled anti-human gamma chain indicated that the autoantibodyreacted in the same distribution on renal basementmembranes as was found using the immuno-fluorescent method. The ferritin-labelled anti-humanIgG was seen along many tubular basement mem-branes and in some endothelial cells of capillaries intubular areas (Fig. 2). There was no ferritin anti-human IgG along glomerular basement membraneof glomerular capillary endothelium. In addition, itwas present in the rat plasma in both tubule andglomerular capillary lumens, indicating that thereagent penetrated to these areas and cross-reactedwith rat IgG. In certain areas activity againstcollagen fibres was demonstrated (Fig. 3).

Fig. 1 Fluorescent photomicrograph of isolated IgG on rat kidney section. Positive stainingof basement membrane ofBowman's capsule and tubules is seen. x 40

640



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Further characterisation ofa spontaneously occurring antibasement membrane antibody

I

Fig. 2 Electron micrographs offerritinmolecules within the basement membrane (B)of this tubular cell can be seen (arrow). Thereis also some ferritin in the endothelial cell (E)lining the capillary lumen (L). The scale is onemicron.

Table 2 summarises the distribution of anti-basement membrane antibody in various clinicalimmunodeficiency disorders. It was not found in 4(0normal adults and children.

Discussion

Spontaneously occurring antibasement membraneantibody has been described with varying histologicaldistributions. In Goodpasture's syndrome, an auto-antibody is found by immunofluorescence whichreacts with the basement membrane of the renalglomerulus and lung (Lerner et al., 1967; McPhauland Dixon, 1969). Autoantibody reacting with thebasement membrane of skin and oesophagus hasbeen found in pemphigoid (Chorzelski et al., 1966;Beutner et al., 1967). The autoantibody described inthis report reacts with the basement membrane ofBowman's capsule and kidney tubules (Fig. 1) butnot with the glomerulus or basement membrane ofoesophagus or cornea. Reactivity was also observedwith basement membrane of the epithelial cells of

villi taken from the ileum, sarcolemmal sheath ofmuscle, and the portal tracts of the liver. In somesections reactivity with collagen fibres could be seen(Fig. 3). From these studies it cannot be determinedif the autoantibodies are reacting with shared anti-genic components or if two distinct autoantibodiesare present. Attempts to absorb the autoantibodywith collagen did not reduce antibasement membraneactivity.The importance of these studies is in the observa-

tion that both basement membrane and collagenreactivity were observed on electron microscopicevaluation whereas dual reactivity could not bedistinguished by immunofluorescent studies. It islikely, therefore, that the antibasement membraneantibody described by us (Ammann and Hong,1971b), the antireticulin antibody described by Seahet al. (1971), and the R1 antibody described byRizzetto and Doniach (1973) have similar, if notidentical, antigenic components, namely, basementmembrane and collagen.The studies of Seah et al. (1971) and the detailed

641J



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

A. J. Ammann and J. R. Goodman

Fig. 3 Electron micrograph ofa smallcluster of collagen fibres (C), which have ascatter offerritin molecules (arrows) onand near their surface. There is also someferritin in the basement membrane (B) ofthis tubule cell. The scale is one micron.

Table 2 Antibasement membrane antibodies inimmunodeficiency disorders

Disorder No. tested No. positive

Coeliac disease with selective IgAdeficiency 3 3

Selective IgA deficiency withoutcoeliac disease 40 1

Hypogammaglobulinaemia 12 0Thymic hypoplasia 6 0Chronic granulomatous disease 10 0Severe combined immunodeficiency 5 0Normals (2 to 40 years) 40 0

investigations of Rizzetto and Doniach (1973) are inagreement as to the tissue localisation of antibodyreactivity. A nodular fluorescence is observed sur-rounding the portal tracts of the liver with discretewave-like bands extending into the sinusoidal walls,a honeycomb appearance on smooth or striatedmuscle, staining of Bowman's capsule, and reactivityin the peritubular area of the kidney. We haveextended the number of tissues examined and findreactivity along the basement membrane of epithelial

cells of ileum and basement membrane of salivaryduct (striated duct) serous cells (Table 1). Noreactivity was observed against basement membranesof cornea, glomerulus or oesophagus, indicating thatthese basement membranes are distinct antigenicallyfrom those of Bowman's capsule, epithelial cells ofgastrointestinal tract and kidney tubule, and theportal tract of the liver. It is of interest that thedistinct distribution of antibasement membrane anti-body detected in Goodpasture's syndrome, pemphi-goid, and in coeliac disease corresponds to distinctdifferences in biochemical composition, as describedby Kefalides (1969).

Several experimental studies suggest that anti-bodies to collagen and basement membrane havesimilar patterns of distribution by immunofluorescenttechniques. Nagasawa and Shibata (1969) usedhuman glomerular basement membrane as antigenand studied the tissue distribution of antibody in rattissue. Antibody activity was demonstrated againstthe basement membrane of Bowman's capsule, renaltubular glomerulus, sarcolemmal muscle sheath, and

642)



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Further characterisation ofa spontaneously occurring antibasement membrane antibody 643

liver sinusoids. However, using histochemical tech-niques, Velican and Velican (1970) demonstratedstructural heterogeneity between glomerular, cap-sular, and tubular basement membranes. Using anti-body to human infant skin collagen, Rothbard andWatson (1972) demonstrated reactivity by immuno-fluorescent methods in a pattern of distributionsimilar to that obtained using antibasement mem-brane antibody. These studies emphasise thedifficulty in distinguishing antibasement membraneactivity from anti-collagen activity by immuno-fluorescent methods alone.

Rizzetto and Doniach (1973) stressed that con-nective tissue elements are a mixture of amorphousproteins and mucopolysaccharides in which there arefibrous proteins related to collagen. They suggestedthat electron microscopic studies would be necessaryto see if the antigens detected by spontaneouslyoccurring autoantibody are fibrillary or belong toamorphous proteins and mucopolysaccharides. Ourstudies clearly demonstrate both basement mem-brane and collagen reactivity. The term 'reticulin'antibody (Seah et al., 1971) or 'R1' antibody(Rizzetto and Doniach, 1973) would, therefore,appear to be appropriate general terms when thisautoantibody is detected by immunofluorescentmethods. The antibody activity may be furtherdefined as against basement membrane and/orcollagen by electron microscopic techniques.

In clinical disorders, the autoantibody has beenfound to be primarily associated with coeliacdisease. Although we originally described the auto-antibody in patients with selective IgA deficiencycoeliac disease (Ammann and Hong, 1971b), Seahet al. (1971) and Alp and Wright (1971) found theautoantibody to be present in patients with coeliacdisease, the majority of whom had normal IgAvalues. In a study of 3628 sera from patients with avariety of diseases, Rizzetto and Doniach (1973)found antibody to reticulin to be most frequentlypresent in dermatitis herpetiformis, coeliac disease,chronic heroin addicts, Sjogren's syndrome, my-asthenia gravis, vitiligo, alopecia, and rheumatoidarthritis. However, they did not distinguish betweenthe various subtypes of antibody to reticulin presentin the various diseases. Spontaneously occurringantitubular basement membrane antibody has beendescribed after renal transplantation (Wilson et al.,1974) and in methicillin-induced interstitial nephritis(Baldwin et al., 1975).We studied the sera of 76 patients with immuno-

deficiency to determine the relationship betweenantibody against reticulin and specific immuno-deficiency disorders. Although the number ofpatients in some categories was small, only onepatient who did not have coeliac disease was found

to be positive. This patient had selective IgAdeficiency and oedema of unknown aetiology. Thus,the autoantibody appears to be confined to immuno-deficient patients with selective IgA deficiency.

This study was supported by John A. HartfordFoundation, Inc., and the Division of ResearchResources, National Institutes of Health 5 MO1-RR00079-12.

References

Alp, M. H., and Wright, R. (1971). Autoantibodies toreticulin in patients with idiopathic steatorrhoea,coeliac disease, and Crohn's disease, and their relationto immunoglobulins and dietary antibodies. Lancet, 2,682-685.

Ammann, A. J., and Hong, R. (1971a). Unique antibodyto basement membrane in patients with selective IgAdeficiency and coeliac disease. Lancet, 1, 1264-1266.

Ammann, A. J., and Hong, R. (1971b). Selective IgAdeficiency. Presentation of 30 cases and a review of theliterature. Medicine, 50, 223-236.

Baldwin, D. S., Levine, B. B., McCluskey, R. T., andGallo, G. R. (1975). Renal failure and interstitialnephritis due to penicillin and methicillin. NewEngland Journal of Medicine, 279, 1245-1252.

Beutner, E. H., Rhodes, E. L., and Holborow, E. J.(1967). Autoimmunity in chronic bullous diseases.Immunofluorescent demonstration of antibodies toskin in sera of patients with pemphigus, bullouspemphigoid and in other human sera. Clinical andExperimental Immunology, 2, 141-151.

Chorzelski, T. P., Von Weiss, J. F., and Lever, W. F.(1966). Clinical significance of autoantibodies inpemphigus. Archives of Dermatology, 93, 570-576.

von Essen, R., Savilahti, E., and Pelkonen, P. (1972).Reticulin antibody in children with malabsorption.Lancet, 2, 1157-1159.

Kefalides, N. A. (1969). The chemistry and structure ofbasement membranes. Arthritis and Rheumatism, 12,427-443.

Lerner, R. A., Glassock, R. J., and Dixon, F. J. (1967).The role of anti-glomerular basement membrane anti-body in the pathogenesis of human glomerulonephritis.Journal ofExperimental Medicine, 126, 989-1004.

McPhaul, J. J., Jr., and Dixon, F. J. (1969). The presenceof anti-glomerular basement membrane antibodies inperipheral blood. Journal of Immunology, 103, 1168-1175.

Nagasawa, T., and Shibata, S. (1969). Immuno-fluorescence studies on the anatomic distribution of thesoluble antigen responsible for the production ofnephrotoxic serum nephritis. Journal of Immunology,103, 736-740.

Rizzetto, M., and Doniach, D. (1973). Types of'reticulin' antibodies detected in human sera by im-munofluorescence. Journal of Clinical Pathology, 26,841-851.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

644 A. J. Ammann and J. R. Goodman

Rothbard, S., and Watson, R. F. (1972). Demonstrationof collagen in human tissues by immunofluorescence.Laboratory Investigation, 27, 76-84.

Seah, P. P., Fry, L., Hoffbrand, A. V., and Holborow,E. J. (1971). Tissue antibodies in dermatitis herpeti-formis and adult coeliac disease. Lancet, 1, 834-836.

Velican, D., and Velican, C. (1970). Structural hetero-geneity of kidney basement membranes. Nature, 226,1259-1261.

Wilson, C. B., Lehman, D. H., McCoy, R. C., Gunnells,J. C., Jr., and Stickel, D. L. (1974). Antitubularbasement membrane antibodies after renal transplanta-tion. Transplantation, 18, 447-452.

Requests for reprints to: Dr Arthur J. Ammann, Depart-ment of Pediatrics, Immunology Section, University ofCalifornia, San Francisco, California 94143, USA.



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

further characterisation of occurring antibasement

Documents