Immunologic Mechanisms of Renal Disease

Review: Immunologic Mechanisms of Renal Disease BY STEPHEN ADLER MD, AND WILLIAM COUSER MD

E

nd-stage renal disease remains the fourth leading cause of death among young adults in this country. About 60% of such patients develop renal failure as a consequence of some form of immunologically mediated glomerular disease resulting in an annual expenditure of over 700 million dollars for chronic dialysis or transplantation therapy. Following the development of immunofluorescence microscopy, and the widespread application of percutaneous renal biopsy as a diagnostic tool in the 1960's, two mechanisms of immune renal disease were recognized. Less than 5% of patients had an uninterrupted, linear pattern of IgG deposition along glomerular capillary walls resulting from deposition of an autoantibody to glomerular basement membrane (GBM) antigens. Most of the remaining patients had a discontinuous, or granular pattern of staining for IgG and C3. Based on the pioneering studies of acute and chronic models ofBSA From the Division ofNephrology, Depart· ment of Medicine, University of Wash· ington, Seattle, Washington. Reprint requests: Dr. William G. Couser, Division ofNephrology, Box RM·ll, Univer· sity of Washington, Seattle, Washington 98195.

serum sickness in rabbits by Germuth and Dixon in the 1950's and 1960's, in which very similar glomerular lesions were produced, it was concluded that these granular deposits all resulted from the passive glomerular trapping of antigen-antibody complexes formed in the circulation. 1 ,2 The principal mechanism of tissue injury was believed to be complement activation by the deposited complexes with generation ofleukochemotactic factors, attraction of neutrophils and release of proteolytic enzymes by these inflammatory effector cells resulting in GBM damage with proteinuria, hematuria, and other clinical and histologic manifestations of glomerulonephritis. In the past several years rapid advances in cell biology, immunology, and physiology have greatly extended these original concepts to a much broader understanding of the pathogenesis of immune renal disease. This article reviews briefly the current status of this rapidly changing area and the implications for understanding several clinically important glomerular diseases in man. More extensive reviews of this subject have been published elsewhere. 3 -5

The American Journal of the Medical Sciences

Mechanisms of Glomerular Immune Deposit Formation Most immune renal diseases are associated with glomerular deposits of IgG and complement. Those that are not are discussed briefly in section III below. It is now clear that separation of immune renal diseases into the two categories of anti-GBM nephritis and nephritis due to circulating immune complex trapping is over-simplified. Both linear and granular types of immune deposits can result from antibody binding to intrinsic glomerular antigens (Figure 1). Granular deposits can also result from antibody binding to antigens that first become localized in the glomerulus by several different mechanisms (in situ immune com(plex formation), as well as from circulating immune complex trapping. Moreover, it now appears that the principal determinant of the type of clinical and histological glomerular lesion that results is the site at which the deposits form in glomeruli, probably because immune deposits at different sites activate different mediators of tissue injury. Immune complexes of antibody bound to glomerular or non-glomerular antigens may form along the subepithelial aspect of the cap-

55

Immunologic Renal Disease

Figure 1: The classical immunofluorescence pattems o( lineal' anti-GBlv! antibody (A) and granular immune complex (8) deposits o( IgG seen in Goodpasture's sYlldrom e alld membranous nephropathy. Previously thought to represellt two di/Terent meclr(/lli.5ms o(iml/llllle renal disease, both o(these lesiolls resulted experimelltally (rom alltibody billdillg to alltigells present in tire Ilormal glomerular capillary wall.

illary wall (post streptococcal glomerulonephritis, membranous nephropathy, SLE), within the GBM itself (Goodpasture's syndrome), along the subendothelial aspect of the capillary wall (SLE, Type I membranoproliferative glomerulonephritis; shunt nephritis) and within the glomerular mesangium (SLE, IgA nephropathy, Henoch-Schonlein purpura with nephritis). Deposits at each of these sites are illustrated schematically in Figure 2. Deposits at different sites often form by different mechanisms. The experimental evidence for this is reviewed briefly below noting the potentially relevant clinical correlates in man_ Subepithelial immune deposit formation. Deposits in a subepithelial distribution are characteristic of idiopathic and other forms of membranous nephropathy (drug induced, tumor associated, etc.), the membranous form of lupus nephritis, and the "humps" that are seen in most cases of post-streptococcal glomerulonephritis. Usually, deposits at this site are associated with a marked increase in glomerular permeability and nephrotic syndrome with relatively little glomerular histologic change and well preserved renal function. Experimentally, deposits identical to those seen in idi-

56

opathic membranous nephropathy in man may be formed by antibody reacting with an antigenic glycoprotein product of epithelial cells (Heymann antigen) resulting in finely granular immune complex deposits along the epithelial cell membrane and in filtration slit pores between the epithelial cell foot processes (Figure 2). 6 This fixed antigen mechanism of in situ immune complex formation was described in 1978 and represented the first recognition that granular as well as linear deposits could result from antibody binding to fixed glomerular antigens. This observation stimulated an intensive reinvestigation of the pathogenesis of immune complex nephritis in general, which has now resulted in recognition of several additional mechanisms by which subepithelial immune complex deposits may form .5 Most of these involve charge-charge interactions between negatively charged glomerular anionic sites, which include sialoglycoproteins ahng the epithelial cell surface and glycosaminoglycans in the lamina rara extern a and interna of the GBM (Figure 2), and positively charged, or cationic, antigens and antibodies. Subepithelial immune complex deposits containing exogenous antigens may form locally, or in situ, by cationic antigens binding

on an electrical basis to glomerular anionic sites followed by interaction with antibody. 7 Relatively cationic antibodies may also localize first and then interact with antigen. It now appears that anionic antigens, such as DNA, may also localize on a charge basis. This sequence may involve initial glomerular localization of non-immune cationic proteins such as neutrophil cationic proteins, cationic products of complement activation, and platelet factor 4. These polycations bind to glomerular anionic sites followed by charge-charge interaction with anionic antigens and subsequent antibody binding to produce local immune complex formation. 5 Other mechanisms of subepithelial immune complex formation involving exogenous antigens include dissociation of subendothelial deposits that may cross the capillary wall and re-form in situ in a subepithelial distribution, and possibly the direct localization of pre-formed cationic immune complexes from the circulation; although the latter mechanism remains con troversia1. 8 ,9 In all forms of experimental membranous nephropathy studied to date, the mechanism of tissue injury appears to be a direct effect of complement that does not involve inflammatory effector cells.lO Thus, membranous nephropathy is a rather bland lesion histologically, with good preservation of renal function. This direct effect appears to involve the terminal complement system and probably reflects the formation of complement membrane attack complexes in some structure on the subepithelial surface of the capillary wall, resulting in a loss of glomerular barrier function. The lack of acute inflammatory change in membranous nephropathy presumably is due to the fact that deposits form at a site separated from the circulation by GBM. This precludes inflammatory cell attraction by chemotactic factor February 1985 Volume 289 Number 2

Adler et 01

generation in the deposits or interaction between Fc or C3b receptors on neutrophils or macrophages and the deposits.lO There are few clues from studies of human diseases to validate these pathogenetic mechanisms derived from experimental studies. In post-streptococcal glomerulonephritis, antibodies to certain cationic extracellular streptococcal products have reacted with antigens in the glomerular deposits of some patients. l l Others have described an altered, positively charged, IgG molecule and antibody to it in these deposits.l 2 Circulating immune complexes are rarely detected in idiopathic membranous nephropathy; an observation more consistent with a fixed antigen than an exogenous antigen mechanism. Idiopathic membranous nephropathy is associated with impaired antibody production. The membranous form of lupus nephritis is often associated with very low levels of anti-DNA antibody. These conditions would favor antigen excess and persistence and thereby allow in situ deposit formation to occur if antigens with an affinity for glomeruli were present. It is likely that both fixed antigen and exogenous antigen planting mechanisms contribute to subepithelial immune deposit formation in membranous nephropathy of different etiologies in man. Intra-membranous immune deposit formation. Anti-GBM disease, considered in light of the mechanisms discussed above, is an example of in situ immune complex formation involving an autoantibody binding to an intrinsic glomerular antigen within the GBM. The spatial distribution of this antigen is such that the deposits appear continuous, or linear, at the magnification of the immunofluorescent microscope (Figure 1). The precise biochemical composition and ultrastructural localization of the nephritogenic GBM antigen(s) remains unclear,

2

.: '-H---

3

LRE LD LRI

Figure 2: Schematic representation of"three glomerular capillary loops showing the sites of glomerular immune complex formation along the subepithelial slllface of the capillary wall as the "humps" of post.streptococcal glomerulonephritis (1) or the diffuse, finely granular deposits of membra/IOUS nephropathy (2, Figure 1). Subendothelial deposits (3) are seen in SLE and '!Ype I membmnoproliferative glomerulonephritis, and deposits ill th e mesangial matrix (MM) (4) are characteristic of SLE, IgA Ilephropathy, Helloch·Sc1lOl/leil/ purpura with nephritis and other diseases. The deposits in mesallgial matrix surroullded mesal/gial cells (Me). Anti·GBM antibody deposits are ill a lillear pattern with in the capillary wall itself (5). The inset illustrates the three layers of the llormal glomerular capillary wall, endoth elial cells (EN), GBM and epithelial cells (EP). The negative cha.rge 011 the capillary wall results from the sialoglycoprotein (S) coatillg elldothelial and epithelial cell surfaces and the hepamn sulfate containillg glycosaminoglycalls (GAG) which are distributed discoll' tinuously in the lamilla mm intema (LRl) and extema (LRE) of the GBM.

The American Journal of the Medical Sciences

but it appears to be one or more non-collagenous glycoproteins localized within the lamina densa of the GBM. Cross-reactive antigenic epitopes are also present in alveolar basement membrane. Patients who develop anti-GBM antibody in the presence ofpulmonary disease (or possibly as a consequence of pulmonary disease) also develop antibody mediated pulmonary hemorrhage as well as nephritis (Goodpasture's syndrome). Patients without pulmonary disease tend to get only glomeruionephritis. 13 Sensitive radioimmunoassays are now available that detect anti-GBM antibodies in the serum with over 95%

sensitivity. In experienced hands, clinically significant antibodies can also be detected in over 80% of cases by indirect immunofluorescence using normal human kidney tissue. Because anti-GBM disease is relatively easy to produce and study experimentally, much of what is known of the mediation of immune renal disease has been defined in models of anti-GBM, or nephrotoxic, nephritis and then extrapolated to other types of immune renal disease. Depending on the source of antibody and the species studied, anti-GBM antibody has been shown to cause glomerular injury by mechanisms

57

Immunologic Renal Disease

that include: a) A direct effect of antibody alone-independent of complement or inflammatory cellson glomerular permeability to protein; b) A direct effect of complement involving terminal complement components; c) A complement-neutrophil mediated form of injury requiring generation of complement chemotactic products. 3 The predominant mechanism by which neutrophils cause glomerular injury appears to be through generation of reactive oxygen species rather than production of proteolytic enzymes. Which of these mechanisms predominates in human disease is unclear, although most cases of anti-GBM nephritis in man do have associated complement deposits. Subendothelial immune deposits. Immune complex deposits in the subendothelial space, between the endothelial cell layer and the GBM, are commonly seen in severe lupus nephritis, Type I membranoproliferative glomerulonephritis, and in other types of immune complex nephritis such as that associated with infected ventriculo-atrial shunts. They are almost invariably accompanied by histologic evidence of a fairly severe inflammatory and proliferative response in the glomerulus, heavy proteinuria, usually in the nephrotic range, and deteriorating renal function. They are thus the most pathogenic of the glomerular immune deposits and usually associated with severe disease and a poor prognosis. Subendothelial immune complex deposits are almost invariably accompanied by immune complex deposits in the mesangium as well. It is likely that both varieties of immune complex deposits form by similar mechanisms. Experimentally, they may be formed in situ following localization of relatively large molecular weight 58

cationic antigens that bind to anionic sites on endothelial cell surfaces and in the lamina rara intern a but do not cross the capillary wall easily. 8 They have also been formed using antigenic lectins such as concanavalin A that bind to glucose and mannose residues in the GBM.l4 Subendothelial deposits may also result from the passive trapping of preformed immune complexes along the capillary wall, particularly when glomerular delivery of immune complexes is high due to increased serum levels and/or decreased mononuclear phagocyte system (MPS) clearing function. I5 A reduction in the glomerular charge barrier and an increase in glomerular permeability induced by circulating polycations also leads to increased subendothelial immune complex localization. I6 Thus, circulating immune complexes may be trapped more readily in the damaged glomerulus than in the normal one. The mechanism of this effect is unclear. At present there is little evidence for participation of fixed glomerular antigens in the pathogenesis of subendothelial immune deposits, although antibodies to several endothelial cell surface antigens have been reported in SLE and could contribute to the process. Because they are in direct contact with the circulation, there is a dynamic equilibrium between circulating and deposited antigens and antibody that is less evident when deposits are in a subepithelial distribution. Thus, circulating antigens and antibodies may bind to their deposited counterparts to enlarge the deposits, or the deposits may be diminished or solubilized by antigen excess that results in smaller, more soluble immune complexes. Other circulating immune reactants such as rheumatoid factors anti-idiotypic antibodies may also bind to complexes in the subendothelial area and increase deposit formation by this local mechanism. The mediation of tissue injury

by subendothelial immune complex deposits has never been studied directly. Various studies of injury induced by antibody binding to exogenous antigens within the GBM itself(in contrast to antiGBM disease) have implicated complement-neutrophil dependent mechanisms and complement independent injury mediated directly by neutrophils (by Fc receptors) or macrophages (via C3b receptors) binding to immune deposits. 3 Because subendothelial deposits are even more accessible to circulating cells, it is presumed that all of these mechanisms are involved. The glomerular inflammatory changes seen histologically support this suggestion. There are few studies in man that clarify the nature of subendothelial deposits in human renal disease. Nuclear antigens and antinuclear antibodies are presumably involved in SLE, and a variety of antigenic components of infectious agents have been demonstrated in glomerular deposits in cases of post-infectious glomerulonephritis. As noted above, however, a variety of proteins may have an affinity for glomerular structures and localize in abnormal glomeruli without necessarily representing the antigenic components of a pathogenic immune complex. Rigorous demonstration of both antigen and specific antibody to it in immune complex form in the circulation and in glomerular immune deposits has virtually never been accomplished in human renal disease. In most diseases there are few clues to what antigens to look for. Until new insights into the events that initiate these diseases are developed and improved techniques become available for more careful studies of human biopsy material, the agents responsible for most cases of immune complex glomerulonephritis in man are likely to remain elusive. Mesangial immune deposits. Immune deposits within the mesangial matrix are characteristic of February 1985 Volume 289 Number 2

Adler et 01

a variety of immune renal diseases including all forms of lupus nephritis, Type I MPGN, IgA nephropathy, Henoch-Schonlein purpura with nephritis, and others. The mesangium is the principal route by which a variety of filtered macromolecules are cleared from the glomerul us, and mesangial immune complex deposits probably occur in most patients with significant levels of immune complexes in the circulation. As evidenced by biopsies from patients with SLE and normal renal function, such deposits may be present in significant amounts without causing any apparent glomerular injury. The process by which mesangial immune complex formation leads to glomerulonephritis is still a subject of active investigation. One possibility is that mesangial immune complexes form locally by antibody reacting with intrinsic mesangial antigens as in anti-GBM disease. However, no evidence of antibody reactivity with intrinsic mesangial antigens has been reported in man. Another possibility is that antibody reacts in situ with previously trapped exogenous antigenic material to produce local immune complex formation with complement activation and inflammatory cell participation. This process has been shown experimentally to produce a focal proliferative glomerular lesion much like that seen in human disease associated with mesangial deposits, although the mechanisms responsible have not been studied. I7 The IgA deposited in IgA nephropathy and HenochSchonlein purpura appears to be an antibody reactive with some exogenous antigen in the mesangium of these patients.l 8 Finally, the deposits may in part represent pre-formed immune complexes passively trapped from the circulation. Pre-formed immune complexes are readily taken up and cleared by the glomerular mesangium, although usually with minimal evidence of

glomerular injury. In IgA nephropathy and Henoch-Schonlein purpura there is considerable evidence for a circulating IgA immune complex trapping mechanism with complexes composed of secretory IgA probably generated in response to a mucosal antigenic challenge by some as yet unidentified antigen. I8 In contrast to capillary wall lesions where injury is caused by non-glomerular immune reactants, mesangial cells themselves have now been shown to produce a variety of inflammatory mediators including proteolytic enzymes, an interleukin I-like molecule, reactive oxygen species and products of arachidonic acid metabolism .I8 Thus, injury or activation of mesangial cells by complex formation or deposition in the mesangium may result in local release ofmediators that cause or exacerbate the damage produced.

The American Journal of the Medical Sciences

Immune Renal Diseases Without Glomerular Immune Deposits

Minimal change nephrotic syndrome disease spectrum . This group of diseases is characterized by a generalized loss of glomerular anionic sites with a diffuse increase in glomerular albumin filtration and nephrotic syndrome associated with varying degrees of structural damage ranging from none to mesangial deposition of IgM and/or varying degrees of mesangial cell proliferation, to focal and segmental glomerular sclerosis. The more structural abnormalities present, the less steroid-responsive the lesion is clinically. The principal pathogenetic event in these diseases appears to be a diffuse loss of the glomerular charge barrier to protein filtration. The prompt recurrence in some transplant patients suggests a circulating mediator of this phenomenon. Since immunoglobulin deposits are absent , attention has focused on the possibility that this material may be a cell-derived lymphokine.l 9 Circumstantial evidence-including

the induction of remission with steroids, cyclophosphamide, and measles infection; the association with Hodgkins disease; and effects of patients' serum on lymphocyte blastogenesis and abnormalities in circulating T-cell subset ratios-suggests that there is Tcell dysfunction in this disease. To date, however, no direct evidence for a cell-derived substance capable of inducing proteinuria has been developed in minimal change nephrotic syndrome, and the pathogenesis of the disease remains obscure. The structural abnormalities that occur are presumably secondary to glomerular hemodynamic adaptations to injury andlor consequences of prolonged increased protein filtration and perhaps mesangial injury due to increased uptake and overload of potentially toxic macromolecules. TYpe II MPGN (dense deposit disease). This disease looks clinically and histologically similar to Type I MPGN, but subendothelial and mesangial immune complex deposits are absent. Nephrotic syndrome occurs secondary to the development of electron dense deposits of nonimmunoglobulin material of uncertain origin and composition within the GBM itself. Although the disease is commonly associated with the presence of an autoantibody to C3 convertase (C3 nephritic factor), hypocomplementemia due to alternate complement pathway activation and complement deposition in glomeruli, the pathogenesis remains obscure. Idiopathic rapidly progressive glomerulonephritis. A substantial number of patients with crescentic glomerulonephritis, including some with various forms of vasculitis, have widespread glomerular crescent formation in the absence of detectable circulating or deposited anti-GBM antibody or immune complexes. 2o Although the crescents are believed to be initiated by capillary wall rupture and fibrin leakage into Bowman's

59

Immunologic Renal Disease

space, glomeruli may not appear severely damaged histologically. There are no pathogenic immunoglobulin deposits, but the disease is believed to have an immune pathogenesis and may represent a form of cell-mediated glomerular injury.2o Summary

Most immune renal diseases are caused by the formation of immune complexes of antibody with either fixed glomerular antigens or exogenous non-renal antigens. Much progress has been made recently in understanding the ways by which these immune deposits form in glomeruli. Immune complex deposits of exogenous antigens may involve prior antigen localization in the glomerulus to initiate immune complex formation locally, or in situ. The type of glomerular lesion produced depends in large part on the site at which deposit formation occurs, which in turn determines what mediators of tissue injury are activated. The nature and quantity of immune reactants are also important. Subepithelial deposits may result from antibody binding to fixed epithelial cell-derived antigens or to exogenous antigens localized by direct interaction with glomerular anionic sites (cationic antigens) or with nonimmune cationic proteins bound to glomerular anionic sites (anionic antigens). Cationic antibody may also localize first, and deposits can form from subendothelial immune complex deposits dissociating to cross the GBM and re-form in a subepithelial distribution. Proteinuria induced by subepithelial immune deposit formation appears to be due to a direct effect of complement, probably involving membrane attack complexes, and independent of inflammatory cells. Intramembranous deposits form from anti-GBM antibody reacting with intrinsic GBM antigens.

60

Subendothelial and mesangial deposits appear not to involve fixed antigens. Rather, they represent .immune complexes containing exogenous antigens and antibody to the antigens. These complexes may result from the passive trapping of pre-formed immune complexes from the circulation or may form in situ by several different mechanisms. The latter process appears to be a more nephitogenic one. With immune complex formation at sites more accessible to circulating effector cells, deposits tend to elicit an inflammatory response mediated by complement, neutrophils, and macrophages. Intrinsic glomerular mesangial cells also produce mediators of tissue injury and may participate in this process. Several glomerular diseases are believed to have an immune basis that are not characterized by glomerular immune deposits. These include the minimal change nephrotic syndrome group of disorders, Type II MPGN, and idiopathic RPGN. The pathogenesis of these diseases remains undefined. References 1. Dixon JF, Feldman JD, Vazquez JJ: Experimental glomerulonephritis: The pathogenesis of a laboratory model resembling the spectrum of human glomerulonephritis. J Exp Med 113:899-920, 1961. 2. Germuth FG Jr, Rodriguez E: Immu· nopathology o{ the Renal Glomerulus: Immune Complex Deposit and AntiBasement Membrane Disease. Boston, Little Brown and Co, 1973. 3. Couser WG, Salant DJ: Immunopathogenesis of glomerular capillary wall injury in nephrotic states. Contemp Isslles NephroI9:47-83, 1982. 4. Couser WG, Sal ant DJ: In situ immune complex formation and glomerulal' injury (Editorial Review). Kidney Int 17:1-13, 1980. 5. Couser WG, Adler S, Baker PJ, Johnson RJ, Perkinson DA: Mechanisms o{ immune complex {ormation and deposition in glomeruli: An overview. Proc of IXth Int Congress of Nephrology, Los Angeles, 1984. New York, Springer-Vel'lag, to be published. 6. Couser WG, Steinmuller DR, Stilmant MM, Sal ant DJ, Lowenstein LM:

Experimental glomerulonephritis in the isolated perfused rat kidney. J Clin Invest 62:1275-1287,1978. 7. Border WA, Ward HJ, Kamil ES, Cohen AH: Induction of membranous nephropathy in rabbits by administration of an exogeneous cationic antigen. Demonstration of a pathogenic role for electrical charge. J Clin Invest 69:451-461, 1982. 8. Vogt A, Rohrbach R, Shimizu F, Takamiya H, Batsford S: Interaction of cationized antigen with rat glomerular basement membrane: In situ immune complex formation. Kidney Int 22:27-35, 1982. 9. Caulin-Glaser T, Gallo GR, Lamm ME: Nondissociating cationic immune complexes can deposit in glomerular basement membrane. J Exp Med 158: 1561-1572, 1983. 10. Sal ant DJ, Belok S, Madaio MP, Couser WG: A new role for complement in experimental membranous nephropathy in rats. J Clin Invest 66:1339-1350, 1980. 11. Vogt A, Batsford S, Rodriguez-Iturbe B, Garcia R : Cationic antigens in poststreptococcal glomerulonephritis. Clin Nephrol 20:271-279, 1983. 12. McIntosh RM, Garcia R, Rabideau D, Rodriguez-ItUl'be B: Evidence for an autologous immune complex pathogenic mechanism in acute poststrepto coccal glomerulonephritis. Kidney Int 14:501-510, 1978. 13. Donaghy M, Rees AJ: Cigarette smoking and lung hemorrhage in glomerulonephritis caused by antibodies to glomerular basement membrane. Lancet 2:1390-1392, 1983. 14. Golbus SM, Wilson CB: Experimental glomerulonephritis induced by in situ formation of immune complexes in the glomerular capillary wall. Kidney Int 16:148-157,1979. 15. Haakenstad AO, Striker GE, Mannik M: The glomerular deposition of soluble immune complexes prepared with reduced and alkylated antibodies and with intact antibodies in mice. Lab Invest 35:293-301,1976. 16. Barnes JL, Venkatachalam MA: Enhancement of glomerular immune complex deposition by a circulating polycation. J Exp Med, to be published. 17. Mauer SM, Sutherland DER, Howard RJ, Fish AJ, Najarian JS, Michael AF: The glomerular mesangium: III. Acute immune mesangial injury : A new model of glomerulonephritis. J Exp Med 137:553-570, 1973. 18. Couser WG: Mesangial IgA nephropathies-steady progress. West J Med 140:89-91, 1984. 19. Shalhoub RJ: Pathogenesis of lipoid nephrosis: A disorder ofT cell function. Lancet 2:556-558, 1974. 20. Couser WG: Idiopathic rapidly progressive glomerulonephritis (Editorial). Am J Nephrol 2:57-69, 1982. February 1965 Volume 269 Number 2