Urinary excretion of kidney antigens in experimental renal diseases of ...

2 downloads 0 Views 2MB Size Report
Jan 14, 1971 - Dishon and Boss, 1969; Rudman, DelRio, Akgun and Frumine, 1969). Small amounts of basement membrane antigens are present in the urine ...
Hr. J. eCx). Path. (1971) 52, 388.

URINARY EXCRETION OF KIDNEY ANTIGENS IN EXPERIMENTAL RENAL DISEASES OF THE RAT E. ROSENMANN, T. DISHON, A. DURST AND J. H. BOSS From the Departments of Pathology and Surgery " B ", Hadassah Medical School, and the Laboratory of Immunology and Virology, Faculty of Dental Medicine, Hebrew University, Jerusalem, Israel Received for publication January 14, 1971

SUMMARY.-Urine samples of healthy rats and animals with experimentally induced nephropathies were tested with precipitating anti-kidney sera, which had been made relatively organ-specific by absorption with diverse tissues. Kidney antigens were not detected in unconcentrated urine of control rats. On the other hand, kidney antigens were demonstrated in the unconcentrated urine of a significant number of animals with toxic nephropathies, produced by mercuric chloride or uranyl nitrate, haemoglobinuric nephrosis, osmotic nephrosis, ischaemic parenchymal damage and acute hydronephrosis. The duration of urinary excretion of renal tissue constituents varied from one animal to another in the same group as well as from one experimental model to another. The described serological test may be of diagnostic and prognostic significance in the evaluation of diseases in which tissue breakdown occurs.

URINARY excretion of tissue constituents in health and disease has beein described in man and experimental animals (Antoine and Neveu, 1968; Dinh, Tremblay and Dufour, 1965; Dishon, Rosenmann, Durst and Boss, 1970; Durst, Dishon, Rosenmann and Boss, 1969; Intorp and Milgrom, 1968; Rosenmann. Dishon and Boss, 1969; Rudman, DelRio, Akgun and Frumine, 1969). Small amounts of basement membrane antigens are present in the urine of normal humans and rabbits (McPhaul and Dixon, 1969a, b). The excretion of basement membrane antigens increases significantly in nephrotoxic serum nephritis (McPhaul and Dixon, 1969b). A kidney-specific antigen, absent from the urine of healthy man and calfs, is excreted in cases of renal tubular necrosis (Intorp and Milgrom, 1968). The purpose of the present communication is to describe certain patterns of urinary excretion kidney-specific antigens in 6 models of induced nephropathies of the rat. MATERIALS AND METHODS

Animals.-Albino rats of both sexes of the Hebrew University strain, weighing 150250 g., and randomly bred local rabbits, weighing about 2 kg., were used. The animals ws ere given regular laboratory chow and water ad lib. Preparation of renal fractions. Rats were exsangtuinated and their viscera and blood collected. The kidneys were homogenized in a Wraring blender in distilled watel, 1 : 10 w/ 3 times for 2 min. The homogenate was centrifuged at 4° at 800 g for 10 mnin. and the sediment was washed twice in distilled water. The supernatant fluid and washings of the 800 g fraction were combined and centrifuged in the cold at 38,000 g for 10 mnin. The precipitate (hereafter referred to as the particulate fraction) was washed twice in distilled water. The sediment, particulate fraction and final supernate of the 38,000 g preparation (hereafter

URINARY EXCRETION OF KIDNEY ANTIGENS

389

referred to as the soluble fraction) were dialysed in the cold against distilled water, lyophilized and stored at -20°. Preparation and testing of antisera.-Rabbits were immunized with each of the 3 renal fractions. One hundred and fifty mg. of lyophilizate suspended in 2 ml. of saline were given i.c. per injection at multiple sites. The first 3 weekly injections contained 2 ml. of complete Freund's adjuvant (Difco). Two further injections were given without adjuvant during the 4th and 5th weeks. The animals were bled by cardiac puncture 10 days after the last injection. Antisera against organ homogenates and fractions thereof contain antibodies to blood components and tissue antigens which are widespread in the organism (Boss, 1965). The antisera to the renal preparations were, therefore, absorbed with lyophilized rat blood and organ pool comprising lungs, hearts, liver and spleens. Twenty ml. of antiserum were incubated with 500 mg. of blood and 500 mg. of organ pool for 2 hr at room temperature and overnight in the refrigerator. Following centrifugation at 10,000 g, the supernatant fluids were tested for blood and non-renal tissue antibodies by Ouchterlony's double immunodiffusion micro-technique in 1 per cent agarose gel in buffered physiological saline, pH 7-2, at room temperature. The plates were observed for 3 days; when absorption was incomplete, the procedure was repeated until no further precipitin lines developed. The 3 absorbed antisera were tested for kidney antibodies to each of the 3 renal preparations by immunodiffusion. One hundred mg. of each lyophilized renal fraction were suspended in 10 ml. of buffered physiologic saline, pH 7-2, and disintegrated by ultrasonic oscillation in the cold in an Ultra-Turrax apparatus, 3 times for 2 min. each, at 170 V., 75 W. and 0 7 A. The supernates were used as the antigenic material for the immunodiffusion tests. Immunofluorescence microscopy.-Kidney and liver samples were rapidly frozen over solid CO2 and 6 ,tm. cryostat sections were incubated with the antisera to the renal fractions, which had been previously absorbed with blood and organ pool, for 30 min. at room temperature in a moist chamber. The sections were washed in cold buffered saline and incubated with fluorescein conjugated goat anti-rabbit y-globulin serum. They were washed and mounted in buffered glycerol. For controls, kidney and liver sections were incubated with normal rabbit serum or anti-renal fraction sera previously absorbed with the corresponding antigenic preparation. The sections were examined with a Zeiss fluorescence microscope, using a BG 12 exciter filter and Sp. yellow plus Sp. blue barrier filters. Urine collection and examination.-Rats were placed in metabolism cages over urine-faeces separators and urine was collected for 6, 16 or 24 hr under a layer of liquid paraffin; food was withheld, but the animals had free access to water. The urine specimens were tested by immunodiffusion against the 3 anti-kidney sera. The plates were observed for 3 days, washed with saline, stained with amido-black and rechecked. Histological examination.-The rats were killed after the last urine collection and the kidneys were removed and fixed in 10 per cent formalin. Sections of paraffin embedded specimens were cut at 6,um. and stained with haematoxylin and eosin and by the periodic acid-Schiff technique. Experimental design.-Urine samples and microscopic sections of the kidneys of the following 7 groups of rats were exmained. Group I. Twenty healthy untreated male and female rats were used as controls. Fifteen urine samples of each animal were tested. In further experiments, urine specimens of 10 rats were concentrated by pervaporation to 1/5, 1/10 and 1/20 of their original volume and examined as described. Group II. Forty-nine rats were injected s.c. with 0-4-0-6 mg. of mercuric chloride per 100 g. of body weight (Rodin and Crowson, 1962). Urine was collected on consecutive days for 2 weeks; 3 or 4 animals were killed on each day. Group III. Forty-four rats were injected i.p. on 2 consecutive days with 0-2-0-3 mg. of uranyl nitrate per 100 g. of body weight (Magee and Foreman, 1958). Urine was collected daily for 2 weeks and 3 or 4 animals were killed on each day. Group I'V. Thirty-nine rats, deprived of water for 24 hr, were injected i.m. with 0-7-0-8 ml. of a 1: 1 glycerol-in-water mixture per 100 g. of body weight (Thiel, Wilson, Arce and Oken, 1967). Urine was tested from the 1st 9th day following the injection; 4 rats were killed on each day. Group V. Thirty-five rats were injected i.p. with 2-0-3-5 ml. of a 50 per cent sucrose solution per 100 g. of body weight (Allen, 1952). Urine was collected for 8 days; 4 rats were killed on each day.

390

E. ROSENMANN, T. DISHON, A. DURST AND J. H. BOSS

Group VI. Midline laparotomy was performed on 20 rats under pentobarbital anaesthesia. The left renal artery was dissected, isolated from the vein and occluded for 30 min. by a free tie; a loop of 000 silk ligature, passing around the artery and through a plastic tube (290 gauge) was tightened down to the vessel and held in place by a clamp. Ischaemia was visualized within a few min. as cyanotic discoloration of the kidney rapidly changing to a diffuse palour. Re-establishment of the blood circulation followed removal of the loop and the colour of the kidney returned to normal. The operation wound was closed in layers. Urine samples were tested for 4 consecutive post-operative days. Group VII. The distal third of the left ureter of 23 rats was ligated. Urine was collected from the 6th-lOth postoperative days. A 2nd laparotomy was performed on the 11th day and urine was aspirated from the hydronephrotic sac. R1ESULTS

The absorbed anti-renal fraction sera did not react with antigens of rat blood, lung, heart, liver and spleen. One, 2 and 3 precipitating antibodies against renal tissue components were found in the antisera to the renal sediment, particulate and soluble fractions, respectively. Following absorption of each antiserum with the 2 non-corresponding renal preparations, 1 and 2 antibodies were still demonstrable in the anti-particulate and anti-soluble fraction serum, respectively, but none was detectable in the anti-sediment serum. Immunohistologically, kidney sections incubated with normal rabbit serum or antisera absorbed with their homologous renal preparations and liver sections incubated with the anti-renal fraction sera followed by fluorescein labelled antirabbit y-globulin serum were devoid of specific fluorescence. Staining of kidney sections which were overlaid with the anti-sediment or anti-particulate fraction serum produced a similar pattern, the cell membranes of the tubular epithelium and brush border of the proximal tubules being specifically fluorescent (Fig. 1).

EXPLANATION OF PLATES FIG. 1. (a) Kidney section incubated with anti-particulate fraction serum and labelled antirabbit y-globulin serum showing specific fluorescence of the cell membranes and brush border of the tubular epithelium. x 200. (b) Non-specific fluorescence of the tubular epithelium in control section incubated with normal rabbit serum and conjugated antiserum. x 200. FIG. 2.-Two precipitin lines between anti-particulate fraction serum (centre well) and urine of rat with mercuric chloride poisoning (A). Precipitin lines not present with urine of another rat with mercuric chloride poisoning (B) and specimens of healthy control rat (C) as well as animals with uranyl nitrate poisoning (D) and osmotic nephrosis (E and F). FIG. 3.-Reaction of identity elicited by anti-soluble fraction serum (centre well) with urine specimens of rats with mercuric chloride poisoning (A), haemogobinuric nephrosis (C) and ischaemic parenchymal damage (D). Precipitation bands did not develop with urine of rats with mercuric ehloride poisoning (B) and osmotic nephrosis (E) as well as with urine of control animals (F). FIG. 4.-Precipitation pattern of anti-soluble fraction serum (centre well) with specimens from urinary bladder (A) and hydronephrotic sac (B) of rat with unilaterally obstructed ureter. Note reaction of identity with urine of rats with osmotic nephrosis (C) and mercuric chloride poisoning (F). No lines are seen opposite wells containing urine of healthy control animals (D) and rats injected with mercuric chloride 11 days prior to collection of sample (E). FIG. 5.-Anti-soluble fraction serum (centre well) giving two precipitin lines with urine of rat with uranyl nitrate poisoning (A). One of these bands shows a reaction of identity with a specimen of an animal with haemoglobinuriC nephrosis (E). Precipitin lines did not develop. in the illustrated instances, with urine of rats with osmotic nephrosis (B), uranyl nitrate poisoning (C) and mercuric chloride poisoning (D) as well as of control rat (F). FIG. 6.-Irregular tubular necrosis and cellular debris in lumina of tubules on the 4th day following two injections of uranyl nitrate. H. and E. x 260. FIG. 7.-Focal tubular necrosis and pigmented casts on the 3rd day following an intramuscular injection of glycerol. H. and E. x 105.

BRITISH JOT-TRNAL

OF

EXPERIMENTAL R'ATHOLOGY.

Vol. LII, No. 4.

( I.\ -1-11..

.. .,

I

...

I

.f

Jr

..

..

:.

.:

:

s. #

:.\

.. ."

*. *.s

'Zf

.R

wo:,.

:.: .z.

,\.,-

Rosenmann, Dishon, Durst and Boss.

BRITISH JOURNAL

OF

EXPERIMENTAL PATHOLOGY.

VOl. LIII, NO. 4

-i

U..

W"f" Asi

IL

aI

.

rn -_

-

0v---m

".

kA

"OP;il

'j.t%

I Rosenmann, Dishon, Durst and Boss.

35

391

URINARY EXCRETION OF KIDNEY ANTIGENS

Sections treated with the latter serum also exhibited weak staining of the mesangial matrix and/or cells of the glomeruli. There was no specific fluorescence of the epithelial and vascular basement membranes, consistent with previous findings that absorption of anti-kidney serum with lung homogenate removes antibodies to the basement membrane antigens (Weinberger and Boss, 1966). Strong diffuse specific fluorescence of the cytoplasm of the tubular epithelial cells was obtained with the anti-soluble fraction serum, which also stained weakly the mesangial matrix and/or cells of the glomeruli. Any urine specimen giving one or more precipitin lines with the anti-renal fraction sera was scored positive; representative patterns are illustrated in Figs 2-5. Positive urine samples of rats with different renal diseases placed in adjacent wells gave reactions of complete identity, indicating that the same antigens were excreted in the urine. The results of urine examinations in Groups II-V are summarized in the accompanying Table. Several rats of Groups II-IV were anuric for 1 or 2 days during the first week. TABLE.-Per cent of Urine Specimens Containing Kidney Antigens in Rats with Experimental Renal Diseases Days following injection

Mercuric chloride

poisoning

Uranyl nitrate

poisoning . 1 55 (12) 91 * 2 31 (26) . . 88 (12) 27 (15) . 3 61 (33) 4 . . 13 63 (3) . 42 (12) . 5 64 (18) 6 17 (17) . . 79 (4) 7 . 16 . 70 (20) 8 5 . . 67 9 0 67 . . 10 . 0 33 11 50 * * . . 12 29 Numbers in parenthesis indicate per cent of anuric rats.

Haemoglobinuric nephrosis . 67 (28) 51 (23) . 68 (13) 33 (15) * . .

.

30 32 13 9 0 0

Osmotic

nephrosis . . . . . . . .

26 35 44 35 32 7 0 0

Group I.-Three hundred urine samples of healthy control animals were negative when tested with the anti-kidney sera. Whereas all 10 urine specimens concentrated 5- and 10-fold were negative, 4 of 10 samples concentrated 20-fold gave one faint precipitin line with the anti-soluble fraction serum. Group II.-The urine of about one half of the rats with mercuric chloride poisoning was positive on the first day. A gradual decline in the incidence of positively reacting specimens occurred during the following week. The pattern of antigen excretion in individual rats was generally constant on consecutive days. The anti-soluble fraction serum gave 1-3 precipitation bands. Antigens were less frequently detected with the anti-particulate fraction and anti-sediment sera. Histologically, severe tubular degeneration and necrosis were evident on the day following injection. Regeneration of the tubular epithelium began on the 2nd day. The parenchyma was practically restituted by the 10th day. Group III.-In rats with uranyl nitrate poisoning, the urine was positive in most cases for the first 3 days. Though the incidence declined thereafter, positive urine findings persisted in some animals till the end of the observation period. Formation of 1-3 and 1 or 2 precipitin lines was found with the anti-soluble and anti-particulate fraction sera, respectively. On the first 2 days, the kidneys showed

392

E. ROSENMANN, T. DISHON, A. DURST ANI) J. H. BOSS

moderate to severe tubular degeneration and necrosis (Fig. 6). The course of the disease was characterized by regeneration of the tubular epithelium, focal calcification and interstitial inflammation and fibrosis. Degenerative and atrophic alterations of the tubular epithelium were still evident by the end of the 2nd week. Group I V. Most urine samples of rats with glycerol induced haemoglobinuric nephrosis reacted positively during the first 3 days. The incidence thereafter declined and all specimens were negative by the 9th day. A single precipitin line developed with the anti-soluble fraction serum. Microscopical examinatioll disclosed severe cortical tubular degeneration and necrosis (Fig. 7). The collectilng tubules were plugged by proteinaceous and pigmented casts. Group V.-About one half of the urine samples of rats with sucrose induced osmotic nephrosis formed one precipitin line with the anti-soluble fraction serum; only in small proportion of cases was a line evident with the anti-particulate fraction serum. By the 7th day all specimens were negative. Histologically, there was widespread vacuolar degeneration of the proximal tubular epithelial cells. Signs of regeneration were present by the 3rd day. Group VI.-The urine of 7 rats reacted with all 3 anti-renal fraction sera onl one or more examinations during the 4 days following ischaemic parenchymal damage. At the time of death, there were widespread degenerative anld discrete regenerative alterations in the tubules. Group VII.-Three positive specimens were obtained by testiing the contenlt of the hydronephrotic sacs. Urine collected from the bladder prior to the seconld laparotomy was negative. DISC TSSION

This investigation demonstrates the presence of kidney-specific alntigenis in the urine of rats with experimentally induced renal tubular damage. The anltirenal fraction sera employed in the immunodiffusion tests do not react withl antigens of the lung, heart, liver and spleen. Though such antisera are ofteintimes referred to in the pertinent literature as organ-specific, this specificity is but a relative one. On the one hand, not all organs have been tested with the aintisera and, on the other hand, antigens may be present in certain organs in amounts below the threshold of the sensitivity of the employed technique. It is for these reasons that the anti-renal fraction sera should be considered as relatively kidneyspecific only. These antisera, which are used in the detection of antigens in the urine, have also been employed to localize, immunohistologically, the respective tissue components in kidney sections. In agreement with the observations of other investigators (Nairn, Ghose, Fothergill and McEntegart, 1962), we have found that the antigens revealed by our antisera reside in the cytoplasm, cell membranes and brush border of the tubular epithelium and in the mesangial matrix and/or cells of the glomeruli. These antigens are not detectable in uncollcentrated urine of healthy rats. In the experimental models studied hereini. which primarily involve the tubular apparatus, damage to and probably breakdown of tubular cells seem to be responsible for liberation of cellular antigens into the urine. It may, therefore, be concluded that demonstration of these antigens in the unconcentrated urine of rats is indicative of kidney disease. Since the same antigens are demonstrable in the tubular cells and mesangial mnatrix and/or cells, a positive test does not differentiate between tubular and glomerular damage. Tissue antigens which are widespread in the organism and shared by

URINARY EXCRETION OF KIDNEY ANTIGENS

393

many organs, e.g. components of basement membranes, connective tissue and cell organelles, are also excreted in the urine in significant amounts under various pathological conditions (Antoine and Neveu, 1968; McPhaul and Dixon, 1969a). Their detection in the urine would obviously not suggest a disorder of one organ or another and, hence, relatively organ-specific antisera are required for diagnostic purposes. Proteins excreted in the urine of healthy animals are of blood plasma and tissue origin (Sellers, 1956; Grant, 1959). Examples of the latter are the TammHorsfall protein (Tamm and Horsfall, 1950), basement membrane antigens (McPhaul and Dixon, 1969a) and constituents of the male accessory genital organs (Rosenmann et al., 1969). As early as the 1930s, Gillman (1935) demonstrated renal components in the urine of patients with glomerulonephritis. His findings were corroborated and extended by other investigators, who showed the excretion of renal tissue constituents in nephrolithiasis (Boyce and King, 1963), tubular disorders (Manual, 1966), toxic nephropathies (Intorp and Milgrom, 1968), lupus nephritis and acute rejection of transplanted kidneys (Antoine and Neveu, 1968). Increased urinary excretion of basement membrane antigens was noted in nephrotoxic serum nephritis (McPhaul and Dixon, 1969b). Thus, it appears that under pathological conditions there is a rise in the amount of renal antigens which are normally present in the urine in minute quantities, while antigens normally absent from the urine make their appearance. The common denominator of the experimental models of rat kidney diseases studied, i.e. toxic nephropathies, haemoglobinuric nephrosis, osmotic nephrosis, ischaemic parenchymal damage and hydronephrosis, is degeneration and necrosis of the tubular epithelium (Allen, 1952; Magee and Foreman, 1958; Rodin and Crowson, 1958; Sheehan and Davies, 1959a, b; Thiel et al., 1967). Kidney antigens are detectable in the unconcentrated urine in each of these disorders, but the duration of antigen excretion, number of excreted antigens as judged by the immunodiffusion test and incidence of positive urine findings vary from one animal to another within the same group as well as from one model to another. It is noteworthy that in uranyl nitrate poisoning, which presents a chronic disease pattern (Black, 1967), urinary excretion of renal antigens often persists for 10 or more days. Though the tubular epithelial changes in sucrose induced osmotic nephrosis are predominantly degenerative in nature (Allen, 1952), evidence indicative of cellular breakdown has been produced (Jannigan and SantaMaria, 1961). Our results confirm this contention by demonstrating kidney constituents in the urine. The main outcome of acute hydronephrosis is tubular atrophy, but vascular impairment with ensuing focal tubular necrosis apparently plays a role as well (Sheehan and Davies, 1959b). The presence of kidney antigens in urine aspirated from hydronephrotic sacs is in favour of such an assumption. The technique used in the present investigation does not distinguish between different types of nephropathies, since similar or identical renal antigens are present in significant amounts in the urine of rats with various tubular injuries as well as nephritides (Dishon, Rosenmann, Durst and Boss, unpublished). On the other hand, the detection of kidney antigens in the unconcentrated urine indicates renal damage, which, in the present experiments, involves mainly the tubular apparatus, as the 6 models were chosen with this view in mind. Furthermore, the continued activity of the underlying process may be assessed by daily examination of urine samples. Tissue antigens of a number of organs are excreted in the urine under

394

E. ROSENMANN, T. DISHON, A. DURST AND J. H. BOSS

physiological and pathological conditions (Antoine and Neveu, 1968, Grant, 1959, McPhaul and Dixon, 1969a, Rosenmann et al., 1969). The diagnostic value of the test depends, therefore, upon the organ-specificity of the employed antisera. This investigation was supported by the S. Lunenfeld and R. Kunin Medical Research Foundation. REFERENCES ALLEN, A. C.-(1952) 'The Kidney, Medical and Surgical Diseases'. New York (Grune and Stratton), p. 324. ANTOINE, B. AND NEvEu, T.-(1968) J. Lab. clin. Med., 71, 101. BLACK, S. A.-(1967) 'Renal Diseases', 2nd edition. Oxford (Blackwell), p. 58!. Boss, J. H.-(1965) Am. J. Obstet. Gynec., 93, 574. BOYCE, W. H. AND KING, J. S.-(1963) Ann. N.Y. Acad. Sci., 104, 563. DINH, B. L., TREMBLAY, A. AND DUFOUR, D.-(1965) J. Immun., 95, 574. DiSHON, T., ROSENMANN, E., DURST, A. AND Boss, J. H.-(1970) Am. J. Physiol., 219, 92. DuRST, A., DISHON, T., ROSENMANN, E. AND Boss, J. H.-(1969) Experientia, 25, 1052. GILLMAN, G.-(1935) J. Urol., 34, 727. GRANT, G. H.-(1959) J. clin. Path., 12, 510. INTORP, W. H. AND MILGROM, F.-(1968) J. Immun., 100, 1195. JANNIGAN, D. T. AND SANTAMARIA, A.-(1961) Am. J. Path., 39, 175. MAGEE, M. AND FOREMAN, H.-(1958) Am. J. Physiol., 195, 354. MANUAL, Y.-(1966) Abstracts of the IlIrd International Congress of Nephrology, free communications, Washington, D.C., p. 236. MCPHAUL, J. J. AND DIXON, J. F.-(1969a) J. exp, Med., 130, 1395.-(1969b) Abstracts of the IVth International Congress of Nephrology, symposia, 29. NAIRN, R. C., GHOSE, T., FOTHERGITL, J. E. AND MCENTEGART, M. G.-(1962) Nature, Lond., 196, 385. RODIN, A. E. AND CROWSON, C. N.-(1962) Am. J. Path., 41, 297. ROSENMANN, E., DISHON, T. AND Boss, J. H.-(1969) J. Lab. clin. Med., 74, 31. RUDMAN, D., DELRIO, A., AKGuN, S. AND FRUMINE, B.-(1969) Am. J. Med., 46, 174. SELLERS, A. L.-(1956) Archs intern. Med., 98, 801. SHEEHAN, H. L. AND DAVIS, J. C.-(1959a) J. Path. Bact., 78, 351.-(1959b) Archs. Path., 68, 185. TAMM, I. AND HORSFALL, F. L.-(1950) Proc. Soc. exp. Biol. Med., 74, 108. THIEL, G., WMSON, D. R., ARcE, M. L. AND OKEN, D. E.-(1967) Nephron, 4, 276. WEINBERGER, N. J. AND Boss, J. H.-(1966) Pathologia Microbiol., 29, 324.