- Email: [email protected]
Urine Autoantibodies in Interstitial Cystitis
0022-5347/97/15731083$03.00/0
DIEJOURNAL
Vol 157. 1083-1087,March 1997 Printed in U.S.A.
OF UR0LCG.Y
Copyright 0 1997 by AMERICAN U R O ~ I C ASS~CIATION, AL INC.
URINE AUTOANTIBODIES IN INTERSTITIAL CYSTITIS S. m A Y , * C.-0. ZHANG, A. L. TRIFILLIS, J. R. HEBEL, S. C. JACOBS AND J. W. WARREN From the Research Service, Department of Veterans Affairs Medical Center, Baltimore, Maryland. and the Division of Infectious Diseases, Department of Medicine, Department of Pathology, Department of Epidemiology, and Division of Urology, Department of Surgery, University of Maryland School of Medicine. Baltimore, Maryland
ABSTRACT
Interstitial cystitis is a chronic bladder disease with certain features that suggest autoimmunity may play a role in initiating or maintaining the disease process. We therefore determined whether immunoglobulin fractions from 14 IC patient and 19 control urine specimens bound in vitro to primary cultures of human bladder epithelial cells, as well as epithelial cells from a variety of other tissues. Urine autoantibodies t h a t bound to normal human bladder epithelial cells were present in 8 of 14 IC specimens (from 6 of 9 IC patients) as compared to 3 of 23 control specimens (from 2 of 17 control patients). These antibodies, which were usually also present at low titers in sera from these persons, bound to at least four nuclear or cytoplasmic antigens, with the specificity of autoantibodies from a given individual varying over time. The autoantibodies were not specific for normal or malignant bladder epithelial cells, but bound to epithelial cells from a variety of tissues. These data show that anti-epithelial cell autoantibodies are present in the urine of IC patients, but suggest that these antibodies are not likely to be a primary cause of this disease. KEY WORDS:cystitis, autoantibodies
Interstitial cystitis is a chronic inflammatory condition of the bladder which affects a t least 20,000-90,000 women in the United States at any given time; approximately onetenth as many men also suffer from this condition.’ This disease, which usually affects people between 30 and 59 years old, causes pain and increased frequency and urgency of urination.l.2 Histopathologic changes evident in IC include submucosal inflammation along with thinning or ulceration of the bladder epithelium.3 No etiology for IC has yet been identified, and no empiric treatment has proven to be generally efficacious for this disorder. Certain aspects of IC suggest that autoimmunity may play a role in initiating or sustaining the chronic inflammatory response evident in this disease.4 Published data indicate that IgM antibodies and complement (C’)have been detected in areas of endothelial cell damage5-7 and IgA antibodies have been detected on the surface of bladder epithelium in IC patients: but the antigenic specificity of these antibodies is unknown. Autoantibodies have been detected in the sera of IC patients to a greater extent than in control^^-^^ but the specificity of these autoantibodies has differed between reports. Whereas one report indicated the presence of antibladder epithelial cell antibodies in the sera of IC patients? other reports have shown no evidence for anti-bladder epithelial cell antibodies12 or have indicated the presence of serum autoantibodies which are not specific for bladder epithelial cells but bind to other cell types,lO.ll intermediate filaments,7 or Tamm-Horsfall protein.l5 In one report, serum autoantibodies in IC patients were directed against a variety of cell proteins which were notably distinct from those recognized by autoantibodies in patients with systemic lupus erythematosis or systemic sclerosis.1° In addition to serum, urine is itself a potential source for autoantibodies which could bind directly to bladder or other urinary tract epithelial cells. We hypothesized that urine
autoantibodies might be important in IC because of their potential to cause direct damage to the bladder epithelium (for example, by causing complement-mediated cytolysis), andor to serve as possible diagnostic tools. We therefore screened both the urine immunoglobulin fraction and whole serum from IC patients and their age, race, and sex-matched controls for binding to bladder epithelial cells (primary adult normal human bladder epithelial cells, T24 bladder carcinoma cells, and cytospun epithelial cells from IC patient urine specimens) as well as epithelial cells from other tissues (A549 human lung carcinoma cells; HEp-2 human laryngeal epidermoid carcinoma cells; and Hep G2 human hepatocarcinoma cells). MATERIALS A N D METHODS
Patients. IC patients were referred by physicians, the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), and the Interstitial Cystitis Association. All patients had previously undergone diagnostic cystoscopy, and fulfilled the NIDDK diagnostic criteria for IC.16 Eight of these patients had evidence for glomerulations at cystoscopy, while one had both Hunner’s ulcers and glomerulations. The average duration of symptoms was 7.4 years (range 1.9 to 14.1 years). Age, race, and sex-matched controls were chosen from volunteers with no history of IC or other urological disease. Urine was initially collected from these patients and their controls at least three months following the most recent known bacterial urinary tract infection and one month following the last antibiotic use. All participating patients were determined to be symptomatic at the time of enrollment; additional specimens were also obtained during periods of IC symptom exacerbation from the patients and their available controls. All participants were a t least 18 years old and enrolled in accordance with guidelines of the Institutional Accepted for publication July 1, 1996. Review Board at the University of Maryland School of Med* Requests for reprints: Veterans Administration Medical Center, icine. The number of IC patients or controls used for each Room 3B184, 10 North Greene Street, Baltimore, MD 21201, Supported by the National Institute of Diabetes and Digestive and experiment was based on the number for whom a sufficient quantity of urine was available for each experiment. HowKidney Diseases, NIH (#R01 DK44818). 1083
1084
URINARY ANTIBODIES IN INTERSTITIAL CYSTITIS
ever, each IC patient urine specimen was studied simulta- Following 3 additional rinses with PBS, SloFade (Molecular neously with urine from one or two age-, race- and sex- Probes, Eugene, OR) was added to the wells and the cells were examined using a Nikon Diaphot inverted phase flu& matched controls for each experiment. Urine and serum specimens. Urine specimens for immuno- rescence microscope, A negative control well was incubated globulin fractionation were collected by t h e clean catch with secondary antibody alone. Statistical analysis. Comparisons of the proportion of Ic method, in which each patient wiped t h e labial area with 10% povidone iodineititratable iodine 1% solution (Clinidine, patient and control specimens that contained anti-epithelial Guilford, CT) then collected a midstream urine into a sterile cell antibodies were made using Fisher’s exact test. container. Serial specimens were obtained over a 24 hour RKSII LI‘S period from each of 9 IC patients ( 14 specimens) and 18 age-, race-, and sex-matched controls (24 specimens), and placed in Binding of IC serum and urine auloantihodics to normal a refrigerated container. Specimens were kept at 4C from the adult human bladder epithelial (HBEJ cells. Ammonium sulbeginning of collection, and transported to the laboratory fate precipitation of immunoglobulins from both IC and conwithin 24 hours of completion; antibodies were precipitated trol urine was shown to result in highly enriched immunofrom each specimen within 6 hours of delivery to the labora- globulin preparations by polyacrylamide gel electrophoresis tory by adding a n equal volume of saturated ammonium (PAGE1 under denaturing conditions (presence of bands corsulfate (pH 7.8) on ice (see below). Sera were collected from 7 responding to molecular mass of immunoglobulin heavy and IC patients and 14 age, race and sex-matched controls and light chains only) (data not shown). The amount of immunostored at -2OC until used. globulin recovered ranged from 0.12-1.3 mg./ml.; there was Human epithelial cells. T24 bladder carcinoma cells (ATCC no significant difference in the mean amount of total immu#4-HTB) (American Type Culture Collection, Rockville, MD) noglobulin between IC patients 10.65 mg./ml.) and controls were grown in McCoy’s medium containing 10% fetal bovine (0.54 mg./ml.). serum (FBS), 1%antibiotidantimycotic solution, and 15%gluAnti-bladder cell~antibodies bound to nuclear or ._ . ~ . ~ ~ which ~ ~ . ~ tamine (Sigma, St. Louis, MO). Normal adult human bladder cytoplasmic antigens in normal adult human primary bladepithelial cells (HBE) were grown from biopsies obtained at der epithelial cells were evident in the urine of 8 of 14 autopsy from patients who had no history of bladder disor- specimens from 6 of 9 IC patients, and 3 of 23 specimens from der.” The explanted cells were grown in Eagle’s minimal 2 of 17 controls. Four of these 9 IC patients (44‘;~ were essential medium (MEM) containing 10% heat inactivated positive for urine autoantibodies a t the first visit 1 # 3 ,#4, #5 FBS, 1%antibiotidantimycotic solution, 1% glutamine, and and #6)as compared to 2 of 17 controls ( p = .14 ). Binding was 1.0 pgiml insulin (all from Sigma). Hep G2 human hepato- evident when urinary immunoglobulin was present a t 0.3cellular carcinoma cells (ATCC #HB 8065) and A549 human 0.03 mg./ml. (data are shown for cells incubated with 0.03 lung carcinoma cells (ATCC #CCL 185)were grown in MEM mg./ml. immunoglobulin). The antibodies from 5 of the 8 containing 10%FBS, 1% antibiotidantimycotic solution and positive IC specimens and both of the positive control speci1% glutamine. All cells were cultured at 37C in a 5% CO, mens bound to the bladder epithelial cells in a fine, diffuse atmosphere. staining pattern (Figure l C ) , while antibodies from 3 IC Urine specimens for uroepithelial cell cytospin prepara- patients had a coarse pattern (Figure 1D1, one had a nucletions were collected by the clean catch method (each patient olar pattern (Figure l B ) , and one had a cytoplasmic binding wiped the labial area with 10% povidone iodineltitratable pattern (Figure 1A) ( 2 patients had 2 types of binding obiodine 17c solution (Clinidine, Guilford, CT), then collected a served). These findings suggest that a t least four different midstream urine into a sterile specimen container); these antigens serve as targets for the anti-bladder epithelial cell were kept a t 4C and transported to the laboratory within one antibodies in IC urine. The use of secondary antibodies spehour of collection. Cytospun cells were prepared as previ- cific for IgG, IgM. or IgA indicated that each positive IC or ously described.19 control urine specimen contained autoantibodies of the IgG Ammonium sulfate precipitation of urinary immunoglobulins. Equivalent volumes of saturated ammonium sulfate and patient urine specimen were mixed together at 4C, placed on ice for 30 minutes, and centrifuged at 4C. The pellet was then resuspended in 0.9% NaCl to half of original urine volume, additional saturated ammonium sulfate was added to this mixture which was stirred a t 4C for one hour and centrifuged at 4C. The second pellet was resuspended in 10 ml. borate buffered saline (BBS), dialyzed against BBS overnight at 4C, then dialyzed against phosphate buffered saline (PBS) at 4C, aliquoted and stored at -8OC until used. The amount of protein in each specimen was determined by the Lowry assay18 with known concentrations of bovine serum albumin as the standard. Immunofluorescence assay. Cells which were grown to confluence on 8 well LabTek tissue culture slides (Nunc, Inc., Naperville, IL), or cytospun preparations of uroepithelial cells from the urine of IC or control patients (prepared as previously described’g), were fixed with acetone/ethanol (1: 1); these cells or fixed HEp-2 cells (purchased from Zeus Scientific, Raritan, N J ) were then incubated in duplicate with varying dilutions of serum (0-1,000) or urinary antibodies (0-0.03 mg./ml.) in PBS. Following 2 hours incubation at 37C, the cells were rinsed with PBS and incubated further Urinary autoantibody binding patterns in IC. Normal adult huat 37C for 2 hours with FITC-conjugated goat anti-human Ig man bladder epithelial (HBE) cells were fixed and incubated with (Southern Biotechnology Associates, Birmingham, AL),or urinary immunoglobulins from IC patients followed by FITCconjugated goat anti-human Ig, as described in Methods. Shon.n are FITC-conjugated goat anti-human antibodies specific for four binding patterns encountered: A ) cytoplasmic; B I nucleolar; C) I@, IgM, or IgA (Vector Laboratories, Burlingame, CA). diffuse nuclear; D) coarse nuclear. Final magnification = 400 x .
URINARY ANTIBODIES IN INTERSTITIAL CYSTITIS
class; 2 IC specimens also contained IgM autoantibodies (#5 and #6a) and 2 also contained IgA autoantibodies (#7 and #8a). Eleven serum specimens from IC patients (7 of whom had urinary autoantibodies) and 16 control specimens (2 of whom had urinary autoantibodies) were also tested. Each patient and control who had urinary autoantibodies also had serum autoantibodies with the same binding pattern; one additional IC patient and one additional control had serum antinuclear antibodies which bound to normal adult human bladder epithelial cells in a diffuse manner. The titer of serum autoantibodies in the IC patients was generally low, with 2 of the 8 positive specimens having a titer <1:40, and only 2 patients having a titer >1:100. For both IC patient urine and serum autoantibodies, there was no obvious correlation between the presence or pattern of antibody binding and duration of symptoms, with patients #1 and 2 having symptoms for 1.9 and 2.3 years prior to the first collection, but patient #8 having symptoms for 14 years at the time of the first collection. Four of the 9 patients returned to give another urine specimen at the time of symptom exacerbation (#la, #3a, #4a, and #8a), but there was also no obvious correlation between an increase in symptoms and antibody positivity or pattern. Two of the IC patients ( # 1 and #8) were initially negative for autoantibodies, but became positive during their exacerbations; one patient (#3) was initially positive but became negative, and another patient (#4j was positive a t both visits but changed in pattern. Binding patterns displayed by these patients during exacerbation also differed from one another (1cytoplasmic, 1coarse nuclear and 1 diffuse nuclear). Serial specimens were also obtained on a fifth patient (#6), although the timing of collection was not correlated with symptom severity; interestingly, the pattern of fluorescence also appeared to change between visits for this patient. Because the IC patients who returned for a second visit often returned with little or no notice during a period of symptom exacerbation, only one second simultaneous repeat control specimen was obtained (#7a). However, we obtained an additional 5 repeat control urine specimens a t random time points, one of which was positive on both visits for diffuse nuclear binding; like #7, the other 4 of these controls were negative on both visits. A comparison of data obtained from the IC patients and these controls indicated that the IC patients were more likely to be positive on one or both visits ( p = .015). Serum and urine autoantibody binding to cells from epithelial malignancies. To determine whether the autoantibodies in IC patient urine and serum specimens were specific for normal adult human bladder epithelial cells, specimens were also tested for binding to T24 human bladder carcinoma cells, as well as Hep G2 human hepatocellular carcinoma cells and A549 human lung carcinoma cells (Tables 1 and 2). Six of the IC urine specimens that had detectable antibodies against HBE cells were also tested for binding to T24 cells; each of these had antibodies with the same binding pattern on T24 cells, indicating that primary bladder epithelial cells are not required for the detection of these antibodies. Antibodies from four of these six specimens also bound to Hep G2 or A549 cells, indicating that a t least some antibodies in IC urine are not specific for cells of bladder origin. In addition, urine antibodies from 5 specimens that were positive for anti-bladder epithelial cell binding and 3 specimens that were negative were tested for binding to HEp-2 human laryngeal carcinoma cells; all 5 of the positive specimens showed evidence for binding in the same pattern displayed on bladder cells ( 3 coarse nuclear and 2 diffuse nuclear), and all 3 negative specimens were negative on these cells. Binding of urine and serum antibodies to cytospun urine cells. To determine whether urine or serum antibodies might display increased or differential binding to autologous epithelial cells, urine specimen cytospin preparations were
TABLE1. Binding
1085
of urinr autoantrhodies t o norrrial epilhelral and
carcrriomn cells Other Epithelial ('ella
Bladder Cells
HBE
T24
Hep (2
A649
~
IC Patients 1 la* 2 3 3a 4 4a 5 6 6a 7 8
8a 9 Controls
1 2 3 4 5 6 7 7a 8 9
-
+Cy"
f
-
DN
I
T
('y
-
DN
f
f
CY DN
+ DN
-
-
-
+ Ncl/DN
n.d. + CN tCN n.d. + CN + DN n.d. + DN n.d.
n.d.
+ CN +CNIDN DN +CN T
-
+ DN -
+ CN
-
n.d.
-
+ DN n.d.
+ DN n.d.
-
+ DN
-
+ DN
10 11 12 13 14 15 16 17 18 * a = results of specimen obtained on a return visit from the same patient. "Cy = cytoplasmic; DN = diffie nuclear; CN = coarse nuclear; Ncl = nucleolar. n.d. = not done.
made on 2 patients (#3 and #4) and their controls (#13,#14, #16 and #17). Approximately 20-35 cells were visible on each slide. Although both patients and one control had evidence for urine and serum antibody binding to cultured bladder cells, only one patient (#4) showed urine antibody binding to her own cytospun epithelial cells, in the same pattern shown on primary HBE and T24 cells; all sera were negative. DISCUSSION
This report presents the first evidence that autoantibodies are present in the urine of IC patients, and shows that these urine autoantibodies are similar in their binding patterns to serum autoantibodies from these patients. Our findings also confirm an earlier report by Ochs, et a1.,10 which described a similar prevalence and variety of IC serum autoantibody binding patterns with HEp-2 cells. Like that study, o u r study found that autoantibodies with a fine speckled pattern of nuclear fluorescence were the most prevalent, with 62% of positive urine immunoglobulin specimens and 75% of positive sera from IC patients having evidence for this type of binding pattern with HBE cells. In addition, our data indicate that neither urine nor serum autoantibodies from IC patients are specific for human bladder cells, nor are they more evident when normal human bladder cells or autologous cytospun cells from the IC patients are used. This nonspecificity is in agreement with previously reported studies which indicate that the serum anti-nuclear or anti-cytoplasmic autoantibodies in IC are generally not specific for a particular cell type.'*.'' Along with the apparent nonspecificity of these autoantibodies, the lack of correlation in the current study between symptom
URINARY ANTIBODIES I N INTERSTITIAL CYSTITIS
1086 TABLE2. Binding
of
serum autoantibodies to normal epithelial and carcinoma cells ___ Other Ep~rhelialCells
Bladder Cells
--
~
HBE
-
T24
.
A549
IC Specimens 1 la' 2 3 4 4a 5 6
-
- DN - NcliDN
nd +Ncl!l)N -CN
+DN NcV- DN + CN DN 7
~
7
8 Ha
DN
- DN
Control Specimens 2 3 4
5 6 ia
8 9 10 11 12 13 15 16 17 18
'a
results of specimen obtained on a return visit from the same patient. c.ytoplasmic: DN = diffuse nuclear; CN = coarse nuclear; Ncl = nucleolar. n.d. = not done.
occurs as a result of chronic inflammation in I c may stimulate the production of autoantibodies directed against several nuclear and cytoplasmic antigens. However, molecular mimicry can also play a role in initiating autoantibody responses, with the first antibody directed against an epitope on a foreign antigen that is homologous to a self antigen. The subsequent antibody response may spread to include other epitopes on the self antigen."7,2* This type of autoimmune stimulation is thought to play a role in some cases of inflammatory bowel disease, in which systemic sensitization to bacterial antigens can lead to ongoing, chronic large-bowel inflammation and in which colonic epithelial cell antigens have been shown to cross-react with bacterial cell wall antigens."" We previously reported an increased incidence of a variety of bladder microorganisms in IC patients a s compared to controls,'" suggesting that urinary tract organisms might be a possible source of foreign antigen that could initiate this type of autoantibody response in the IC patient. Additional studies to determine whether bacterial antigen(s) on microorganisms present in the urine of IC patients crossreact with epithelial cell nuclear or cytoplasmic antigens may be helpful for further determining the mechanism of this disease process. Acknowledgments. The authors thank Edward W. Campbell, Jr. M.D., Michael J. Naslund, M.D., Ms. Linda Horne and Terry Wilson, R.N. for their assistance with patient enrollment and specimen acquisition; Marcia Fullem for technical assistance; and Barbara White, M.D. for helpful discussions.
=
" Cy
=
duration or exacerbation and presence or pattern of urine autoantibodies, as well as the generally low titer of the serum autoantibodies, suggest that these antibodies are probably not causally related to the epithelial cell damage evident in IC. However, prospective studies with defined criteria for disease and symptom severity may be helpful for determining the role of serum and/or urine autoantibodies in IC. Like autoantibodies present in patients with various other conditions characterized by chronic inflammation, the autoantibodies in IC patients appear to be directed against several different cellular antigen^.^, lo. 13. Is Autoantibodies against multiple nuclear and cytoplasmic antigens have been described in inflammatory bowel diseases such as ulcerative colitis2'-23 or infectious diseases such as chronic active viral hepatitis,ze the presence or amount of which does not generally appear to be related to disease a ~ t i v i t y .At ~ ~least . ~ ~two distinct neutrophil cytoplasmic antigens have been found in patients with any of several diverse inflammatory conditions including ulcerative colitis, Crohn's disease, autoimmune liver disease and rheumatoid arthritis."4 These data suggest that the autoantibodies associated with each of these conditions arise as a result of chronic inflammation but may not be causally related to the disease processes. The fact that the binding patterns of patients who were positive a t both of 2 visits (#4 and #6)were not constant, and that 2 patients (specimens #4 and # 5 ) appeared simultaneously to have more than one type of binding present, underscores the variable nature of autoantibodies in IC. However, the autoantibody binding patterns displayed by serum or urine immunoglobulins obtained from our IC patients were similar regardless of the type of epithelial cell used, allowing the use of a variety of epithelial cells for further studies on autoantibodies in IC. Autoantibodies may arise through several mechanisms.4.2" The variety of antigens recognized by autoantibodies in IC suggests that the degeneration of bladder epithelial cells that
REFERENCES
1. Hanno, P. M., Staskin, D. R., Krane, R. J. and Wein, A. J., eds.: Interstitial Cystitis. London: Springer-Verlag, 1990.
2. Oravisto, K. J.: Epidemiology of interstitial cystitis. Ann. Chir. Gynaecol. Fenn., 75, 1975. 3. Lynes, W. L., Flynn, S. D., Shortliffe, L. D. and Stamey, T. A,: The histology of interstitial cystitis. Amer. J. Surg. Pathol., 14: 969, 1990. 4. Ratliff, T. L., Klutke, C. G., Hofmeister, M., He, F., Russell, J. H.
and Becich, M. J.: Role of the immune response in interstitial cystitis. Clin. Immunol. Immunopathol., 7 4 209, 1995. 5. Johansson, S. L. and Fall, M.: Clinical features and spectrum of light microscopic changes in interstitial cystitis. J. Urol., 1 4 3 1118, 1990. 6. Matilla, J.: Vascular immunopathology in interstitial cystitis, Clin. Immunol. Immunopathol., 23: 648, 1982. 7. Matilla, J. and Linder, E.: Immunoglobulin deposits in bladder
epithelium and vessels in interstitial cystitis: possible relationship to circulating anti-intermediate filament autoantibodies. Clin. Immunol. Immunopathol., 32: 81, 1984. 8. Said, J . W., Van deVelde, R. and Gillespie, L.: Immunopathology of interstitial cystitis. Mod. Pathol., 2: 593, 1989. 9. Silk, M.: Bladder antibodies in interstitial cystitis. J. Urol., 103: 307, 1970. 10. Ochs, R. L., Stein, T. W., Peebles, C. L., Gittes, R. F. and Tan, E. M.: Autoantibodies in interstitial cystitis. J . Urol., 151: 587, 1994. 11. Oravisto, K. J.: Interstitial cystitis as an autoimmune disease. A review. Eur. Urol., 6: 10, 1980. 12. Jokinen, E. J., Alfthan, 0. S.and Oravisto, K. J.: Antitissue antibodies in interstitial cystitis. Clin. Exper. Immunol., 11: 333, 1972. 13. Anderson, J. B., Parivar, F., Lee, G., Wellington, T. B., MacIver,
A. G., Bradbrook, R. A. and Gingell, J . C.: The enigma of interstitial cystitis-an autoimmune disease? Brit. J . Urol., 63: 58, 1989. 14. Gordon, H., Rossen, R., Hersh, E. and Yium, J. J.: Immunologic aspects of interstitial cystitis. J. Urol., 109: 228, 1973. 15. Durwood. E., Dilworth, J. and Kaack, M.: Tamm-Horsfall autoantibodies in interstitial cystitis. J. Urol., 145: 37, 1991. 16. Division of Kidney, Urologic, and Hematologic Diseases (DKUHD) of the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK).: Diagnostic criteria for research studies (interstitial cystitis). Am. J. Kidney Dis., 1 3 353, 1989.
URINARY ANTIBODIES IN INTERSTITIAL CYSTITIS
' 17. Trifillis, A. L., Cui, X., Jacobs, S. and Warren, J. W.: Culture and
'
characterization of normal epithelium from cystoscopic biopsies of human bladder. In Vitro Cell Dev. Biol., 29A: 908, 1993. 1 18. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. and Randall, R. R.: , Protein measurement with the Folin phenol reagent. J. Biol. Cheni., 193: 265, 1951. 19. Keay, S., Schwalbe, R. S., Trifillis, A. L., Lovchik, J. C., Jacobs, S., and Warren, J. W.: A prospective study of microorganisms in urine and bladder biopsies from interstitial cystitis patients and controls. Urology, 45: 223, 1995. 120. Reeves, W. H., Satoh, M., Wang, J., Chou, C. H. and Ajmani, A. K.: Antibodies to DNA, DNA-binding proteins, and histones. Rheum. Dis. Clin. N. Am., 2 0 1, 1994. i 21. Mulder, A. H., Broekroelofs, J., Horst, G. and Limburg, P. C . : Antineutrophil cytoplasmic antibodies (ANCA) in inflammatory bowel disease: characterization and clinical correlates. , Clin. Exp. Immunol., 95: 490, 1994. I 22. Aitola, P., Miettinen, A., Mattila, A., Matikainen, M. and Soppi, E.: Effect of proctocolectomy on serum antineutrophil cytoplasI mic antibodies in patients with chronic ulcerative colitis. I J. Clin. Pathol., 4 8 645, 1995. I 23. Dalekos, G. N., Manoussakis, M. N., Goussia, A. C., Tsianos, E. V. and Moutsopoulos, H. M.: Soluble interleukin-2 recep-
,
,
,
1087
tors, antineutrophil cytoplasmic antibodies. and other autoantibodies in patients with ulcerative colitis. Gut, 34: 658, 1993.
24. Lee, J. C., Lennard-Jones, J. E. and Cambridge, G.: Antineutrophil antibodies in familial inflammatory bowel disease. Gastroenterology, 108 428, 1995. 25. Mulder, A. H., Broekroelofs, J.. Horst, G . and Limburg, P. C.: Antineutrophil antibodies in inflammatory bowel disease recognize different antigens. Adv. Exp. Med. Biol., 336 519, 1993. 26. Clifford, B. D., Donahue, D., Smith, L., Cable, E., Luttig, B., Manns, M. and Bonkovsky, H. L.: High prevalence of serological markers of autoimmunity in patients with chronic hepatitis C. Hepatology, 21: 613, 1995. 27. Query, C. C. and Keene, J. D.: A human autoimmune protein associated with U1 RNA contains a region of homology that is crossreactive with retroviral p30gag antigen. Cell, 51: 211, 1988. 28. Adams, T. E., Alpert, S. and Hanahan, D.: Non-tolerance and autoantibodies to a transgenic self antigen expressed in pancreatic beta cells. Nature, 325 223, 1987. 29. Strober, W. and James, S. P.: The immunopathogenesis of gastrointestinal and hepatobiliary diseases. J.A.M.A., 268: 2910, 1992.
DIEJOURNAL
Vol 157. 1083-1087,March 1997 Printed in U.S.A.
OF UR0LCG.Y
Copyright 0 1997 by AMERICAN U R O ~ I C ASS~CIATION, AL INC.
URINE AUTOANTIBODIES IN INTERSTITIAL CYSTITIS S. m A Y , * C.-0. ZHANG, A. L. TRIFILLIS, J. R. HEBEL, S. C. JACOBS AND J. W. WARREN From the Research Service, Department of Veterans Affairs Medical Center, Baltimore, Maryland. and the Division of Infectious Diseases, Department of Medicine, Department of Pathology, Department of Epidemiology, and Division of Urology, Department of Surgery, University of Maryland School of Medicine. Baltimore, Maryland
ABSTRACT
Interstitial cystitis is a chronic bladder disease with certain features that suggest autoimmunity may play a role in initiating or maintaining the disease process. We therefore determined whether immunoglobulin fractions from 14 IC patient and 19 control urine specimens bound in vitro to primary cultures of human bladder epithelial cells, as well as epithelial cells from a variety of other tissues. Urine autoantibodies t h a t bound to normal human bladder epithelial cells were present in 8 of 14 IC specimens (from 6 of 9 IC patients) as compared to 3 of 23 control specimens (from 2 of 17 control patients). These antibodies, which were usually also present at low titers in sera from these persons, bound to at least four nuclear or cytoplasmic antigens, with the specificity of autoantibodies from a given individual varying over time. The autoantibodies were not specific for normal or malignant bladder epithelial cells, but bound to epithelial cells from a variety of tissues. These data show that anti-epithelial cell autoantibodies are present in the urine of IC patients, but suggest that these antibodies are not likely to be a primary cause of this disease. KEY WORDS:cystitis, autoantibodies
Interstitial cystitis is a chronic inflammatory condition of the bladder which affects a t least 20,000-90,000 women in the United States at any given time; approximately onetenth as many men also suffer from this condition.’ This disease, which usually affects people between 30 and 59 years old, causes pain and increased frequency and urgency of urination.l.2 Histopathologic changes evident in IC include submucosal inflammation along with thinning or ulceration of the bladder epithelium.3 No etiology for IC has yet been identified, and no empiric treatment has proven to be generally efficacious for this disorder. Certain aspects of IC suggest that autoimmunity may play a role in initiating or sustaining the chronic inflammatory response evident in this disease.4 Published data indicate that IgM antibodies and complement (C’)have been detected in areas of endothelial cell damage5-7 and IgA antibodies have been detected on the surface of bladder epithelium in IC patients: but the antigenic specificity of these antibodies is unknown. Autoantibodies have been detected in the sera of IC patients to a greater extent than in control^^-^^ but the specificity of these autoantibodies has differed between reports. Whereas one report indicated the presence of antibladder epithelial cell antibodies in the sera of IC patients? other reports have shown no evidence for anti-bladder epithelial cell antibodies12 or have indicated the presence of serum autoantibodies which are not specific for bladder epithelial cells but bind to other cell types,lO.ll intermediate filaments,7 or Tamm-Horsfall protein.l5 In one report, serum autoantibodies in IC patients were directed against a variety of cell proteins which were notably distinct from those recognized by autoantibodies in patients with systemic lupus erythematosis or systemic sclerosis.1° In addition to serum, urine is itself a potential source for autoantibodies which could bind directly to bladder or other urinary tract epithelial cells. We hypothesized that urine
autoantibodies might be important in IC because of their potential to cause direct damage to the bladder epithelium (for example, by causing complement-mediated cytolysis), andor to serve as possible diagnostic tools. We therefore screened both the urine immunoglobulin fraction and whole serum from IC patients and their age, race, and sex-matched controls for binding to bladder epithelial cells (primary adult normal human bladder epithelial cells, T24 bladder carcinoma cells, and cytospun epithelial cells from IC patient urine specimens) as well as epithelial cells from other tissues (A549 human lung carcinoma cells; HEp-2 human laryngeal epidermoid carcinoma cells; and Hep G2 human hepatocarcinoma cells). MATERIALS A N D METHODS
Patients. IC patients were referred by physicians, the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), and the Interstitial Cystitis Association. All patients had previously undergone diagnostic cystoscopy, and fulfilled the NIDDK diagnostic criteria for IC.16 Eight of these patients had evidence for glomerulations at cystoscopy, while one had both Hunner’s ulcers and glomerulations. The average duration of symptoms was 7.4 years (range 1.9 to 14.1 years). Age, race, and sex-matched controls were chosen from volunteers with no history of IC or other urological disease. Urine was initially collected from these patients and their controls at least three months following the most recent known bacterial urinary tract infection and one month following the last antibiotic use. All participating patients were determined to be symptomatic at the time of enrollment; additional specimens were also obtained during periods of IC symptom exacerbation from the patients and their available controls. All participants were a t least 18 years old and enrolled in accordance with guidelines of the Institutional Accepted for publication July 1, 1996. Review Board at the University of Maryland School of Med* Requests for reprints: Veterans Administration Medical Center, icine. The number of IC patients or controls used for each Room 3B184, 10 North Greene Street, Baltimore, MD 21201, Supported by the National Institute of Diabetes and Digestive and experiment was based on the number for whom a sufficient quantity of urine was available for each experiment. HowKidney Diseases, NIH (#R01 DK44818). 1083
1084
URINARY ANTIBODIES IN INTERSTITIAL CYSTITIS
ever, each IC patient urine specimen was studied simulta- Following 3 additional rinses with PBS, SloFade (Molecular neously with urine from one or two age-, race- and sex- Probes, Eugene, OR) was added to the wells and the cells were examined using a Nikon Diaphot inverted phase flu& matched controls for each experiment. Urine and serum specimens. Urine specimens for immuno- rescence microscope, A negative control well was incubated globulin fractionation were collected by t h e clean catch with secondary antibody alone. Statistical analysis. Comparisons of the proportion of Ic method, in which each patient wiped t h e labial area with 10% povidone iodineititratable iodine 1% solution (Clinidine, patient and control specimens that contained anti-epithelial Guilford, CT) then collected a midstream urine into a sterile cell antibodies were made using Fisher’s exact test. container. Serial specimens were obtained over a 24 hour RKSII LI‘S period from each of 9 IC patients ( 14 specimens) and 18 age-, race-, and sex-matched controls (24 specimens), and placed in Binding of IC serum and urine auloantihodics to normal a refrigerated container. Specimens were kept at 4C from the adult human bladder epithelial (HBEJ cells. Ammonium sulbeginning of collection, and transported to the laboratory fate precipitation of immunoglobulins from both IC and conwithin 24 hours of completion; antibodies were precipitated trol urine was shown to result in highly enriched immunofrom each specimen within 6 hours of delivery to the labora- globulin preparations by polyacrylamide gel electrophoresis tory by adding a n equal volume of saturated ammonium (PAGE1 under denaturing conditions (presence of bands corsulfate (pH 7.8) on ice (see below). Sera were collected from 7 responding to molecular mass of immunoglobulin heavy and IC patients and 14 age, race and sex-matched controls and light chains only) (data not shown). The amount of immunostored at -2OC until used. globulin recovered ranged from 0.12-1.3 mg./ml.; there was Human epithelial cells. T24 bladder carcinoma cells (ATCC no significant difference in the mean amount of total immu#4-HTB) (American Type Culture Collection, Rockville, MD) noglobulin between IC patients 10.65 mg./ml.) and controls were grown in McCoy’s medium containing 10% fetal bovine (0.54 mg./ml.). serum (FBS), 1%antibiotidantimycotic solution, and 15%gluAnti-bladder cell~antibodies bound to nuclear or ._ . ~ . ~ ~ which ~ ~ . ~ tamine (Sigma, St. Louis, MO). Normal adult human bladder cytoplasmic antigens in normal adult human primary bladepithelial cells (HBE) were grown from biopsies obtained at der epithelial cells were evident in the urine of 8 of 14 autopsy from patients who had no history of bladder disor- specimens from 6 of 9 IC patients, and 3 of 23 specimens from der.” The explanted cells were grown in Eagle’s minimal 2 of 17 controls. Four of these 9 IC patients (44‘;~ were essential medium (MEM) containing 10% heat inactivated positive for urine autoantibodies a t the first visit 1 # 3 ,#4, #5 FBS, 1%antibiotidantimycotic solution, 1% glutamine, and and #6)as compared to 2 of 17 controls ( p = .14 ). Binding was 1.0 pgiml insulin (all from Sigma). Hep G2 human hepato- evident when urinary immunoglobulin was present a t 0.3cellular carcinoma cells (ATCC #HB 8065) and A549 human 0.03 mg./ml. (data are shown for cells incubated with 0.03 lung carcinoma cells (ATCC #CCL 185)were grown in MEM mg./ml. immunoglobulin). The antibodies from 5 of the 8 containing 10%FBS, 1% antibiotidantimycotic solution and positive IC specimens and both of the positive control speci1% glutamine. All cells were cultured at 37C in a 5% CO, mens bound to the bladder epithelial cells in a fine, diffuse atmosphere. staining pattern (Figure l C ) , while antibodies from 3 IC Urine specimens for uroepithelial cell cytospin prepara- patients had a coarse pattern (Figure 1D1, one had a nucletions were collected by the clean catch method (each patient olar pattern (Figure l B ) , and one had a cytoplasmic binding wiped the labial area with 10% povidone iodineltitratable pattern (Figure 1A) ( 2 patients had 2 types of binding obiodine 17c solution (Clinidine, Guilford, CT), then collected a served). These findings suggest that a t least four different midstream urine into a sterile specimen container); these antigens serve as targets for the anti-bladder epithelial cell were kept a t 4C and transported to the laboratory within one antibodies in IC urine. The use of secondary antibodies spehour of collection. Cytospun cells were prepared as previ- cific for IgG, IgM. or IgA indicated that each positive IC or ously described.19 control urine specimen contained autoantibodies of the IgG Ammonium sulfate precipitation of urinary immunoglobulins. Equivalent volumes of saturated ammonium sulfate and patient urine specimen were mixed together at 4C, placed on ice for 30 minutes, and centrifuged at 4C. The pellet was then resuspended in 0.9% NaCl to half of original urine volume, additional saturated ammonium sulfate was added to this mixture which was stirred a t 4C for one hour and centrifuged at 4C. The second pellet was resuspended in 10 ml. borate buffered saline (BBS), dialyzed against BBS overnight at 4C, then dialyzed against phosphate buffered saline (PBS) at 4C, aliquoted and stored at -8OC until used. The amount of protein in each specimen was determined by the Lowry assay18 with known concentrations of bovine serum albumin as the standard. Immunofluorescence assay. Cells which were grown to confluence on 8 well LabTek tissue culture slides (Nunc, Inc., Naperville, IL), or cytospun preparations of uroepithelial cells from the urine of IC or control patients (prepared as previously described’g), were fixed with acetone/ethanol (1: 1); these cells or fixed HEp-2 cells (purchased from Zeus Scientific, Raritan, N J ) were then incubated in duplicate with varying dilutions of serum (0-1,000) or urinary antibodies (0-0.03 mg./ml.) in PBS. Following 2 hours incubation at 37C, the cells were rinsed with PBS and incubated further Urinary autoantibody binding patterns in IC. Normal adult huat 37C for 2 hours with FITC-conjugated goat anti-human Ig man bladder epithelial (HBE) cells were fixed and incubated with (Southern Biotechnology Associates, Birmingham, AL),or urinary immunoglobulins from IC patients followed by FITCconjugated goat anti-human Ig, as described in Methods. Shon.n are FITC-conjugated goat anti-human antibodies specific for four binding patterns encountered: A ) cytoplasmic; B I nucleolar; C) I@, IgM, or IgA (Vector Laboratories, Burlingame, CA). diffuse nuclear; D) coarse nuclear. Final magnification = 400 x .
URINARY ANTIBODIES IN INTERSTITIAL CYSTITIS
class; 2 IC specimens also contained IgM autoantibodies (#5 and #6a) and 2 also contained IgA autoantibodies (#7 and #8a). Eleven serum specimens from IC patients (7 of whom had urinary autoantibodies) and 16 control specimens (2 of whom had urinary autoantibodies) were also tested. Each patient and control who had urinary autoantibodies also had serum autoantibodies with the same binding pattern; one additional IC patient and one additional control had serum antinuclear antibodies which bound to normal adult human bladder epithelial cells in a diffuse manner. The titer of serum autoantibodies in the IC patients was generally low, with 2 of the 8 positive specimens having a titer <1:40, and only 2 patients having a titer >1:100. For both IC patient urine and serum autoantibodies, there was no obvious correlation between the presence or pattern of antibody binding and duration of symptoms, with patients #1 and 2 having symptoms for 1.9 and 2.3 years prior to the first collection, but patient #8 having symptoms for 14 years at the time of the first collection. Four of the 9 patients returned to give another urine specimen at the time of symptom exacerbation (#la, #3a, #4a, and #8a), but there was also no obvious correlation between an increase in symptoms and antibody positivity or pattern. Two of the IC patients ( # 1 and #8) were initially negative for autoantibodies, but became positive during their exacerbations; one patient (#3) was initially positive but became negative, and another patient (#4j was positive a t both visits but changed in pattern. Binding patterns displayed by these patients during exacerbation also differed from one another (1cytoplasmic, 1coarse nuclear and 1 diffuse nuclear). Serial specimens were also obtained on a fifth patient (#6), although the timing of collection was not correlated with symptom severity; interestingly, the pattern of fluorescence also appeared to change between visits for this patient. Because the IC patients who returned for a second visit often returned with little or no notice during a period of symptom exacerbation, only one second simultaneous repeat control specimen was obtained (#7a). However, we obtained an additional 5 repeat control urine specimens a t random time points, one of which was positive on both visits for diffuse nuclear binding; like #7, the other 4 of these controls were negative on both visits. A comparison of data obtained from the IC patients and these controls indicated that the IC patients were more likely to be positive on one or both visits ( p = .015). Serum and urine autoantibody binding to cells from epithelial malignancies. To determine whether the autoantibodies in IC patient urine and serum specimens were specific for normal adult human bladder epithelial cells, specimens were also tested for binding to T24 human bladder carcinoma cells, as well as Hep G2 human hepatocellular carcinoma cells and A549 human lung carcinoma cells (Tables 1 and 2). Six of the IC urine specimens that had detectable antibodies against HBE cells were also tested for binding to T24 cells; each of these had antibodies with the same binding pattern on T24 cells, indicating that primary bladder epithelial cells are not required for the detection of these antibodies. Antibodies from four of these six specimens also bound to Hep G2 or A549 cells, indicating that a t least some antibodies in IC urine are not specific for cells of bladder origin. In addition, urine antibodies from 5 specimens that were positive for anti-bladder epithelial cell binding and 3 specimens that were negative were tested for binding to HEp-2 human laryngeal carcinoma cells; all 5 of the positive specimens showed evidence for binding in the same pattern displayed on bladder cells ( 3 coarse nuclear and 2 diffuse nuclear), and all 3 negative specimens were negative on these cells. Binding of urine and serum antibodies to cytospun urine cells. To determine whether urine or serum antibodies might display increased or differential binding to autologous epithelial cells, urine specimen cytospin preparations were
TABLE1. Binding
1085
of urinr autoantrhodies t o norrrial epilhelral and
carcrriomn cells Other Epithelial ('ella
Bladder Cells
HBE
T24
Hep (2
A649
~
IC Patients 1 la* 2 3 3a 4 4a 5 6 6a 7 8
8a 9 Controls
1 2 3 4 5 6 7 7a 8 9
-
+Cy"
f
-
DN
I
T
('y
-
DN
f
f
CY DN
+ DN
-
-
-
+ Ncl/DN
n.d. + CN tCN n.d. + CN + DN n.d. + DN n.d.
n.d.
+ CN +CNIDN DN +CN T
-
+ DN -
+ CN
-
n.d.
-
+ DN n.d.
+ DN n.d.
-
+ DN
-
+ DN
10 11 12 13 14 15 16 17 18 * a = results of specimen obtained on a return visit from the same patient. "Cy = cytoplasmic; DN = diffie nuclear; CN = coarse nuclear; Ncl = nucleolar. n.d. = not done.
made on 2 patients (#3 and #4) and their controls (#13,#14, #16 and #17). Approximately 20-35 cells were visible on each slide. Although both patients and one control had evidence for urine and serum antibody binding to cultured bladder cells, only one patient (#4) showed urine antibody binding to her own cytospun epithelial cells, in the same pattern shown on primary HBE and T24 cells; all sera were negative. DISCUSSION
This report presents the first evidence that autoantibodies are present in the urine of IC patients, and shows that these urine autoantibodies are similar in their binding patterns to serum autoantibodies from these patients. Our findings also confirm an earlier report by Ochs, et a1.,10 which described a similar prevalence and variety of IC serum autoantibody binding patterns with HEp-2 cells. Like that study, o u r study found that autoantibodies with a fine speckled pattern of nuclear fluorescence were the most prevalent, with 62% of positive urine immunoglobulin specimens and 75% of positive sera from IC patients having evidence for this type of binding pattern with HBE cells. In addition, our data indicate that neither urine nor serum autoantibodies from IC patients are specific for human bladder cells, nor are they more evident when normal human bladder cells or autologous cytospun cells from the IC patients are used. This nonspecificity is in agreement with previously reported studies which indicate that the serum anti-nuclear or anti-cytoplasmic autoantibodies in IC are generally not specific for a particular cell type.'*.'' Along with the apparent nonspecificity of these autoantibodies, the lack of correlation in the current study between symptom
URINARY ANTIBODIES I N INTERSTITIAL CYSTITIS
1086 TABLE2. Binding
of
serum autoantibodies to normal epithelial and carcinoma cells ___ Other Ep~rhelialCells
Bladder Cells
--
~
HBE
-
T24
.
A549
IC Specimens 1 la' 2 3 4 4a 5 6
-
- DN - NcliDN
nd +Ncl!l)N -CN
+DN NcV- DN + CN DN 7
~
7
8 Ha
DN
- DN
Control Specimens 2 3 4
5 6 ia
8 9 10 11 12 13 15 16 17 18
'a
results of specimen obtained on a return visit from the same patient. c.ytoplasmic: DN = diffuse nuclear; CN = coarse nuclear; Ncl = nucleolar. n.d. = not done.
occurs as a result of chronic inflammation in I c may stimulate the production of autoantibodies directed against several nuclear and cytoplasmic antigens. However, molecular mimicry can also play a role in initiating autoantibody responses, with the first antibody directed against an epitope on a foreign antigen that is homologous to a self antigen. The subsequent antibody response may spread to include other epitopes on the self antigen."7,2* This type of autoimmune stimulation is thought to play a role in some cases of inflammatory bowel disease, in which systemic sensitization to bacterial antigens can lead to ongoing, chronic large-bowel inflammation and in which colonic epithelial cell antigens have been shown to cross-react with bacterial cell wall antigens."" We previously reported an increased incidence of a variety of bladder microorganisms in IC patients a s compared to controls,'" suggesting that urinary tract organisms might be a possible source of foreign antigen that could initiate this type of autoantibody response in the IC patient. Additional studies to determine whether bacterial antigen(s) on microorganisms present in the urine of IC patients crossreact with epithelial cell nuclear or cytoplasmic antigens may be helpful for further determining the mechanism of this disease process. Acknowledgments. The authors thank Edward W. Campbell, Jr. M.D., Michael J. Naslund, M.D., Ms. Linda Horne and Terry Wilson, R.N. for their assistance with patient enrollment and specimen acquisition; Marcia Fullem for technical assistance; and Barbara White, M.D. for helpful discussions.
=
" Cy
=
duration or exacerbation and presence or pattern of urine autoantibodies, as well as the generally low titer of the serum autoantibodies, suggest that these antibodies are probably not causally related to the epithelial cell damage evident in IC. However, prospective studies with defined criteria for disease and symptom severity may be helpful for determining the role of serum and/or urine autoantibodies in IC. Like autoantibodies present in patients with various other conditions characterized by chronic inflammation, the autoantibodies in IC patients appear to be directed against several different cellular antigen^.^, lo. 13. Is Autoantibodies against multiple nuclear and cytoplasmic antigens have been described in inflammatory bowel diseases such as ulcerative colitis2'-23 or infectious diseases such as chronic active viral hepatitis,ze the presence or amount of which does not generally appear to be related to disease a ~ t i v i t y .At ~ ~least . ~ ~two distinct neutrophil cytoplasmic antigens have been found in patients with any of several diverse inflammatory conditions including ulcerative colitis, Crohn's disease, autoimmune liver disease and rheumatoid arthritis."4 These data suggest that the autoantibodies associated with each of these conditions arise as a result of chronic inflammation but may not be causally related to the disease processes. The fact that the binding patterns of patients who were positive a t both of 2 visits (#4 and #6)were not constant, and that 2 patients (specimens #4 and # 5 ) appeared simultaneously to have more than one type of binding present, underscores the variable nature of autoantibodies in IC. However, the autoantibody binding patterns displayed by serum or urine immunoglobulins obtained from our IC patients were similar regardless of the type of epithelial cell used, allowing the use of a variety of epithelial cells for further studies on autoantibodies in IC. Autoantibodies may arise through several mechanisms.4.2" The variety of antigens recognized by autoantibodies in IC suggests that the degeneration of bladder epithelial cells that
REFERENCES
1. Hanno, P. M., Staskin, D. R., Krane, R. J. and Wein, A. J., eds.: Interstitial Cystitis. London: Springer-Verlag, 1990.
2. Oravisto, K. J.: Epidemiology of interstitial cystitis. Ann. Chir. Gynaecol. Fenn., 75, 1975. 3. Lynes, W. L., Flynn, S. D., Shortliffe, L. D. and Stamey, T. A,: The histology of interstitial cystitis. Amer. J. Surg. Pathol., 14: 969, 1990. 4. Ratliff, T. L., Klutke, C. G., Hofmeister, M., He, F., Russell, J. H.
and Becich, M. J.: Role of the immune response in interstitial cystitis. Clin. Immunol. Immunopathol., 7 4 209, 1995. 5. Johansson, S. L. and Fall, M.: Clinical features and spectrum of light microscopic changes in interstitial cystitis. J. Urol., 1 4 3 1118, 1990. 6. Matilla, J.: Vascular immunopathology in interstitial cystitis, Clin. Immunol. Immunopathol., 23: 648, 1982. 7. Matilla, J. and Linder, E.: Immunoglobulin deposits in bladder
epithelium and vessels in interstitial cystitis: possible relationship to circulating anti-intermediate filament autoantibodies. Clin. Immunol. Immunopathol., 32: 81, 1984. 8. Said, J . W., Van deVelde, R. and Gillespie, L.: Immunopathology of interstitial cystitis. Mod. Pathol., 2: 593, 1989. 9. Silk, M.: Bladder antibodies in interstitial cystitis. J. Urol., 103: 307, 1970. 10. Ochs, R. L., Stein, T. W., Peebles, C. L., Gittes, R. F. and Tan, E. M.: Autoantibodies in interstitial cystitis. J . Urol., 151: 587, 1994. 11. Oravisto, K. J.: Interstitial cystitis as an autoimmune disease. A review. Eur. Urol., 6: 10, 1980. 12. Jokinen, E. J., Alfthan, 0. S.and Oravisto, K. J.: Antitissue antibodies in interstitial cystitis. Clin. Exper. Immunol., 11: 333, 1972. 13. Anderson, J. B., Parivar, F., Lee, G., Wellington, T. B., MacIver,
A. G., Bradbrook, R. A. and Gingell, J . C.: The enigma of interstitial cystitis-an autoimmune disease? Brit. J . Urol., 63: 58, 1989. 14. Gordon, H., Rossen, R., Hersh, E. and Yium, J. J.: Immunologic aspects of interstitial cystitis. J. Urol., 109: 228, 1973. 15. Durwood. E., Dilworth, J. and Kaack, M.: Tamm-Horsfall autoantibodies in interstitial cystitis. J. Urol., 145: 37, 1991. 16. Division of Kidney, Urologic, and Hematologic Diseases (DKUHD) of the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK).: Diagnostic criteria for research studies (interstitial cystitis). Am. J. Kidney Dis., 1 3 353, 1989.
URINARY ANTIBODIES IN INTERSTITIAL CYSTITIS
' 17. Trifillis, A. L., Cui, X., Jacobs, S. and Warren, J. W.: Culture and
'
characterization of normal epithelium from cystoscopic biopsies of human bladder. In Vitro Cell Dev. Biol., 29A: 908, 1993. 1 18. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. and Randall, R. R.: , Protein measurement with the Folin phenol reagent. J. Biol. Cheni., 193: 265, 1951. 19. Keay, S., Schwalbe, R. S., Trifillis, A. L., Lovchik, J. C., Jacobs, S., and Warren, J. W.: A prospective study of microorganisms in urine and bladder biopsies from interstitial cystitis patients and controls. Urology, 45: 223, 1995. 120. Reeves, W. H., Satoh, M., Wang, J., Chou, C. H. and Ajmani, A. K.: Antibodies to DNA, DNA-binding proteins, and histones. Rheum. Dis. Clin. N. Am., 2 0 1, 1994. i 21. Mulder, A. H., Broekroelofs, J., Horst, G. and Limburg, P. C . : Antineutrophil cytoplasmic antibodies (ANCA) in inflammatory bowel disease: characterization and clinical correlates. , Clin. Exp. Immunol., 95: 490, 1994. I 22. Aitola, P., Miettinen, A., Mattila, A., Matikainen, M. and Soppi, E.: Effect of proctocolectomy on serum antineutrophil cytoplasI mic antibodies in patients with chronic ulcerative colitis. I J. Clin. Pathol., 4 8 645, 1995. I 23. Dalekos, G. N., Manoussakis, M. N., Goussia, A. C., Tsianos, E. V. and Moutsopoulos, H. M.: Soluble interleukin-2 recep-
,
,
,
1087
tors, antineutrophil cytoplasmic antibodies. and other autoantibodies in patients with ulcerative colitis. Gut, 34: 658, 1993.
24. Lee, J. C., Lennard-Jones, J. E. and Cambridge, G.: Antineutrophil antibodies in familial inflammatory bowel disease. Gastroenterology, 108 428, 1995. 25. Mulder, A. H., Broekroelofs, J.. Horst, G . and Limburg, P. C.: Antineutrophil antibodies in inflammatory bowel disease recognize different antigens. Adv. Exp. Med. Biol., 336 519, 1993. 26. Clifford, B. D., Donahue, D., Smith, L., Cable, E., Luttig, B., Manns, M. and Bonkovsky, H. L.: High prevalence of serological markers of autoimmunity in patients with chronic hepatitis C. Hepatology, 21: 613, 1995. 27. Query, C. C. and Keene, J. D.: A human autoimmune protein associated with U1 RNA contains a region of homology that is crossreactive with retroviral p30gag antigen. Cell, 51: 211, 1988. 28. Adams, T. E., Alpert, S. and Hanahan, D.: Non-tolerance and autoantibodies to a transgenic self antigen expressed in pancreatic beta cells. Nature, 325 223, 1987. 29. Strober, W. and James, S. P.: The immunopathogenesis of gastrointestinal and hepatobiliary diseases. J.A.M.A., 268: 2910, 1992.