- Email: [email protected]
Dormant Microbes in Interstitial Cystitis
Vol. 153,1321-1326,April 1995 Printed in U.S.A.
DORMANT MICROBES IN INTERSTITIAL CYSTITIS GERALD J. DOMINGUE", GAMAL, M. GHONIEM, KENNETH L. BOST, CESAR FERMIN LISET G. HUMAN
AND
From the Departments of Urology, Microbiology and Immunology and Pathology, Tulane University, New Orleans, Louisiana
ABSTRACT
Interstitial cystitis (IC) is a n inflammatory disease of the urinary bladder that has no known etiology. A microbial association with this disease has not been supported since routine cultures of urine from IC patients are usually negative. However, we have demonstrated the presence of bacterial 1 6 s rRNA genes in bladder biopsies from 29%of patients with IC, but not from control patients with other urological diseases. The ability to identify the presence of bacterial DNA in these patients was accomplished using a sensitive and specific nested PCR method capable of amplifylng 16s rRNA genes from a wide variety of bacterial genera. Cloning and sequencing of 16s rRNA gene fragments amplified from bladder tissue of IC patients showed that these genes were derived from genera representing Gram-negative bacteria. In addition to the molecular data, a novel finding of 0.22 pm. filterable forms has been isolated in culture from the biopsy tissue of 14 of 14 IC patients and from 1 of 15 controls. The forms contain nucleic acids and resemble cell wall-deficient bacteria in gross morphology; however, their swirled myelin-like ultrastructure is unusual and suggests a heretofore unclassified microbe. These results demonstrate for the first time an association of Gram-negative bacterial DNA and filterable forms with affected bladder tissue from patients with IC. KEY WORDS:bladder, cystitis, bacteria, inflammation, etiology
Interstitial cystitis (IC), a n inflammatory disease of the bladder, has a n unknown etiology. Disease states predominate in women and have varying degrees of morbid it^.^-^ In the most severe cases the disease is debilitating and characterized by small bladder capacity, ulcers, severe frequency, nocturia, dysuria and pain. Quality of life scores in these patients have been measured to be lower than in chronic renal disease patients. Treatment for the disease is varied and includes bladder distension therapy and sometimes surgical removal of the bladder. The limitations of current therapies result in part from the undetermined cause of the disease. It has been suggested that IC may be an autoimmune however, no organ-specific immune response has been demonstrated to induce or be responsible for the d i s e a ~ e . ~ Furthermore, .'~ recent investigations argue against immune abnormalities in these A microbial association with IC has not been demonstrated.' This disease is distinct from other bladder infections in which microbes can be readily cultured from urine. Thus the inability to routinely culture bacteria has been suggested to indicate that bacteria are not involved in the etiology or morbidity associated with the disease. The present study uses the polymerase chain reaction (PCR)I4 to detect the presence of bacterial DNA in bladder biopsies from patients with IC and also uses specialized culture techniques to grow fastidious organisms that may be difficult to culture. 's2
MATERIALS AND METHODS
Patients, biopsies and cultures. Informed consent was obtained from all patients and control subjects participating in this study. The exact nature of the study and any possible consequences were explained to each participant. AU patients were female except for R.O.P. and M.I.L. Patients were diag-
nosed with IC according to the criteria set forth by the National Institute of Arthritis,Diabetes, Digestive and Kidney Disease ~Two. separate ~ ~ cold cup biopsies (approximately 1mg.) were surgically removed from intlamed sites in the bladder wall of IC patients or from the normal bladder wall of control subjects using sterile techniques. One set of biopsy samples was homogenized immediately and was exhaustively evaluated for the presence of bacteria using culture techniques described be10w.l~ The second biopsy specimen was stored as a coded sample at -7OC until total DNA was extracted. Catheterized urine specimens and biopsy bladder tissue from each of these subjects were cultured for ordinary and fastidious bacteria on the following media: 5%sheep blood agar and MacConkey agar incubated aerobically, anaerobically and under a 5% CO, atmosphere;'6 tryptme broth for revealing
Accepted for publication November 30, 1994. * Requests for reprints: .Department of Urology, SL42, Tulane University School o Medicine, 1430 Tulane Avenue, New Orleans, Louisiana 70112. This work was supported by N.I.H. Grant DK44812-03 awarded to G m l d J. Domingue and donations to Gerald J. Domingue by the Interstitial Cystitis Association and Mr. and Mrs. W. Russell Harp. 1321
DORMANT MICROBES I N INTERSTITIAL CYSTITIS
1322
bladder tissue. Deoxyribonucleicacid was isolated using aseptic conditionsby lysing cells with SDS followed by sequential treatment of the sample with pronase and RNase as described by the manufacturer (Stratagene, La Jolla, California). Precipitated DNA was resuspended in water and quantified using absorbance at 260 nm. Using a method which assured that exogenous DNA was not introduced into the sample, total DNA extracted from the biopsies was amplified by PCR using nested primers. The appropriate DNA was added to a PCR14 containing 5 units Thermus aquaticus DNA polymerase (Boehringer Mannheim, Indianapolis, Indiana), 2 mM. dNTPs, 1pg. of each primer and PCR buffer (10 mM. Tris-HC1, 50 mM. KCl, 1 pg./ml. gelatin and 2 mM. IvIgC1,). Positive displacement pipetters were used to ensure that no crosscontamination occurred. The PCR buffer, water and mineral oil were exposed to W light to destroy any contaminating DNA that might be present in these it was of critical importance to treat ~ o i u t i o nFurthermore, ~.~~ the Thermus aquaticus DNA polymerase with bovine pancreatic DNase (Promega, Madison, Wisconsin) to remove contaminating bacterial DNA present in commercial preparations of this e n ~ y m e . ' ~ "Thus ~ , prior to inclusion in PCR, 1 unit of DNase was used to treat 20 units of Thermus aquaticus DNA polymerase in PCR buffer for 1 hour at 37C. The DNase was heat inactivated at 95C for 10 minutes prior to using the DNase-treated DNA polymerase in PCR. Reactions were brought to 70C prior to addition of 5 units of DNase-treated Thermus aquaticus DNA polymerase to minimize nonspecific priming. Reactions were taken through 25 cycles using a BIOCYCLER thermal cycler (BIOS, New Haven, Connecticut) programmed for 95C denaturation, 58C annealing and 72C extension temperatures, with the first three cycles having extended denaturation and annealing times. The primers used are shown in table 1.The initial PCR was performed using the 5' primer, lA,and the 3' primer, 5. ARer completion of the first PCR, 10% of each reaction was used for a second PCR using the same conditions described above with the 5' primer, 3, and the 3' primer, 4. Primers were selected for 1)their ability to amplify the 16s rRNA gene from a wide variety of bacterial genera, 2) the lack of significant complementarity between primers, 3)the lack of significant secondary structure, 4) optimal annealing temperatures and 5)significant homology of the 5' primers with 16s rFWA genes, but not with other microbial or human genes contained in the Genbank database version 69. ARer the second PCR, 2 W of the reaction was electrophoresedon 1.3% agarowl ethidium bromide gels and the presence of a 467 base pair fragment was detected by U V fluorescence. Southern blot analyses to demonstrate hybridization of the amplified DNA Fagments with an oligonucleotide specific for bacterial 16s rRNAgenes. Southern blot analyses were performed essentially as described.20 Briefly, duplicate agarose gels were denatured, neutralized and transferred to nylon membranes by positive pressure (Possiblot, Stratagene). The DNA was fixed by exposure to W light (Stratalinker, Stratagene), and membranes were prehybridized overnight at
45C. Hybridization with a 32P end-labelled oligonucleotide probe (CYT GGA GGA AGG TGG GGA TGA C, where Y equals C or T) specific for bacterial 1 6 s rRNA genes was performed overnight a t 45C in the presence of 4X SSC. After washing, hybridization was detected by autoradiography. Bacterial identitkatwn. To identify the type(s) of bacteria present in IC patients having 16s rRNA genes amplified from their bladder biopsies, these gene fragments were cloned and sequenced. Gene fragments were isolated from agarose gels using Prep-a-gene (Bio-Rad, Richmond, California) and ligated into the pGEM-3Z vector (Promega). This ligation was accomplished by incorporatmg, respectively, EcoRl and BamHl restriction endonuclease sites into the 5' and 3' primers used in the secondary PCR. Vector and amplified DNA fragments were cleaved with these enzymes prior to ligation. Escherichia coli (strain JM83) made competent with calcium chloride were transformed and plated onto Luria agar plates containing ampicillin, X-gal and ETG. White colonies were selected and expanded, and plasmids were isolated from these clones by minipreping (Promega). Sequencing of plasmids containing inserts was accomplished by the dideoxy chain termination method using Squenase 2.0 and instructions supplied by the manufacturer (United States Biochemical). Sequencing primers used are given in table 1. At least two W e r e n t clones were sequenced for each of the 4 isolated fragments. Sequences of approximately 400 base pairs were analyzed for optimal alignments against 750 different 16s rRNA gene sequences contained in the Ribosomal RNA database using alignment softKnown bacterial ware programs available at this facility.21*22 sequences having the most similar alignments with DNA fragments amplified from the biopsy tissue were furnished by the Ribosomal RNA database project. Homology of DNA fragments amplified from the bladder tissue with known bacterial 16s rRNA genes identified by the Ribosomal RNA database project as having the best alignments were determined using MacVector sequence analysis software (IBI, New Haven, Connecticut). RESULTS
A sensitive and specific PCR capable of amplifying the 1 6 s rRNA genes from a wide variety of bacterial genera using nested primers was developed. As shown in figure 1, nested priming amplified a 467 base pair fragment when 100 pg. of input DNA from Salmonella typhi (lane A), Staphylococcus haemolyticus (lane C), Pseudomonas aeruginosa (lane D), or Escherichia coli (lane G)were used. Conversely, amplification was not detected with the addition of no input DNA (lane B), Candida albicans DNA (lane E) or human DNA (lanes F and HI. Identity of the appropriate bacterial genus was confirmed by cloning and sequencing the PCR amplified bacterial 16s rRNA gene fragments. To develop this nested PCR, several significant technical problems had to be overcome. First, we and have
TABLE1. Oligonucleotide primers Oligonucleotides (Primers, p)
lA
Sequence 5' to 3'
CAC AAG CGG TGG AGC ATG TGG 'IT
16s rR.NA Position
Strand Orientation
932-955
+
Reference Domingue et. a1 (this publication)
CCT ACG CYT ACC TX 'ITA CGA CT Y equals C or T 1514-1492 GGA A'IT CTG CAA CGC GAA CCT TAC TA 1168-1194 + 1392-1371 GCG GAT CCT GGT KTG ACG GGC GGT TG TA K equals G or T GAACCTTACCTGGKYTXACATCCCAAGGCCCGGGA ACG TAT TCA CCG CYT GGA GGA AGG TGG GGA TGAC _. Y equals C or T and K equals G or T 246 AGA G'IT TGA TCC TGG CTC AG 8-28 + _7 4_ 240 GGG TCA A'IT CCT 'ITG AGT 'IT 928-908 244 CCC ACT GCT GCC TCC CGT AG 361-342 .-a .ry. F. wnLtna nrimnr. The primera selected for the ability to amplifythel6S rRNA gene were plAL l..u ra.yL=.u p3 and p4. Nested PCR with p3 and p4 produced amplification of a 460-bp fragment of bacterial DNA. Additional primers amplifyingthe 16s rRNA gene were used with DNA extracted from tissue of 6 controls:x p246 and p240 yield a 900-bp amplified produet, p246 and p244 amplify a 300-bp bacterial DNA fragment 5 3 4 Sequencing Primers
~
..--_..
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
r A B C D E F G H
1 13 07 5 8 3 872-
603-
- 4 67
3 1 0 -
= FIG.1. Polymerase chain reaction amplification of bacterial 16s rRNA genes using nested primers. Amplification of 467 base pair fragment was observed when 100 pg. of DNA from Salmonella tYPb (lane A), Staph. haemolyticus (lane C), Pseud. aerugmosa (lane D) and E, coli (lane G ) was added to reaction. N~ amplification was observed with addition of no input DNA (lane B), 1 pg. of DNA from Candida dbi-s (lane E), 1 a.of purchased human DNA (Promega, lane F), Or a.Of human DNA OUT method (lane H). Numbers to left of gel represent migration of DNA standards in base -, Number to right of gel shows migration position of 467 base fragment.
LNA
found that commercial preparations of DNA polymerase isolated from Thermus aquaticus are contaminated to varying degrees with DNA from these thermophiles. Using our nested primer PCR procedure, a 467 base pair gene fragment could be readily amplified if no attempt was made to remove the contaminating Thermus aquaticus DNA from the commercial enzyme preparations. Identity of the amplified fragment as being derived from Thermus aquaticus 16s rRNA gene was confirmed by DNA sequencing. However, if the DNA polymerase preparations were treated with DNase prior to use in PCR, amplification of contaminating bacterial DNA was eliminated (for example fig. 1, lane B). Second, it was important to use DNA polymerase isolated from Thermus aquaticus and not recombinant Taq DNA polymerase isolated from E. coli. This assured that if E. coli DNA was amplified from patient biopsies, the presence of such amplified fragments could not be due to E. coli DNA contamination of the polymerase preparations. Finally due to the sensitivity of this technique, DNA extractions and assembly of PCR were carried out in separate laboratories using sterile conditions t o assure that no contaminating bacterial DNA was introduced. The sensitivity of nested PCR coupled with the
1323
ability to eliminate exogenously introduced bacterial DNA permitted an evaluation of the presence of bacterial DNA in the bladders of patients with IC. Fourteen patients (table 2) with IC were identified based on the criteria set forth by the National Institute of Arthritis, Diabetes, Digestive and Kidney Diseases." These patients had no ordinary bacteria in their urine as judged by routine laboratory cultures in accordance with the established criteria for patients with IC. The control subjects are listed in table 3. Figure 2 shows 4 reactions containing an amplified fragment of approximately 467 base pairs (fig. 2, panel A) that hybridized with a radiolabelled probe specific for bacterial 16s rRNA genes (fig. 2, panel B). Upon breaking the sample code, we discovered that 4 of 14 (table 4) patients (approximately 29%) with IC showed the presence of bacterial 1 6 s rRNA genes, whereas no amplification was seen in the reactions containing input DNA from control subjects or in 4 reactions containing no input DNA. To further assure that bacterial 16s rRNA genes had been amplified and to identify the bacterial genera present in the 4 patients' biopsies, the amplified gene fragments were cloned and sequenced. Cloning of amplified fragments into the pGEM-3Z vector was facilitated by the incorporation of restriction endonuclease sites into the second set of nested PCR primers. At least 2 different clones from each of the 4 amplified fragments were sequenced, and the results of the analysis for homology with other bacterial 1 6 s rRNA gene sequences are shown in table 5. Using the Ribosomal RNA database as a source of known bacterial gene it was clear that amplified fragments from all 4 patients represented Gram-negative bacteria. When screened for homology against 750 16s rRNA gene sequences in this database, 2 of the amplified fragments were most similar to E. coli, and 2 were most similar to members of the genus, Pseudomonas. The fluorescent chlamydial monoclonal antibody test was negative on all urine specimens studied. Although the routine Mycoplasma cultures (M. hominis and Ureaplasma urealyticum) on urine were negative on all patients and control subjects, studies are presently in progress with specific molecular probes for difficult-to-culture genital mycoplasmas (M. genitalium) which could have been missed in the amplification studies utilizing primers amplifying the 16s rRNA gene. Only patients with negative routine urine cultures are reported in the molecular results. However, there were 2 patients with a clinical diagnosis of IC who had evidence of urinary tract infection: patient B.O.M., a 74-year-old female had >100,000 cfdml. E. coli in the urine and 1,000 cWml. (containing approximately 1 mg. of tissue) E. coli in the biopsy specimen cultured. Patient M.C.M., a 40-year-old female had 1,000 c f d d . Serratia marcescens in the urine and 1,400 cfdml. Serr. marcescens in the bladder tissue cultured.
TABLE2. Characteristics of interstitial cystitis patients Patient (he)
PCR Results
Gel Lane in Fie. 2
History of UTI
KEJ (35) CAC (60) MOR (86) RHE (62) MCD (36) SES (33) SAS (57) LED (57)
+ +
4 27 9 2 10 3 14 13
No No Yes No No No No Yes
PIG (37) WER (38) INR (71) JEE (79) JAS (45) WAL (25)
-
21 20 7 23 15 25
Yes Yes No No
-
-
-
+ +
Organism cfu/ml.
Time of Culture: DayslMonths PreA'ostbiopsy
Streptococcus 10,000
13 mos. postbiopsy
Entemoccus >100.OOO E. cob >100,ooO Anaerobes 10,OOO Candida albicans >lOO,OOO
18 mos. postbiopsy 21 mos. postbiopsy 5 days postbiopsy
No Yes
E. coli >lOO,OO
6 mos. postbiopsy
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
1324
TABLE3. Characteristics of control subjects Patient
(Age)
Diagnosis
Organism
History of UTI
Time of Culture: Months Mostbiopsy
&ml
No Cystocele 12 mos. postbiopsy Lactobacillus >100,000 Yes Urethral Diverticulum 31 mo8. postbiopsy Entemcoceus >100,000 Yes urethritis No Hematuria 11 mos. postbiopsy Corynebacteria 50,000 Yes MIL (69) Bladder Diverticulum 1 mo. postbiopsy Staphylococcus >100,000 Yes Prostatodynia ROP (72)** No Bladder Instability TRJ (42) No Urethral Syndrome WIA (29)** No Painful bladder HEE (36)** 2 mos. prebiopsy Yes Stress Incontinence LES (37) 12 mos. postbiopsy Yes Stress Incontinence Juc (56) No HOP (54) Stress Incontinence No Stress Incontinence BRJ (64) No Pelvic Prolapse GEL (55) TOJ (44) Stress Incontinence No * All bladder biopsies from 15 subjects were negative by PCR. ** These patients did not fulfillthe NIH criteria for an IC diagnosis. The endoscopic and pathological findings were completely normal.
BRL (61)
FIC (51) FRL (44)** GIM (41)
Although not qualifying for analysis in the molecular data, a control subject's biopsy tissue grew 400 colonies of Staph. epidermidis and was positive by PCR. These data further attest to the sensitivity of the PCR methodology for detection of bacteria in the bladder tissue. Unusual 0.22 Fm. filterable forms were isolated from the modified PPLO solid medium and broth cultures inoculated
TABLE4. PCR amplification of the 16s rRNA gene from urinary bladder tissue PCR Positive Negative
Total
#I.C. Patients 4 10 14
#Controls 0 15 15
with biopsy tissue from 14 of 14 IC patients and 1 of 15 controls. These forms resembled cell wall deficient bacteria in gross colony morphology on solid media and in liquid culture (table 6). Interestingly these forms were also demonstrated in the serum and urine specimens of 14 of 14 IC patients and in 3 of 15 urine specimens and 1 of 10 sera from controls. These forms contain DNA as detected by the fluorescent acridine orange stain and by the phenol-chloroform method for chemical extraction of DNA. The structures do not appear to be osmotically fragile, as they can grow in nonosmotically stabilized media (approximately 300 m0sm.l kg. H,O). They do not Gram stain, but are stained by malachite green. These forms have a n absolute requirement of serum for growth. This growth requirement is not fulfilled by substituting cholesterol for serum in the medium, and they are difficult to propagate in subculture. The electron photomicrograph composite of the filterable forms found in culture from the biopsy tissue of patient W.A.L. is shown in figure 3 (see detailed description of these forms in the photo legend). The cultured form's unusual ultrastructure suggests a heretofore unclassified organism. DISCUSSION
It is likely that the number of IC patients determined to be positive for the presence of bacterial 1 6 s rRNA genes using this methodology is a minimal estimate. This likelihood is based upon several technical considerations which include: 1) the limited amount of biopsy tissue available for study (approximately 1 mg.); 2) the limitation of using a single biopsy as representative of bacterial DNA in the entire bladder; 3) FIG.2. Amplification of bacterial 16s rRNA genes from intersti- the need for bacterial DNA in the tissues to be sufficiently tial cystitis patients' bladder biopsies. Every slxth reaction (lanes 6, 12, 18, and 24) was negative control containing no i n r t DNA. Nine intact so that PCR amplification can occur; 4) the necessity, reactions included in ut DNA from patients with uro ogwal dweases in treating Taq DNA polymerase preparations with DNase, other than interstitiar cystitis (lanes 1,5,8, 11,16,17,19,22and 26). to eliminate Thermus aquaticus DNA contamination prior to Fkmainin 14 reactions included input DNA from patients with use in PCR, which reduced the sensitivity of the assay; and 5) and interstitia? cystitis (lanes 2,3,4,7,9,10,13,14,15,20,21,23,25 27). Four reactions (lanes 4, 9, 15 and 25) showed amplification of the possibility that the primers selected may not be suitable 467 base pair fragment, and all of these positive reactions were for PCR amplification of all possible bacterial species that result of using input DNA from patients with interstitial cystitis. might be present. Despite these technical limitations, 4 of 14 Numbers to right of gel represent ositions of DNA standards of IC patients were positive for Gram-negative bacterial 16s indicated number of base pairs. Eac! PCR am lified fragment was hybridized with this radiolabelled oligonucleoti$e (lanes 4.9, 15 and rRNA genes; all 15 controls were negative. It is of interest that patient W.A.L. developed a urinary 25). Number to right of Southern blot represents migration position of standard 467 base pair DNA fragment. infection caused by E. coli 6 months after the biopsy was
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
1325
TABLE5. Sequence analysis of PCR amplified DNA from bladder biopsies of interstitial cystitis patients with known 16s rRNA genes Patient Code
Presence of' Bacterial 16s rRNA Genes in Bladder Biopsies
Alignment of Amplified DNA Sequenees with Known Bacterial 16s rRNA Genes
Homology of Amplified DNA Sequences with the Indicated Gram-Negative 16s rRNA Gene
KEJ Yes (fig. 2, lane 4 ) E. coli 98.5% MOR Yes (fig. 2, lane 9) Pseud. maltophilia2 96.2% JAS Yes (fig. 2, lane 15) 96.7% Pseud. mendocina WAL Yes (fig. 2, lane 25) E. coli 95.5% 'See figure 2. * It has been suggested that the taxonomic placement of Pseud. maltophilia should be with the genus. Xanthomonaa, to indicate differences with other Paeudomonad~.~~
TABLE6. Isolation of filterable forms from IC patients and control subjects Number Positivflotal Number Cultured Subjects Bladder Biopsy
1
IC Patients 14/14 Controls 1"/15 .Subject: WIA Subjects: GIM, MIL, FLOP Subject: FIC
Urine
serum
14/14 3b115
14/14 1'110
obtained. This patient's bladder tissue previously indicated the presence of the 1 6 s rRNA gene whose alignment of amplified DNA sequences most closely resembled E. coli. It is unlikely that the amplified DNA represents ''left over" DNA from a previous infection. For example, control subject L.E.S. had a Proteus infection 2.5months prior to the biopsy; no Proteus DNA could be amplified from the biopsy tissue specimen, which suggests that bacterial DNA does not persist for prolonged periods after an infection has been cleared (antibiotic therapy). One might argue that damaged urothelium such as is found in IC would be less resistant to the adherence of bacteria and that the bacterial DNA found is uninvolved in the IC process. However, the sensitive culture methods for detecting
FIG.3. Electronmicrograph composite of filterable forms grown in culture bio sy tissue of patient W.A.L. Magnification indicated by calibrated gar. Forms consist of myelin-like swirl forming central core-likestructure, and a tail. Core diameter ranges between 40 to 80 nm. and can be electron lucent or electron dense. Electron lucent cores display packing of swirls in various configurations, some of which seem geometrically organized (single arrow). Tail is continuation of core and usually has same number of myelin-like sheets as core. Arrangement of sheets around core is similar to afferent myelinated s o n s (small arrows).Electron dense material around forms corresponds to a ar used for holding them together during dehydration and embedgng. When core becomes electron dense (u per left insert), sheets are difficult to discern (asterisk); spacing &tween individual sheets remains in register of approximately 5 to 7 nm. (arrowheads).Periodicity of sheets in tail remains in register as well (upperright insert), and distancebetween individual lamella is same as core (arrowheads).True myelin in nervous system also has periodicity of ap roximately 5 to 7 nm. However, contrary to true myelin double memtrane spacing, spacing of form's membrane-likesheets is tighter (arrowheadsin both inserts). True myelin is mentioned here as guide for spacing between lamellae. No molecular relationship between myelin structure and these cultured Nterable forms is implied.
indicate that these organisms are not viable, perhaps because they have been destroyed by the resulting immune response. Either of these two possibilities could account, not only for the lack of culturable bacteria, but also for the presence of an idammatory response in this tissue. Structurally, aberrant bacteria, including cell wall deficient or nutritionally deficient forms, require extraordinary efforts to culture in vitr0,2~and our preliminary results suggests that such forms may be present in the tissues of IC patient^?^.^' The ultrastructure of the filterable forms resembles myelin. Although no molecular relationship is implied between true myelin and these forms, it would nevertheless be of interest to determine whether these cultured forms have a tropism for nervous tissue. The relationship (if any) of the amplified Gram-negative bacterial DNA to the unidentified 0.22 pm. filterable form remains to be determined. The unusual, filterable forms isolated from the bladder tissue of 14 of 14 IC patients and from 1 of 15 controls (nonIC subjects) represent novel and provocative findings. The extent of the presence of unusual filterable forms in the sera of healthy and diseased humans (other than those with
1326
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
IC) is presently under investigation. Although specialized culture methods demonstrated their presence i n bladder tissue, their DNA was not amplified by the primers utilized for amplification of nucleic acids extracted from the biopsy tissue. This suggests that the 16s rRNA gene may not be present in these forms or, if present, that the experimental conditions used for amplification may have to be modified. It remains to be determined whether these bizarre structures are symbionts found in a variety of hosts. It is therefore necessary to characterize these unusual forms biochemically a n d immunologically and study the host response in an attempt to determine their exact role in the host-microbial interaction. If the filterable structures are symbionts (with the potential of becoming pathogens) and exist in various forms in diseased and healthy subjects, defining their role as infectious agents will require information as to which mechanisms allow these organisms to be tolerated by the host during the healthy state a n d which are responsible for producing disease. It is of interest that we have shown filterable forms of similar morphology, yet antigenically dissimilar, in supposedly sterile sera of some nonhuman animals. The presence of Gram-negative and/or filterable forms in tissues a n d body fluids does not prove that they are the cause of IC. However, it would be difficult to imagine that such a microbial presence in the bladder has no effect o n the disease and its resulting morbidity. By identifying patients having microbes in the bladder, i t should be possible to determine if antimicrobials aid in the treatment of the disease. It is logical to suggest that even if organisms are not causative agents, their presence may lead to immune and host cell responses that could exacerbate the inflammatory state. Thus it would be predicted that appropriate treatments to minimize microbial presence in this tissue may significantly improve the morbidity associated with interstitial cystitis. Acknowledgement. Data used in preparing table 2 and figure 3 were derived h m the Ribosomal Database Project (RDP) email updated on February 18,1993with d a t a release 2.1.We thank G.J. Olson, B. Maidak and M. McCaughey at the RDP (University of Illinois, Urbana, Illinois) for help with sequence alignment and construction of the phylogenetic tree. D. Neal, former faculty member at Tulane Medical Center (presently at University of Texas Medical School, Galveston, Texas), provided specimens from 1 interstitial cystitis patient and 1 control patient. We thank Dale S. Martin for expert technical assistance in electron microscopy. REFERENCES
1. Parivar, F. and Bradbmk, R. A.: Interstitial cystitis. Br. J. Urol., MI: 239, 1986. 2. Holm-Bentzen,M.: Pathology, pathophysiology, and pathogenesis of painful bladder disease. Urol. Res., 17: 203, 1989. 3. Messing, E. M.: The diagnosis of interstitial cystitis. Urology, suppl. 4,29: 4, 1987. 4. Hanno, P., Levin, R. M., Monson, F. C., Teuscher, C., Zhou, Z. Z., Ruggieri, M., Whitmore, K and Wein, A. J.: Diagnosis of interstitial cystitis. J. Urol.. 143: 278. 1990. 5. Koziol, J. A., Clark, D. C., Gittes, R. F. and Eng, M. T.: The
natural history of interstitial cystitis: a survey of 374 patients. J. Urol., 149 465, 1993. 6. Messing, E. M. and Stamey, T. A.: Interstitial cystitis: early diagnosis, pathology, and treatment. Urology, 1 2 381,1978. 7. Oravisto, K J.: Interstitial cystitis as an autoimmune disease: a review. Eur. Urol., 6 10, 1980. 8. Fall, M., Johansson, S. L. and Aldenborg, F.: Chronic interstitial cystitis: a heterogeneous syndrome. J. Urol., 137: 35,1987. 9. Mattila, J., Harmoinen, A. and Hallstorm, 0.: Serum immunoglobulin and complement alterations in interstitial cystitis. Eur. Urol., 9 350, 1983. LO. Aldenborg, F., Fall, M. and Enerback, L.: Proliferation and transepithelial migration of mucosal mast cells in interstitial cystitis. Immunology, 58: 411,1986. 11. Anderson, J. B., Parivar, F., Lee, G., Wallington, T. B., MacGiver, A. G., Bradbrook, R. A. and Gingell, J. C.: The enigma of interstitial cystitis-an autoimmune disease. Br. J. Urol., 63:58, 1989. 12. MacDermott, J.P., Miller, C. H., Levy, N. and Stone, A. R.: Cellular immunity in interstitial cystitis. J. Urol., 145:274,1991. 13. Miller, C. H.,MacDermott, J. P., Quatrocchi, K. B., Broderick, G. A. and Stone, A. R.: Lymphocyte function in patients with interstitial cystitis. J. Urol., 147: 592, 1992. 14. Mullis, K B., and Faloona, F.: Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods Enzymol. 155:335, 1987. 15. Gillenwater,J. Y.and Wein, A. J.: Summary of the National Institute of Arthritis,Diabetes, Digestive and Kidney Disease workshop on interstitial cystitis, National Institutes of Health, Bethesda, Maryland, August 28-29,1987.J. Urol., 140:203,1988. 16. Domingue, G. J., Thomas, R., Walters, F., Serrano, A. and Heidger, P. M., Jr.: Cell wall deficient bacteria as a cause of idiopathic hematuria. J. Urol., 150 483,1993. 17. Sarkar, G. and Sommer, S. S.: Shedding light on PCR contamination. Nature, 343:27, 1990. 18. Schmidt, T. M., Pace, B. and Pace, N. R.: Detection of DNA contamination in taq polymerase. BioTechniques, 11: 176,1991. 19. Rochelle, P. A,, Weightman, A. J. and Fry, J. C.: DNase I treatment of taq DNA polymerase for complete PCR decontamination. BioTechniques, 13: 520, 1992. 20. Selden, R. F.: Analysis of DNA sequences by blotting and hybridization. In: Current Protocols in Molecular Biology. Edited by F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith and K. Struhl. New York: Wiley, vol. 1, unit 2.9,1989. 21. Olsen, G. J., Larsen, N. and Woese, C. R.: The ribosomal RNA database project. Nucleic Acids Res., suppl., 1 9 2017,1991. 22. Olsen, G. J.,Overbeck, R., Larsen, N., Marsh, T. L., McGaughey, M. J., Maciukenas, M. A., Kuan, W. M., Macke, T. J., Xing, Y. and Woese, C. R.: The ribosomal database project. Nucleic Acids Res., suppl., 2 0 2199,1992. 23. Domingue, G. J.: Cell Wall-Deficient Bacteria: Basic Principles and Clinical Significance. Reading Massachusetts: AddisonWesley Publishing Co., 1982. 24. Meir, A, Swings, J., De Vos, M., M e n , M. and Bottger, E. C.: Elimination of contaminating DNA within polymerase reaction reagents: implications for a general approach to detection of uncultured pathogem. J. Clin. Micmbiol., 31: 646,1993. 25. Swings, J., Meier, A., Persing, D. H., Finken, M. and Bottger, E. c.:Transfer of Pseudomonas maltoDhilia Hueh 1981 to the genus Xanthomonas as Xanthomoias maltoihilia (Hugh 1981)comb. Nov. Int. J. Syst. Bacteriol., 33: 409,1983.
DORMANT MICROBES IN INTERSTITIAL CYSTITIS GERALD J. DOMINGUE", GAMAL, M. GHONIEM, KENNETH L. BOST, CESAR FERMIN LISET G. HUMAN
AND
From the Departments of Urology, Microbiology and Immunology and Pathology, Tulane University, New Orleans, Louisiana
ABSTRACT
Interstitial cystitis (IC) is a n inflammatory disease of the urinary bladder that has no known etiology. A microbial association with this disease has not been supported since routine cultures of urine from IC patients are usually negative. However, we have demonstrated the presence of bacterial 1 6 s rRNA genes in bladder biopsies from 29%of patients with IC, but not from control patients with other urological diseases. The ability to identify the presence of bacterial DNA in these patients was accomplished using a sensitive and specific nested PCR method capable of amplifylng 16s rRNA genes from a wide variety of bacterial genera. Cloning and sequencing of 16s rRNA gene fragments amplified from bladder tissue of IC patients showed that these genes were derived from genera representing Gram-negative bacteria. In addition to the molecular data, a novel finding of 0.22 pm. filterable forms has been isolated in culture from the biopsy tissue of 14 of 14 IC patients and from 1 of 15 controls. The forms contain nucleic acids and resemble cell wall-deficient bacteria in gross morphology; however, their swirled myelin-like ultrastructure is unusual and suggests a heretofore unclassified microbe. These results demonstrate for the first time an association of Gram-negative bacterial DNA and filterable forms with affected bladder tissue from patients with IC. KEY WORDS:bladder, cystitis, bacteria, inflammation, etiology
Interstitial cystitis (IC), a n inflammatory disease of the bladder, has a n unknown etiology. Disease states predominate in women and have varying degrees of morbid it^.^-^ In the most severe cases the disease is debilitating and characterized by small bladder capacity, ulcers, severe frequency, nocturia, dysuria and pain. Quality of life scores in these patients have been measured to be lower than in chronic renal disease patients. Treatment for the disease is varied and includes bladder distension therapy and sometimes surgical removal of the bladder. The limitations of current therapies result in part from the undetermined cause of the disease. It has been suggested that IC may be an autoimmune however, no organ-specific immune response has been demonstrated to induce or be responsible for the d i s e a ~ e . ~ Furthermore, .'~ recent investigations argue against immune abnormalities in these A microbial association with IC has not been demonstrated.' This disease is distinct from other bladder infections in which microbes can be readily cultured from urine. Thus the inability to routinely culture bacteria has been suggested to indicate that bacteria are not involved in the etiology or morbidity associated with the disease. The present study uses the polymerase chain reaction (PCR)I4 to detect the presence of bacterial DNA in bladder biopsies from patients with IC and also uses specialized culture techniques to grow fastidious organisms that may be difficult to culture. 's2
MATERIALS AND METHODS
Patients, biopsies and cultures. Informed consent was obtained from all patients and control subjects participating in this study. The exact nature of the study and any possible consequences were explained to each participant. AU patients were female except for R.O.P. and M.I.L. Patients were diag-
nosed with IC according to the criteria set forth by the National Institute of Arthritis,Diabetes, Digestive and Kidney Disease ~Two. separate ~ ~ cold cup biopsies (approximately 1mg.) were surgically removed from intlamed sites in the bladder wall of IC patients or from the normal bladder wall of control subjects using sterile techniques. One set of biopsy samples was homogenized immediately and was exhaustively evaluated for the presence of bacteria using culture techniques described be10w.l~ The second biopsy specimen was stored as a coded sample at -7OC until total DNA was extracted. Catheterized urine specimens and biopsy bladder tissue from each of these subjects were cultured for ordinary and fastidious bacteria on the following media: 5%sheep blood agar and MacConkey agar incubated aerobically, anaerobically and under a 5% CO, atmosphere;'6 tryptme broth for revealing
Accepted for publication November 30, 1994. * Requests for reprints: .Department of Urology, SL42, Tulane University School o Medicine, 1430 Tulane Avenue, New Orleans, Louisiana 70112. This work was supported by N.I.H. Grant DK44812-03 awarded to G m l d J. Domingue and donations to Gerald J. Domingue by the Interstitial Cystitis Association and Mr. and Mrs. W. Russell Harp. 1321
DORMANT MICROBES I N INTERSTITIAL CYSTITIS
1322
bladder tissue. Deoxyribonucleicacid was isolated using aseptic conditionsby lysing cells with SDS followed by sequential treatment of the sample with pronase and RNase as described by the manufacturer (Stratagene, La Jolla, California). Precipitated DNA was resuspended in water and quantified using absorbance at 260 nm. Using a method which assured that exogenous DNA was not introduced into the sample, total DNA extracted from the biopsies was amplified by PCR using nested primers. The appropriate DNA was added to a PCR14 containing 5 units Thermus aquaticus DNA polymerase (Boehringer Mannheim, Indianapolis, Indiana), 2 mM. dNTPs, 1pg. of each primer and PCR buffer (10 mM. Tris-HC1, 50 mM. KCl, 1 pg./ml. gelatin and 2 mM. IvIgC1,). Positive displacement pipetters were used to ensure that no crosscontamination occurred. The PCR buffer, water and mineral oil were exposed to W light to destroy any contaminating DNA that might be present in these it was of critical importance to treat ~ o i u t i o nFurthermore, ~.~~ the Thermus aquaticus DNA polymerase with bovine pancreatic DNase (Promega, Madison, Wisconsin) to remove contaminating bacterial DNA present in commercial preparations of this e n ~ y m e . ' ~ "Thus ~ , prior to inclusion in PCR, 1 unit of DNase was used to treat 20 units of Thermus aquaticus DNA polymerase in PCR buffer for 1 hour at 37C. The DNase was heat inactivated at 95C for 10 minutes prior to using the DNase-treated DNA polymerase in PCR. Reactions were brought to 70C prior to addition of 5 units of DNase-treated Thermus aquaticus DNA polymerase to minimize nonspecific priming. Reactions were taken through 25 cycles using a BIOCYCLER thermal cycler (BIOS, New Haven, Connecticut) programmed for 95C denaturation, 58C annealing and 72C extension temperatures, with the first three cycles having extended denaturation and annealing times. The primers used are shown in table 1.The initial PCR was performed using the 5' primer, lA,and the 3' primer, 5. ARer completion of the first PCR, 10% of each reaction was used for a second PCR using the same conditions described above with the 5' primer, 3, and the 3' primer, 4. Primers were selected for 1)their ability to amplify the 16s rRNA gene from a wide variety of bacterial genera, 2) the lack of significant complementarity between primers, 3)the lack of significant secondary structure, 4) optimal annealing temperatures and 5)significant homology of the 5' primers with 16s rFWA genes, but not with other microbial or human genes contained in the Genbank database version 69. ARer the second PCR, 2 W of the reaction was electrophoresedon 1.3% agarowl ethidium bromide gels and the presence of a 467 base pair fragment was detected by U V fluorescence. Southern blot analyses to demonstrate hybridization of the amplified DNA Fagments with an oligonucleotide specific for bacterial 16s rRNAgenes. Southern blot analyses were performed essentially as described.20 Briefly, duplicate agarose gels were denatured, neutralized and transferred to nylon membranes by positive pressure (Possiblot, Stratagene). The DNA was fixed by exposure to W light (Stratalinker, Stratagene), and membranes were prehybridized overnight at
45C. Hybridization with a 32P end-labelled oligonucleotide probe (CYT GGA GGA AGG TGG GGA TGA C, where Y equals C or T) specific for bacterial 1 6 s rRNA genes was performed overnight a t 45C in the presence of 4X SSC. After washing, hybridization was detected by autoradiography. Bacterial identitkatwn. To identify the type(s) of bacteria present in IC patients having 16s rRNA genes amplified from their bladder biopsies, these gene fragments were cloned and sequenced. Gene fragments were isolated from agarose gels using Prep-a-gene (Bio-Rad, Richmond, California) and ligated into the pGEM-3Z vector (Promega). This ligation was accomplished by incorporatmg, respectively, EcoRl and BamHl restriction endonuclease sites into the 5' and 3' primers used in the secondary PCR. Vector and amplified DNA fragments were cleaved with these enzymes prior to ligation. Escherichia coli (strain JM83) made competent with calcium chloride were transformed and plated onto Luria agar plates containing ampicillin, X-gal and ETG. White colonies were selected and expanded, and plasmids were isolated from these clones by minipreping (Promega). Sequencing of plasmids containing inserts was accomplished by the dideoxy chain termination method using Squenase 2.0 and instructions supplied by the manufacturer (United States Biochemical). Sequencing primers used are given in table 1. At least two W e r e n t clones were sequenced for each of the 4 isolated fragments. Sequences of approximately 400 base pairs were analyzed for optimal alignments against 750 different 16s rRNA gene sequences contained in the Ribosomal RNA database using alignment softKnown bacterial ware programs available at this facility.21*22 sequences having the most similar alignments with DNA fragments amplified from the biopsy tissue were furnished by the Ribosomal RNA database project. Homology of DNA fragments amplified from the bladder tissue with known bacterial 16s rRNA genes identified by the Ribosomal RNA database project as having the best alignments were determined using MacVector sequence analysis software (IBI, New Haven, Connecticut). RESULTS
A sensitive and specific PCR capable of amplifying the 1 6 s rRNA genes from a wide variety of bacterial genera using nested primers was developed. As shown in figure 1, nested priming amplified a 467 base pair fragment when 100 pg. of input DNA from Salmonella typhi (lane A), Staphylococcus haemolyticus (lane C), Pseudomonas aeruginosa (lane D), or Escherichia coli (lane G)were used. Conversely, amplification was not detected with the addition of no input DNA (lane B), Candida albicans DNA (lane E) or human DNA (lanes F and HI. Identity of the appropriate bacterial genus was confirmed by cloning and sequencing the PCR amplified bacterial 16s rRNA gene fragments. To develop this nested PCR, several significant technical problems had to be overcome. First, we and have
TABLE1. Oligonucleotide primers Oligonucleotides (Primers, p)
lA
Sequence 5' to 3'
CAC AAG CGG TGG AGC ATG TGG 'IT
16s rR.NA Position
Strand Orientation
932-955
+
Reference Domingue et. a1 (this publication)
CCT ACG CYT ACC TX 'ITA CGA CT Y equals C or T 1514-1492 GGA A'IT CTG CAA CGC GAA CCT TAC TA 1168-1194 + 1392-1371 GCG GAT CCT GGT KTG ACG GGC GGT TG TA K equals G or T GAACCTTACCTGGKYTXACATCCCAAGGCCCGGGA ACG TAT TCA CCG CYT GGA GGA AGG TGG GGA TGAC _. Y equals C or T and K equals G or T 246 AGA G'IT TGA TCC TGG CTC AG 8-28 + _7 4_ 240 GGG TCA A'IT CCT 'ITG AGT 'IT 928-908 244 CCC ACT GCT GCC TCC CGT AG 361-342 .-a .ry. F. wnLtna nrimnr. The primera selected for the ability to amplifythel6S rRNA gene were plAL l..u ra.yL=.u p3 and p4. Nested PCR with p3 and p4 produced amplification of a 460-bp fragment of bacterial DNA. Additional primers amplifyingthe 16s rRNA gene were used with DNA extracted from tissue of 6 controls:x p246 and p240 yield a 900-bp amplified produet, p246 and p244 amplify a 300-bp bacterial DNA fragment 5 3 4 Sequencing Primers
~
..--_..
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
r A B C D E F G H
1 13 07 5 8 3 872-
603-
- 4 67
3 1 0 -
= FIG.1. Polymerase chain reaction amplification of bacterial 16s rRNA genes using nested primers. Amplification of 467 base pair fragment was observed when 100 pg. of DNA from Salmonella tYPb (lane A), Staph. haemolyticus (lane C), Pseud. aerugmosa (lane D) and E, coli (lane G ) was added to reaction. N~ amplification was observed with addition of no input DNA (lane B), 1 pg. of DNA from Candida dbi-s (lane E), 1 a.of purchased human DNA (Promega, lane F), Or a.Of human DNA OUT method (lane H). Numbers to left of gel represent migration of DNA standards in base -, Number to right of gel shows migration position of 467 base fragment.
LNA
found that commercial preparations of DNA polymerase isolated from Thermus aquaticus are contaminated to varying degrees with DNA from these thermophiles. Using our nested primer PCR procedure, a 467 base pair gene fragment could be readily amplified if no attempt was made to remove the contaminating Thermus aquaticus DNA from the commercial enzyme preparations. Identity of the amplified fragment as being derived from Thermus aquaticus 16s rRNA gene was confirmed by DNA sequencing. However, if the DNA polymerase preparations were treated with DNase prior to use in PCR, amplification of contaminating bacterial DNA was eliminated (for example fig. 1, lane B). Second, it was important to use DNA polymerase isolated from Thermus aquaticus and not recombinant Taq DNA polymerase isolated from E. coli. This assured that if E. coli DNA was amplified from patient biopsies, the presence of such amplified fragments could not be due to E. coli DNA contamination of the polymerase preparations. Finally due to the sensitivity of this technique, DNA extractions and assembly of PCR were carried out in separate laboratories using sterile conditions t o assure that no contaminating bacterial DNA was introduced. The sensitivity of nested PCR coupled with the
1323
ability to eliminate exogenously introduced bacterial DNA permitted an evaluation of the presence of bacterial DNA in the bladders of patients with IC. Fourteen patients (table 2) with IC were identified based on the criteria set forth by the National Institute of Arthritis, Diabetes, Digestive and Kidney Diseases." These patients had no ordinary bacteria in their urine as judged by routine laboratory cultures in accordance with the established criteria for patients with IC. The control subjects are listed in table 3. Figure 2 shows 4 reactions containing an amplified fragment of approximately 467 base pairs (fig. 2, panel A) that hybridized with a radiolabelled probe specific for bacterial 16s rRNA genes (fig. 2, panel B). Upon breaking the sample code, we discovered that 4 of 14 (table 4) patients (approximately 29%) with IC showed the presence of bacterial 1 6 s rRNA genes, whereas no amplification was seen in the reactions containing input DNA from control subjects or in 4 reactions containing no input DNA. To further assure that bacterial 16s rRNA genes had been amplified and to identify the bacterial genera present in the 4 patients' biopsies, the amplified gene fragments were cloned and sequenced. Cloning of amplified fragments into the pGEM-3Z vector was facilitated by the incorporation of restriction endonuclease sites into the second set of nested PCR primers. At least 2 different clones from each of the 4 amplified fragments were sequenced, and the results of the analysis for homology with other bacterial 1 6 s rRNA gene sequences are shown in table 5. Using the Ribosomal RNA database as a source of known bacterial gene it was clear that amplified fragments from all 4 patients represented Gram-negative bacteria. When screened for homology against 750 16s rRNA gene sequences in this database, 2 of the amplified fragments were most similar to E. coli, and 2 were most similar to members of the genus, Pseudomonas. The fluorescent chlamydial monoclonal antibody test was negative on all urine specimens studied. Although the routine Mycoplasma cultures (M. hominis and Ureaplasma urealyticum) on urine were negative on all patients and control subjects, studies are presently in progress with specific molecular probes for difficult-to-culture genital mycoplasmas (M. genitalium) which could have been missed in the amplification studies utilizing primers amplifying the 16s rRNA gene. Only patients with negative routine urine cultures are reported in the molecular results. However, there were 2 patients with a clinical diagnosis of IC who had evidence of urinary tract infection: patient B.O.M., a 74-year-old female had >100,000 cfdml. E. coli in the urine and 1,000 cWml. (containing approximately 1 mg. of tissue) E. coli in the biopsy specimen cultured. Patient M.C.M., a 40-year-old female had 1,000 c f d d . Serratia marcescens in the urine and 1,400 cfdml. Serr. marcescens in the bladder tissue cultured.
TABLE2. Characteristics of interstitial cystitis patients Patient (he)
PCR Results
Gel Lane in Fie. 2
History of UTI
KEJ (35) CAC (60) MOR (86) RHE (62) MCD (36) SES (33) SAS (57) LED (57)
+ +
4 27 9 2 10 3 14 13
No No Yes No No No No Yes
PIG (37) WER (38) INR (71) JEE (79) JAS (45) WAL (25)
-
21 20 7 23 15 25
Yes Yes No No
-
-
-
+ +
Organism cfu/ml.
Time of Culture: DayslMonths PreA'ostbiopsy
Streptococcus 10,000
13 mos. postbiopsy
Entemoccus >100.OOO E. cob >100,ooO Anaerobes 10,OOO Candida albicans >lOO,OOO
18 mos. postbiopsy 21 mos. postbiopsy 5 days postbiopsy
No Yes
E. coli >lOO,OO
6 mos. postbiopsy
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
1324
TABLE3. Characteristics of control subjects Patient
(Age)
Diagnosis
Organism
History of UTI
Time of Culture: Months Mostbiopsy
&ml
No Cystocele 12 mos. postbiopsy Lactobacillus >100,000 Yes Urethral Diverticulum 31 mo8. postbiopsy Entemcoceus >100,000 Yes urethritis No Hematuria 11 mos. postbiopsy Corynebacteria 50,000 Yes MIL (69) Bladder Diverticulum 1 mo. postbiopsy Staphylococcus >100,000 Yes Prostatodynia ROP (72)** No Bladder Instability TRJ (42) No Urethral Syndrome WIA (29)** No Painful bladder HEE (36)** 2 mos. prebiopsy Yes Stress Incontinence LES (37) 12 mos. postbiopsy Yes Stress Incontinence Juc (56) No HOP (54) Stress Incontinence No Stress Incontinence BRJ (64) No Pelvic Prolapse GEL (55) TOJ (44) Stress Incontinence No * All bladder biopsies from 15 subjects were negative by PCR. ** These patients did not fulfillthe NIH criteria for an IC diagnosis. The endoscopic and pathological findings were completely normal.
BRL (61)
FIC (51) FRL (44)** GIM (41)
Although not qualifying for analysis in the molecular data, a control subject's biopsy tissue grew 400 colonies of Staph. epidermidis and was positive by PCR. These data further attest to the sensitivity of the PCR methodology for detection of bacteria in the bladder tissue. Unusual 0.22 Fm. filterable forms were isolated from the modified PPLO solid medium and broth cultures inoculated
TABLE4. PCR amplification of the 16s rRNA gene from urinary bladder tissue PCR Positive Negative
Total
#I.C. Patients 4 10 14
#Controls 0 15 15
with biopsy tissue from 14 of 14 IC patients and 1 of 15 controls. These forms resembled cell wall deficient bacteria in gross colony morphology on solid media and in liquid culture (table 6). Interestingly these forms were also demonstrated in the serum and urine specimens of 14 of 14 IC patients and in 3 of 15 urine specimens and 1 of 10 sera from controls. These forms contain DNA as detected by the fluorescent acridine orange stain and by the phenol-chloroform method for chemical extraction of DNA. The structures do not appear to be osmotically fragile, as they can grow in nonosmotically stabilized media (approximately 300 m0sm.l kg. H,O). They do not Gram stain, but are stained by malachite green. These forms have a n absolute requirement of serum for growth. This growth requirement is not fulfilled by substituting cholesterol for serum in the medium, and they are difficult to propagate in subculture. The electron photomicrograph composite of the filterable forms found in culture from the biopsy tissue of patient W.A.L. is shown in figure 3 (see detailed description of these forms in the photo legend). The cultured form's unusual ultrastructure suggests a heretofore unclassified organism. DISCUSSION
It is likely that the number of IC patients determined to be positive for the presence of bacterial 1 6 s rRNA genes using this methodology is a minimal estimate. This likelihood is based upon several technical considerations which include: 1) the limited amount of biopsy tissue available for study (approximately 1 mg.); 2) the limitation of using a single biopsy as representative of bacterial DNA in the entire bladder; 3) FIG.2. Amplification of bacterial 16s rRNA genes from intersti- the need for bacterial DNA in the tissues to be sufficiently tial cystitis patients' bladder biopsies. Every slxth reaction (lanes 6, 12, 18, and 24) was negative control containing no i n r t DNA. Nine intact so that PCR amplification can occur; 4) the necessity, reactions included in ut DNA from patients with uro ogwal dweases in treating Taq DNA polymerase preparations with DNase, other than interstitiar cystitis (lanes 1,5,8, 11,16,17,19,22and 26). to eliminate Thermus aquaticus DNA contamination prior to Fkmainin 14 reactions included input DNA from patients with use in PCR, which reduced the sensitivity of the assay; and 5) and interstitia? cystitis (lanes 2,3,4,7,9,10,13,14,15,20,21,23,25 27). Four reactions (lanes 4, 9, 15 and 25) showed amplification of the possibility that the primers selected may not be suitable 467 base pair fragment, and all of these positive reactions were for PCR amplification of all possible bacterial species that result of using input DNA from patients with interstitial cystitis. might be present. Despite these technical limitations, 4 of 14 Numbers to right of gel represent ositions of DNA standards of IC patients were positive for Gram-negative bacterial 16s indicated number of base pairs. Eac! PCR am lified fragment was hybridized with this radiolabelled oligonucleoti$e (lanes 4.9, 15 and rRNA genes; all 15 controls were negative. It is of interest that patient W.A.L. developed a urinary 25). Number to right of Southern blot represents migration position of standard 467 base pair DNA fragment. infection caused by E. coli 6 months after the biopsy was
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
1325
TABLE5. Sequence analysis of PCR amplified DNA from bladder biopsies of interstitial cystitis patients with known 16s rRNA genes Patient Code
Presence of' Bacterial 16s rRNA Genes in Bladder Biopsies
Alignment of Amplified DNA Sequenees with Known Bacterial 16s rRNA Genes
Homology of Amplified DNA Sequences with the Indicated Gram-Negative 16s rRNA Gene
KEJ Yes (fig. 2, lane 4 ) E. coli 98.5% MOR Yes (fig. 2, lane 9) Pseud. maltophilia2 96.2% JAS Yes (fig. 2, lane 15) 96.7% Pseud. mendocina WAL Yes (fig. 2, lane 25) E. coli 95.5% 'See figure 2. * It has been suggested that the taxonomic placement of Pseud. maltophilia should be with the genus. Xanthomonaa, to indicate differences with other Paeudomonad~.~~
TABLE6. Isolation of filterable forms from IC patients and control subjects Number Positivflotal Number Cultured Subjects Bladder Biopsy
1
IC Patients 14/14 Controls 1"/15 .Subject: WIA Subjects: GIM, MIL, FLOP Subject: FIC
Urine
serum
14/14 3b115
14/14 1'110
obtained. This patient's bladder tissue previously indicated the presence of the 1 6 s rRNA gene whose alignment of amplified DNA sequences most closely resembled E. coli. It is unlikely that the amplified DNA represents ''left over" DNA from a previous infection. For example, control subject L.E.S. had a Proteus infection 2.5months prior to the biopsy; no Proteus DNA could be amplified from the biopsy tissue specimen, which suggests that bacterial DNA does not persist for prolonged periods after an infection has been cleared (antibiotic therapy). One might argue that damaged urothelium such as is found in IC would be less resistant to the adherence of bacteria and that the bacterial DNA found is uninvolved in the IC process. However, the sensitive culture methods for detecting
FIG.3. Electronmicrograph composite of filterable forms grown in culture bio sy tissue of patient W.A.L. Magnification indicated by calibrated gar. Forms consist of myelin-like swirl forming central core-likestructure, and a tail. Core diameter ranges between 40 to 80 nm. and can be electron lucent or electron dense. Electron lucent cores display packing of swirls in various configurations, some of which seem geometrically organized (single arrow). Tail is continuation of core and usually has same number of myelin-like sheets as core. Arrangement of sheets around core is similar to afferent myelinated s o n s (small arrows).Electron dense material around forms corresponds to a ar used for holding them together during dehydration and embedgng. When core becomes electron dense (u per left insert), sheets are difficult to discern (asterisk); spacing &tween individual sheets remains in register of approximately 5 to 7 nm. (arrowheads).Periodicity of sheets in tail remains in register as well (upperright insert), and distancebetween individual lamella is same as core (arrowheads).True myelin in nervous system also has periodicity of ap roximately 5 to 7 nm. However, contrary to true myelin double memtrane spacing, spacing of form's membrane-likesheets is tighter (arrowheadsin both inserts). True myelin is mentioned here as guide for spacing between lamellae. No molecular relationship between myelin structure and these cultured Nterable forms is implied.
indicate that these organisms are not viable, perhaps because they have been destroyed by the resulting immune response. Either of these two possibilities could account, not only for the lack of culturable bacteria, but also for the presence of an idammatory response in this tissue. Structurally, aberrant bacteria, including cell wall deficient or nutritionally deficient forms, require extraordinary efforts to culture in vitr0,2~and our preliminary results suggests that such forms may be present in the tissues of IC patient^?^.^' The ultrastructure of the filterable forms resembles myelin. Although no molecular relationship is implied between true myelin and these forms, it would nevertheless be of interest to determine whether these cultured forms have a tropism for nervous tissue. The relationship (if any) of the amplified Gram-negative bacterial DNA to the unidentified 0.22 pm. filterable form remains to be determined. The unusual, filterable forms isolated from the bladder tissue of 14 of 14 IC patients and from 1 of 15 controls (nonIC subjects) represent novel and provocative findings. The extent of the presence of unusual filterable forms in the sera of healthy and diseased humans (other than those with
1326
DORMANT MICROBES IN INTERSTITIAL CYSTITIS
IC) is presently under investigation. Although specialized culture methods demonstrated their presence i n bladder tissue, their DNA was not amplified by the primers utilized for amplification of nucleic acids extracted from the biopsy tissue. This suggests that the 16s rRNA gene may not be present in these forms or, if present, that the experimental conditions used for amplification may have to be modified. It remains to be determined whether these bizarre structures are symbionts found in a variety of hosts. It is therefore necessary to characterize these unusual forms biochemically a n d immunologically and study the host response in an attempt to determine their exact role in the host-microbial interaction. If the filterable structures are symbionts (with the potential of becoming pathogens) and exist in various forms in diseased and healthy subjects, defining their role as infectious agents will require information as to which mechanisms allow these organisms to be tolerated by the host during the healthy state a n d which are responsible for producing disease. It is of interest that we have shown filterable forms of similar morphology, yet antigenically dissimilar, in supposedly sterile sera of some nonhuman animals. The presence of Gram-negative and/or filterable forms in tissues a n d body fluids does not prove that they are the cause of IC. However, it would be difficult to imagine that such a microbial presence in the bladder has no effect o n the disease and its resulting morbidity. By identifying patients having microbes in the bladder, i t should be possible to determine if antimicrobials aid in the treatment of the disease. It is logical to suggest that even if organisms are not causative agents, their presence may lead to immune and host cell responses that could exacerbate the inflammatory state. Thus it would be predicted that appropriate treatments to minimize microbial presence in this tissue may significantly improve the morbidity associated with interstitial cystitis. Acknowledgement. Data used in preparing table 2 and figure 3 were derived h m the Ribosomal Database Project (RDP) email updated on February 18,1993with d a t a release 2.1.We thank G.J. Olson, B. Maidak and M. McCaughey at the RDP (University of Illinois, Urbana, Illinois) for help with sequence alignment and construction of the phylogenetic tree. D. Neal, former faculty member at Tulane Medical Center (presently at University of Texas Medical School, Galveston, Texas), provided specimens from 1 interstitial cystitis patient and 1 control patient. We thank Dale S. Martin for expert technical assistance in electron microscopy. REFERENCES
1. Parivar, F. and Bradbmk, R. A.: Interstitial cystitis. Br. J. Urol., MI: 239, 1986. 2. Holm-Bentzen,M.: Pathology, pathophysiology, and pathogenesis of painful bladder disease. Urol. Res., 17: 203, 1989. 3. Messing, E. M.: The diagnosis of interstitial cystitis. Urology, suppl. 4,29: 4, 1987. 4. Hanno, P., Levin, R. M., Monson, F. C., Teuscher, C., Zhou, Z. Z., Ruggieri, M., Whitmore, K and Wein, A. J.: Diagnosis of interstitial cystitis. J. Urol.. 143: 278. 1990. 5. Koziol, J. A., Clark, D. C., Gittes, R. F. and Eng, M. T.: The
natural history of interstitial cystitis: a survey of 374 patients. J. Urol., 149 465, 1993. 6. Messing, E. M. and Stamey, T. A.: Interstitial cystitis: early diagnosis, pathology, and treatment. Urology, 1 2 381,1978. 7. Oravisto, K J.: Interstitial cystitis as an autoimmune disease: a review. Eur. Urol., 6 10, 1980. 8. Fall, M., Johansson, S. L. and Aldenborg, F.: Chronic interstitial cystitis: a heterogeneous syndrome. J. Urol., 137: 35,1987. 9. Mattila, J., Harmoinen, A. and Hallstorm, 0.: Serum immunoglobulin and complement alterations in interstitial cystitis. Eur. Urol., 9 350, 1983. LO. Aldenborg, F., Fall, M. and Enerback, L.: Proliferation and transepithelial migration of mucosal mast cells in interstitial cystitis. Immunology, 58: 411,1986. 11. Anderson, J. B., Parivar, F., Lee, G., Wallington, T. B., MacGiver, A. G., Bradbrook, R. A. and Gingell, J. C.: The enigma of interstitial cystitis-an autoimmune disease. Br. J. Urol., 63:58, 1989. 12. MacDermott, J.P., Miller, C. H., Levy, N. and Stone, A. R.: Cellular immunity in interstitial cystitis. J. Urol., 145:274,1991. 13. Miller, C. H.,MacDermott, J. P., Quatrocchi, K. B., Broderick, G. A. and Stone, A. R.: Lymphocyte function in patients with interstitial cystitis. J. Urol., 147: 592, 1992. 14. Mullis, K B., and Faloona, F.: Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods Enzymol. 155:335, 1987. 15. Gillenwater,J. Y.and Wein, A. J.: Summary of the National Institute of Arthritis,Diabetes, Digestive and Kidney Disease workshop on interstitial cystitis, National Institutes of Health, Bethesda, Maryland, August 28-29,1987.J. Urol., 140:203,1988. 16. Domingue, G. J., Thomas, R., Walters, F., Serrano, A. and Heidger, P. M., Jr.: Cell wall deficient bacteria as a cause of idiopathic hematuria. J. Urol., 150 483,1993. 17. Sarkar, G. and Sommer, S. S.: Shedding light on PCR contamination. Nature, 343:27, 1990. 18. Schmidt, T. M., Pace, B. and Pace, N. R.: Detection of DNA contamination in taq polymerase. BioTechniques, 11: 176,1991. 19. Rochelle, P. A,, Weightman, A. J. and Fry, J. C.: DNase I treatment of taq DNA polymerase for complete PCR decontamination. BioTechniques, 13: 520, 1992. 20. Selden, R. F.: Analysis of DNA sequences by blotting and hybridization. In: Current Protocols in Molecular Biology. Edited by F. M. Ausubel, R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith and K. Struhl. New York: Wiley, vol. 1, unit 2.9,1989. 21. Olsen, G. J., Larsen, N. and Woese, C. R.: The ribosomal RNA database project. Nucleic Acids Res., suppl., 1 9 2017,1991. 22. Olsen, G. J.,Overbeck, R., Larsen, N., Marsh, T. L., McGaughey, M. J., Maciukenas, M. A., Kuan, W. M., Macke, T. J., Xing, Y. and Woese, C. R.: The ribosomal database project. Nucleic Acids Res., suppl., 2 0 2199,1992. 23. Domingue, G. J.: Cell Wall-Deficient Bacteria: Basic Principles and Clinical Significance. Reading Massachusetts: AddisonWesley Publishing Co., 1982. 24. Meir, A, Swings, J., De Vos, M., M e n , M. and Bottger, E. C.: Elimination of contaminating DNA within polymerase reaction reagents: implications for a general approach to detection of uncultured pathogem. J. Clin. Micmbiol., 31: 646,1993. 25. Swings, J., Meier, A., Persing, D. H., Finken, M. and Bottger, E. c.:Transfer of Pseudomonas maltoDhilia Hueh 1981 to the genus Xanthomonas as Xanthomoias maltoihilia (Hugh 1981)comb. Nov. Int. J. Syst. Bacteriol., 33: 409,1983.