POLYMERASE CHAIN REACTION AMPLIFICATION OF BACTERIAL 16S rRNA GENES FROM COLD-CUP BIOPSY FORCEPS

0022-5347/98/1606-2229$03.00/0

THEJOGRNAL OF UROLOGY

Vol. 160, 2229-2231. December 1998 Prrnted 111 U.S.A.

Copyright 0 1998 by AMERICAN UROLOGICAL ASSOCIATION, INC.

POLYMERASE CHAIN REACTION AMPLIFICATION OF BACTERIAL 16s rRNA GENES FROM COLD-CUP BIOPSY FORCEPS S. KEAY,* C-0. ZHANG, B. R. BALDWIN, R. B. ALEXANDER

AND

J. W. WARREN

From the Division of Infectious Diseases, Department of Medicine, the Department of Molecular and Cellular Biology, the Division of Urology, Department of Surgery, University of Maryland School of Medicine, and the Research Service and Surgical Service, Section of Urology, Veterans Administration Maryland Health Care System, Baltimore, Maryland

ABSTRACT

Purpose: In looking for a possible infectious cause for interstitial cystitis (IC), we previously determined that bladder tissue specimens from both I C patients a n d controls were uniformly positive by polymerase chain reaction assay (PCR) for bacterial 1 6 s ribosomal RNA genes from various genera including Escherichia, Propionobacterium, Acinetobacter, a n d Salmonella. We therefore determined whether the biopsy forceps might be contaminated with bacterial DNA. Materials and Methods: A total of 23 samples were obtained following disinfection of 6 cold-cup bladder biopsy forceps (2 to 5 specimens from each forceps over a period of 19 months). DNA w a s extracted from each sample, and PCR performed using nested primers from a highly conserved region of the bacterial 16s rRNA gene. Amplified DNA w a s purified and sequenced, a n d the sequences obtained were compared with bacterial rRNA gene sequences recorded in GenBank. Results: Thirteen of 23 forceps specimens were positive by PCR for bacterial DNA, including at least one rinse from each of the 6 forceps. In comparison, none of 9 negative control specimens (sterile distilled water put into tubes a n d processed in the same m a n n e r as forceps rinses) had detectable bacterial DNA. Sequence data indicated the presence of a predominant organism in 12 of the 13 positive specimens, with > 95% homology t o DNA from several different genera of bacteria including Escherichia, Propionobacterium, Stenotrophomonas and Pseudomonas. Conclusions: These d a t a indicate that reusable bladder biopsy forceps are frequently contaminated with bacterial DNA. Tissue specimens procured with such instruments therefore are inappropriate sources to look for the presence of bacterial pathogens by PCR. KEY Woms: polymerase chain reaction, bacteria, biopsy, forceps

1

Interstitial cystitis (IC) is a chronic bladder disease of unknown etiology with several characteristics which suggest that it may be caused by an infectious organism.'-6 However, attempts to identify a n infectious etiology for IC have yielded conflicting result^.^-'^ Using sensitive culture methods, we previously determined that the prevalence of microorganisms (mostly bacteria) was significantly greater in urine of IC patients (6/11) than in urine of controls (0/7) (p <0.05)? However, because various genera and species of microorganisms were present in these specimens,no single microorganism could be identified as a possible cause for IC; moreover, these findings suggested to us that abnormalities of the bladder epithelium in IC may predispose patients to colonization with a variety of bacteria. This report was followed by three subsequent studies in which polymerase chain reaction (PCR) technology was used to search for evidence in IC patient biopsy specimens of the genes that encode bacterial ribosomal RNA.'-'' These studies yielded conflicting results; one report indicated no evidence for bacterial DNA in biopsy specimens from 11 IC patients and 3 controls using a single set of PCR primers ', another indicated that 4 of 14 (29%) IC patients but 0 of 15 controls had evidence for the presence of bacterial DNA by nested PCR? and our own study found that bladder biopsies obtained from 6 IC patients and 6 controls were uniformly positive for bacterial DNA (for various

bacterial genera including Escherichia, Propionobacterium, Acinetobacter, and Salmonella).'o The finding of bacterial DNA in all of our control and IC biopsy specimens for genera different from the bacteria cultured from these same specimens suggested that nonviable bacteria were in the biopsied bladder tissue. We hypothesized that this phenomenon most likely resulted either from povidone iodine-killed organisms being introduced into the bladder during the biopsy procedure or from nonviable organisms remaining on the reusable biopsy forceps after prior routine disinfection. To understand whether bacterial DNA could be introduced into tissue specimens by reusable biopsy instruments, we determined whether the genes for bacterial 16s ribosomal RNA could be detected on bladder biopsy forceps which were ready for use in the operating room. MATERIALS AND METHODS

Biopsy forceps. Reusable cold-cup biopsy forceps (3 manufactured by Olympus, Chicago, IL, and 3 manufactured by Circon ACMI, Stamford, CT) were tested at two hospitals for the presence of bacterial DNA. Forceps were routinely hand cleaned with a soap solution (Endozyme solution, Ruhof Corp., Valley Stream, NY) and washed additionally in a commercial washer prior to disinfection or sterilization. At the University of Maryland Hospital, the Olympus forceps were then disinfected by soaking in Cidex (2.4% glutaraldehyde) solution (Johnson & Johnson, Arlington, TX) over 20 Accepted for publication June 10, 1998. * Requests for reprints: VA Medical Center, Room 3B184, 10 N. minutes, while at the Veterans Administration Maryland Greene St., Baltimore MD 21201. Health Care System the ACMI forceps were autoclaved Supported by the National Institute of Diabetes and Digestive and (steam sterilized at 30 psi, 270F for 3 minutes) prior to use. fidney Diseases, National Institutes of Health ( R 0 1 DK44818) and Specimens used for PCR analysis were obtained at periodic the Interstitial Cystitis Association. 2229

BACTERIAL 16s RRNA GENES FROM FORCEPS BY PCR

2230

intervals over a period of 19 months. Each disinfected or rRNA genes in a dilution of purified bacterial DNA from autoclaved forceps was rinsed in 200 p1. sterile double dis- approximately 0.5 colony forming units of a laboratory strain tilled water in a sterile test tube at room temperature, and of Escherichia coli after the second round of arnplification.'O the rinses immediately processed upon arrival in the labora- One, two or all three forceps available in either of two operating rooms were sampled on each of 10 occasions (for a total tory for isolation of DNA. Deoxyribonucleic acid extraction and PCR amplification. of 16 samples from disinfected forceps and 7 samples from DNA was extracted from the forceps' rinses by incubation in autoclaved forceps, with each forceps being tested between 2 0.4 ml. digestion buffer overnight at 5OC1O followed by mix- and 5 times). A negative control (sterile double distilled waing with a n equal volume of phenollchlorofodisoamyl alco- ter in a sterile test tube transported to the operating room hol (25/24/1) and centrifuging at 1700 g for 10 minutes at 4C. but not used t o rinse forceps) and a positive control (23. coli The DNA was precipitated from the aqueous layer by adding strain Y1090r- DNA) were run with each set of samples. Results of these experiments indicated the presence of a 0.5 volume of 7.5 M ammonium acetate and 2.0 volume of absolute ethanol for 30 minutes at -80C and centrifuged band of the appropriate size for the bacterial rRNA gene 1700 g for 10 minutes at 4C. Following a rinse with 70% fragment in 13 of 23 rinse specimens, including at least one ethanol, the pellet was lyophilized, resuspended in 20 ml. TE rinse from each of the 6 forceps tested (fig.). Because the buffer (10 mM Tris-HC1 pH 8.0, 1.0 mM EDTA pH 8.01, and individual forceps were not labeled, the exact number of incubated for 1 hour at 65C. Purified DNA specimens were times a particular forceps was positive is unknown. However, since all 3 forceps from each operating room were positive stored at -2OC. PCR amplification to amplify the bacterial 1 6 s rRNA gene from the same visit, we know that all 6 forceps were posiwas carried out using nested primers, according to previously tive on at least one occasion. In comparison, none of the 9 published methods.', lo To prevent contamination by exoge- negative control specimens was positive. There was no disnous DNA, 5 units Thermus aquaticus DNA polymerase cernible difference in the rate of PCR positivity between the (AmpliTaq Polymerase, Perkin Elmer, Foster City, CA) were Cidex disinfected and autoclaved forceps, with Cidex preincubated with 0.2 units bovine pancreatic DNase I disinfected forceps being positive 9 of 16 times (56%), and (Boehringer Mannheim, Indianapolis, IN) for 1hour at 37C, autoclaved forceps being positive 4 of 7 times (57%). Sequence analysis of amplified DNA was performed on all and the DNase then heat inactivated for 10 minutes at 9%. Each DNA specimen was then mixed with 0.2 mM dNTPs, 1 specimens. In 12 of the 13 positive specimens, a predominant pg. of each primer, and 5 units DNase-treated AmpliTaq organism was present which allowed identification of the polymerase in PCR buffer (50mM KC1, 1.5 mM MgCl,, and bacterium present. We compared t h e sequences obtained 10 mM Tris-HC1 pH 8.31, and amplification performed using from these amplified DNA segments with sequences from a Perkin Elmer DNA Thermal Cycler programmed for 95C various bacterial species recorded in t h e GenBank database, denaturation, 58C annealing, and 72C extension tempera- and found evidence for the presence of DNA with > 95% tures for 25 cycles (first 3 cycles had extended denaturation homology to several bacteria, as shown in the table. and annealing times). Primers used for the first round of DISCUSSION amplification were 1A [5'-CACAAGCGGTGGAGCATGTGGTT-3'1 and 5 [5'-CCTACGGYTACC'M'GTTACCACT-3'] Our findings indicate that reusable bladder biopsy forceps, where Y equals C or T; those used for the second round were whether disinfected or sterilized, are frequently contami3 [5'-GGAATTCTGCAACGCGAAGAACCTTACCT-3'] and 4 nated with bacterial DNA that can be amplified by PCR. [5'-GCGGATCCTGGTKTGACGGGCGGTGTGTA-3'] where Bladder tissue specimens procured with such instruments K equals G or T. Following PCR amplification, approximately therefore may also be contaminated by PCR-amplifiable bac20% of the reaction mixture was run on a 1% agaroseiTAE terial DNA, making them inappropriate sources to look for buffer gel which was stained with ethidium bromide, and the presence of bacterial pathogens using this sensitive techthe presence of a 467 base pair fragment detected by U V nique. fluorescence. Similar PCR amplification techniques have previously Deoxyribonucleic acid sequencing. Amplified DNA was pu- noted contamination of reusable surgical instruments with rified by electrophoresis using a 1.4%NuSieve GTG gel (FMC DNA from both bacteria and viruses. Kaul et a1 noted the BioProducts, Rockland, ME) and the Wizard PCR DNA purification resin kit (Promega, Madison, WI1 according to the detection of residual Mycobacterium tuberculosis DNA on manufacturer's instructions, and stored at -2OC. The gene 3.6% of washes obtained from sterile bronchoscopes, which fragments were sequenced directly using the Sequenase Ver- they hypothesized led to false-positive PCR results in two sion 2.0 PCR Product Sequencing Kit (Amersham, Arlington patient specimens;" the same study noted that 20% of their Heights, IL) and the primers 3 or 4 (as used in the PCR bronchoscopes were contaminated with human DNA, suggesting that DNA contamination in general can occur even reaction). more frequently. A similar contamination was noted to occur on endoscopes by Roosendaal et al, who reported that 8 of 23 RESULTS rinse samples obtained from 5 fiberoptic endoscopes which We previously determined that the nested PCR method had undergone standard cleaning and disinfection were reused in these studies was of sufficient sensitivity to detect ported to be positive for Helicobacter pylori DNA;'" in 5 of

0.5 kb + -2

g$

22

1 2

3

4 5 6

7 8 9 101112 1314 1516

Cidex disinfected forceps

17 18 19 20 21 22 23

Autoclaved forceps

Polymerase chain reaction amplification of bacterial 16s rRNA genes from bladder biopsy forceps using nested primers and two round nmplification Following second round of amplification, 467 base pair fragment was evident in 13 of 23 samples. Far left lane shows locatlon of DNA standards, with labeling of 500 bp standard for reference,

BACTERIAL 16s RRNA GENES FROM FORCEPS BY PCR Identity of bacteria present on cold cup biopw forceDs Olympus (chemically disinfected) [51 E. coli (strain #1)* [ll Stenotrophomonas sp. [l]E. coli (strain #2) I21 Propionobacterium sp. ACMI (heat sterilized) [ZIE . coli (strain #3) [11 Pseudornonas sp. * Three strains of E. coli were evident by identical base pair changes at specific alleles.

2231

interpret the results of such analyses if a control specimen is also obtained from each cold cup biopsy forceps prior to biopsy. In addition, the extent of DNA contamination of other types of reusable disinfected or sterilized surgical instruments should also be determined. Because reusable surgical instruments can be contaminated with human as well as microbial DNA,l’ a prebiopsy control specimen should probably also be obtained from these instruments prior to molecular diagnostic or staging procedures to look for the presence of specific human genes. Finally, the usefulness of disposable or previously unused forceps t o obtain tissue for PCR analysis should also be evaluated. Acknowledgments. The authors gratefully acknowledge the assistance of Nancy Schmidt, R.N. and Steven Hullett, R.N. in obtaining the forceps specimens.

these 8 cases, a tissue specimen obtained during the previous endoscopic procedure was PCR positive for H . pylori DNA, These findings raised questions about the origin of the H . pylori DNA detected in patient tissue specimens, as well as the appropriateness of using PCR to detect bacterial pathoREFERENCES gens in specimens obtained with reusable surgical instruments. In addition, using a duck hepatitis B virus model to 1. Warren, J . W.: Interstitial cystitis as an infectious disease. Urol. determine the effectiveness of disinfection of surgical laparoClin. North Am., 21: 31, 1994. 2. Messing, E. M. and Stamey, T. A,: Interstitial cystitis: early scopes, Deva et a1 noted evidence of residual viral DNA in 10 diagnosis, pathology, and treatment. Urology, 1 2 381,1978. of 12 laparoscopes which did not correlate with transmissi3. Oravisto, K. J.: Epidemiology of interstitial cystitis. Ann. Chir. bility of virus infection in vivo, suggesting that PCR was also Gynaecol. Fenn., 64.75, 1975. inappropriate for evaluating disinfection of these surgical 4. Warren, J. W.: Clinical presentation and epidemiology of urinary instruments.’“ tract infections. In: Urinary Tract Infections. Edited by MobIt is unknown whether contamination of the biopsy forceps ley, H. L. T. and Warren, J . W. Washington, D. C.:ASM Press, (as well as the bronchoscopes and endoscopes) played a role 1996. in bacterial DNA contamination of the tissue specimens in 5. Held, P. J., Hanno, P. M. and Wein, A. J.: Epidemiology of interstitial cystitis: 2. In: Interstitial Cystitis. Edited by the first two reports, since neither study evaluated those Hanno, P. M., Staskin, D. R., Krane, R. J. and Wein, A. J. forceps. However, like the surgical instruments previously London: Springer-Verlag, 1990. noted to be contaminated with bacterial or viral DNA, the 6. Said, J. W, Van de Velde, R. and Gillespie, L.: Immunopathology biopsy forceps used to obtain bladder tissue specimens from of interstitial cystitis. Mod. Pathol., 2 593,1989. our IC and control patients had been cleaned and disinfected 7. Keay, S.,Schwalbe, R. S., Trifillis, A. L., Lovchik, J. C., Jacobs, using standard chemicals a t the University of Maryland HosS. and Warren, J . W.: A prospective study of microorganisms pital; it is therefore possible that the bacterial DNA present in urine and bladder biopsies from interstitial cystitis patients in the bladder tissue from our IC and control patients reand controls. Urology, 45:223, 1995. sulted from contamination of the chemically treated bladder 8. Domingue, G. J., Ghoniem, G. M., Bost, K. L., Fermin, C. and Human, L. G.: Dormant microbes in interstitial cystitis. biopsy forceps used to obtain those specimens. In addition, J . Urol., 153 1321,1995. our current report notes that PCR amplification also allows 9. Haarala, M., Jalava, J., Laato, M., Kiilholma, P., Nurmi, M. and recovery of bacterial DNA following heat sterilization of reAlanen, A,: Absence of bacterial DNA in the bladder of pausable surgical instruments. tients with interstitial cystitis. J. Urol., 156 1843,1996. Two studies which used single round PCR amplification 10. Keay, S., Zhang, C.-O., Baldwin, B. R., Jacobs, S. C. and Warren, had previously reported finding no evidence for bacterial J . W.: Polymerase chain reaction amplification of bacterial 16s DNA in IC patient bladder tissue specimen^.^.^^ However, a rRNA genes in interstitial cystitis and control patient bladder more sensitive nested PCR method with 2 round amplificabiopsies. J . Urol., 159 280, 1998. tion has been able to detect bacterial DNA in both IC and 11. Kaul, K., Luke, S., McGurn, C., Snowden, N., Monti, C. and Fry, W. A.: Amplification of residual DNA sequences in sterile control patient bladder biopsy specimens, as shown in our bronchoscopes leading to false-positive PCR results. J. Clin. previous report as well as a recent report by Domingue e t Microbiol., 34: 1949,1996. al,lO. 14 Our current findings suggest that the bacterial DNA R., Kuipers, E. J., van den Brule, A. J., Pena, A. S., found in these specimens may have resulted from contami- 12. Roosendaal, Meuwissen, S. G., Walboomers, J . M. and de Graaff, J.: Detecnation of the biopsy forceps used to obtain the specimens, tion of Helicobacter pylori DNA by PCR in gastrointestinal since the types of bacteria were similar in both studies, not equipment Petted. Lancet, 341(8849): 900,1993. fastidious, and, if alive, should have been (but were not) 13. Deva, A. K.,Vickery, K., Zou, J., West, R. H., Harris, J. P. and Cossart, Y. E.: Establishment of an in-use testing method for culturable. Furthermore, the bacteria detected in those bievaluating disinfection of surgical instruments using the duck opsy specimens were also similar to those detected on the hepatitis B model. J. Hosp. Infect., 3 3 119,1996. biopsy forceps, including primarily gram negative rods such as E. coli and Pseudornonas. It remains difficult to explain, 14. Cardenas Freytag, L., Ghoniem, G. M., Miyazaki, S., Human, L., Hurley, D. and Domingue, G. J.: Bacterial DNA detection and however, the difference noted by Domingue’s group between cytokine activation in interstitial cystitis. Proceedings of the IC patients and controls. 1997 International Research Symposium on Interstitial CysBased on data presented in this report, we conclude that titis, p. 79, 1997. the use of tissue procured with reusable bladder biopsy for- 15. Duncan, J . L., Schaeffer, A. J. and Dardick, F. B.: Analysis of ceps is generally inappropriate for PCR analysis of potential specimens from interstitial cystitis patients for infectious agents. Proceedines bacterial pathogens for bladder disease. However, as pro--.. of the 1997 International Research Svmposium on Interstitial Cystitis, p. 85, 1997. posed by Kaul et al” for bronchoscopes, it may be possible to ~