VOL. 29, 1991
trying to view the question in a scientifically objective way. I am neither in favor of nor against this organism actually being a pathogen. To support his cause, Dr. Zierdt states that there is a 20:1 preponderance of affirmative literature as compared to "anti" literature concerning pathogenicity. Does he mean that we should base scientific credibility on the quantity of papers written on a subject as opposed to the quality and significance of the content? He also says that ". . . the two most quoted 'anti' papers are basically flawed," without citing the references or indicating how they are flawed. The second paragraph of Dr. Zierdt's letter seems to suggest that we should assume that B. hominis is a pathogen until it is proven not to be one. Somehow, I always thought that in science things should be the other way around. Since we live in a "sea" of microorganisms, most of which are nonpathogenic for humans, the presence of one of them in our flora should not be equated with pathogenicity. The reference to Entamoeba histolytica is inappropriate since that organism is a well-established pathogen (4) and has been shown to exist in both pathogenic and nonpathogenic forms
(5).
Dr. Zierdt states that fulfillment of Koch's postulates is unnecessary without telling us why. Have they been shown to be invalid? He also refers to "at least a dozen articles describing epidemics . . ." without providing the references or any other specific information. I have reviewed the articles cited by Dr. Zierdt as providing evidence of gastrointestinal pathology except number 10, which is from the Polish literature and not accessible. Reference 1 indicates that no relevant pathology was found in the four patients who underwent endoscopy. Reference 2 indicates that of 14 patients who had B. hominis infection only, none had an abnormal endoscopy and 13 had normal tissue biopsies; 1 showed mild inflammation, but no organisms were present. References 3, 7, and 9 consist of single case histories associating the presence of B. hominis with rectal ulcers, colitis, and ileitis, respectively. Two patients seemed to respond to metronidazole. While there may be a cause-andeffect relationship here, the absence of full microorganism stool workups and the lack of long-term follow-up information on the patients, as well as the dissimilar nature of the lesions in the three patients, leave room for doubt. Perhaps these patients had inflammatory bowel diseases, such as regional enteritis or ulcerative colitis. Moreover, it is very unlikely that the disease in these three patients is representative of the gastroenteritis present in the relatively large number of patients whose stools contain B. hominis when examined in the routine clinical parasitology laboratory.
LETTERS TO THE EDITOR
663
Reference 5 provides no information on gastrointestinal pathology in humans and indicates that when germfree guinea pigs were given B. hominis, infection occurred only in those that also received a bacterial inoculum. Those that received B. hominis only were not infected. Even if one accepts the concept of pathogenicity of B. hominis only when it is given together with bacterial cohorts to germfree guinea pigs, this is a far cry from pathogenicity in the human intestine. Dr. Zierdt's reference to the requirement for bacterial cohorts for pathogenesis of E. histolytica is not completely accurate. Although amebic virulence, as determined by in vitro assays, is stimulated by association with bacteria, axenically grown trophozoites are capable alone of virulence (1). I do not believe that this same capability has been demonstrated for B. hominis. I am eager to see the as-yet-unpublished results of B. hominis and the rabbit ileal loop assay to which Dr. Zierdt refers in his paragraph iii. I am puzzled by his comments on giardia and cryptosporidium, since both have been shown to attach to the intestinal epithelium (2, 3). As I stated in my original letter, perhaps B. hominis will someday be demonstrated to be pathogenic for humans. I certainly have nothing against it being a pathogen. It just has yet to be proven adequately. By the way, there are two more articles which can be thrown on the "anti" pile (6, 7). REFERENCES 1. Bracha, R., and D. Mirelman. 1984. Virulence of Entamoeba histolytica: effects of bacteria, microaerobic conditions, and metronidazole. J. Exp. Med. 160:357-361. 2. Erlandsen, S. L., and D. E. Feely. 1984. Trophozoite motility and the mechanism of attachment, p. 33-63. In S. L. Erlandsen and E. A. Meyer (ed.), Giardia and giardiasis: biology, pathogenesis, and epidemiology. Plenum Press, New York. 3. Janoff, E. N., and L. B. Reller. 1987. Cryptosporidium species. a protean protozoan. J. Clin. Microbiol. 25:967-975. 4. Ravdin, J. L. 1988. Pathogenesis of amebiasis: an overview, p. 166-176. In J. L. Ravdin (ed.), Amebiasis: human infection by Entamoeba histolytica. John Wiley & Sons, Inc., New York. 5. Sargeaunt, P. G. 1988. Zymodemes of Entamoeba histolvtica, p. 370-387. In J. L. Ravdin (ed.), Amebiasis: human infection by Entamoeba histolytica. John Wiley & Sons, Inc., New York. 6. Senay, H., and D. MacPherson. 1990. Blastocystis hominis: epidemiology and natural history. J. Infect. Dis. 162:987-990. 7. Sun, T., S. Katz, B. Tanenbaum, and C. Schenone. 1989. Questionable clinical significance of Blastocystis hominis infection. Am. J. Gastroenterol. 84:1543-1547.
Jon E. Rosenblatt Section of Clinical Microbiology The Mayo Clinic Rochester, Minnesota 55905
Genetic Heterogeneity in Strains of Pseudomonas aeruginosa from Patients with Cystic Fibrosis The article by Hjelm et al. (2) contains interesting observations. However, we question the authors' interpretation of data published previously by our group and others regarding the number of strains of Pseudomonas aeruginosa present in patients with cystic fibrosis (CF) and the genetic events related to the changes in the restriction patterns reported by Hjelm et al. in their studies.
The authors state: "Ogle et al.... and Pasloske et al. using separate DNA probes to test restriction fragment length polymorphisms (RFLPs) within serial isolates of P. aeruginosa from cystic fibrosis patients, demonstrate that cystic fibrosis patients are infected by only one strain of P. aeruginosa, which displays many different phenotypes." Although CF patients are usually infected with only one . . .,
664
LETTERS TO THE EDITOR
strain, we did note exceptions in our published reports (3, 8). We subsequently have documented CF patients with up to four different strains (4). Speert et al. (7) reported changes in the RFLP type and lipopolysaccharide serotype in sequential isolates from 10 of 23 patients. Grothues et al. (1) reported that 36% of patients in a German CF clinic were infected with more than one strain and that one patient was infected with five different strains. Thus, there are strong data in the literature for the presence of multiple strains in this patient population. Pritchard and Vasil (6) investigated the molecular basis for heterogeneity in the 5' region of toxA by sequencing DNA from three strains. A mosaic organization of the genome was shown to extend several thousand base pairs upstream of toxA. Alternating segments of homologous and nonhomologous sequences were shown by comparison of the DNA sequences. The nonhomologous segments have several characteristics of IS elements. Neither the extensive heterogeneity 5' of toxA reported by Ogle et al. (3) nor the evidence of Pritchard and Vasil (6) suggests that simple excisions or insertions account for all the variability in this region of the chromosome. This is in variance with the findings of Hjelm et al. Hjelm et al. state that "The random probe suggested that insertion events of similar sizes were occurring throughout the genome of the phenotypic variants and that the area surrounding the 800-bp region was not unique for DNA rearrangement." Could the result shown in Fig. 5 (an 2.7-kb increase in the XhoI fragment in WSU3860-10) be the result of methylation of the XhoI site (an enzyme known to be affected by methylation) and not a DNA rearrangement? Regarding Fig. 6, although the region around the 800-bp insertion is undoubtedly not unique, Hjelm et al. do not clarify how commonly genomic rearrangements do occur in P. aeruginosa. For example, sequences around the phospholipase C structural gene are conserved from strain to strain (6), as are sequences 3' to toxA. Pulsed-field electrophoresis of Dral and SpeI fragments shows that related strains differ by six or fewer bands (1). It would be interesting to know whether Hjelm et al. now have additional data regarding the extent of variability seen throughout the chromosome. Most importantly, it is unclear whether isolates with different RFLP patterns were derived from a single clone in these studies. We reported use of a 741-bp PstI-NruI fragment from the 5' region of toxA as an epidemiologic marker to investigate clonal relationships (3, 8). We believe that typing using this fragment correlates with clonal relationships, and our subsequent studies using pulsed-field electrophoresis support our earlier findings. Although definitions of clonal relatedness are not established and agreed upon, others have frequently used several techniques to establish relatedness (1, 5) among isolates. Additional data in this regard would have been of value in clarifying the data
presented. REFERENCES 1. Grothues, D., U. Koopmann, H. von der Hardt, and B. Tummler. 1988. Genome fingerprinting of Pseudomonas aeruginosa indicates colonization of cystic fibrosis siblings with closely related strains. J. Clin. Microbiol. 26:1973-1977. 2. Hjelm, L. N., A. A. Branstrom, and R. L. Warren. 1990. Detection of restriction fragment length polymorphisms in clinical isolates and serially passaged Pseudomonas aeruginosa strains. J. Clin. Microbiol. 28:2178-2182. 3. Ogle, J. W., J. M. Janda, D. E. Woods, and M. L. Vasil. 1987. Characterization and use of a DNA probe as an epidemiological
J. CLIN. MICROBIOL.
marker for Pseudomonas aeruginosa. J. Infect. Dis. 155:119126. 4. Ogle, J. W., A. I. Vasil, and H. R. Rabin. 1988. Abstr. Annu. Meet. Am. Soc. Microbiol. 1988, B141, p. 53. 5. Pasloske, B. L., A. M. Joffe, Q. Sun, K. Volpel, W. Paranchych, F. Eftekhar, and D. P. Speert. 1988. Serial isolates of Pseudomonas aeruginosa from a cystic fibrosis patient have identical pilin sequences. Infect. Immun. 56:665-672. 6. Pritchard, A. E., and M. L. Vasil. 1990. Possible insertion sequences in a mosaic genome organization upstream of the exotoxin A gene in Pseudomonas aeruginosa. J. Bacteriol.
172:2020-2028. 7. Speert, D. P., M. E. Campbell, S. W. Farmer, K. Volpel, A. M. Joffe, and W. Paranchych. 1989. Use of a pilin gene probe to study molecular epidemiology of Pseudomonas aeruginosa. J. Clin. Microbiol. 27:2589-2593. 8. Vasil, M. L., C. Chamberlain, and C. C. R. Grant. 1986. Molecular studies of Pseudomonas exotoxin A gene. Infect. Immun. 52:538-548.
John W. Ogle Department of Pediatrics Michael L. Vasil Department of Microbiology and Immunology University of Colorado School of Medicine 4200 East Ninth Avenue Denver, Colorado 80262
Author's Reply The comments made by Drs. Ogle and Vasil need to be addressed. The strongest evidence for the colonization of cystic fibrosis patients by more than one Pseudomonas aeruginosa strain is the field inversion gel electrophoresis analysis presented by Grothues et al. (1). Both Speert et al. (7) and Ogle et al. (4) have presented data that demonstrate the presence of multiple restriction fragment length polymorphisms (RFLPs) in a single patient. In both instances, the DNA from the isolates had different restriction endonuclease digestion profiles. More convincingly, Ogle and Vasil mention in their letter that subsequent studies using pulsed-field electrophoresis supported the clonal relatedness observed with the RFLPs. We do not question the ability of more than one strain of P. aeruginosa to colonize cystic fibrosis patients. The article by Hjelm et al. (2) emphasizes the point that RFLPs were detected in serially passaged isolates from a single colony as well as isolates from a single patient. As mentioned in Results, the increase observed in the XhoI fragment shown in Fig. 5 was from DNA isolated from serial passages of a single isolate, WSU3860. The RFLPs observed could not be due to methylation alone, since Fig. 6 clearly depicts a restriction map comparing the parent from which the serial passages were made and the subclone exhibiting the increased size of the XhoI fragment. Altered restriction fragments were also generated from BglII and KpnI digests. The data presented in our article demonstrate that RFLPs alone are not sufficient to establish clonal relatedness when the 800-bp HindIII-PstI fragment is used as a marker. Comparing our results to those of Ogle et al. (3) and Pasloske et al. (5), we found that clonally related strains had different RFLPs when probed with the 800-bp HindIII-PstI probe. This demonstrates intrastrain genomic variability that was not previously reported when either the upstream portion of the exotoxin A gene or the pilin gene probe was used. We showed that the 800-bp HindIII-PstI fragment can detect RFLPs with XhoI-, BglII-, and KpnI-digested chromosomal DNA and highly conserved regions of DNA can be detected with DdeI, PstI, and Hinfl chromosomal digests.
VOL. 29, 1991
LETTERS TO THE EDITOR
The variability seen within the region surrounding our 800-bp probe may be the result of IS elements, as was discussed. We did not imply that this intrastrain variability was the result of simple excisions or insertions. The frequency with which genomic rearrangements occurred was enough to detect 2 of 22 phenotypic variants examined. It is interesting that the mosaic genome organization observed by Pritchard and Vasil (6) may be the result of IS-like elements, which does not conflict with our results. However, do the mechanisms that gave rise to the interstrain genomic organizations upstream of the exotoxin A gene also give rise to the intrastrain variability that we observe using the 800-bp probe? The insertion resulting in the increased XhoI fragment needs to be further characterized to make that determination.
665
strains. J. Clin. Microbiol. 28:2178-2182. 3. Ogle, J. W., J. M. Janda, D. E. Woods, and M. L. Vasil. 1987. Characterization and use of a DNA probe as an epidemiological marker for Pseudomonas aeruginosa. J. Infect. Dis. 155:119126. 4. Ogle, J. W., A. I. Vasil, and H. R. Rabin. 1988. Abstr. Annu. Meet. Am. Soc. Microbiol. 1988, B141, p. 53. 5. Pasloske, B. L., A. M. Joffe, Q. Sun, K. Volpel, W. Paranchych, F. Eftekhar, and D. P. Speert. 1988. Serial isolates of Pseudomonas aeruginosa from a cystic fibrosis patient have identical pilin sequences. Infect. Immun. 56:665-672. 6. Pritchard, A. E., and M. L. Vasil. 1990. Possible insertion sequences in a mosaic genome organization upstream of the exotoxin A gene in Pseudomonas aeruginosa. J. Bacteriol. 172:2020-2028. 7. Speert, D. P., M. E. Campbell, S. W. Farmer, K. Volpel, A. M. Joffe, and W. Paranchych. 1989. Use of a pilin gene probe to study molecular epidemiology of Pseudomonas aeruginosa. J. Clin. Microbiol. 27:2589-2593.
REFERENCES 1. Grothues, D., U. Koopmann, H. von der Hardt, and B. Tummier. 1988. Genome fingerprinting of Pseudomonas aeruginosa indicates colonization of cystic fibrosis siblings with closely related strains. J. Clin. Microbiol. 26:1973-1977. 2. Hjelm, L. N., A. A. Branstrom, and R. L. Warren. 1990. Detection of restriction fragment length polymorphisms in clinical isolates and serially passaged Pseudomonas aeruginosa
Arthur A. Branstrom Department of Bacterial Immunology Richard L. Warren Department of Bacterial Diseases Division of Communicable Disease and Immunology Walter Reed Army Institute of Research Washington, D.C. 20307
trying to view the question in a scientifically objective way. I am neither in favor of nor against this organism actually being a pathogen. To support his cause, Dr. Zierdt states that there is a 20:1 preponderance of affirmative literature as compared to "anti" literature concerning pathogenicity. Does he mean that we should base scientific credibility on the quantity of papers written on a subject as opposed to the quality and significance of the content? He also says that ". . . the two most quoted 'anti' papers are basically flawed," without citing the references or indicating how they are flawed. The second paragraph of Dr. Zierdt's letter seems to suggest that we should assume that B. hominis is a pathogen until it is proven not to be one. Somehow, I always thought that in science things should be the other way around. Since we live in a "sea" of microorganisms, most of which are nonpathogenic for humans, the presence of one of them in our flora should not be equated with pathogenicity. The reference to Entamoeba histolytica is inappropriate since that organism is a well-established pathogen (4) and has been shown to exist in both pathogenic and nonpathogenic forms
(5).
Dr. Zierdt states that fulfillment of Koch's postulates is unnecessary without telling us why. Have they been shown to be invalid? He also refers to "at least a dozen articles describing epidemics . . ." without providing the references or any other specific information. I have reviewed the articles cited by Dr. Zierdt as providing evidence of gastrointestinal pathology except number 10, which is from the Polish literature and not accessible. Reference 1 indicates that no relevant pathology was found in the four patients who underwent endoscopy. Reference 2 indicates that of 14 patients who had B. hominis infection only, none had an abnormal endoscopy and 13 had normal tissue biopsies; 1 showed mild inflammation, but no organisms were present. References 3, 7, and 9 consist of single case histories associating the presence of B. hominis with rectal ulcers, colitis, and ileitis, respectively. Two patients seemed to respond to metronidazole. While there may be a cause-andeffect relationship here, the absence of full microorganism stool workups and the lack of long-term follow-up information on the patients, as well as the dissimilar nature of the lesions in the three patients, leave room for doubt. Perhaps these patients had inflammatory bowel diseases, such as regional enteritis or ulcerative colitis. Moreover, it is very unlikely that the disease in these three patients is representative of the gastroenteritis present in the relatively large number of patients whose stools contain B. hominis when examined in the routine clinical parasitology laboratory.
LETTERS TO THE EDITOR
663
Reference 5 provides no information on gastrointestinal pathology in humans and indicates that when germfree guinea pigs were given B. hominis, infection occurred only in those that also received a bacterial inoculum. Those that received B. hominis only were not infected. Even if one accepts the concept of pathogenicity of B. hominis only when it is given together with bacterial cohorts to germfree guinea pigs, this is a far cry from pathogenicity in the human intestine. Dr. Zierdt's reference to the requirement for bacterial cohorts for pathogenesis of E. histolytica is not completely accurate. Although amebic virulence, as determined by in vitro assays, is stimulated by association with bacteria, axenically grown trophozoites are capable alone of virulence (1). I do not believe that this same capability has been demonstrated for B. hominis. I am eager to see the as-yet-unpublished results of B. hominis and the rabbit ileal loop assay to which Dr. Zierdt refers in his paragraph iii. I am puzzled by his comments on giardia and cryptosporidium, since both have been shown to attach to the intestinal epithelium (2, 3). As I stated in my original letter, perhaps B. hominis will someday be demonstrated to be pathogenic for humans. I certainly have nothing against it being a pathogen. It just has yet to be proven adequately. By the way, there are two more articles which can be thrown on the "anti" pile (6, 7). REFERENCES 1. Bracha, R., and D. Mirelman. 1984. Virulence of Entamoeba histolytica: effects of bacteria, microaerobic conditions, and metronidazole. J. Exp. Med. 160:357-361. 2. Erlandsen, S. L., and D. E. Feely. 1984. Trophozoite motility and the mechanism of attachment, p. 33-63. In S. L. Erlandsen and E. A. Meyer (ed.), Giardia and giardiasis: biology, pathogenesis, and epidemiology. Plenum Press, New York. 3. Janoff, E. N., and L. B. Reller. 1987. Cryptosporidium species. a protean protozoan. J. Clin. Microbiol. 25:967-975. 4. Ravdin, J. L. 1988. Pathogenesis of amebiasis: an overview, p. 166-176. In J. L. Ravdin (ed.), Amebiasis: human infection by Entamoeba histolytica. John Wiley & Sons, Inc., New York. 5. Sargeaunt, P. G. 1988. Zymodemes of Entamoeba histolvtica, p. 370-387. In J. L. Ravdin (ed.), Amebiasis: human infection by Entamoeba histolytica. John Wiley & Sons, Inc., New York. 6. Senay, H., and D. MacPherson. 1990. Blastocystis hominis: epidemiology and natural history. J. Infect. Dis. 162:987-990. 7. Sun, T., S. Katz, B. Tanenbaum, and C. Schenone. 1989. Questionable clinical significance of Blastocystis hominis infection. Am. J. Gastroenterol. 84:1543-1547.
Jon E. Rosenblatt Section of Clinical Microbiology The Mayo Clinic Rochester, Minnesota 55905
Genetic Heterogeneity in Strains of Pseudomonas aeruginosa from Patients with Cystic Fibrosis The article by Hjelm et al. (2) contains interesting observations. However, we question the authors' interpretation of data published previously by our group and others regarding the number of strains of Pseudomonas aeruginosa present in patients with cystic fibrosis (CF) and the genetic events related to the changes in the restriction patterns reported by Hjelm et al. in their studies.
The authors state: "Ogle et al.... and Pasloske et al. using separate DNA probes to test restriction fragment length polymorphisms (RFLPs) within serial isolates of P. aeruginosa from cystic fibrosis patients, demonstrate that cystic fibrosis patients are infected by only one strain of P. aeruginosa, which displays many different phenotypes." Although CF patients are usually infected with only one . . .,
664
LETTERS TO THE EDITOR
strain, we did note exceptions in our published reports (3, 8). We subsequently have documented CF patients with up to four different strains (4). Speert et al. (7) reported changes in the RFLP type and lipopolysaccharide serotype in sequential isolates from 10 of 23 patients. Grothues et al. (1) reported that 36% of patients in a German CF clinic were infected with more than one strain and that one patient was infected with five different strains. Thus, there are strong data in the literature for the presence of multiple strains in this patient population. Pritchard and Vasil (6) investigated the molecular basis for heterogeneity in the 5' region of toxA by sequencing DNA from three strains. A mosaic organization of the genome was shown to extend several thousand base pairs upstream of toxA. Alternating segments of homologous and nonhomologous sequences were shown by comparison of the DNA sequences. The nonhomologous segments have several characteristics of IS elements. Neither the extensive heterogeneity 5' of toxA reported by Ogle et al. (3) nor the evidence of Pritchard and Vasil (6) suggests that simple excisions or insertions account for all the variability in this region of the chromosome. This is in variance with the findings of Hjelm et al. Hjelm et al. state that "The random probe suggested that insertion events of similar sizes were occurring throughout the genome of the phenotypic variants and that the area surrounding the 800-bp region was not unique for DNA rearrangement." Could the result shown in Fig. 5 (an 2.7-kb increase in the XhoI fragment in WSU3860-10) be the result of methylation of the XhoI site (an enzyme known to be affected by methylation) and not a DNA rearrangement? Regarding Fig. 6, although the region around the 800-bp insertion is undoubtedly not unique, Hjelm et al. do not clarify how commonly genomic rearrangements do occur in P. aeruginosa. For example, sequences around the phospholipase C structural gene are conserved from strain to strain (6), as are sequences 3' to toxA. Pulsed-field electrophoresis of Dral and SpeI fragments shows that related strains differ by six or fewer bands (1). It would be interesting to know whether Hjelm et al. now have additional data regarding the extent of variability seen throughout the chromosome. Most importantly, it is unclear whether isolates with different RFLP patterns were derived from a single clone in these studies. We reported use of a 741-bp PstI-NruI fragment from the 5' region of toxA as an epidemiologic marker to investigate clonal relationships (3, 8). We believe that typing using this fragment correlates with clonal relationships, and our subsequent studies using pulsed-field electrophoresis support our earlier findings. Although definitions of clonal relatedness are not established and agreed upon, others have frequently used several techniques to establish relatedness (1, 5) among isolates. Additional data in this regard would have been of value in clarifying the data
presented. REFERENCES 1. Grothues, D., U. Koopmann, H. von der Hardt, and B. Tummler. 1988. Genome fingerprinting of Pseudomonas aeruginosa indicates colonization of cystic fibrosis siblings with closely related strains. J. Clin. Microbiol. 26:1973-1977. 2. Hjelm, L. N., A. A. Branstrom, and R. L. Warren. 1990. Detection of restriction fragment length polymorphisms in clinical isolates and serially passaged Pseudomonas aeruginosa strains. J. Clin. Microbiol. 28:2178-2182. 3. Ogle, J. W., J. M. Janda, D. E. Woods, and M. L. Vasil. 1987. Characterization and use of a DNA probe as an epidemiological
J. CLIN. MICROBIOL.
marker for Pseudomonas aeruginosa. J. Infect. Dis. 155:119126. 4. Ogle, J. W., A. I. Vasil, and H. R. Rabin. 1988. Abstr. Annu. Meet. Am. Soc. Microbiol. 1988, B141, p. 53. 5. Pasloske, B. L., A. M. Joffe, Q. Sun, K. Volpel, W. Paranchych, F. Eftekhar, and D. P. Speert. 1988. Serial isolates of Pseudomonas aeruginosa from a cystic fibrosis patient have identical pilin sequences. Infect. Immun. 56:665-672. 6. Pritchard, A. E., and M. L. Vasil. 1990. Possible insertion sequences in a mosaic genome organization upstream of the exotoxin A gene in Pseudomonas aeruginosa. J. Bacteriol.
172:2020-2028. 7. Speert, D. P., M. E. Campbell, S. W. Farmer, K. Volpel, A. M. Joffe, and W. Paranchych. 1989. Use of a pilin gene probe to study molecular epidemiology of Pseudomonas aeruginosa. J. Clin. Microbiol. 27:2589-2593. 8. Vasil, M. L., C. Chamberlain, and C. C. R. Grant. 1986. Molecular studies of Pseudomonas exotoxin A gene. Infect. Immun. 52:538-548.
John W. Ogle Department of Pediatrics Michael L. Vasil Department of Microbiology and Immunology University of Colorado School of Medicine 4200 East Ninth Avenue Denver, Colorado 80262
Author's Reply The comments made by Drs. Ogle and Vasil need to be addressed. The strongest evidence for the colonization of cystic fibrosis patients by more than one Pseudomonas aeruginosa strain is the field inversion gel electrophoresis analysis presented by Grothues et al. (1). Both Speert et al. (7) and Ogle et al. (4) have presented data that demonstrate the presence of multiple restriction fragment length polymorphisms (RFLPs) in a single patient. In both instances, the DNA from the isolates had different restriction endonuclease digestion profiles. More convincingly, Ogle and Vasil mention in their letter that subsequent studies using pulsed-field electrophoresis supported the clonal relatedness observed with the RFLPs. We do not question the ability of more than one strain of P. aeruginosa to colonize cystic fibrosis patients. The article by Hjelm et al. (2) emphasizes the point that RFLPs were detected in serially passaged isolates from a single colony as well as isolates from a single patient. As mentioned in Results, the increase observed in the XhoI fragment shown in Fig. 5 was from DNA isolated from serial passages of a single isolate, WSU3860. The RFLPs observed could not be due to methylation alone, since Fig. 6 clearly depicts a restriction map comparing the parent from which the serial passages were made and the subclone exhibiting the increased size of the XhoI fragment. Altered restriction fragments were also generated from BglII and KpnI digests. The data presented in our article demonstrate that RFLPs alone are not sufficient to establish clonal relatedness when the 800-bp HindIII-PstI fragment is used as a marker. Comparing our results to those of Ogle et al. (3) and Pasloske et al. (5), we found that clonally related strains had different RFLPs when probed with the 800-bp HindIII-PstI probe. This demonstrates intrastrain genomic variability that was not previously reported when either the upstream portion of the exotoxin A gene or the pilin gene probe was used. We showed that the 800-bp HindIII-PstI fragment can detect RFLPs with XhoI-, BglII-, and KpnI-digested chromosomal DNA and highly conserved regions of DNA can be detected with DdeI, PstI, and Hinfl chromosomal digests.
VOL. 29, 1991
LETTERS TO THE EDITOR
The variability seen within the region surrounding our 800-bp probe may be the result of IS elements, as was discussed. We did not imply that this intrastrain variability was the result of simple excisions or insertions. The frequency with which genomic rearrangements occurred was enough to detect 2 of 22 phenotypic variants examined. It is interesting that the mosaic genome organization observed by Pritchard and Vasil (6) may be the result of IS-like elements, which does not conflict with our results. However, do the mechanisms that gave rise to the interstrain genomic organizations upstream of the exotoxin A gene also give rise to the intrastrain variability that we observe using the 800-bp probe? The insertion resulting in the increased XhoI fragment needs to be further characterized to make that determination.
665
strains. J. Clin. Microbiol. 28:2178-2182. 3. Ogle, J. W., J. M. Janda, D. E. Woods, and M. L. Vasil. 1987. Characterization and use of a DNA probe as an epidemiological marker for Pseudomonas aeruginosa. J. Infect. Dis. 155:119126. 4. Ogle, J. W., A. I. Vasil, and H. R. Rabin. 1988. Abstr. Annu. Meet. Am. Soc. Microbiol. 1988, B141, p. 53. 5. Pasloske, B. L., A. M. Joffe, Q. Sun, K. Volpel, W. Paranchych, F. Eftekhar, and D. P. Speert. 1988. Serial isolates of Pseudomonas aeruginosa from a cystic fibrosis patient have identical pilin sequences. Infect. Immun. 56:665-672. 6. Pritchard, A. E., and M. L. Vasil. 1990. Possible insertion sequences in a mosaic genome organization upstream of the exotoxin A gene in Pseudomonas aeruginosa. J. Bacteriol. 172:2020-2028. 7. Speert, D. P., M. E. Campbell, S. W. Farmer, K. Volpel, A. M. Joffe, and W. Paranchych. 1989. Use of a pilin gene probe to study molecular epidemiology of Pseudomonas aeruginosa. J. Clin. Microbiol. 27:2589-2593.
REFERENCES 1. Grothues, D., U. Koopmann, H. von der Hardt, and B. Tummier. 1988. Genome fingerprinting of Pseudomonas aeruginosa indicates colonization of cystic fibrosis siblings with closely related strains. J. Clin. Microbiol. 26:1973-1977. 2. Hjelm, L. N., A. A. Branstrom, and R. L. Warren. 1990. Detection of restriction fragment length polymorphisms in clinical isolates and serially passaged Pseudomonas aeruginosa
Arthur A. Branstrom Department of Bacterial Immunology Richard L. Warren Department of Bacterial Diseases Division of Communicable Disease and Immunology Walter Reed Army Institute of Research Washington, D.C. 20307
