JOURNAL OF CLINICAL , Apr. 1984, p. 541-542 Vol. 19, No. 4 0095-1137/84/040541-02$02.00/0 Copyright © 1984, American Society for Microbiology penneri and Urinary Calculi Formation SIGMUND KRAJDEN,* MILAN FUKSA, WALTER LIZEWSKI, LESLEY BARTON, AND ALBERT LEE Department of Microbiology, Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada Received 3 October 1983/Accepted 16 December 1983

The clinical significance of Proteus penneri, a newly described species, is unknown. A case report is presented, which is to the best of our knowledge the first description of this organism causing a and bladder calculi.

Recent revisions in the classification of the Proteeae, a nonhemolytic, were identified (isolates 2 and 3, respective- tribe of belonging to the family Enterobacteriaceae, ly). recognized three genera, namely, Proteus, Providencia, and Once again antimicrobial therapy was ineffective in clear- Morganella (3). In the genus Proteus, three species were ing the infection, since on 4 November 1982, midstream identified. These included the well-known P. vulgaris and P. urine culture revealed three strains of >105 CFU of P. mirabilis species. A third, P. myxofaciens, was proposed for penneri per ml (isolates no. 4, 5, 6, respectively). A kidney- an organism isolated from gypsy moth larvae (Porthetria ureter-bladder X ray demonstrated three stones in the blad- dispar), but is not of clinical importance. Bacteria formerly der. These were removed surgically, and a pure growth of in the genus Proteus and previously classified as Proteus P. penneri was isolated from the center of one stone (isolate rettgeri and Proteus morganii were reassigned to the genera no. 7). Subsequent urine cultures were negative for P. penneri, Providencia and Morganella, respectively. Subsequent stud- and the hematuria resolved. The stones were composed of a ies of 20 indole-negative strains generally grouped with P. mixture of uric acid and magnesium ammonium phosphates. vulgaris led to the finding that they constituted a separate In an attempt to identify the reservoir of P. penneri in the species (8). The name given to this species was Proteus patient, swabs of the rectum, perineum, scrotum, penile penneri (10). Although strains of this species have been shaft, and abdominal skin were cultured. These were nega- reported to be isolated from urine, stool, and blood speci- tive for P. penneri. Urine cultured from each of the five mens, their clinical significance is largely unknown (8). With other patients that shared the same room were negative for the report of this new species, more attention was given to P. penneri. the precise classification of the Proteus species in our Urinary tract infections caused by P. mirabilis and P. laboratory. This species may be differentiated by tests for vulgaris have been described on numerous occasions (1, 5, esculin hydrolysis, salicin, and indole (all negative at 48 h) 12-14). Some of these may have been due to organisms that which are readily performed in clinical laboratories. This would have been classified as indole-negative P. vulgaris. approach has led to the diagnosis of an infection caused by Little, if any, significance would have been attached to the P. penneri. We present in this paper a case report that, to our negative indole reaction as it would most likely have been knowledge, describes the first detailed account of such an considered to be a biochemical variation. However, since infection. these bacteria are now known to constitute a separate Case report. An 84-year-old man was admitted to the species (10), it may be expected that certain biological hospital in 1974 after a cerebrovascular accident. He devel- properties, other than the ones that are defined readily in the oped a nosocomial urinary tract infection with Escherichia microbiology laboratory, may be different from those of P. coli and Klebsiella pneumoniae. Subsequently, he was trahs- vulgaris or P. mirabilis, such as differences in pathogenicity. ferred to a chronic care facility where he remained for the More infections need to be assessed and compared with P. next 8 years. His course was punctuated by symptomatic vulgaris infections to demonstrate such differences. urinary tract infections with a variety of gram-negative production is a characteristic of P. penneri (8) bacilli, managed with appropriate antimicrobial therapy. For which is therefore similar to other species in the genus the last 3 years, he had no bacteriuria. Proteus. The enzymatic breakdown of urea results in in- In August 1982, the patient was noted to have painless, creased urinary alkalinity and bicarbonate and ammonia foul-smelling hematuria. Otherwise, he was asymptomatic. concentrations, all of which favor the precipitation of mag- Urinalysis revealed 30 to 40 erythrocytes and 3 leukocytes nesium, ammonium, phosphate, and carbonate apatite (6, 7, per high-power field, trace protein, no glucose, pH 9, and a 10). specific gravity of 1.011. Urine culture demonstrated a pure Urinary tract infections range in severity from simple growth of >105 CFU of P. penneri per ml (isolate no. 1). bacterial growth in the urine without symptoms (asympto- Treatment was instituted with appropriate antibiotics; how- matic or covert bacteriuria) to massive bilateral renal infec- ever, hematuria persisted, and a urological consult was tion resulting in chronic renal failure (4). In the untreated obtained. On 24 September 1982, the patient underwent patient, whether symptomatic or asymptomatic, counts of cystoscopy which revealed a severely inflamed and hemor- more than 105 CFU per ml of urine are indicative of infection rhagic bladder mucosa with diverticuli. On 21 October 1982, (2). urine culture again yielded pure growth of >105 CFU of P. Our patient in whom P. penneri was isolated in pure penneri per ml. Two variants, one hemolytic and the other culture on three separate occasions (> 105 CFU per ml) would be classified as having a urinary tract infection (2). Therefore, urinary tract stone formation as found in our * Corresponding author. patient infected with P. penneri is consistent with struvite 541 542 NOTES J. CLIN. MICROBIOL. stone formation with other members of the Proteeae group penneri infections need to be studied to define the full of organisms. The isolation of P. penneri from the center of spectrum of disease. the stone in our patient suggests that a urinary tract infection by this urea-splitting organism preceded the formation of We thank John Penner for critical review of the manuscript and bladder calculi. These calculi then traumatized the bladder Pat Catterall for typing the manuscript. mucosa, resulting in hematuria. The paucity of leukocytes LITERATURE CITED on urinalysis requires further comment. Pyuria (the presence of at least 3 to 10 leucocytes per high-power field) has not 1. Adler, J. L., J. P. Burke, D. F. Martin, and M. Finland. 1971. been found to be a reliable guide to the presence of infection; Proteus infections in a general hospital. II. Some clinical and indeed, only about epidemiological characteristics with an analysis of 71 cases of 50% of patients with significant bacteri- Proteus bacteremia. Ann. Intern. Med. 75:531-536. uria will have the usual 3 to 10 leucocytes per high-power 2. Barry, A. L., P. B. Smith, and M. Turck. 1975. Cumitech 2, field that are often considered meaningful (9, 11). In addi- Laboratory diagnosis of urinary tract infections. Coordinating tion, it is well known that alkaline urine, as found in our ed., T. L. Gavan. American Society for Microbiology, Washing- patient, very rapidly lyses leucocytes (15). ton, D.C. We could not identify any reservoir of P. penneri in our 3. Brenner, D. J., J. J. Farmer III, G. R. Fanning, A. G. Steiger- patient or his environment. Clearly, further studies with walt, P. Klykken, H. G. Wathen, F. W. Hickman, and W. H. larger groups of patients are needed to elucidate the ecology Ewing. 1978. Deoxyribonucleic acid relatedness of Proteus and of this organism. Providencia species. Int. J. Syst. Bacteriol. 28:269-282. 4. Cotran, R. S., and J. E. Pennington. 1981. Urinary tract Biochemically, all seven isolates from our patient were infection, pyelonephritis, and reflux nephropathy, p. 1571-1632. negative for the following reactions: indole (24 h, 48 h), In B. M. Brenner and F. C. Rector (ed.), The kidney. The W. B. Voges-Proskauer, Simmons citrate, L-lysine, L-arginine, L- Saunders Co., Philadelphia, Pa. ornithine, malonate, arabinose, adonitol, cellobiose, dulci- 5. Evans, A. T. 1974. Nosocomial infections and the urologist. J. tol, inositol, lactose, mannitol, melibiose, raffinose, rham- Urol. 111:813-816. nose, salicin, sorbitol, trehalose, amygdalin, esculin (24 h, 48 6. Griffith, D. P., and D. M. Musher. 1973. Prevention of infected h), oxidase, and o-nitrophenyl-p-D-galactopyranoside. The urinary stones by urease inhibition. Invest. Urol. 11:228-233. following reactions were positive for all seven strains: meth- 7. Griffith, D. P., D. M. Musher, and C. Itin. 1976. Urease: the primary cause of infection-induced urinary stones. Invest. Urol. yl red (24 h, 48 h), hydrogen sulfide, urea, phenylalanine, 13:346-350. maltose, and sucrose. 8. Hickman, F. W., A. G. Steigerwalt, J. J. Farmer III, and D. J. These biochemical characteristics are identical with those Brenner. 1982. Identification of Proteus penneri sp. nov., for- of the strain of P. penneri in the American Type Culture merly known as indole negative or as Proteus Collection (ATCC 33519) (8). There were strain variations vulgaris biogroup 1. J. Clin. Microbiol. 15:1097-1102. among our seven isolates. On sheep blood agar, isolates 3 9. Kunin, C. M. 1979. An overview of urinary tract infections, p. and 6 were nonhemolytic whereas isolates 2, 4, 5, and 7 were 1-56. In C. M. Kunin (ed.), Detection, prevention and manage- beta-hemolytic (37°C, 24 h). Isolates 1, 2, and 6 were xylose ment of urinary tract infections. 3rd ed. Lea & Febiger, Phila- positive, with the others being xylose negative. The gelatin delphia, Pa. reaction was for all of our seven 10. List No. 11. 1983. Validation of the publication of new names negative isolates but was and new combinations previously effectively published outside positive for strain ATCC 33519 (8). All seven isolates of P. the IJSB. Int. J. Syst. Bacteriol. 33:672-674. penneri were inhibited by at a concentration > 16 11. McGukin, M., L. Cohen, and R. R. McGregor. 1978. Signifi- p.g/ml, at -32 [Lg/ml, at >64 ciance of pyuria in urinary sediment. J. Urol. 120:452-454. Ftg/ml, and cefamandole at >64 p.g/ml. All strains were 12. Resnick, M. I. 1981. Evaluation and management of infection inhibited by at concentrations ranging from 1 to 4 stones, p. 265-276. Urologic clinics of North America, vol. 8, Vjg/ml, at <0.5 to 8 VLg/ml, and amikacin at 4 to no. 2. The W. B. Saunders Co., Philadelphia, Pa. 32 ,ug/ml. The minimal inhibitory concentration against 13. Sullivan, N. M., V. L. Sutter, M. M. Mims, V. H. Marsh, and ranged from 2 to 8 p.g/ml; however, a third- S. M. Finegold. 1973. Clinical aspects of bacteriuria after manipulation of the genitourinary tract. J. Infect. Dis. 127:49- generation , cefotaxime, was less active with 55. minimal inhibitory concentrations 2 32 ,ug/ml. 14. Wallace, J. J., and R. G. Petersdorf. 1971. Urinary tract Therefore, asymptomatic bacteriuria, with bladder calculi infections. Postgrad. Med. 50:138-144. formation causing traumatic hematuria, may be manifesta- 15. Wilson, D. M. 1975. Urinalysis and other tests of renal function. tions of infection with P. penneri. Additional patients with P. Minn. Med. 58:9-17.