Adherence of bacteria to human foreskins

Circumcision is one of if not the oldest operation known to mankind. It has been documented as part of religious rites and tribal customs in sun worshiping cultures of Egypt and Greece as early as 12,000 B.C. Certain religions continue the procedure as a part of their belief. Since the nineteenth century this religious custom has become a common surgical procedure, often without religious association and mostly in the English-speaking world.[1]

Whether circumcision has a medical indication has been questioned in recent years. In fact the American Academy of Pediatrics, and the American College of Obstetricians and Gynecologists stated that "There is no absolute medical indication for the routine circumcision of the newborn".[2] Proponents believe that circumcision prevents infection, venereal disease and penile cancer but it has been suggested that good hygienic standards prevent these penile problems as well as circumcision.[1]

Lincoln and Winberg in 1964[3] and Littlewood in 1972[4] found a higher incidence of urinary tract infections in the male than in the female neonate. Ginsburg and McCracken reported that of the male infants with urinary tract infection 95 per cent were uncircumcised.[5] Wiswell and associates reported a 10-fold increase of urinary tract infection in uncircumcised male infants over that in female infants and circumcised male infants.[6] Since this is so, bacterial colonization of the foreskin could be the cause. Indeed, Enterobacteriaceae that produce ascending urinary tract infection have been found to colonize the foreskin of male neonates.[7] Once the foreskin has been colonized these bacteria could ascend the urinary tract via the urethra, producing cystitis, pyelonephritis, sepsis and/or septic death. Bacterial fimbrial adhesion to the mucosal surface of the foreskin could aid in the colonization of the foreskin.

Most pathogenic bacteria have the ability to adhere to epithelial cells. These adhesins often are fimbrial in nature. Fimbriae are hair-like projections from bacteria that adhere to cells by specific and nonspecific means.[8] Nonspecific adherence may be owing to hydrophobic or electrostatic factors. The more specific fimbrial adherence is that which occurs because of receptors of glycoprotein and glycolipid nature on the mucosa of the urogenital tract.[9] There are various bacterial fimbrial types but types 1 and P are those most understood,in urinary tract infections. Type 1 adherence is mannose sensitive, as these fimbriae adhere to mannose molecules of glycoproteins that comprise mucus, acting as receptors.[9] However, P fimbriae, named after the glycolipid antigens of the human P-blood group and the fact that they are strongly associated with pyelonephritis,[10] are mannose resistant and adherence is to a specific glycolipid receptor.[11]

Roberts proposed a hypothesis that if circumcision could prevent urinary tract infection it might do so by removing the substrate (the foreskin) for the initial attachment and later colonization with nephropathogenic bacteria.[12] To test this hypothesis human foreskins from male neonates were incubated in vitro with different Enterobacteriaceae (pathogenic and nonpathogenic strains) and observed in the electron micro-scope for bacterial attachment, fimbriae and colonization of the skin compared to the mucosa of the foreskin.

Materials and methods

Bacteria. The bacteria used were strains of Escherichia coli, Serratia oderiferous, Klebsiella pneumoniae, Proteus mirabilis and Pseudomonas aeruginosa. These strains of bacteria came from patients with acute pyelonephritis (strain JR1 and nosocomial strain G211), prostatitis (strain JR340, and nosocomial strains K3 and JR500), cystitis (strain G114 and nosocomial strain K2) or asymptomatic bacteriuria (strain JR7, and nosocomial strains G87, P2 and Gl). They were cultured on either colonization factor antigen agar plates for 24 hours to encourage P fimbriation or in tryptose broth for 2, 48-hour periods to encourage expression of type 1 fimbriae. The type of fimbriae on each bacterial strain was determined by the P-fimbriae specific particle agglutination test for P fimbriae,[13] and the mannose inhibitable agglutination of yeast cells for type 1 fimbriae.[14]

The hydrophobicity and electrostatic charges of the bacterial strains were determined according to the method of Hermansson and associates.[15] Briefly, bacteria were cultured overnight at 37C in tryptose broth to which had been added tritiated (³H)-leucine at 0.35 μCi./ml. The bacteria then were washed twice with phosphate buffered saline at pH 7.5 and resuspended in the same buffer to a concentration of 1 x 109 bacteria per ml. phosphate buffered saline. Either Octyl-Sepharose CL-4B* (for hydrophobicity) or diethyl amino ethyl-Sepharose CL-6B* (anionic exchange) was packed into small columns and washed with phosphate buffered saline, and a 1 ml. sample of the bacteria was added to each column. The columns were washed 4 times with 2 ml. fractions of phosphate buffered saline that were collected into scintillation vials as was the gel after the last eluent. Ten ml. scintillation fluor were added to each vial and the vials were counted for 10-minute periods in a Beckman LS 7800 liquid scintillation counter. The counts per minute of the 4 elution fractions were totaled. The ratio of counts per minute of the bacteria attached to the gel to counts per minute of bacteria in eluent fractions was calculated. The higher the ratio the higher the tendency toward hydrophobicity or electrostatic interaction.

* Pharmacia, Fine Chemicals, Inc., Uppsala, Sweden.

Incubation and fixation. The foreskin tissue samples were obtained from 24 specimens taken during routine circumcision at 2 community hospitals and transported to the laboratory in sterile saline. The patients most often were 2 days old. The tissue was then trimmed into pieces containing skin and mucosal (foreskin in opposition to the glans) surfaces. The specimens were incubated with bacteria at a concentration of 1 x 109 cells per ml. at 37C with constant rotation for 1, 2 or 4 hours. Then, the samples were fixed in modified Karnovsky's fixative[16] for scanning and transmission electron microscopy, and histological study by light microscopy.

Samples for scanning electron microscopy were washed 3 times in 0.2 M. sodium cacodylate buffer, 3 times in distilled water, dehydrated in ascending ethanols and critical point dried in absolute ethanol. They were then sputter coated with gold, examined and photographed in a JEOL-T300 scanning electron microscope. Tissues for transmission electron microscopy were washed 6 times with 0.2 M. sodium cacodylate buffer, post-fixed in 1.0 per cent osmium tetroxide, dehydrated in ascending concentrations of ethanol and embedded in Spurr's epoxy resin.[17] A Siemens Elmiskop 101 electron microscope was used to observe the sectioned samples. To compare adhesion between strains the number of bacteria adhering to each 7,500x field was determined.

To determine whether the attachment of the JR1 E. coli strain was owing to type P or 1 fimbriae 1 x 10⁹ bacteria per ml. were incubated at room temperature with the receptor α-Galp-1-4-β-Galp O methyl (30 mmol.) and/or 5 per cent man-nose for 30 minutes. Foreskin samples then were added to the suspension and incubated with rotation at 37C for 2 hours. Then, they were processed for electron microscopy. Samples for histological study were processed and stained in our histology laboratory.

Results

The human foreskin is a thin layer of skin that folds down over the glans penis. This layer consists of a typical stratified squamous epithelial epidermis and an underlying dermis as shown by Rhodin.[18] Around the area of the fold there is a transition between the keratinized epithelium of the skin with a rough-appearing surface, and the nonkeratinized epithelium of the mucosa with a smooth-appearing surface (fig. 1). The mucosal surface surrounds the glans penis from the fold down to its attachment at the base of the glans.

Fig. 1. Scanning electron micrograph of transition between keratinized epithelium (K) of skin surface and nonkeratinized epithelium (N) of mucosal surface of human foreskin. Note that bacteria (E. coli strain G211) attach to mucosal surface (arrows) and not skin surface. Reduced from x1,800.

The keratinized epithelial cells of the skin become filled with keratin filaments that replace most of the cellular organelles, forming the horny layer or stratum corneum. This layer is only a few cells thick (fig. 2, A and B). The outer layer, the stratum disjunctum (also a few cells thick), becomes detached giving the surface its rough appearance (figs. 1, and 2, A and B). The nonkeratinized epithelium of the mucosa has keratin filaments within the superficial cells but they do not become as compacted as those in the keratinized epithelium (fig. 2, C and D). Al-though these mucosal epithelial cells are sloughed they are not shed at the same rate as the skin epithelium and, thus, the surface is smooth (figs. 1, and 2, C and D). With the addition of ruthenium red no ultrastructural difference in the glycocalix of either the keratinized or nonkeratinized superficial epithelial cells could be determined.

FIG. 2. A, light microscopic section of thin skin surface of human foreskin. 1, skin surface. 2, keratinized squamous epithelial cells. 3, stratified squamous epithelium. 4, dermis. Reduced from x750. B, transmission electron micrograph of keratinized skin surface of human foreskin. 1, stratum disjunctum gives surface rough-looking appearance. 2, stratum comeum. 3, nucleus of squamous cell. Reduced from x12,500. C, light microscopic section of mucosal surface of human foreskin. 1, area between mucosal surface and glans penis. 2, nonkeratinized squamous epithelium. 3, dermis. Reduced from x750. D, transmission electron micrograph of nonkeratinized mucosal surface of human foreskin. Note keratin filaments are not condensed and desmosome junctions (arrowheads) are still intact, producing smooth mucosal surface. Reduced from x12,500.

Bacteria adhered to the cellular membranes of the mucosg cells but not to the skin cells. Figure 1 shows the difference between the skin and mucosal surfaces. There are no bacteria on the skin (rough cells) but many are attached to the mucosal surface (the smooth cells). The control foreskin before experimental incubation with bacteria after saline washing did not show bacteria attached to either surface. A progression of the number of attached E. coli strain JR1 was noted with time, with 13, 32 and 66 bacterial cells per 7,500x power field at 1, 2 and 4 hours, respectively. The addition of the receptor Gal-Gal and/or 5 per cent mannose did not prevent the attachment of strains JR1 or G211 to foreskin mucosal surfaces.

Fig. 3. A, transmission electron micrograph of adherence of E. coli JR1 strain to mucosal surface of human foreskin. Reduced from x2,000. B, enlargement of part A with attachment of E. coli to mucosal surface by fimbriae (arrowheads). C, cell division. F, flagellum. Reduced from x10,000. C, enlargement of part 13 with attachment of E. coli by fimbriae (arrowheads). c, cell division. Reduced from x20,000. Inset, some bacteria (asterisk) appear to attach without fimbriae. Distance (arrowheads) between bacteria (asterisk) and epithelial cellular membrane (Ep) is 550 to 600 A. Reduced from x32,000.

Figure 3, A demonstrates the colonization of E. coli strain JR1 on the mucosal surface at low magnification. At a higher magnification the fimbriae on JR1 cells can be observed while flagella are noted on some of the cells (fig. 3, B). Different stages of cellular development also can be seen. There are large cells, dividing cells and daughter cells (fig: 3, B). On closer examination the fimbriae can be seen attaching to the mucosal surface with a cell division line in some of the larger cells (fig. 3, C). Fimbriae can be observed on the upper surface of the bacteria as white dots (fig. 3, C). When the G114 strain of P. mirabilis was incubated with the foreskin for 2 hours there was massive colonization (100 cells per 7,500x) of th'e mucosal surface (fig. 4). As with E. coli strain JR1, many fimbriae and flagella were observed. The characteristics of the bacterial strains studied and the relative values for hydrophobic inter-action and negative electrostatic charge are shown in the table. Some of the bacteria that attached to the mucosal epithelial cells did so without noticeable fimbriae. Upon closer examination with transmission electron microscopy these bacteria are at a distance of about 550 to 570 A. from the epithelial cellular membrane (fig. 3, C, inset). Fimbriated and nonfimbriated adherent strains showed a significant association with hydrophobicity (fig. 5, A) and tended to be more negatively charged (fig. 5, B).

Fig. 4. Massive colonization of P. mirabilis G114 strain to mucosal epithelium. Fimbriae (arrowheads). F, flagellum. Reduced from x15,000. Inset, attachment (arrowheads) of P. mirabilis to mucosal epithelium (Ep). Distance between bacteria and epithelium is 550 to 600 A (similar to figure 3, inset). Reduced from x38,000.

The receptors for types P and 1 fimbriae found on E. coli are well known and were tested not only for hemagglutination but for inhibition by either Gal-Gal in the case of P fimbriae or mannose in the case of type 1 fimbriae. The Proteus strains were tested with human red cells. Strain G1 did not agglutinate these cells but strain G114 did in a manner not inhibited by either Gal-Gal or mannose, which suggests a different fimbrial receptor for G114. Pseudomonas strain P2 agglutinated yeast but the agglutination was not inhibited by mannose, which suggests nonspecific adherence for this nonfimbriated strain. Serratia exhibited a positive reaction to the P-fimbriae specific particle agglutination test but also to the control test (not P) and a mannose-inhibited fimbrial agglutination of yeast, suggesting type 1 fimbriae. Klebsiella K2 showed mannose-inhibited agglutination of yeast only, which suggests type 1 fimbriae, while K3 did not agglutinate yeast and had no fimbriae on transmission electron microscopy.

Fig. 5. A, correlation between hydrophobicity and bacterial adhesion is depicted. Bacteria adherent to cells within 7,500x field were counted. Note significant correlation coefficient. B, correlation between negative electrostatic charge and bacterial adherence (7,500x field) is depicted with correlation coefficient of 0.57.

Table: Characteristics of bacteria.

Discussion

Urinary tract infections are rare in men until old age but they are more frequent in the male than in the female newborn.[3,4] 4 However, this higher incidence of urinary tract infection is true only in the uncircumcised male neonate.[5,6] The mucosal surface of the foreskin was colonized readily by pathogenic E. coli, fimbriated strains of P. mirabilis, and nonfimbriated Pseudomonas, Klebsiella and Serratia. They did not adhere to the keratinized skin surface of the foreskin. We believe that once the bacteria attach and colonize the tissue near the urethra they can ascend the urethra easily to produce urinary tract infection.

The ability of bacteria to attach to uroepithelial cells has been well documented. Kallenius and Winberg reported that bacterial adherence to periurethral cells preceded urinary tract infections in girls.[19] That same year Smith and Roberts found bacterial attachment to kidney tubular cells via fimbriae in the monkey model of pyelonephritis.[20] Fussell and Roberts reported that after inoculation with E. coli attachment to kidney tubular cells can be seen as early as 6 hours and by 12 hours attachment to ureteral epithelial cells has been observed. In both cases fimbriae are prominent.[21] In vitro attachment of bacteria to the mucosal surface of foreskins by fimbriae occurred during 1 hour of incubation.

Adhesion appears to correlate with fimbrial and nonspecific adherence factors. The presence of fimbriae seemed to be the only factor responsible for adherence in the case of P. mirabilis and S. odoriferous, while in the case of P. aeruginosa and the Klebsiella strains hydrophobicity and electrostatic charge correlated more with adherence than the presence of fimbriae. In the case of the E. coli strains there seems to be several adherence mechanisms. Electrostatic charge was not always a deciding factor but it seemed to require some additional factor as noted in strain JR340 in which charge was high but adherence low. Over-all, adherence related best to relative hydrophobicity as shown by a correlation coefficient of 0.97 (fig. 5, A ). Strains JR1 and G211, the strains showing the greatest hydrophobicity ratio and adherence, in addition were derived from patients with acute pyelonephritis unlike the other E. coli strains.

Hermansson and associates reported the importance of hydrophobic and electrostatic forces in adhesion to cell surfaces.[15] Both of the Proteus strains tested are highly negatively charged but show little hydrophobicity. Proteus G1, which did not adhere, has no fimbriae but G114 does. Therefore, in this case fimbrial attachment seems to be important. Thus, all 3 characteristics, fimbriae, hydrophobic interaction and electrostatic charges, may be involved. Therefore, our thesis that circumcision might prevent urinary tract infection by removing the substrate for adhesion and colonization by pathogenic bacteria is supported.

Ms. Beth Underwood, Louisiana State University Dental School, provided technical assistance with the scanning electron microscope.

References

1. Gairdner, D.: The fate of the foreskin: a study of circumcision. In: Current Operative Urology, 2nd ed. Edited by E. D. Whitehead and E. Leiter. Philadelphia: Harper & Row, part 36, chapt. 92, pp. 1189-1195, 1984.
2. American Academy of Pediatrics and the American College of Obstetricians and Gynecologists: Guidelines for Perinatal Care. Edited by A. W. Brann and R. C. Defalo. Evanston, Illinois: American Academy of Pediatrics, p. 87, 1983.
3. Lincoln, K. and Winberg, J : Studies of urinary tract infections in infancy and childhood. IL Quantitative estimation of bacteriuria in unselected neonates with special reference to the occurrence of asymptomatic infections. Acta Paed., 53: 307, 1964.
4. Littlewood, J. M.: 66 infants with urinary tract infections in first month of life. Arch. Dis. Child., 47: 218, 1972.
5. Ginsburg, C. M. and McCracken, G. H., Jr.: Urinary tract infections in young infants. Pediatrics, 69: 409, 1982
6. Wiswell, T. E., Smith, F. R. and Bass, J. W.: Decreased incidence of urinary tract infections in circumcised male infants. Pediatrics, 75: 901, 1985.
7. Winberg, J., Andersen, H. J., Bergstrom, T., Jacobsson, B., Larson, H. and Lincoln, K.: Epidemiology of symptomatic urinary tract infection in childhood. Acts Paed. Scand., suppl. 252, p. 1, 1974.
8. Duguid, J. P., Clegg, S. and Wilson, M. I.: The fimbrial and nonfimbrial haemagglutinins of Escherichia coli. J. Med. Microbiol., 12: 213, 1978.
9. Ofek, I., Mirelman, D. and Sharon, N.: Adherence of Escherichia coli to human mucosal cells mediated by mannose receptors. Nature, 265: 623, 1977
10. Kallenius, G., Mollby, R., Svenson, S. B., Winberg, J., Lundblad, A., Svensson, S. and Cedergren, B.: The pk antigen as receptor for the haemagglutinin of pyelonephritic Escherichia coli. FEMS Microbiol. Lett., 7: 297, 1980.
11. Kallenius, G., Mollby, R., Svenson, S. B., Winberg, J. and Hultberg, H.: Identification of a carbohydrate receptor recognized by uropathogenic Escherichia coli. Infection, suppl. 3, 8: 288, 1980.
12. Roberts, J. A.: Does circumcision prevent urinary tract infection? J. Urol., 135: 991, 1986.
13. Svenson, S. B., Kallenius, G., Mollby, R., Hultberg, H. and Winberg, J.: Rapid identification of P-fimbriated Escherichia coli by a receptor-specific particle agglutination test. Infection, , 10: 209, 1982.
14. Roberts, J. A., Suarez, G. M., Kaack, B., Domingue, G. J. and Svenson, S. B.: Host-parasite relationships in acute pyelonephritis. Amer. J. Kidney Dis., 8: 139, 1986.
15. Hermansson, M., Kjelleberg, S., Korhonen, T. K. and Stenstrom, T. A.: Hydrophobic and electrostatic characterization of surface structures of bacteria and its relationship to adhesion to an air-water interface. Arch. Microbiol., 131: 308, 1982.
16. Karnovsky, M. J.: A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy. J. Cell. Biol., 27: 137A, 1965.
17. Spurt, A. R.: A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res., 26: 31, 1969.
18. Rhodin, J. A. G.: Histology: A Text and Atlas. New York: Oxford University Press, pp. 68-72 and 476-482, 1974.
19. 19. Kallenius, G. and Winberg, J.: Bacterial adherence to periurethral epithelial cells in girls prone to urinary-tract infections. Lancet, 2: 540, 1978.
20. 20. Smith, T. W., Jr. and Roberts, J.' A.: Chronic pyelonephritis: an electron microscope study in nonhuman primates. Invest. Urol., 16: 148, 1978.
21. Fussell, E. N. and Roberts, J. A.: The ultrastructure of acute pyelonephritis in the monkey. J. Urol., 132: 179, 1984.