Adhesion and Internalization of E.coli Strains Expressing Various Pathogenicity Determinants

Zbl. Bakt. 278, 218-228 (1993) © Gustav Fischer Verlag, StuttgartlNew York

Adhesion and Internalization of E. coli Strains Expressing Various Pathogenicity Determinants E. STRAUBE!, G. SCHMIDT2, R. MARRE\ and J. HACKER 4 lInstitut fiir Medizinische Mikrobiologie der Friedrich-Schiller-Universitiit Jena, 0-6900 Jena, Germany 2 Institut fiir Experimentelle Biologie und Medizin, W-2061 Borstel, Germany 3 Institut fiir Medizinische Mikrobiologie der Universitiit VIm, W-7900 VIm, Germany 4 Theodor-Boveri-Institut fiir Biowissenschaften, Universitiit Wiirzburg, W-8700 Wiirzburg, Germany

With 2 Figures

Summary The adhesion of Escherichia coli to host epithelial cells is the very first step of urinary tract infections followed by the internalization of some bacteria into these cells. These steps are influenced by several surface antigens or products of the pathogen, e.g. fimbriae or adhesins, K antigen, and hemolysin. The bacterial adherence and the internalization of several mutants of an E. coli 018:KS strain differing in the expression of KS antigen, hemolysin, and fimbriae were measured using a permanent line of porcine tubuloepithelial cells (LLC-PK 1 ). Strains with KS antigen were reduced in adherence and internalization in comparison to the K-negative strains. The expression of hemolysin by these strains lead to an increase of adherence and internalization. The internalization of bacteria is influenced mainly by their adherence to the epithelial cells. Thus, the internalization of attached bacteria is rather a kind of endocytosis than an invasion of the bacteria. To confirm this thesis, we investigated the influence of cytoskeletal inhibitors (cytochalasine B, cytochalasine D, colchicine, and chloroquine) on bacterial adherence and internalization. The cytoskeletal inhibitors lead to a significant inhibition of internalization of the bacteria tested. The receptor-mediated endocytosis of bacteria by tubuloepithelial cells may be of importance in the pathogenesis of recurrent urinary tract infection.

Zusammenfassung Die Adhiision von Escherichia coli an Wirtszellen ist ein sehr friiher Schritt bei der Harnwegsinfektion. Er wird gefolgt von der Internalisierung der Bakterien durch diese Zellen. Diese Vorgiinge werden durch verschiedene Faktoren, wie Fimbrien, K-Antigen und Hiimolysin beeinflulk Wir untersuchten verschiedene Mutanten eines Escherichia coli 018:KS Stammes, die sich in der Expression von Fimbrien, des KS-Antigens und des Hiimolysins unterschieden. Die Untersuchungen wurden mit Hilfe der Tubulusepithel-Zellinie

Adhesion and Internalization of E. coli Strains

219

LLC-PK1 vorgenommen. Die Stamme mit K5-Antigen zeigten eine verminderte Adharenz und auch eine verminderte Internalisierung im Vergleich zu den K5-neg. Stammen. Stamme mit Hamolysin wiesen hier eine hohere Aktivitat auf. Da die Internalisierung offenbar vorwiegend durch das Ausmag der Adharenz beeinflugt wird, mug angenommen werden, dag die Internalisierung durch Endozytose zustandekommt. Urn diese Annahme zu priifen, untersuchten wir den Einflug verschiedener Inhibitoren zytoskelettaler Funktionen (Cytochalasin B, Cytochalasin D, Kolchizin und Chloroquin) auf die bakterielle Adharenz und Internalisierung. Diese Inhibitoren beeinflugten die Adharenz nicht, fiihrten aber zu einer drastischen Reduktion der Internalisierung. Die rezeptorvermittelte Endozytose konnte somit von Bedeutung bei der Pathogenese der rezidivierenden Harnwegsinfektion sein.

Introduction Escherichia coli is the most common pathogen in the human urinary tract. Infection is brought about by means of several virulence factors. A very first step of infection i; the adherence of the bacteria to the surface of the uroepithelial cells (3, 7, 14, 18,25). The adherence is mediated by adhesins which are incorporated in fimbrial structures (7, 8). According to the membrane structure recognized as a receptor, the bacterial adhesins are divided into different groups (8, 10, 16, 22). The adherence property of bacteria, however, does not only depend on the type of fimbriae or adhesins and the number of fimbriae expessed, but is also influenced by other surface structures as e.g. K antigens. A further step of infection is the invasion of the epithelial cells by some of the bacteria. Especially hemolysin commonly produced by uropathogenic E. coli strains is believed to contribute to the invasions process (7). The epithelial cells may also contribute to the internalisation of bacteria by an endocytosis-like process (24). While a bulk of information exists on the adherence process, only few data are available about the internalisation of E. coli strains in uroepithelial cells. We therefore investigated the adherence of several E. coli 018 strains differing in their expression of virulence factors, together with their internalisation in porcine tubuloepithelial cells. In additional experiments, the influence of endocytosis inhibitors on the endocytosis of E. coli 06 strains with different adhesin determinants was investigated.

Material and Methods

Strains. The E. coli 018 strains used have been listed in Table 1. Capsule-negative strains were selected using the bacteriophage 0 K5 (26). Mutants of E. coli strain 2980 were selected following mutagenesis by N-methyl-N-nitroso guanidine (NTG) (27). The E. coli 06 K15 strains as listed in Table 2 have been described previously (15); they were used for experiments with inhibitors of endocytosis. The bacteria were cultured on Loeb agar plates (14) for 24 h, washed twice and suspended in medium M 199 (Gibco-BRL, Eggenstein, Germany) at a final concentration of 108 cfu/ml for adhesion experiments and 106 cfulml, respectively, when used for invasion experiments. Eucaryotic cells. A porcine tubuloepithelial cell line (LLC-PK1/ATCC CL 101) was used for the experiments. The cells were cultured in medium M 199 (Gibco) supplemented with 2 mmol L-Glutamine and 5% fetal calf serum. For the adherence assay, the cells were grown in 24-well plates (Greiner, Niirtingen, Germany) until a confluent cell layer was obtained.

220

E. Straube, G. Schmidt, R. Marre, and J. Hacker

Table 1. Escherichia coli 018 strains used Antigenic Characteristics

Strain No.

B2052* B2332 B2062* B2063 20565* B2020 20887* B2021 20383* B2037 2980* 20485** 20469** 20470** B2313** B2314** 20397** B2789** B2791** B2792**

018

K5

+ + + + + + + + + + + + + + + + + + + +

+ +

H

1 1

+

P-fim (F-type)

Type I fim FIA FIC

14 14

+ + + + + +

14 14

12 12

+ 7 7

+

12

+ + + +

5 5 5 5 5 5 5 5 5 5

+ + + +

12 12

12

+

Hemolysin

8 8 8 8 8 8

+ + + + + + + + + + + +

+ + + + + + + + + +

+ + + + + + + + + + + +

+ +

* Wild type strains. The K5-negative strains were selected using the bacteriophage OK5 (26). ** Mutants derived from E. coli 2980 selected following NTG mutagenesis (27).

Table 2. Escherichia coli 06: K15 strains used in the experiments with the endocytosis inhibitors Strain

Plasmid

Fimbrial antigen

536-21 536-21 536-21 536-21 536-21

non pANN801-4 pRHU845 pGB30 pPilll0-54

non Sfa+ Pap+ Msh+ FIC+

Adherence assay. For determination of adherent bacteria, the bacteria were incubated with the cells in the COrincubator for 1 h at 37°C. After washing the cells five times with medium M 199, distilled water at 4°C was added for lysis of the cell monolayer. The lysate was suspended and diluted 1: 10 in distilled water. 0.1 ml of the suspension was plated on an agar plate for colony counting.

Adhesion and Internalization of E. coli Strains

221

Internalization assay: The bacteria were incubated with LLC PK1 cells for 1 h at 37°C in a CO 2 incubator in medium M 199 lacking phenol red. Nonadherent bacteria were removed by washing with medium M 199. The extracellular and adherent bacteria were killed by the use of gentamicin (final concentration 100 mg/l medium M 199) for 1 h at room temperature. After washing, the cells were lysed by adding distilled water at 4°C for 15 min. The suspension was mixed and plated on an agar plate for counting the number of intracellular bacteria. Endocytosis inhibition experiments. The tubuloepithelial cells prepared as above were treated for 45 min with medium M 199 free of antibiotics, and containing cytochalasin B (5 mg!l), or cytochalasin D (5 mg/l), or colchicine (5 mg!l), or chloroquine (200 mg/l). After washing the LLC-PK 1 cells, bacteria (10 6 cfu/ml) were added. The number of intracellular bacteria was counted as described in the internalization assay. Statistics. The tests for adherence and internalization were made at least ten times per strain. The statistical evaluation was done by the use of the parameter-free U test (23).

Results Influence of fimbriae expressed on adhesion and internalization The numbers of bacteria per cell adherent to LLC-PK 1 cells are shown in Table 3. Strains expressing F14, F1A and Fl C fimbriae showed the strongest adhesion, whereas strains expressing only F12 fimbriae showed a low degree of adhesion to the LLC-Pk 1 cells.

Influence of the capsule and hemolysin on adhesion and internalization Strains expressing K5 antigen showed a remarkably reduced adhesion in comparison to the capsule negative isolates. These strains were internalized by LLC-PK 1 cells as Table 3. Adhesion and internalization of Escherichia coli 018 strains differing in the types of fimbriae expressed and in the production of K5 antigen Strain No.

P-Fim.

K5

Adhesion (cfu±SD)

Internalization (cfu±SD)

B2052" B2332

14 14

+

2471 ±312 5621 ± 604

123± 32 361± 62

B2062* B2063

14 14

+

2733 ±444 3603 ±598

94± 39 289 ± 123

20565* B2020

12 12

+

2376±657 2667±960

73± 19 124± 25

20887* B2021

12 12

+

1470±167 1400 ± 263

54± 24 87± 22

20383* B2037

12 12

+

1149±435 1310±344

50± 21 67± 15

2980* 20485

8 8

+

1834±448 2052±487

68± 24 152± 21

* Wild type strains. The K5-negative strains were selected using the bacteriophage OKS (26).

222

E. Straube, G. Schmidt, R. Marre, and J. Hacker.

Table 4. Adhesion and internalization of Escherichia coli 018 strains differing in the expression of K5 antigen, F8 fimbriae, and hemolysin Strain No.

P-Fim. (F-type)

K5

Hly

Adhesion (Cfu±SD)

Internalization (cfu±SD)

2980 20485* 20469* 20470* B2313* B2314* 20397* B2789* B2791* B2792 *

8 8 8 8 8 8

+

+ +

16921 ± 4471 22100± 7953 15792± 3732 15742± 2282 60510± 19433 42140± 12328 4950± 1916 16000± 816 6050± 3514 15700± 4191

811± 423 1304± 524 549± 338 707± 343 1024± 370 1272± 653 992± 326 1202± 118 405± 97 2355± 1786

+ + + +

+ + +

* Mutants derived from E. coli 2980 (27).

given in Tables 3 and 4. The number of internalized bacteria corresponded to the number of adherent bacteria. As shown in Table 4, hemolysin contributed to the internalization only to a small extent (statistically not significant). .

Effect of endocytosis inhibitors on adhesion and internalization To resolve the question whether the bacteria invade the cells or the cells take them up by endocytosis, we investigated the influence of different endocytosis inhibitors on adherence and internalization. The results are shown in Figs. 1 A-D (adherence) and Fig. 2A-D (internalization). The cytoskeletal inhibitors, cytochalasine B, cytochalasine D, colchicine, and chloroquine led to a statistically significant inhibition of the internalization of the bacteria tested into the LLC-PK1 cells, whereas the adherence remained nearly unaffected. The most effective one of the substances tested was chloroquine.

Discussion The adhesion of Escherichia coli to host cells is an early step of urinary tract infection (1, 13, 18,21). Bacterial adherence to epithelial cells is influenced mainly by the type of adhesin or fimbriae, the amount of fimbriae, and the number of fimbrial types expressed (19,24). As shown in Table 3, the strains expressing F14, FIA, and FIC were the most active ones in adhesion and internalization. It seems that the expression of FlA or Fl C fimbriae has only a marginal effect on the adhesion to LLCPK 1 cells, when strains 20887 and 20383 are compared. The lack of expression of P fimbriae reduces the adherence to the LCC-PK1 cells significantly (Table 4). The difference of the adh~sion rate in strain 2980 (Table 3/Table 4) was due to different series of the adhesion test. The most impressing result was the fact that strains expressing K5 antigen were less adherent to LLC-PK1 cells at a significant level as compared to K+ isolates. These strains were internalized by LCC-PK1 cells significantly less, indepen-

Adhesion and Internalization of E. coli Strains

223

(non)

(Sfa+)

(Pap+)

(Msh+)

I ffi2l without. with

(F1C+)

o

10

20

30

40

50

60

thousand (cfu)

Fig. lA. Influence of cytochala&in B on the adherence of E. coli 536-21 containing different adhesin plasmids ot LLC-PKl cells.

(non)

(Msh+)

I ~ without •

(F1C+)

o

20

40

60

80

100

with

I

120

thousand (cfu)

Fig. lB. Influence of cytochalasin D on the adherence of E. coli 536-21 containing different adhesin plasmids to LLC-PKl cells. 15 Zbl. Bakt. 278/2-3

224

E. Straube, G. Schmidt, R. Marre, and ]. Hacker (non)

o

40

20

60

80

100

120

thousand (cfu)

Fig. 1 C. Influence of colchicine on the adherence of E. coli 536-21 containing different adhesin plasmids to LLC-PK1 cells.

(non)

(Sfa+ )

(Pap+) >""

I~

(F1C+)

o

20

40

without

60

.with

I 80

thousand (cfu)

Fig. 1D. Influence of chloroquine on the adherence of E. coli 536-21 containing different adhesin plasmids to LLC-PK1 cells.

Adhesion and Internalization of E. coli Strains

225

without • with

o

2

3

4

5

thousand (cfu)

Fig. 2A. Influence of cytochalasin B on the internalization of E. coli 536-21 containing different adhesin plasmids into LLC-PKI cells.

(non)

(Sfa+)

(Pap+) .

I Cd without •

(F1C+)

o

2

4

6

8

with I 10

thousand (cfu)

Fig. 2B. Influence of cytochalasin D on the internalization of E. coli 536-21 containing different adhesin plasmids into LLC-PKI cells.

226

E. Straube, G. Schmidt, R. Marre, and J. Hacker (non)

(Sfa+) without (Pap+)

•

with

(Msh+)

(F1 C+)

o

2

3

4

5

6

thousand (cfu)

Fig. 2e. Influence of colchicin on the internalization of E. coli 536-21 containing different adhesin plasmids into LLC-PK1 cells.

Ilillilllwithout

o

2

3

.with I

456

thousand (cfu)

Fig. 2D. Influence of chloroquine on the internalization of E. coli 536-21 containing different adhesin plasmids into LLC-PK1 cells.

Adhesion and Internalization of E. coli Strains

227

dent of the fimbriae expressed. This confirms results published previously (24). These results were the reason for the thesis that the E. coli strains reported were internalized by an mechanism of endocytosis in dependency on the degree of adhesion to the LLCPK 1 cells. To confirm this thesis, we studied the influence of cytoskeletal inhibitors on bacterial adherence and internalization using E. coli strains listed in Table 2. It has been shown that cytoskeletal inhibitors impair the penetration of varions EIEC (5), EPEC strains (4, 20), and other bacteria (11) into HEp-2 cells. Cytochalasin D disturbes the formation of micro filaments whereas cytochalasin B affects the function of microfilaments (2). Colchicin is thought to affect the microtubule function and is able to influence the penetration of EPEC into HEp-2 cells (4, 6). Chloroquine is well known as an inhibitor of phagocytic activity of macrophages and fibroblasts and as an inhibitor of toxin internalization (12, 17). As shown in the Figs. 1A-D, none of the inhibitors of endocytosis (cytochalasine B, cytochalasine D, colchicine, and chloroquine) did affect the adherence of the bacteria to the cells. In contrast, there was a significant inhibition of internalization of the five E. coli strains tested, as shown in Figs.2A-D. This supports the thesis that even tubuloepithelial cells internalize E. coli cells by endocytosis. The precondition of the internalization is the interaction between the bacterial adhesin and a cell-derived receptor. This receptor-mediated endocytosis of bacteria by tubuloepithelial cells may be of importance in the pathogenesis of recurrent urinary tract infection. Acknowledgements. This work was kindly supported by grants of the Deutsche Forschungsgemeinschaft Ma 864/3-2 and Str 383/2-1.

References 1. Abraham, S. N. and E. H. Beachey: Binding of bacteria to mucosal surfaces. Monogr. Allergy 24 (1988) 38--43 2. Cooper, J. A.: Effect of cytochalasin and phalloidin on actin. J. Cell BioI. 104 (1987) 1473-1478 3. de Man, P.: Bacterial adherence, inflammation and renal scarring in urinary tract infection. Thesis, Goteborg (1990) 4. Donnenberg, M. S., A. Donohue-Rolfe, and G. T. Keusch: A comparison of HEp-2 cell invasion by enteropathogenic and enteroinvasive Escherichia coli. FEMS Microbiol. Lett. 69 (1990) 83-86 5. Finlay, B. B. and S. Falkow: Common themes in microbial pathogenicity. Microbiol. Rev. 53 (1989) 210-230 6. Geme III, J. w. S. and S. Falkow: Haemophilus influenzae adheres to and enters cultured human epithelial cells. Infect. Immun. (1990) 4036--4044 7. Hacker, J., L. Bender, M. Ott, J. Wingender, B. Lund, R. Marre, and W. Goebel: Deletions of chromosomal regions coding for fimbriae and hemolysins occur in vitro and in vivo in various extra intestinal Escherichia coli isolates. Microbial Pathogen. 8 (1990) 213-225 8. Hacker, J.: Genetic determinants coding for fimbriae and adhesins of extraintestinal Escherichia coli. Curr. Top. Microbiol. 151 (1990) 1-27 9. Hull, R. N., W. R. Cherry, and G. W. Weaver: The origin and characteristics of a pig kidney cell strain, LLC-PK 1 . In Vitro 12 (1976) 67Q-677 10. Hultgren, S. J., S. Normak, and S. N. Abraham: Chaperone-Assisted assembly and molecular architecture of adhesive pili. Annu. Rev. Microbiol. 45 (1991) 383--415

228

E. Straube, G. Schmidt, R. Marre, and J. Hacker

11. Janda, J. M., S. L. Abbou, and L. S. Oshiro: Penetration and replication of Edwardsiella spp. inHEp-2 cells. Infect. Immun. 59 (1991) 154-161 12. Janicot, M., F. Fouque, and B. Dessbuquois: Activation of rat liver adenylate cyclase by cholera toxin requires toxin interalization and processing in endosomes. J. BioI. Chern. 266 (1991) 1258-1265 13. Johnson, J. R.: Virulence factors in Escherichia coli urinary tract. infection. Clin. MicrobioI. Rev. 4 (1991) 80-128 14. Kreft, B., O.Carstensen, E. Straube, S. Bohnet, J. Hacker, and R. Marre: Adherence and cytotoxicity of Escherichia coli to eucariotic cell lines quantified by MIT (3-[4, 5-dimethylthiazol-2-yl 1-2, 5-diphenyl-tetrazolium bromide). Zbl. Bakt. 276 (1991) 231242 15. Kreft, B., N. Sloot, U. Schramm, K. Sack, J. Hacker, and R. Marre: Untersuchungen zur Interaktion von uropathogenen Escherichia coli mit renalen Tubuluszellen. Focus MUL 8 (1991) 82-89 16. Korhonen, T. K., R. Virkola, B. Westerlund, A.-M. Tarkkanen, K. Liitheenmiiki, T. Sareneva, J. Parkkinen, P. Kuusela, and H. Holthofe: Tissue interactions of Escherichia coli adhesins. Antonie v. Leeuwenhoek 54 (1988) 411-420 17. Lee, P. K., G. M. Vercellotti, J. R. Deringer, and P. M. Schlievert: Effects of staphylococcal toxic shock syndrome toxin 1 on aortic endothelial cells. J. Infect. Dis. 164 (1991) 711-719 18. Lomberg, H.: Host-parasite interaction in recurrent urinary tract infection. Thesis, G6teborg (1986) 19. Marre, R., B. Kreft, and J. Hacker: Genetically engineered Sand FIC fimbria differ in their contribution to adherence of Escherichia coli to cultured renal tubular cells. Infect. Immun. 58 (1990) 3434-3437 20. Miliotis, M. D., H. J. Koornhof, and J. I. Phillips: Invasive potential of noncytotoxic enteropathogenic Escherichia coli in an in vitro Henle 407 cell modell. Infect. Immun. 57 (1990) 1928-1935 21. Orndorff, P. E. and C. A. Bloch: The role of type 1 pili in the pathogenesis of Escherichia coli infections: A short review and some new ideas. Microbial Pathogen. 9 (1990) 75-79 22. 0rskov, 1. and F. 0rskov: Serology of Escherichia coli fimbriae. Progr. Allergy 33 (1983) 80-105 23. Sachs, L.: Angewandte Statistik. Springer-Verlag, Berlin-Heidelberg-New York (1984) 24. Straube, E., M. Kretschmar, G. Schmidt, R. Marre, B. Kreft, and G. Zingler: Is Escherichia coli invading tubuloepithelial cells? Zbl. Bakt. (1992) in press 25. Svanborg-Eden, c., S. Hausson, U. Jodal, G. Lidin-Janson, K. Lincoln, H. Linder, H. Lomberg, P. deMan, S. Marild, J. Martinelli, K. Plos, T. Sandberg, and K. Stenqvist: Host-Parasite interaction in the urinary tract. J. Infect. Dis. 157 (1988) 421-426 26. Ulmer, E., J. Hacker, E. Fasske, and G. Schmidt: Isyolation and characterization of coliphage 18A specific for Escherichia coli 018ac strains. Zbl. Bakt. Hyg. A 266 (1987) 403-411 27. Wevers, P., R. Picken, G. Schmidt, B. Jann, K. Jann, R. Golecki, and M. Kist: Characterization of pili associated with Escherichia coli 018ac. Infect. Immun. 29 (1980) 685-691

Professor Dr. Eberhard Straube, Institut fur Med. Mikrobiologie der Friedrich-SchillerUniversitat, Semmelweisstr. 4, D-O 6900 Jena.