Rho GTPase is activated by cytotoxic necrotizing factor 1 in peripheral blood T lymphocytes: potential cytotoxicity for intestinal epithelial cells

Abstract

Some strains of Escherichia coli related to acute cystitis or colitis produce a toxin named cytotoxic necrotizing factor 1 (CNF-1). CNF-1 mediates its effects on epithelial cells or phagocytes via the permanent activation of small GTP-binding proteins, caused by the toxin-induced deamidation of Glu(63) of p21 Rho. The behavior of peripheral blood T lymphocytes during the acute phase of bacterial colitis has been poorly investigated. Our study was conducted to test whether (i) peripheral blood T lymphocytes can be activated by CNF-1 and (ii) CNF-1-activated T lymphocytes are cytotoxic against intestinal epithelial cells. Activation of T lymphocytes by CNF-1 was assessed by electrophoresis, flow cytometry, confocal microscopy, and electron microscopy studies. Assays for migration and adherence of CNF-1-treated T lymphocytes were performed in Transwell chambers with T84 intestinal epithelial cells grown on polycarbonate semipermeable filters. CNF-1 induced a decrease in the electrophoretic mobility of the GTP-binding protein Rho in treated T lymphocytes. CNF-1 provoked an increase in the content of actin stress fibers and pseudopodia in T lymphocytes. Several adherence molecules were clustered into cytoplasmic projections in CNF-1-treated T lymphocytes and adherence of such lymphocytes on the basolateral pole of T84 was increased, resulting in cytotoxicity toward epithelial cells. Such enhanced adherence in response to CNF-1 was dependent on p42-44(MAP) kinase activation of T lymphocytes. Taken together, these results suggest that CNF-1, by acting on T lymphocytes, may increase in an important fashion the virulence of certain strains of E. coli against the intestinal epithelia.

FIG. 1.

Mobility shift of Rho in lymphoid cells upon treatment with CNF-1. (A) Jurkat T cells (10⁷ cells/assay) treated or not treated for 4 or 24 h with different concentrations of CNF-1 (1, 10, 100, and 300 pM and 1, 1.5, and 3 nM) were lysed. Rho proteins were [³²P]ADP ribosylated for 90 min with C3 exoenzyme. Deamidation of Rho was visualized by SDS-12% PAGE, followed by blotting and autoradiography as described in Materials and Methods. (B) Densitometric scanning of the deamidated Rho protein. (C) Upshift of Rho in T lymphocytes and Jurkat and HEp-2 cells treated or not treated for 4 h with 3 nM of CNF-1 was analyzed by [³²P]ADP ribosylation.

FIG. 2.

Morphological modifications and F-actin reorganization induced by CNF-1 (10⁻⁹ M, 24 h) in lymphoid cells. Transmission electron microscopic photographs of control (A) and CNF-1-treated (B) Jurkat T cells and of control (C) and CNF-1-treated (D) peripheral blood T lymphocytes (magnification, ×2,500). F-actin distribution stained with rhodamine-phalloidin (500 nM) in control (E) and CNF-1-treated (F) peripheral blood T lymphocytes were observed by confocal microscopy (magnification, ×2,000). Determination of lymphocyte F-actin by flow cytometry incubated with CNF-1 (H) exhibited a shift in the fluorescence peak, indicating polymerization of actin compared to control cells (G). One of five experiments is shown (each condition performed in triplicate). P < 0.05.

FIG. 3.

Effect of CNF-1 on CD29 and CD11a expression. CNF-1 did not cause an increase in CD11a expression (B) or in CD29 expression (D) as assessed by flow cytometry, in comparison with control peripheral blood T lymphocytes (A and C). For CD11a and CD29 staining, numerous beads were regrouped on filopodia in CNF-1-treated cells (anti-CD11a MAb [F]; anti-CD29 MAb [H]) (arrows). No beads were observed in nonfilopodial plasma membrane (arrowheads). Beads were evenly distributed along the plasma membrane in control cells (anti-CD11a MAb [E]; anti-CD29 MAb [G]) (electron microscopy magnification, ×4,500).

FIG. 4.

(A) Effect of CNF-1 on T lymphocytes migration across acellular filters. T lymphocytes present in the lower reservoirs were determined after 6 h of transmigration with or without SDF-1α as described in Materials and Methods. Data are pooled from 6 to 12 individual monolayers for each condition and results are means + SE (error bars) of five experiments (*, P < 0.05). Also shown is the effect of CNF-1 on Jurkat T cells (B) or T-lymphocyte (C) adherence to T84 monolayers. Lymphoid cells associated with monolayers were counted after 6 h of transmigration (basolateral-to-apical direction) in the absence or the presence of SDF-1α as described in Materials and Methods; data are pooled from 6 to 12 individual monolayers for each condition, and results are means + SE (error bars) of five different experiments (*, P < 0.05).

FIG. 5.

Electron micrographs showing CNF-1-induced-epithelial damages in T84 monolayers cocultured for 24 h with CNF1-treated T lymphocytes. CNF1-treated lymphocytes provoked in T84 epithelial cells intracytoplasmic vacuolization (A) (arrows) (magnification, ×2,500), and loss of microvilli (B) (magnification, ×1,800 [magnification for inset, ×2,800]). Contact (24 h) with control T lymphocyte did not affect the morphological features of the T84 monolayers (C) (magnification, ×1,800) and the brush border (inset [magnification, ×2,800]) Abbreviations: Ap, apical side; Bl, basolateral side; LT, lymphocyte.

FIG. 6.

Time course of p42-44^MAPk and Jun kinase activation in T lymphocytes after incubation with CNF-1. CNF-1 effect was compared to the effect of PMA. Lysates from untreated (lane 1) or CNF-1-treated (1 nM) cells treated for different times or PMA (5 ng/ml)-treated cells were prepared and blotted with phosphospecific MAPK antibodies. Blots were stripped and reprobed for total ERK2 and JNK1.

FIG. 7.

Effect of the MEK inhibitor PD 98059 on the adherence of CNF-1-treated T lymphocytes to T84 monolayers. T lymphocytes adhering to the T84 monolayers were counted as described in the Material and Methods after 6 h of transmigration (basolateral-to-apical direction) in the absence or the presence of SDF-1α. In some experiments, T lymphocytes were preincubated with the MEK inhibitor PD 98059 and then treated with CNF-1 before transmigration. Data are pooled from 6 to 12 individual monolayers for each condition, and results are means + SE (error bars) of five experiments (*, P < 0.05).

FIG. 8.

Expression of IL-2, IL-4, IL-8, IFN-γ, TNF-α, TGF-β1 TGF-β2, and TGF-β3 mRNA in CNF-1-treated peripheral T lymphocytes (24 h, 1 nM). RNA expression was analyzed by RPA as described in Materials and Methods. *, P < 0.01.

FIG. 9.

(A) p42-44^MAPk and JNK activation in CNF-1-treated T lymphocytes are linked neither to TNF-α nor to TGF-β1 production. p42-44^MAPk and JNK activation in CNF-1-treated T lymphocytes was analyzed by Western blotting in cells preincubated with anti-TNF-α (lane 3) or anti-TGF-β1 (lane 4) MAbs. T lymphocytes were preincubated for 1 h at 37°C with either anti-TNF-α or anti-TGF-β1 antibodies before addition of 1 nM CNF-1 for 3 h. Lane 1, untreated cells; lane 2, CNF-1-treated cells (3 h, 1 nM); lane 5, PMA-treated cells. (B) Increased adherence of CNF-1-treated T lymphocytes to T84 monolayers is not modified in T lymphocytes pretreated with anti-TNF-α or with anti-TGF-β1 MAbs. T lymphocytes associated with monolayers were counted after 6 h of transmigration (basolateral-to-apical direction) induced by SDF-1α as described in Materials and Methods. Data are pooled from 6 to 12 individual monolayers for each condition, and results are means + SE (error bars) of five different experiments.