Activation of the Classical Mitogen-Activated Protein Kinases Is Part of the Shiga Toxin-Induced Ribotoxic Stress Response and May Contribute to Shiga Toxin-Induced Inflammation

Abstract

Infection with enterohemorrhagic Escherichia coli (EHEC) can result in severe disease, including hemorrhagic colitis and the hemolytic uremic syndrome. Shiga toxins (Stx) are the key EHEC virulence determinant contributing to severe disease. Despite inhibiting protein synthesis, Shiga toxins paradoxically induce the expression of proinflammatory cytokines from various cell types in vitro, including intestinal epithelial cells (IECs). This effect is mediated in large part by the ribotoxic stress response (RSR). The Shiga toxin-induced RSR is known to involve the activation of the stress-activated protein kinases (SAPKs) p38 and JNK. In some cell types, Stx also can induce the classical mitogen-activated protein kinases (MAPKs) or ERK1/2, but the mechanism(s) by which this activation occurs is unknown. In this study, we investigated the mechanism by which Stx activates ERK1/2s in IECs and the contribution of ERK1/2 activation to interleukin-8 (IL-8) expression. We demonstrate that Stx1 activates ERK1/2 in a biphasic manner: the first phase occurs in response to StxB1 subunit, while the second phase requires StxA1 subunit activity. We show that the A subunit-dependent ERK1/2 activation is mediated through ZAK-dependent signaling, and inhibition of ERK1/2 activation via the MEK1/2 inhibitors U0126 and PD98059 results in decreased Stx1-mediated IL-8 mRNA. Finally, we demonstrate that ERK1/2 are activated in vivo in the colon of Stx2-intoxicated infant rabbits, a model in which Stx2 induces a primarily neutrophilic inflammatory response. Together, our data support a role for ERK1/2 activation in the development of Stx-mediated intestinal inflammation.

FIG 1

Western blot for phosphorylated ERK1/2 in whole-cell lysates of HCT-8 cells treated with Stx1 for zero minutes (C₀) to 6 h. Lanes are labeled C (untreated), S₁ (Stx1), or HI (heat-inactivated Stx1). Ladder refers to the molecular mass ladder. Anti-total ERK1/2 (α-ERK1/2) were used for loading controls. The graph displays the relative phospho-ERK1/2 band intensity normalized to the average from all untreated controls (no treatment). The error bar represents the standard deviations for the untreated controls. Stx1 and heat-inactivated Stx1 were used at a final concentration of 1 μg/ml.

FIG 2

(A) Western blot of phosphorylated Elk-1 fusion protein as an indicator of ERK1/2 activity in immunoprecipitates from whole-cell lysates of HCT-8 cells treated from 5 min to 60 min with Stx1, StxB1 subunit only, heat-inactivated Stx1 (HI), or the ribotoxic stressor anisomycin (500 ng/ml). Briefly, phosphorylated-ERK1/2 was immunoprecipitated from 200 μg of whole-cell lysate. To each immunoprecipitation sample, buffer containing 0.5 μg Elk-1 fusion protein (i.e., ERK1/2 substrate) and 200 μM ATP was added, followed by incubation at 30°C for 30 min. (B) Western blot of phosphorylated ERK1/2 in whole-cell lysates from untreated samples (C) or samples treated with Stx1 (S) or StxA_Y77S,E167QB1 (M). Growth medium was replaced with with fresh medium containing 60% FBS (GF₃₀) as a positive control. Cells were treated with the MEK1/2 inhibitor U0126 (10 μM) and placed in fresh medium containing 60% fresh FBS as a negative control. Cells were incubated for 6 h posttreatment for samples C, S, and M and for 30 min posttreatment for samples GF₃₀ and U0126 plus GF₃₀. Graphs beneath blots show the relative band intensity for corresponding blot lanes. Stx1, mutant, and HI treatments were applied at a final protein concentration of 1 μg/ml for all experiments.

FIG 3

(A) Western blot for phosphorylated ERK1/2 in whole-cell lysates from HCT-8 cells treated with 300 μg/ml puromycin (P) or 10 ng/ml ricin (R) for 2, 4, and 6 h or Stx1 for 4 h. As a positive control, media were replaced with fresh media containing 60% FBS (GF₃₀). (B) Western blot for phosphorylated mTOR in whole-cell lysates from HCT-8 cells treated with 1 μg/ml Stx1 following a pretreatment with vehicle (dimethyl sulfoxide [DMSO]) or the MEK1/2 inhibitor U0126 (10 μM) or PD98059 (20 μM). (C) Western blot for Phospho-ERK1/2 in whole-cell HCT-8 lysates with no pretreatment, pretreatment for 1 h with rapamycin (10 nM), or pretreatment with the MEK1/2 inhibitor U0126 (10 nM) for 1 h followed by no treatment or treatment with Stx1 (1 μg/ml) for 6 h.

FIG 4

Western blot of whole-cell lysates of HCT-8 cells pretreated with the vehicle DMSO (−) or the specific ZAK inhibitor DHP-2 at 200 nM (+), followed by treatment with Stx1 1 μg/ml or ricin 100 ng/ml for 6 h (A) or Stx2 1 μg/ml for 5 h (B). The graph beneath each blot shows the relative band intensity for corresponding lanes.

FIG 5

(A) qRT-PCR for zak expression in HCT-8 cells that were not transduced or were transduced with the lentiviral vector pGipZ expressing shRNAs against gapdh or zak (zak 69). Expression was normalized to the expression of the gene for β-actin. Error bars represent the standard errors of the means (SEM). (B) Western blot of phosphorylated c-jun, which is indicative of JNK activation, in immunoprecipitates from HCT-8 cells transduced with pGipZ expressing shRNAs against gapdh or zak (zak 69) and following no treatment or treatment with 1 μg/ml Stx2. (C) Phospho-ERK1/2 in whole-cell lysates from HCT-8 cells transduced with lentiviral pGipZ vectors expressing shRNAs for gapdh or zak (zak 69) following no treatment or treatment with 1 μg/ml Stx2. The graph beneath each blot shows the relative band intensity for corresponding lanes.

FIG 6

(A) Northern blot of IL-8 mRNA from HCT-8 cells following no treatment or treatment with 1 μg/ml Stx1, StxA_Y77S,E167QB1, StxB1 subunit, or 1 μg/ml Stx1 following pretreatment with DMSO or the MEK1/2 inhibitor PD98059 (20 μM) or U0126 (10 μM). The graph beneath each blot shows the relative band intensity for corresponding lanes. (B) Immunohistochemistry for phosphorylated ERK1/2 in sections of distal colon from infant rabbits treated twice over 2 days with 1.0 mg/kg heat-inactivated Stx2, active-site mutant StxA_Y77S,E167QB2, or wild-type Stx2.

FIG 7

(A) Shiga toxins depurinate the 28S rRNA subunit, which results in inhibition of protein synthesis (61). (B) Following Stx-mediated 28S rRNA damage on actively translating ribosomes, one or more isoforms of ZAK are activated. (C) Stx-activated ZAK results in activation of MAP2Ks that subsequently phosphorylate and activate JNK and p38 (19). (D) Phosphorylation of the MAP2Ks MEK1/2 results in activation of ERK1/2. (E) Activation of MAPKs ERK1/2, JNK, and p38 contributes to increased proinflammatory cytokine production (6, 10–14, 19, 27).