Enterohemorrhagic E. coli (EHEC)-Secreted Serine Protease EspP Stimulates Electrogenic Ion Transport in Human Colonoid Monolayers

Abstract

One of the characteristic manifestations of Shiga-toxin-producing Escherichia coli (E. coli) infection in humans, including EHEC and Enteroaggregative E. coli O104:H4, is watery diarrhea. However, neither Shiga toxin nor numerous components of the type-3 secretion system have been found to independently elicit fluid secretion. We used the adult stem-cell-derived human colonoid monolayers (HCM) to test whether EHEC-secreted extracellular serine protease P (EspP), a member of the serine protease family broadly expressed by diarrheagenic E. coli can act as an enterotoxin. We applied the Ussing chamber/voltage clamp technique to determine whether EspP stimulates electrogenic ion transport indicated by a change in short-circuit current (Isc). EspP stimulates Isc in HCM. The EspP-stimulated Isc does not require protease activity, is not cystic fibrosis transmembrane conductance regulator (CFTR)-mediated, but is partially Ca²⁺-dependent. EspP neutralization with a specific antibody reduces its potency in stimulating Isc. Serine Protease A, secreted by Enteroaggregative E. coli, also stimulates Isc in HCM, but this current is CFTR-dependent. In conclusion, EspP stimulates colonic CFTR-independent active ion transport and may be involved in the pathophysiology of EHEC diarrhea. Serine protease toxins from E. coli pathogens appear to serve as enterotoxins, potentially significantly contributing to watery diarrhea.

Keywords: CFTR; EHEC; SPATEs; diarrhea; human colonoid monolayers; intracellular Ca2+; serine protease EspP; short circuit current.

Figure 1

EspP stimulates Isc in proximal colonic HCM and its action is independent of its serine protease activity. (A,B) Representative Isc traces show that apically added EspP increases active ion transport in (A) undifferentiated and (B) differentiated HCM. (C) Representative Isc trace shows that EspP S263A also stimulates active ion transport in HCM. (D) Quantitative analysis of the increase in Isc shows that both EspP and EspP S263A–stimulated currents are of similar magnitude regardless of HCM differentiation status. Blue arrows indicate time that EspP or EspP S263A were added. N is the number of monolayers studied.

Figure 2

Isc stimulated by EspP in proximal HCM is concentration dependent and EspP-specific. (A) EspP-stimulated Isc increases with each sequential 6 µg/mL apical addition of EspP (blue arrows). (B) Representative Isc traces show that apically added purified recombinant EspP (6 µg/mL; red traces) as well as lysate from AD202 bacteria containing EspP plasmid (30 µL, blue traces), but not lysate from AD202 parental strain lacking either EspP or EspP S263A plasmid (30 µL, black traces), rapidly increase Isc. (C) Representative Isc traces show that pre-absorbing EspP toxin with specific anti-EspP antibodies reduced the EspP-increase in Isc by ~70% (red traces); black trace—time control; blue trace—purified EspP. Blue arrows indicate the time point when toxins or lysates were added apically.

Figure 3

EspP-stimulated Isc in proximal HCM is CFTR-independent. Representative (A) trace of EspP-stimulated Isc shows that apical addition of CFTR inhibitor CFTR_inh-172 (25 µM) neither inhibits Isc nor prevents further Isc increase following second EspP addition (each EspP addition is 6 µg/mL). (B) Traces of Isc increase following the escalating apical addition of EspP S263A (each addition is 6 µg/mL) in the presence (red) and the absence of (blue) combination of CFTR inhibitors, CFTR_inh-172 (25 µM) and BPO-27 (5 µM). Time control (black) in the absence of toxin. (C) Forskolin (10 µM) rapidly generates Isc, which is completely inhibited by a combination of CFTR inhibitors, used in (B). Blue arrows in all panels indicate the time-point of corresponding apical treatment.

Figure 4

EspP S263A-stimulated Isc in proximal HCM is Ca²⁺-dependent. (A) Representative traces and (B) quantitative analysis of BAPTA-AM effect on EspP S263A-stimulated Isc. BAPTAM-AM (25 µM pretreatment for up to 30 min) significantly inhibits EspP S263A-stimulated Isc. (C) EspP S263A-stimulated Isc is not inhibited by CaCC_inh-A01. (D) Esp S263A-stimulated Isc is similar in the buffer containing Cl⁻ (black trace) and without Cl⁻ (red trace). The inset shows that ΔIsc is quantitatively similar in Cl⁻ and Cl⁻-free buffer; n is the number of monolayers studied.

Figure 5

Effect of other SPATEs on active ion transport in HCM and human enteroid monolayers (HEM). (A) SepA—stimulated ion secretion is CFTR-mediated and completely inhibited by CFTR_inh-172 (25 µM). Blue arrows show the time points of each SepA addition (10 µg/mL) as well as addition of CFTR_inh-172. Neither Pic (10 µg/mL) (B) nor EatA (10 µg/mL) (C) stimulated Isc in proximal UD HCM or jejunal HEM, respectively. Blue arrows indicate the time points of Pic or EatA addition.