Proteolytic cleavage of pertussis toxin S1 subunit is not essential for its activity in mammalian cells

Abstract

Background: Pertussis toxin (PT) is an exotoxin virulence factor produced by Bordetella pertussis, the causative agent of whooping cough. PT consists of an active subunit (S1) that ADP-ribosylates the alpha subunit of several mammalian G proteins, and a B oligomer (S2-S5) that binds glycoconjugate receptors on cells. PT appears to enter cells by endocytosis, and retrograde transport through the Golgi apparatus may be important for its cytotoxicity. A previous study demonstrated that proteolytic processing of S1 occurs after PT enters mammalian cells. We sought to determine whether this proteolytic processing of S1 is necessary for PT cytotoxicity.

Results: Protease inhibitor studies suggested that S1 processing may involve a metalloprotease, and processing does not involve furin, a mammalian cell protease that cleaves several other bacterial toxins. However, inhibitor studies showed a general lack of correlation of S1 processing with PT cellular activity. A combination of replacement, insertion and deletion mutations in the C-terminal region of S1, as well as mass spectrometry data, suggested that the cleavage site is located around residue 203-204, but that cleavage is not strongly sequence-dependent. Processing of S1 was abolished by each of 3 overlapping 8 residue deletions just downstream of the putative cleavage site, but not by smaller deletions in the same region. Processing of the various mutant forms of PT did not correlate with cellular activity of the toxin, nor with the ability of the bacteria producing them to infect the mouse respiratory tract. In addition, S1 processing was not detected in transfected cells expressing S1, even though S1 was fully active in these cells.

Conclusions: S1 processing is not essential for the cellular activity of PT. This distinguishes it from the processing of various other bacterial toxins, which has been shown to be important for their cytotoxicity. S1 processing may be mediated primarily by a metalloprotease, but the cleavage site on S1 is not sequence-dependent and processing appears to depend on the general topology of the protein in that region, indicating that multiple proteases may contribute to this cleavage.

Figure 1

S1 processing and detergent fractionation in CHO cells. (A) Cells were incubated with PT (20 nM) for 4 h followed by lysis with RIPA or Triton X-100 (TX100) buffer on ice or at room temperature (RT). A western blot of the detergent-insoluble (I) and -soluble (S) fractions is shown, with bands corresponding to full length S1 or processed S1 (S1p) indicated. Lane marked C is PT (200 ng) loading control. (B) Western blot of Triton X-100 lysate fractions of CHO cells treated with PT or PT* (catalytically inactive mutant of PT) for 4 h. (C) Western blot of Triton X-100 lysate fractions of untreated CHO cells after addition of 500 ng PT to the lysis mixture.

Figure 2

Effect of inhibitors on S1 processing and PT activity in CHO cells. (A, B) CHO cells were preincubated with the indicated inhibitor before addition of PT (20 nM) and then analyzed for S1 processing as before. Western blots are shown with unprocessed and processed (S1p) forms of S1 indicated. Lanes marked C are PT (100–200 ng) loading controls. DMSO = dimethylsulfoxide (solvent control). Metalloprotease inhibitor concentrations are indicated (mM). 0 = no inhibitor. BafA1 = bafilomycin A₁. (C) Effect of inhibitors on ADP-ribosylation of CHO cell G proteins by PT. Autoradiogram of cellular ADP-ribosylation assay after treatment of CHO cells with the indicated inhibitor and either PT or PT* (last lane). Radiolabeled band (approximately 41 kDa) corresponds presumably with G_iα2 and G_iα3 [9].

Figure 3

Diagram of the S1 protein and the putative cleavage site. The region of investigation between A200 and E221 is highlighted and details of some of the mutant constructs are shown. The line with a Δ symbol indicates the deleted amino acids.

Figure 4

Localization of the processing site on S1. (A) CHO cells were incubated with PT or PT*-CSP/N (PT*/N) for 4 h and analyzed for S1 processing as before. Western blots are shown of the processing samples (proc) or purified protein loading controls (C). (B) Western blot comparing S1p after processing of PT in CHO cells (proc) or trypsin cleavage (tryp) of PT (200 or 100 ng).

Figure 5

Processing of mutant PT proteins in CHO cells. (A) Western blot showing purified protein (lanes marked [C]) or processing of S1 (lanes marked [proc]) after addition of various mutant PT proteins (20 nM) to CHO cells. (B) Western blot showing processing of S1 (lanes marked [proc]) after addition of the 210–218/R replacement mutant PT proteins or PT and PT* controls to CHO cells.

Figure 6

ADP-ribosylation assay of mutant PT constructs. (A) Autoradiogram of cellular ADP-ribosylation assay after treatment of CHO cells with the indicated PT protein. The upper radiolabeled band (approximately 41 kDa) corresponds presumably with G_iα2 and G_iα3, while the lower band is PT-independent endogenously labeled band often seen in this assay with CHO cell lysates [9]. (B) Graphed data showing the mean (n = 3) cellular activity of the 2 deletion constructs that are not processed (activity of wild type PT was set at 100%) and the in vitro ADP-ribosylation activity of the same proteins. (C) Autoradiogram of cellular ADP-ribosylation assay after treatment of CHO cells with the indicated PT protein, demonstrating the retention of activity of the replacement mutant PT constructs.

Figure 7

S1 processing in CHO cell transfectants expressing S1. (A) Western blot showing lack of processing of endogenous S1 in transfectants expressing S1 in the cytosol (S1-SP) or in the ER (S1+SP). (B) Western blot showing EDTA-inhibitable processing of S1 after exogenous addition of purified PT (20 nM) to these transfectants.