OZITX, a pertussis toxin-like protein for occluding inhibitory G protein signalling including Gαz

Abstract

Heterotrimeric G proteins are the main signalling effectors for G protein-coupled receptors. Understanding the distinct functions of different G proteins is key to understanding how their signalling modulates physiological responses. Pertussis toxin, a bacterial AB₅ toxin, inhibits Gα_i/o G proteins and has proven useful for interrogating inhibitory G protein signalling. Pertussis toxin, however, does not inhibit one member of the inhibitory G protein family, Gα_z. The role of Gα_z signalling has been neglected largely due to a lack of inhibitors. Recently, the identification of another Pertussis-like AB₅ toxin was described. Here we show that this toxin, that we call OZITX, specifically inhibits Gα_i/o and Gα_z G proteins and that expression of the catalytic S1 subunit is sufficient for this inhibition. We identify mutations that render Gα subunits insensitive to the toxin that, in combination with the toxin, can be used to interrogate the signalling of each inhibitory Gα G protein.

Fig. 1. Identification of Gα carboxy-tail amino acid residues that are putatively ADP-ribosylated by OZITX.

a Amino acid sequence alignment of carboxy-termini residues of heterotrimeric Gα proteins. Sequences were aligned with Clustal Omega version 1.2.4. ‘*’ represents a completely conserved residue. ‘:’ represents a conserved residue (>0.5 in the Gonnet PAM 250 matrix). ‘.’ represents a weakly conserved residue (≤0.5 and >0 in the Gonnet PAM 250 matrix). Cysteine residues ADP ribosylated by PTX are indicated in red. Putative lysine and asparagine residues ADP ribosylated by OZITX identified by Littler and colleagues are indicated in yellow. The asparagine residue that is a putative substrate is conserved across many Gα subunits. b The location of OZTX’s and PTX’s substrate amino acid sites within a GPCR–G protein complex. The structure of rhodopsin bound to Gα_i1β₁γ₂ is depicted in the cartoon (PDB code 6CMO). Rhodopsin is shown in dark blue, Gα_i1 in light blue, Gβ₁ in green and Gγ₂ in light green. The carboxy-terminal Cys³⁵¹ residue ADP ribosylated by PTX is shown in red spheres. Lys³⁴⁵ and Asn³⁴⁷, the putative residues ADP ribosylated by OZITX, are highlighted in yellow spheres. Graphic constructed using UCSF chimera.

Fig. 2. Activation of members of the Gαi subfamily in the presence of OZITX and PTX.

a Representation of the BRET sensors used for detection of G-protein activation. The Gαβγ heterotrimer is activated through agonist binding to the GPCR and the Gα and Gβγ-venus dissociate. Free Gβγ-venus is bound by masGRK3ct-Nluc that serves as a BRET donor resulting in non-radiative energy transfer from Nluc to venus. b Dopamine D₂receptor (D₂R)-mediated activation or (c) μ opioid receptor (MOPR)-mediated activation of Gα_i subfamily members in the presence of OZITX or PTX. HEK293A CRISPR/Cas ΔGα-all cells expressing the D₂R or MOPR were pre-treated with either vehicle (black), OZITX (blue) or PTX (red) for 24 h. Cells were stimulated with 1 μM ropinirole (D₂R) or 1 μM DAMGO (MOPR) for 2.5 min followed by BRET detection. Data represent the mean drug-induced increase in BRET ratio from vehicle ± SEM from six independent experiments (D₂R) or three independent experiments (MOPR). *Represents where the response is significantly different (P < 0.001) from the respective vehicle toxin untreated control condition (black bar) as determined by a one-way ANOVA with Dunnett’s multiple comparisons test. d Time course of OZITX treatment on G-protein activation. HEK 293 ΔGα-all cells were transfected with cDNA encoding the D_2LR, Gα_i2 and G-protein activation sensors. Cells were pre-treated with OZITX for the indicated times. BRET was measured 2.5 min after stimulation with 1 µM ropinirole (blue open circles) or vehicle (blue filled circles). The basal BRET ratio prior to agonist stimulation has been subtracted to give the drug-induced ΔBRET ratio. Data represent the mean ± SD from three separate experiments. Individual replicates are shown as small circles.

Fig. 3. Gαs, Gαq and Gα12 subfamily activation in presence of OZITX and PTX.

a Activation of Gα_s subfamily members by the dopamine D₁ receptor (D₁R) in the presence of OZITX and PTX. b Activation of Gα_q subfamily members by NTS₁R in the presence of OZITX and PTX. c Activation of Gα₁₂ subfamily members by the neurotensin receptor 1 (NTS₁R) in the presence of OZITX and PTX. HEK 293 ΔGα-all CRISPR cells were transfected with cDNA encoding the particular GPCR and Gα together with the G-protein activation sensors as described in “Methods”. The cells were pre-treated with either vehicle (black), OZITX (blue) or PTX (red) for 24 h before stimulation with the agonists 100 nM SKF83822 (D₁R)/1 μM NT8-13 (NTS₁R) for 2.5 min followed by BRET detection. The data are represented as the mean drug-induced increase in BRET ratio from vehicle control ± SEM from three separate experiments. *Represents where the OZITX or PTX-treated condition is significantly different (P < 0.001) from the vehicle-treated condition (black) as determined by a one-way ANOVA with Dunnett’s multiple comparisons test.

Fig. 4. The effect of OZITX on Gαi2-, GαoA- and Gαz-mediated inhibition of cAMP production.

Inhibition of forskolin-stimulated cAMP production was detected in live cells using CAMYEL; a conformational BRET sensor based on EPAC. HEK 293 ΔGα_i/o CRISPR cells were transfected with DNA encoding the D₂R, CAMYEL and either (a) pcDNA3.1+ control, (b) Gα_i2, (c) Gα_oA or (d) Gα_z. Transfected cells were then incubated with either vehicle (black) or OZITX (blue) for 24 h. Cells were then pre-stimulated with 10 µM forskolin for 10 min before stimulation with either vehicle control (filled circles) or 1 μM ropinirole (open circles). Data are baseline-corrected to the cells not treated with OZITX or ropinirole and is shown as the mean ± SEM from four separate experiments. Measurements of cAMP inhibition between vehicle and OZITX-treated conditions were compared using an unpaired Student’s t test * represents statistical significance P < 0.05 (pcDNA –P= 0.700, Gα_i2 −P = 0.008, Gα_oA – P = 0.004_, Gα_z – 0.019).

Fig. 5. OZITX sensitivity of Gαi subfamily carboxy-tail Asn347/348 mutants.

a Gα_i3-WT activation, n = 11. b Gα_oA-WT activation, n = 4. c Gα_z-WT activation, n = 5. d Gα_i3-N347A activation, n = 4. e Gα_oA-N347A activation, n = 4. f Gα_z-N347A activation, n = 6. g Gα_i3-C351 activation, n = 4. h Gα_oA-C351 activation, n = 4. The G-protein activation assay was performed on WT, Asn347Ala/Asn348Ala (putative OZITX site) and Cys351Ile (PTX-insensitive) mutants. Cells were transfected with the D₂R, the particular Gα mutant, the G-protein activation sensors and either a pcDNA3.1+ control (black open circles), OZITX-S1 cDNA (blue open circles) or PTX-S1 cDNA (red open circles). Cells were then stimulated with increasing concentrations of quinpirole before BRET detection. Data represent the mean drug-induced increase in BRET ratio from vehicle ± SEM. Individual replicates are shown as small circles.