Table 2.

Effect of AIB2 substitution in GLP-1, glucagon, or OXM on mini-Gs and β-arrestin-2 recruitment responses. Mean ± SEM parameter estimates from 3-parameter fitting of AUC data from Figure 1A and association rate constants at maximal agonist stimulation (K@[max]). Statistical comparisons performed by paired t-tests (GLP-1 and GLP-1-AIB2) or randomised block one-way ANOVA with Dunnett's test (glucagon analogues). Note that, in general, if > 1 ligand was a full agonist, Emax values were compared after normalisation to the globally fitted maximum response, whereas if only one ligand was a full agonist, statistical comparison was performed prior to normalisation, but the numerical results are presented after normalisation to the full agonist response. See Supplementary Fig. 1 for further analysis of β-arrestin-2 recruitment using a different system. ∗p < 0.05 indicated by the statistical test.

GLP-1R
Mini-Gs				β-arrestin-2
	pEC50 (M)	Emax (% max)	K@[max] (min-1)	pEC50 (M)	Emax (% max)	K@[max] (min-1)
GLP-1	7.7 ± 0.1	105 ± 2	0.30 ± 0.04	7.3 ± 0.2	100 ± 0	1.25 ± 0.31
GLP-1-AIB2	7.9 ± 0.0	96 ± 2	0.21 ± 0.01∗	7.7 ± 0.1∗	68 ± 4∗	0.82 ± 0.10
GCGR
Mini-Gs				β-arrestin-2
	pEC50 (M)	Emax (% max)	K@[max] (min-1)	pEC50 (M)	Emax (% max)	K@[max] (min-1)
GCG	6.7 ± 0.0	100 ± 0	0.17 ± 0.03	6.1 ± 0.1	100 ± 0	0.89 ± 0.18
GCG-AIB2	6.7 ± 0.1	54 ± 4∗	0.11 ± 0.01	6.2 ± 0.1	25 ± 2∗	0.29 ± 0.03∗
GCG-H3	6.8 ± 0.1	66 ± 5∗	0.13 ± 0.04	6.3 ± 0.1	54 ± 6∗	0.51 ± 0.06∗
GCG-AIB2H3	7.0 ± 0.1∗	74 ± 5∗	0.13 ± 0.01	6.6 ± 0.1∗	53 ± 2∗	0.56 ± 0.06∗