Endogenous agonist–bound S1PR3 structure reveals determinants of G protein–subtype bias

See allHide authors and affiliations

Science Advances  09 Jun 2021:
Vol. 7, no. 24, eabf5325
DOI: 10.1126/sciadv.abf5325
  • Fig. 1 Overall structure of d18:1 S1P–bound S1PR3.

    (A) Overall structure of the d18:1 S1P–S1PR3–Fab AS55 complex. d18:1 S1P is shown as a space-filling model, with carbon atoms in yellow, nitrogen in blue, oxygen in red, and phosphorus in orange. S1PR3 is colored dark salmon. Fab light chain is colored blue, and Fab heavy chain is colored orange. (B and C) Extracellular view of the d18:1 S1P–bound S1PR3 structure (B) (PDB ID: 7C4S) and antagonist (ML056)–bound S1PR1 structure (C) (PDB ID: 3V2Y). Each molecule is colored as follows: S1PR3, dark salmon; d18:1 S1P, yellow; S1PR1, dark gray; and ML056, light blue. TM, helical transmembrane domain; ECL, extracellular loop. (D to H) Cross section of the following structures: (D) d18:1 S1P–bound S1PR3, (E) antagonist (ML056)–bound S1PR1 structure (22), (F) AM11542 (agonist)–bound CNR1 (23), (G) WIN 55,212-2 (agonist)–bound CNR2 (12), and (H) misoprostol (agonist)–bound PTGER3 (25).

  • Fig. 2 d18:1 S1P binding mode.

    (A and B) Views of residues interacting with d18:1 S1P in the d18:1 S1P–S1PR3–Fab AS55 complex (A) with views rotated 90° counterclockwise (B). The ligand-binding residues in S1PR3 are shown as a licorice model in dark salmon. Hydrogen bonds are shown as black dashed lines. (C) Comparison of pEC50 values of d18:1 S1P in TGFα shedding assay with HEK293 parental cells expressing mutant and wild-type S1PR3. pEC50 values of mutant S1PR3 N95A, R114A, and E115A were estimated to be very low according to the dose-response curve in fig. S4. ****P < 0.0001, mutant versus wild-type pEC50 values according to one-way analysis of variance (ANOVA) and Dunnett’s post test. (D) Schematic of the interaction between d18:1 S1P and S1PR3. Hydrogen bonds and salt bridges are shown as black dashed lines. Critical residues for d18:1 S1P binding according to mutation analyses are shown in blue boxes.

  • Fig. 3 Short sphingolipids represent G12/13- and Gi/o-biased agonists for S1PR3.

    (A) Chemical structures of ligands used for signaling assays. (B) Normalized G protein signaling efficacy. Emax of each G protein subtype–specific signaling in response to d18:1 S1P was normalized to 100%. Each bar is colored as follows: Gq/11, yellow; G12/13, salmon; and Gi/o, light blue. ****P < 0.0001, d18:1 S1P versus S1P analogs % G protein signaling according to two-way ANOVA followed by Dunnett’s post test. (C) Schematic summary of biased agonism in S1PR3. Major S1P, d18:1 S1P, induces Gq/11-, G12/13-, and Gi/o-dependent signaling via S1PR3. On the other hand, S1P analogs with a short lipidic tail, such as d16:1 S1P and FTY720-P, induce G12/13- and Gi/o-biased signaling via S1PR3.

  • Fig. 4 S1PR-activation mechanism.

    (A) Superposition of the S1PR3 (dark salmon) and S1PR1 (3V2Y, dark gray) structures. Each molecule is colored as follows: S1PR3, dark salmon; S1PR1, light gray; d18:1 S1P, yellow; and ML056, light blue. (B) Comparison of residues comprising the quartet core in active S1PR3 and inactive S1PR1 structures. Extracellular view of S1PR3 superposed on S1PR1. ECL, extracellular loop. (C) Intracellular view of S1PR3 superposed on S1PR1. ICL, intracellular loop. (D) Comparison of maximal response (Emax) of d18:1 S1P in TGFα shedding assay with HEK293 parental cells expressing mutant and wild-type S1PR3. **P < 0.01, ****P < 0.0001, mutant versus wild-type (WT) Emax values according to one-way ANOVA and Dunnett’s post test. n.s., not significant.

  • Fig. 5 The quartet core and biased agonism.

    (A) WT and mutant S1PR3 were examined for d18:1 S1P–induced AP-TGFα shedding responses in the G12/13-knockout HEK293 cell line, reflecting Gq/11-mediated signaling. (B) WT and mutant S1PR3 were examined for d18:1 S1P–induced AP-TGFα shedding responses in the Gq/11-knockout HEK293 cell line, reflecting G12/13-mediated signaling. Symbols and error bars in (A) and (B) represent mean and SEM, respectively, of three to six independent experiments with each performed in triplicate. Each line represents WT and mutant S1PR3 and is colored as follows: WT, gray; Leu122Ala, orange; Phe204Ala, light blue; Trp256Ala, pink; and Phe260Ala, green. (C) Bias plot of mutant and WT S1PR3 of maximal Gq/11-dependent % AP-TGFα shedding versus maximal G12/13-dependent % AP-TGFα shedding. Symbols and error bars represent mean of maximal response and SEM in TGFα shedding assay with HEK293 ΔG12/13 or ΔGq/11 cell lines. Leu122Ala mutant shows G12/13-biased activity. Phe204Ala mutant shows similar response to WT in both assays. Phe260Ala and Trp256Ala mutants show lower activity in both assays than WT. (D) Schematic models of S1PR3 activation and G protein–subtype bias. S1PR3 cannot couple to any G proteins in the antagonist-bound state (left). S1PR3 couples only to G12/13 and Gi/o upon binding to an agonist with a shorter lipidic chain, such as d16:1 S1P or FTY720-P (middle). S1PR3 couples with Gq/11, Gi/o, and G12/13 upon binding to the major natural ligand 18:1 S1P (right).

Supplementary Materials

Stay Connected to Science Advances

Navigate This Article