Report

Resolution of a Signal Transfer Region from a General Binding Domain in Gβ for Stimulation of Phospholipase C-β2

See allHide authors and affiliations

Science  26 Feb 1999:
Vol. 283, Issue 5406, pp. 1332-1335
DOI: 10.1126/science.283.5406.1332

Abstract

Signaling by guanine nucleotide–binding proteins (G proteins) involves sequential protein-protein interactions. G protein–βγ subunit (Gβγ) interactions with phospholipase C–β2 (PLC-β2) were studied to determine if all Gβ contacts are required for signaling. A peptide encoding Gβ amino acid residues 86 to 105 stimulated PLC-β2. Six residues (96 to 101) within this sequence could transfer signals and thus constitute a core signal transfer region. Another peptide, encoding Gβ amino acid residues 115 to 135, did not substantially stimulate PLC-β2 by itself but inhibited Gβγ stimulation, indicating that residues 115 to 135 constitute a general binding domain. Resolution of signal transfer regions from general binding domains indicates that all protein-protein contacts are not required for signal transfer and that it may be feasible to synthesize agonists and antagonists that regulate intracellular signal flow.

Transmembrane signaling in G protein–coupled systems occurs through protein-protein interactions. Agonist-occupied receptors interact with G proteins to promote nucleotide exchange and subunit dissociation. The Gα subunits and the Gβγ complex interact with and regulate effectors (1). The Gβγ complex regulates numerous effectors including K+ channels, adenylyl cyclase 2, PLC-β2, and Ca2+ channels. We designed experiments to test whether signal transfer through protein-protein interactions requires all of the contacts between protein partners for information flow.

Residues including amino acids 60 to 150 of Gβ have been implicated in effector interactions (2–5). Because a relatively large area of Gβ participates in effector interactions, we chose one effector, PLC-β2, and determined a minimal region of Gβ required for stimulation. We also determined whether there are regions of Gβ that take part in effector interactions but are not required for signal transfer. Two peptides (6) encoding amino acids 86 to 105 and 115 to 135 of Gβ can inhibit Gβγ regulation of AC1 and AC2 (3). We tested the effect of the Gβ 86-105 peptide on Gβγ stimulation of PLC-β2 (7). The Gβ 86-105 peptide stimulated PLC-β2 in both the absence and presence of Gβγ. The stimulation by maximal concentration of peptide was not additive with that by Gβγ (Fig. 1A). Substitution of Met for Asp at position 101 (M101N) renders this peptide inactive for interactions with AC2 and AC1 (3). The M101N peptide did not activate PLC-β2 (Fig. 1A), indicating that the 101 position could be important for interactions with PLC-β2. To determine if the stimulation resulted from direct interactions between the peptide and PLC-β2, we tested the binding of the Gβ 86-105 and Gβ 86-105 M101N peptides to PLC-β2 by fluorescent resonance energy transfer (FRET) (8). The Gβ 86-105 peptide bound to PLC-β2 with a dissociation constant of about 1 μM (Fig. 1B), whereas the M101N peptide had no measurable binding. The difference in the affinities for binding and activation is probably due to differences between the two assay systems. The activity measurements were done with PLC from cytosolic lysates, whereas the binding measurements were done with purified PLC. The binding experiment was done in the presence and absence of phospholipids with identical results, indicating that the binding of the peptide to PLC-β2 is independent of the binding of substrate.

Figure 1

(A) Effects of various concentrations of Gβ 86-105 peptide on PLC-β2 activity. (Top) Effects of Gβ 86-105 peptide on basal and Gβγ (100 nM)–stimulated PLC-β2 activity. (Bottom) Effects of Gβ 86-105 peptide and M101N Gβ 86-105 peptide on basal activity of PLC-β2. IP3, inositol trisphophate. (B) Binding of Gβ 86-105 peptide and M101N Gβ 86-105 peptide to PLC-β2. They axis is expressed as relative energy transfer, which was monitored by the loss of the Cascade Blue fluorescence by transfer to the nonfluorescent acceptor DABMI.

Complementary charge interactions are often key determinants for protein-protein interactions. The Gβ86-105 peptide contains two charged residues, Lys89 and Arg96, and one histidine, His91. We evaluated the importance of each of these residues for the Gβ86-105 peptide stimulation of PLC-β2. Substitution of either Lys89 or Arg96, but particularly Arg96, decreased the affinity of the peptide for PLC-β2 but did not affect maximal stimulation (Fig. 2A). The Lys89 → Ala (K89A) substitution has a similar effect when made in the Gβ subunit through site-directed mutagenesis (9). When all three residues were substituted, the peptide did not stimulate PLC-β2 (Fig. 2B, top panel) and did not affect Gβγ stimulation of PLC-β2 (Fig. 2B, bottom panel). These results indicate that charge interactions may be crucial for both interactions and signal transfer from Gβγ to PLC-β2. To test whether charged peptides might nonspecifically activate PLC-β2, we measured the effects on PLC-β2 of an unrelated peptide, FLLT, which encodes region 660 to 688 of adenylyl cyclase 6 but has the same overall charge (+2 at pH 6.8 to 6.9) as the Gβ 86-105 peptide. The FLLT peptide had no measurable stimulatory effect (Fig. 2C). These results demonstrate that the stimulatory effects of the Gβ 86-105 peptide on PLC-β2 are not solely due to the charge of the peptide. To ascertain that the Gβ 86-105 peptide stimulation of PLC-β2 was selective, we tested stimulation of PLC-Xβ, an isoform of PLC-β from Xenopus that is stimulated poorly by Gβγ subunits under our assay conditions. The Gβ 86-105 peptide stimulated PLC-β2 robustly, but it had relatively little ability to stimulate PLC-Xβ (Fig. 2D). Thus, the Gβ 86-105 peptide selectively stimulates an isoform of PLC-β that is regulated by Gβγ subunits.

Figure 2

Effects of various concentrations of Gβ 86-105 peptide and (A) Lys89 → Ala (K89A)–, His91 → Ala (H91A)–, and Arg96 → Ala (R96A)–substituted peptides on PLC-β2 activity and (B) peptide with K89A, H91A, and R96A substitutions on basal (top) and Gβγ (100 nM)-stimulated (bottom) PLC-β2 activity. (C) Effects of various concentrations of Gβ 86-105 peptide and FLLT peptide on PLC-β2 activity. (D) Effects of 100 nM Gβγ and various concentrations of Gβ 86-105 peptide on PLC-β2 and PLC-Xβ activity.

Mutants of Gβ in which Ser98 is mutated to Ala (S98A) stimulate PLC-β2 more extensively (5). We studied the effects of four types of substitutions at this position. When Ser98 was substituted with Ala (Fig. 3, top panel), there was about a twofold increase in the affinity for PLC-β2. This increase could be consistent with the site-directed mutagenesis experiment in Gβ (5). When both the serines were substituted with Arg (S97R and S98R), there was a fivefold increase in affinity of the peptide (Fig. 3, middle panel). In contrast, substitution with Asp (S97D and S98D) resulted in an inactive peptide, whereas substitution with Cys (S97C and S98C) resulted in reduced affinity (Fig. 3, bottom panel). These experiments indicated that the region around amino acids 96 to 101 was crucial for signal transfer. We tested several short peptides, including a 3–amino acid peptide encoding residues 96 to 98, a 6–amino acid peptide encoding residues 96 to 101, and a 13–amino acid peptide encoding residues 89 to 101. The 3–amino acid peptide did not stimulate PLC-β2, but the 6– and the 13–amino acid peptides did stimulate it (Fig. 4A). The Gβ 96-101 6–amino acid peptide had lower affinity than the Gβ 86-105 peptide (Fig. 4B). However, when the serines corresponding to positions 97 and 98 were substituted by Arg, Gβ 96-101 stimulated with an apparent activation constant (K act) of 30 μM (Fig. 4B, top panel), as compared with 5 to 10 μM K actfor the Gβ 86-105 peptide (Figs. 1 to 3). When the serines were substituted with Asp, Gβ 86-105 did not stimulate PLC-β2 (Fig. 4B, bottom panel). Substitution of either Ser with Arg increased affinity of stimulation, and substitution at position 97 resulted in stimulation of PLC with a small (less than twofold) but reproducible and higher affinity than the substitution at position 98 (Fig. 4C). The efficacy of the S97R-substituted peptide appears to be similar to that for the full-length Gβ 86-105 peptide (Fig. 4C, bottom panel), albeit with lower affinity.

Figure 3

Effects of various concentrations of Gβ 86-105 peptide and (top) S98A Gβ 86-105 peptide, (middle) S97R and S98R Gβ 86-105 peptides, and (bottom) S97D and S98D and S97C and S98C Gβ 86-105 peptides on PLC-β2 activity.

Figure 4

Effects of shorter peptides from Gβ 86-105 region on PLC-β2 activity. (A) Effects of 600 μM Gβ 96-98, Gβ 96-101, and Gβ 89-101 peptides on PLC-β2 activity. Values are given as mean ± SEM of three experiments. (B) Effects of various concentrations of Gβ 96-101 peptide and S97R and S98R (top) and S97D and S98D (bottom) Gβ 96-101 peptides on PLC-β2 basal activity. (C) Effects of various concentrations of Gβ 96-101 peptide and Gβ 96-101 S97R and S98R peptides (top) and of Gβ 86-105 peptide and Gβ 96-101 S97R and S98R peptides (bottom) on PLC-β2 basal activity.

Amino acids 96 to 101 of Gβ constitute a core signal transfer region (STR) for activation of PLC-β2. Other regions of Gβ that interact with PLC-β2 may contribute to the overall affinity of the interactions but not be involved in signal transfer. If this were the case, then a peptide encoding such a region should inhibit Gβγ stimulation of PLC-β2 but not stimulate PLC-β2 by itself. The Gβ 115-135 peptide inhibits Gβγ modulation of both AC2 and AC1 (3). This peptide evoked a small (∼20% over basal) stimulation of PLC-β2 at saturating concentrations. But, when added with Gβγ, the peptide substantially (∼80%) inhibited Gβγ stimulation of PLC-β2. When the conserved tyrosine at position 124 was substituted, the peptide was inactive (Fig. 5A). The Gβ 115-135 peptide inhibited PLC-β2 with an apparent K act of 5 μM (Fig. 5B). Thus, we conclude that the region of Gβ containing amino acids 115 to 135 constitutes a general binding domain (GBD) that is not required for signal transfer.

Figure 5

Effects of Gβ 115-135 peptide on PLC-β2 activity. (A) Effects of 30 μM Gβ 115-135 peptide and Tyr124 → Val (Y124V) Gβ 115-135 peptide on basal and Gβγ (100 nM)-stimulated PLC-β2 activity. Values are given as mean ± SEM of three experiments. (B) Effect of various concentrations of Gβ 115-135 peptide on Gβγ-stimulated PLC-β2 activity.

Our studies demonstrate that all of the contacts between two proteins are not required for signal transfer. In the case of Gβγ and PLC-β2, our data show that a relatively short stretch of six amino acids, 96 to 101, appears to be sufficient to transfer the signal, that is, activate the enzyme. Substitution of residues within the six–amino acid peptide produces a more potent peptide than the naturally occurring sequence. The naturally occurring residues in STRs may not be optimized for this particular set of interactions. Such suboptimal interactions may be one mechanism to achieve regulated reversibility. It should also be noted that the Gβ 86-105 peptide does not stimulate AC2 in the presence of Gαs(3), suggesting that there may be different STRs on Gβ for different effectors. Gβ 115-135 minimally stimulates PLC-β2 but is effective in inhibiting Gβγ stimulation of PLC-β2. This indicates that the 115 to 135 region of Gβ is part of a GBD that participates in interactions with PLC-β2.

What is the relevance of such a functional resolution between STRs and GBDs within the overall interactions area? From the perspective of protein engineering, it offers a built-in capability to regulate the affinity of interaction between the protein partners and thus make reversibility feasible. Peptide hormones have distinct address and message regions (10) that mediate binding interactions with receptors and activation of intracellular signaling pathways, respectively. This functional resolution of peptide hormones has been used for the design of peptidomimetic antagonists (11). Similarly, our resolution of an STR from GBD for interactions between intracellular proteins provides an approach to identifying molecular interactions relevant for the development of agonists and antagonists at intracellular protein interaction sites. The interactions between the STR peptide and PLC-β2 could form the basis for synthesis of agonists that mimic receptor-dependent activation of PLC-β2. In contrast, the interactions between the GBD peptide and PLC-β2 would form the basis for synthesis of antagonists that block receptor-dependent activation of PLC-β2. Signaling pathways are major targets for therapeutic agents. Up to now, agonists and antagonists have largely been focused on extracellular receptor sites; our studies indicate that it may be feasible to design agonists and antagonists directed at the interface between signaling components inside the cell.

  • * To whom correspondence should be addressed. E-mail: iyengar{at}msvax.mssm.edu

REFERENCES AND NOTES

View Abstract

Stay Connected to Science

Navigate This Article