Molecular determinants and signaling effects of PKA RIα phase separation

Affiliations

¹ Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA; Shu Chien-Gene Lay Department of Bioengineering, University of California, San Diego, La Jolla, CA 92093, USA.
² Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA; Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093, USA.
³ Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA.
⁴ Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, CA 92093, USA.
⁵ Department of Pharmacology, University of Iowa, Iowa City, IA 52242, USA.
⁶ Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093, USA; Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, CA 92093, USA.
⁷ Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA; Shu Chien-Gene Lay Department of Bioengineering, University of California, San Diego, La Jolla, CA 92093, USA; Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093, USA; Moores Cancer Center, University of California, San Diego, La Jolla, CA 92093, USA. Electronic address: jzhang32@health.ucsd.edu.

PMID: 38537638
PMCID: PMC11031308
DOI: 10.1016/j.molcel.2024.03.002

Molecular determinants and signaling effects of PKA RIα phase separation

Julia C Hardy et al. Mol Cell. 2024.

. 2024 Apr 18;84(8):1570-1584.e7.

doi: 10.1016/j.molcel.2024.03.002. Epub 2024 Mar 26.

Authors

Affiliations

¹ Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA; Shu Chien-Gene Lay Department of Bioengineering, University of California, San Diego, La Jolla, CA 92093, USA.
² Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA; Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093, USA.
³ Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA.
⁴ Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, CA 92093, USA.
⁵ Department of Pharmacology, University of Iowa, Iowa City, IA 52242, USA.
⁶ Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093, USA; Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, CA 92093, USA.
⁷ Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093, USA; Shu Chien-Gene Lay Department of Bioengineering, University of California, San Diego, La Jolla, CA 92093, USA; Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093, USA; Moores Cancer Center, University of California, San Diego, La Jolla, CA 92093, USA. Electronic address: jzhang32@health.ucsd.edu.

PMID: 38537638
PMCID: PMC11031308
DOI: 10.1016/j.molcel.2024.03.002

Abstract

Spatiotemporal regulation of intracellular signaling molecules, such as the 3',5'-cyclic adenosine monophosphate (cAMP)-dependent protein kinase (PKA), ensures proper cellular function. Liquid-liquid phase separation (LLPS) of the ubiquitous PKA regulatory subunit RIα promotes cAMP compartmentation and signaling specificity. However, the molecular determinants of RIα LLPS remain unclear. Here, we reveal that two separate dimerization interfaces, combined with the cAMP-induced unleashing of the PKA catalytic subunit (PKA-C) from the pseudosubstrate inhibitory sequence, drive RIα condensate formation in the cytosol of mammalian cells, which is antagonized by docking to A-kinase anchoring proteins. Strikingly, we find that the RIα pseudosubstrate region is critically involved in forming a non-canonical R:C complex, which recruits active PKA-C to RIα condensates to maintain low basal PKA activity in the cytosol. Our results suggest that RIα LLPS not only facilitates cAMP compartmentation but also spatially restrains active PKA-C, thus highlighting the functional versatility of biomolecular condensates in driving signaling specificity.

Keywords: biomolecular condensates; intrinsically disordered region; protein kinase; signaling compartmentation.

PubMed Disclaimer

Conflict of interest statement

Declaration of interests The authors declare no competing interests.

Figures

**Figure 1.. RIα D/D domain promotes LLPS via dimerization.**
**(A)** PKA RIα domain structure, including the Dimerization/Docking (D/D) domain, intrinsically disordered linker region, which includes the Inhibitory Sequence (IS), and the cyclic nucleotide-binding domains, CNB-A and CNB-B. **(B)** Amino acid sequence **(top)** and structural models (PBD: 3IM4, L50R model) **(bottom)** of the RIα D/D domain. Highlighted: two disulfide bonds and residues involved in D/D dimerization, including L50 or L50R (red) and clashes (arrows). **(C)** Representative maximum intensity projections from confocal Z-stacks showing GFP fluorescence from HEK293T cells expressing RIα-GFP2 plus Cα-mCherry before (−) and after (+) Fsk (50 μM) and IBMX (100 μM) stimulation. **(D)** Average time-course **(left)** and summary **(right)** of puncta number per cell before (−) and after (+) Fsk/IBMX stimulation in HEK293T cells expressing Cα-mCherry plus either wild-type (WT) RIα-GFP2 (WT; n = 157 cells from 4 experiments) or RIα^L50R-GFP2 (L50R; n = 92 cells from 2 experiments). Inset shows a representative confocal fluorescence image of an RIα^L50R-GFP2-expressing cell. Time-courses show the mean (solid line) and 95% confidence intervals (CI, shading). Error bars in summary quantification indicate median ± 95% CI. ††††P < 1 × 10⁻¹⁵ (WT+ vs WT−), paired Wilcoxon rank sum test; ***P = 0.000221 (L50R− vs WT−), ****P < 1 × 10⁻¹⁵ (L50R+ vs WT+), unpaired Komogorov-Smirnov test. **(E)** BRET2 assay for monitoring RIα dimerization. **(F)** Maximum GFP2/RLuc8 emission ratio in HEK293T cells expressing RLuc8 and GFP2 fused to RIα WT or L50R: WT-RLuc8 with WT-GFP2 (WT/WT; n = 7 experiments), L50R-RLuc8 with L50R-GFP2 (L50R/L50R; n = 3 experiments), and RLuc8 alone (Ctrl) with L50R-GFP2 (Ctrl/L50R; n = 5 experiments). ***P = 0.000227, ****P = 1.67 × 10⁻⁵ vs WT/WT; ns, not significant; ordinary one-way ANOVA followed by Tukey’s multiple comparisons test. Error bars indicate mean ± SD. All scale bars, 10 μm. See also Figure S1 and Table S1.

**Figure 2.. AKAP docking disrupts RIα condensate formation in cytosol.**
**(A)** Representative confocal fluorescence images and line-intensity profiles of the indicated regions of HEK293T cells co-expressing EGFP-tagged wild-type (WT) smAKAP or smAKAP^L2P2 (L2P2) plus mRuby2-tagged WT or L50R-mutant RIα (n = 10 cells from 4 independent experiments). **(B)** Recruitment of RIα-bound A-kinase-binding (AKB) domain to the plasma membrane using the rapamycin (Rapa)-induced dimerization of FKBP and FRB. **(C-D)** Representative confocal fluorescence images and line-intensity profiles of the indicated regions of HEK293T cells co-expressing **(C)** WT or **(D)** L2P2-mutant AKB-mVenus-FKBP plus RIα-mRuby2 and Lyn-FRB before (t = 0), as well as 1 and 2 min after addition of 1 μM rapamycin (n = 10 cells from 4 independent experiments). All scale bars, 10 μm. See also Figure S2.

**Figure 3.. Secondary dimerization via the CNB-A N3A motif promotes RIα LLPS.**
**(A)** Amino acid sequence **(top)** and crystal structures (PBD: 6BYS, Y122A model) **(bottom)** of the RIα CNB-A N3A motif (green) dimerization at the C-terminus of linker region (yellow). RIα CNB-A: light blue; RIα CNB-B: dark blue. Highlighted: residues involved in CNB-A N3A motif dimerization, including Y122 and Y122A (red). **(B)** Maximum GFP2/RLuc8 emission ratio in HEK293T cells expressing RLuc8 and GFP2 fused to RIα WT or Y122A: WT-RLuc8 with WT-GFP2 (WT/WT; data reproduced from Fig. 1), Y122A-RLuc8 with Y122A-GFP2 (Y122A/Y122A; n = 4 experiments), and RLuc8 alone (Ctrl) with L50R-GFP2 (Crtl/L50R; data reproduced from Fig. 1). **P = 0.00656, ****P = 4.24 × 10⁻⁵ vs WT/WT; ns, not significant; ordinary one-way ANOVA followed by Tukey’s multiple-comparisons test. Error bars indicate mean ± SD. **(C)** Average time-course (left) and summary (right) of puncta number per cell before (−) and after (+) Fsk/IBMX stimulation in HEK293T cells co-expressing Cα-mCherry plus either RIα-GFP2 (WT, reproduced from Fig. 1) or RIα^Y122A-GFP2 (Y122A, n = 95 cells from 2 experiments). Inset shows a representative confocal fluorescence image of an RIα^Y122A-GFP2-expressing cell. ††††P < 1 × 10⁻¹⁵ (WT+ vs WT−), paired Wilcoxon rank-sum test; ****P = 2.48 × 10⁻¹² (Y122A+ vs WT+), unpaired Komogorov-Smirnov test. Time-courses show the mean (sold line) ± 95% CI (shading). Error bars in summary quantification indicate median ± 95% CI. Scale bar, 10 μm. **(D)** Fsk-stimulated change in the GFP2/RLuc8 emission ratio in HEK293T cells co-expressing GFP2-Cα plus RLuc8 fused to either RIα^WT or RIα^Y122A(n = 3 experiments each), ns, not significant; unpaired, two-tailed Student’s t-test. Error bars indicate mean ± SD. See also Figure S3.

**Figure 4.. Both CNB domains are required for cAMP-induced RIα LLPS.**
Crystal structures (PBD: 4JV4) of cAMP bound to the **(A)** CNB-A and **(B)** CNB-B of RIα, highlighting hydrogen bonds formed with residues **(A)** E202 and **(B)** E326. **(C)** Fsk-stimulated change in the GFP2/RLuc8 emission ratio in HEK293T cells co-expressing GFP2-Cα plus RLuc8 fused to WT RIα, RIα^E202A, or RIα^E326A (n = 3 experiments each). ****P = 5.97 × 10⁻⁵ (E202A vs WT), 8.08 × 10⁻⁵ (E326A vs WT); ns, not significant; ordinary one-way ANOVA followed by Tukey’s multiple-comparisons test. Error bars indicate mean ± SD. **(D)** Average time-course (left) and summary (right) of puncta number per cell before (−) and after (+) Fsk/IBMX stimulation in HEK293T cells co-expressing Cα-mCherry plus RIα^WT (WT; reproduced from Fig. 1), RIα^E202A (E202A; n = 118 cells from 2 experiments), and RIα^E326A (E326A; n = 89 cells from 2 experiments). Insets show representative confocal fluorescence images of HEK293T cells expressing RIα^E202- or RIα^E326A-GFP2. Time-courses show the mean (solid lines) ± 95% CI (shading). Error bars in summary quantification indicate median ± 95% CI. ††††P < 1 × 10⁻¹⁵ (WT+ vs WT−); paired Wilcoxon rank-sum test and ***P = 0.000317 (E202A− vs WT−), ****P < 1 × 10⁻¹⁵ (E202A+ vs WT+), ****P < 1 × 10⁻¹⁵ (E326A+ vs WT+); unpaired Komogorov-Smirnov test. Scale bars, 10 μm. See also Figure S4.

**Figure 5.. RIα LLPS is a function of IS binding state.**
**(A)** Amino acid sequence of the RIα linker region, including the linker region charged residues (coral), Inhibitory Sequence (IS) (bold), and the A99 “P”-site (red) **(top)** and crystal structure of PKA-C-(left; PBD: 6NO7) or cAMP-bound (right; PBD: 4MX3) RIα, highlighting RIα residues that interact with Cα residues **(bottom)**. In the structures, the subunits and domains are colored as follows: PKA-C: grey and RIα linker: yellow, IS: red, CNB-A N3A motif: green, CNB-A: light blue, and CNB-B: dark blue. **(B)** Fsk-stimulated change in the GFP2/RLuc8 emission ratio in HEK293T cells co-expressing GFP2-Cα plus RLuc8 fused to either RIα^WT (WT; data reproduced from Fig. 3) or RIα^A99S (A99S; n = 3 experiments). ****P = 7.33 × 10⁻⁵ vs WT; unpaired, two-tailed Student’s t-test. Error bars indicate mean ± SD. **(C)** Representative maximum intensity projections from confocal z-stacks of HEK293T cells co-expressing Cα-mCherry and RIα^A99S-GFP2 in HEK293T cells. Scale bar, 10 μm. **(D)** Average time-course (left) and summary (right) of puncta number per cell before (−) and after (+) Fsk/IBMX stimulation in HEK293T cells co-expressing Cα-mCherry plus either RIα-GFP2 (WT; data reproduced from Fig. 1) or RIα^A99S-GFP2 (A99S; n = 49 cells from 2 experiments). Time-courses show the mean (solid lines) ± 95% CI (shading). Error bars in summary quantification indicate median ± 95% CI. ††††P < 1 × 10⁻¹⁵ (WT+ vs WT−), ††††P = 1.86 × 10⁻⁸ (A99S+ vs A99S−); paired Wilcoxon rank-sum test. ****P = 4.71 × 10⁻¹³ (A99S− vs WT−), ****P = 7.83 × 10⁻⁶ (A99S+ vs WT+); unpaired Komogorov-Smirnov test. See also Figure S5.

**Figure 6.. Non-canonical recruitment of PKA catalytic subunit to RIα condensates.**
**(A)** Representative confocal images and line-intensity profiles of the indicated regions of HEK293T cells co-expressing RIα-GFP2plus Cα-mRuby2 (n = 10 cells from 4 independent experiments). Scale bars, 10 μm. **(B)** Pearson’s coefficient per RIα puncta in HEK293T cells co-expressing RIα-GFP2 (n = 2288 puncta before and n = 2502 puncta after stimulation) or RIα^A99S-GFP2 (n = 1055 puncta before and n = 1253 after stimulation) with Cα-mRuby2. Effect size was calculated using Cohen’s d. Error bars indicate mean ± SD. **(C)** Schematic illustrating the classical model of dissociation of the canonical PKA holoenzyme (top) and the proposed model of the non-canonical R:C interaction (bottom). In the classical model, binding of cAMP (blue dots) to the CNB domains (open circles) causes complete dissociation of PKA-C (triangles), in contrast to the proposed non-canonical interaction in which active PKA-C remains tethered to RIα via the linker region. **(D)** Recovery kinetics (t_1/2) of RIα puncta in HEK293T cells expressing RIα-GFP2 alone (RIα/−; n = 35 puncta) or co-expressing Cα-mCherry plus either RIα-GFP2 (RIα/Ca; n = 51 puncta), or RIα^A99S-GFP2 (A99S/Cα; n = 42 puncta) measured using FRAP. Error bars indicate median ± 95% CI. ****P = 6.15 × 10⁻⁶ (RIα vs RIα/Cα), ***P = 0.000845 (RIα/Cα vs A99S/Cα); ns, not significant; Brown-Forsythe and Welch ANOVA followed by Dunnett’s T3 multiple-comparisons test. **(E)** Average amplitude-weighted lifetimes measured from RIα puncta in HEK293T cells expressing RIα-GFP2 alone (RIα/−; n = 39 cells) or co-expressing Cα-mCherry plus either RIα-GFP2 (RIα/Ca; n = 54 cells) or RIα^A99S-GFP2 (A99S/Cα; n = 50 cells). Error bars indicate mean ± SD. **P = 0.0066 (RIα/− vs RIα/Cα), ****P = 2.91 × 10⁻⁵ (RIα/Cα vs A99S/Cα); ns, not signficant; Brown-Forsythe and Welch ANOVA followed by Dunnett’s T3 multiple-comparisons test. **(F)** Normalized FluoSTEP-AKAR R/G emission ratio when RIα-FP11 (RIα; n = 41 cells), RIα^A99S-FP11 (A99S; n = 42 cells), or RIα-FP11 and RIα(62-113)-mTagBFP (RIα/Linker; n = 38 cells) are expressed in RIα-KO HEK293T cells treated with 10 μM H89, followed by washout and addition of Fsk/IBMX. ****P = 1.98 × 10⁻⁵ (RIα vs A99S), ****P = 6.95 × 10⁻⁵ (RIα vs RIα/Linker); ns, not significant; unpaired, two-tailed Student’s t-test. **(G)** Normalized AKAR4 Y/C emission ratio change in RIα-KO HEK293T cells treated with 10 μM H89, followed by washout and addition of Fsk/IBMX in the absence (KO; n = 69 cells) or presence of RIα-mRuby2 (RIα; n = 53 cells) or RIα^A99S-mRuby2 (A99S; n = 87 cells) expression. Error bars indicate median ± 95% CI. ***P = 6.16 × 10⁻⁴ (KO vs RIα), **** P = 1.23 × 10⁻⁵ (RIα vs A99S); ns, not significant; Brown-Forsythe and Welch ANOVA followed by Dunnett’s T3 multiple-comparisons test. See also Figure S6.

See this image and copyright information in PMC

Update of

Molecular Determinants and Signaling Effects of PKA RIα Phase Separation.
Hardy JC, Pool EH, Bruystens JGH, Zhou X, Li Q, Zhou DR, Palay M, Tan G, Chen L, Choi JLC, Lee HN, Strack S, Wang D, Taylor SS, Mehta S, Zhang J. Hardy JC, et al. bioRxiv [Preprint]. 2023 Dec 11:2023.12.10.570836. doi: 10.1101/2023.12.10.570836. bioRxiv. 2023. Update in: Mol Cell. 2024 Apr 18;84(8):1570-1584.e7. doi: 10.1016/j.molcel.2024.03.002. PMID: 38168176 Free PMC article. Updated. Preprint.

References

1. Taylor SS, Ilouz R, Zhang P, and Kornev AP (2012). Assembly of allosteric macromolecular switches: lessons from PKA. Nature Reviews Molecular Cell Biology 13, 646–658. 10.1038/nrm3432. - DOI - PMC - PubMed
1. Taylor SS, Zhang P, Steichen JM, Keshwani MM, and Kornev AP (2013). PKA: lessons learned after twenty years. Biochimica et biophysica acta 1834, 1271–1278. 10.1016/j.bbapap.2013.03.007. - DOI - PMC - PubMed
1. Caretta A, and Mucignat-Caretta C (2011). Protein kinase a in cancer. Cancers (Basel) 3, 913–926. 10.3390/cancers3010913. - DOI - PMC - PubMed
1. Kandel ER (2012). The molecular biology of memory: cAMP, PKA, CRE, CREB-1, CREB-2, and CPEB. Molecular brain 5, 14. 10.1186/1756-6606-5-14. - DOI - PMC - PubMed
1. Boularan C, and Gales C (2015). Cardiac cAMP: production, hydrolysis, modulation and detection. Frontiers in pharmacology 6, 203. 10.3389/fphar.2015.00203. - DOI - PMC - PubMed

MeSH terms

Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions
Actions

Save citation to file

Email citation

Add to Collections

Add to My Bibliography

Your saved search

Create a file for external citation management software

Your RSS Feed

Molecular determinants and signaling effects of PKA RIα phase separation

Affiliations

Molecular determinants and signaling effects of PKA RIα phase separation

Authors

Affiliations

Abstract

Conflict of interest statement

Figures

Update of

References

MeSH terms

Substances

Grants and funding

LinkOut - more resources

Full Text Sources

Research Materials