Protein kinase A catalytic-α and catalytic-β proteins have nonredundant regulatory functions

Abstract

Vasopressin regulates osmotic water transport in the renal collecting duct by protein kinase A (PKA)-mediated control of the water channel aquaporin-2 (AQP2). Collecting duct principal cells express two seemingly redundant PKA catalytic subunits, PKA catalytic α (PKA-Cα) and PKA catalytic β (PKA-Cβ). To identify the roles of these two protein kinases, we carried out deep phosphoproteomic analysis in cultured mpkCCD cells in which either PKA-Cα or PKA-Cβ was deleted using CRISPR-Cas9-based genome editing. Controls were cells carried through the genome editing procedure but without deletion of PKA. TMT mass tagging was used for protein mass spectrometric quantification. Of the 4,635 phosphopeptides that were quantified, 67 phosphopeptides were significantly altered in abundance with PKA-Cα deletion, whereas 21 phosphopeptides were significantly altered in abundance with PKA-Cβ deletion. However, only four sites were changed in both. The target proteins identified in PKA-Cα-null cells were largely associated with cell membranes and membrane vesicles, whereas target proteins in PKA-Cβ-null cells were largely associated with the actin cytoskeleton and cell junctions. In contrast, in vitro incubation of mpkCCD proteins with recombinant PKA-Cα and PKA-Cβ resulted in virtually identical phosphorylation changes. In addition, analysis of total protein abundances in in vivo samples showed that PKA-Cα deletion resulted in a near disappearance of AQP2 protein, whereas PKA-Cβ deletion did not decrease AQP2 abundance. We conclude that PKA-Cα and PKA-Cβ serve substantially different regulatory functions in renal collecting duct cells and that differences in phosphorylation targets may be due to differences in protein interactions, e.g., mediated by A-kinase anchor proteins, C-kinase anchoring proteins, or PDZ binding.

INTRODUCTION

The actions of vasopressin in collecting duct principal cells (24) are mediated by a G-protein-coupled receptor, vasopressin receptor 2 (gene symbol: Avpr2), which couples to the heterotrimeric G protein stimulatory α-subunit (G_sα) with activation of adenylyl cyclase 6 (23). Increased intracellular cAMP levels result in physiological effects in collecting duct principal cells, largely through activation of protein kinase A (PKA) (5, 6, 9, 12, 13, 18, 20–22). Two of the most important physiological end effects are 1) membrane trafficking changes that increase the abundance of the water channel protein aquaporin-2 (AQP2) in the plasma membrane (19) and 2) increased transcription of the Aqp2 gene (8, 15, 25), both of which contribute to vasopressin-induced increases in osmotic water transport across the collecting duct epithelium. PKA is a widely studied protein that has been viewed by most investigators as a single entity, although its catalytic subunits are coded in mammalian genomes by two separate genes, PKA catalytic α (PKA-Cα; gene symbol: Prkaca) and PKA catalytic β (PKA-Cβ; gene symbol: Prkacb). At the amino acid level, the two are 91% identical and the catalytic domains are very similar. In native collecting duct cells and mpkCCD cells, PKA-Cα and PKA-Cβ are expressed at comparable levels (1, 10). (A third entity, PKA catalytic γ, is not widely expressed and will not be considered in this paper.) We have recently succeeded in using CRISPR-Cas9 to create disruptive mutations in both PKA catalytic genes [PKA double knockout (PKA dKO)] in vasopressin-responsive kidney epithelial cells (mpkCCD cells) (10). We then used phosphoproteomics to identify a large number of novel PKA targets as well as many secondary changes in phosphorylation due to loss of PKA-mediated regulation of other kinases and phosphatases (10). Whether the two PKA catalytic proteins have redundant regulatory functions, as is implicitly assumed in many studies involving PKA-mediated regulation, has not been tested. Here, we carried out mass spectrometry-based quantitative proteomics and phosphoproteomics in both PKA-Cα and PKA-Cβ single knockouts (KOs) in mpkCCD collecting duct cells to address this issue. Thus, the goal of the present study was to identify the relative roles of PKA-Cα and PKA-Cβ in the signaling processes triggered by vasopressin.

METHODS

Cell culture.

The present study used immortalized mpkCCD cells in which either Prkaca or Prkacb gene expression was deleted (“PKA-Cα-null” and “PKA-Cβ-null”) by introducing mutations using CRISPR-Cas9 (10). Control cell lines (“PKA intact”) were carried through the CRISPR-Cas9 protocol but did not show deletion of either PKA catalytic gene. We used three independent PKA-Cα-null lines and three independent PKA-intact controls. Similarly, we used three independent PKA-Cβ-null lines and three different independent PKA-intact controls, giving three biological replicates for each kinase. Experiments with these lines were performed in duplicate (i.e., 2 technical replicates each), giving 24 samples in total. The overall scheme is shown in Fig. 1. Cells were cultured as previously described (14). Briefly, cells were initially maintained in complete medium, DMEM-F-12 containing 2% serum and other supplements (5 μg/mL insulin, 50 nM dexamethasone, 1 nM triiodothyronine, 10 ng/mL epidermal growth factor, 60 nM sodium selenite, and 5 μg/mL transferrin, all from Sigma). Cells were seeded on permeable membrane supports (Transwell) and grown in complete medium containing 0.1 nM 1-desamino-8-d-arginine-vasopressin (dDAVP; basal side only) for 4 days, during which time cells formed confluent monolayers. The medium was then changed to simple medium (DMEM-F-12 with dexamethasone, sodium selenite, and transferrin and no serum) with 0.1 nM dDAVP and maintained for 3 days, at which time cells were harvested for proteomic analysis. Because dDAVP was present in the culture medium under all conditions, it was not an experimental variable in the present study.

Fig. 1.

Experimental method. Cells were grown on permeable supports with three biological replicates as indicated. There were two technical replicates each, resulting in a total of 24 samples. Proteins were isolated, trypsinized, and labeled using isotopic tags (TMT-11 plex). The 24 samples were combined into 3 batches of 8 each with a shared sample pooled from all 24 samples. The combined samples were processed for proteomic and phosphoproteomic analysis as described in methods.

Sample preparation for total and phospho-proteomics.

Cells were washed three times with ice-cold PBS and lysed with TEAB buffer (ThermoFisher) with SDS (1%) containing protease and phosphatase inhibitors (Halt, ThermoFisher). Membranes were scraped, and samples were homogenized using a QIAshredder (Qiagen). Protein concentrations were measured using the Pierce BCA Protein Assay Kit. Protein lysates were reduced with 20 mM DTT for 1 h at 25°C and then alkylated with 40 mM iodoacetamide for 1 h at 25°C in the dark. Proteins were acetone precipitated before digestion with recombinant trypsin [1:50 (wt/wt), ThermoFisher] overnight at 37°C. The resulting peptides were quantified using the Pierce Quantitative Colorimetric Peptide Assay. For each replicate, 250 μg of peptide were labeled using the TMT11Plex Mass Tag Labeling Kit (lot no. UE283355, Thermo Scientific) following the manufacturer’s instructions, and they were combined as shown in Fig. 1. A total of 3 labeling batches using the same TMT11 Plex Mass Tag kit were run to quantify all 24 samples. Each batch included a common pooled sample containing a mixture of all 24 experimental samples. Samples were combined and desalted using hydrophilic-lipophilic-balanced extraction cartridges (Oasis) and then fractionated into 12 fractions using high pH reverse-phase chromatography (Agilent 1200 HPLC System). After an aliquot was taken from each fraction for total proteomics, phosphopeptides were sequentially enriched following the Sequential Enrichment from Metal Oxide Affinity Chromatography protocol (Thermo Scientific) using High-Select TiO₂ and then High-Select Fe-NTA Phosphopeptide enrichment kits (Thermo Scientific). Samples were then vacuum dried and stored at −80°C until analysis.

Dried peptides were resuspended with 0.1% formic acid and 2% acetonitrile in LC-MS grade water (J.T. Baker) before mass spectrometry analysis. Peptides (total and phospho-) were analyzed using a Dionex UltiMate 3000 Nano LC system connected to an Orbitrap Fusion Lumos mass spectrometer equipped with an EASY-Spray ion source (ThermoFisher Scientific). Peptides were introduced into a peptide nanotrap at a flow rate of 5 μL/min. The trapped peptides were fractionated with a reverse-phase EASY-Spray PepMap column (C18, 75 μm × 50 cm) using a linear gradient of 4−32% acetonitrile in 0.1% formic acid (120 min at 0.3 μL/min). The Thermo Scientific Synchronous Precursor Selection-MS3 (SPS-MS3) workflow (17) was selected on the mass spectrometer for TMT quantification. The main settings for MS3 were as follows: HCD activation, 65% normalized collision energy, 2 m/z isolated window, 50,000 Orbitrap resolution, AGC target of 15,000, and 120-ms maximum injection time.

Mass spectrometry data processing.

Raw mass spectra were searched against the Mus musculus UniProtKB (27) reference proteome (proteome ID: UP000000589, release 2019_06, plus contaminant database) using MaxQuant (3) 1.6.7.0. Reporter ion MS3 with TMT10plex was specified as the labeling type, and lot-specific TMT isotopic impurity correction factors were used as recommended in the TMT product data sheets. Carbamidomethyl (C) was configured as fixed modifications. Variable modifications included phospho (STY), oxidation (M), and acetyl (Protein-N-terminal). The false discovery rate was controlled at 1% (target-decoy). “Trypsin/P” was set as the digestion enzyme with up to two missed cleavages allowed. Other parameters were set to the defaults. We used the “proteinGroups.txt” output file as the input data for total proteome analyses. Both “Phospho (STY)Sites.txt” and “evidence.txt” output files were used for phosphoproteome analyses. For each TMT batch, corrected reporter ion intensities were first normalized to make the total sum intensities for each channel equal. The average of the common pooled channel from three TMT batches was then used to normalize the batch effects.

In vitro phosphorylation experiments.

Three independent PKA dKO cell lines (10) were grown on permeable membrane supports as described above. Confluent monolayers were washed three times with ice-cold PBS and lysed with TEAB buffer (ThermoFisher) with SDS (1%) containing protease and phosphatase inhibitors. Membranes were scraped, samples were homogenized using a QIAshredder (Qiagen), and proteins were precipitated with acetone. The protein pellet was resuspended in TEAB buffer, and protein concentration was determined by the Pierce BCA Protein Assay Kit.

Equal quantities of proteins from the three PKA dKO cell lines were pooled together, and 500 μg pooled protein extract was mixed with either recombinant PKA-Cα (1.5 μg) or recombinant PKA-Cβ (1.5 μg) enzymes obtained from Genetex (PRKACA- GTX65206-pro and PRKACB-GTX65207-pro). Samples were incubated at 30°C for 24 h in buffer containing 50 mM Tris·HCl, 10 mM MgCl₂, 0.1 mM EDTA, 2 mM DTT, and 250 μM ATP. Samples without any added kinases were used as controls. Following the in vitro kinase incubation, proteins underwent standard phosphoproteomic analysis as previously described.

Bioinformatics and data sharing.

Technical replicates were summarized by taking the median value. All analyses were performed using Perseus, Excel, and R software. Moderated P (P_mod) values were calculated using the empirical Bayes method, which integrates variance information from all peptides measured in the same LC-MS/MS run (11). Data have been deposited in PRIDE (as part of ProteomeXchange) with Accession No. PXD015050.

RESULTS

Previously, PKA-Cα-null and PKA-Cβ-null cells were characterized by immunoblot analysis, which showed that PKA-Cα and PKA-Cβ proteins were undetectable in the corresponding single KO lines (10). Both single KO lines grew well and formed confluent monolayered epithelial sheets when grown on permeable supports. Here, we used protein mass spectrometry to quantify changes in total protein abundances and changes in phosphorylation in PKA-Cα-null versus PKA-Cα-intact cells and PKA-Cβ-null versus PKA-Cβ-intact cells across the proteome.

Proteome-wide quantification of protein abundances.

Quantitative data for the total abundance of all individual proteins are online at https://hpcwebapps.cit.nih.gov/ESBL/Database/PKA-singleKO-total/ and as Supplemental Data Set S1 in the Supplemental Material (provided online at https://hpcwebapps.cit.nih.gov/ESBL/Database/Supplemental_Data_PKA_sKO/). Figure 2 shows changes in protein abundances corresponding to 4,691 different genes in both PKA-Cα-null cells and PKA-Cβ-null cells versus their respective PKA-intact control cells. Each point shows values for a different protein. Proteins whose abundances were previously found to be significantly altered in PKA-Cα/PKA-Cβ dKO cells are marked by red or green points depending on the direction of change in dKO. Interestingly, Aqp2, which had previously been seen to be almost completely ablated in PKA dKO cells, was decreased in only PKA-Cα-null cells and not in PKA-Cβ-null cells. Two other proteins that were strongly decreased in abundance in PKA dKO experiments, namely, complement factor C3 (C3) and mucin-4 (Muc4) were differentially affected in the two single KO cells. C3 was selectively decreased in PKA-Cα-null cells, whereas mucin-4 was selectively decreased in PKA-Cβ-null cells. These data support the view that PKA-Cα and PKA-Cβ do not have the same physiological roles in collecting duct cells. Interestingly, PKA-Cβ was substantially increased in abundance in PKA-Cα-null cells, suggesting a compensatory response.

Fig. 2.

Effect of PKA-Cα (left) and PKA-Cβ (right) deletion on protein abundances in mouse mpkCCD cells. Red points indicate proteins decreased in PKA-Cα/PKA-Cβ double knockout cells (10). Green points indicate proteins increased in PKA-Cα/PKA-Cβ double knockout cells (10). Arrows to these red and green points show official gene symbols for the specific proteins. P_mod value, moderated P value.

We plotted the changes in protein abundances in PKA-Cα-null cells versus those in PKA-Cβ-null cells for those proteins that changed significantly in either PKA-KO line (Fig. 3A). As can be seen, some proteins (yellow) underwent parallel changes in abundance in both single KO cell lines, whereas others underwent selective, significant changes in either PKA-Cα-null (red) or PKA-Cβ-null (green) cells. Control versus control comparisons for the same proteins are shown in Fig. 3B, illustrating the magnitude of changes expected from random variability in the mass spectrometric quantification. [The 95% confidence interval for log₂(control/control) was 0.382.] A listing of the proteins that most convincingly changed [P_mod < 0.02 and |log(ratio)| > 0.5] in PKA-Cα-null cells is shown in Table 1, whereas changes in PKA-Cβ-null cells are shown in Table 2. Overall, we conclude that PKA-Cα and PKA-Cβ are not redundant with respect to regulation of protein abundances.

Fig. 3.

Comparison of effects of PKA-Cα and PKA-Cβ deletion on protein abundances. A: red points show proteins changed with PKA-Cα deletion but not PKA-Cβ deletion. Green points show proteins changed with PKA-Cβ deletion but not PKA-Cα deletion. Yellow points are changed in both, but not necessarily in the same direction. B: intrinsic variability of the data estimated by comparing values for PKA-intact controls versus other controls (see text).

Table 1.

Proteins with substantial changes in abundance in response to PKA-Cα deletion [P_mod < 0.02 and |log(ratio)| > 0.5]

UniProt ID	Gene Symbol	Annotation	Log2 (PKA-Cα Null/PKA-Cα Intact)	Pmod Value	Log2 (PKA-Cβ Null/ PKA-Cβ Intact)	Pmod Value
P01027	C3	Complement C3	−1.427	0.006	0.195	0.655
P56402	Aqp2	Aquaporin-2	−1.286	0.009	0.694	0.115
Q9QYI5	Dnajb2	DnaJ homolog subfamily B member 2	−0.850	0.012	0.056	0.848
P97742	Cpt1a	Carnitine O-palmitoyltransferase 1, liver isoform	−0.799	0.002	−0.091	0.647
Q8BH86	Dglucy	d-glutamate cyclase, mitochondrial	−0.798	0.003	−0.094	0.666
P24472	Gsta4	Glutathione S-transferase A4	−0.684	0.006	−0.451	0.049
Q9EPK8	Trpv4	Transient receptor potential channel subfamily V member 4	−0.682	4.93 × 10−4	0.067	0.647
Q9WTQ5	Akap12	A-kinase anchor protein 12	−0.669	0.004	0.070	0.714
Q99J39	Mlycd	Malonyl-CoA decarboxylase, mitochondrial	−0.602	0.016	0.087	0.690
Q8BLF1	Nceh1	Neutral cholesterol ester hydrolase 1	−0.551	0.010	−0.346	0.077
P63011	Rab3a	Ras-related protein Rab-3A	−0.550	4.05 × 10−4	−0.115	0.324
P26443	Glud1	Glutamate dehydrogenase 1, mitochondrial	−0.543	0.003	−0.057	0.692
G3X9Y6	Akr1c19	Aldo-keto reductase family 1, member C19	−0.534	0.014	−0.307	0.123
Q8BG05	Hnrnpa3	Heterogeneous nuclear ribonucleoprotein A3	0.507	0.010	0.033	0.845
D3YYU8	Obsl1	Obscurin-like protein 1	0.512	0.016	0.024	0.897
P62482	Kcnab2	Voltage-gated potassium channel subunit β2	0.518	0.005	0.366	0.033
Q8VI84	Noc3l	Nucleolar complex protein 3 homolog	0.524	0.002	0.369	0.016
Q5DU09	Znf652	Zinc finger protein 652	0.527	0.006	−0.275	0.108
O88508	Dnmt3a	DNA (cytosine-5)-methyltransferase 3A	0.562	0.005	−0.034	0.838
O54786	Dffa	DNA fragmentation factor subunit α	0.576	0.003	0.255	0.127
Q9D187	Ciao2b	Cytosolic iron-sulfur assembly component 2B	0.596	0.006	−0.206	0.268
Q61180	Scnn1a	Amiloride-sensitive sodium channel subunit α	0.637	0.017	−0.016	0.945
P0DOV2	Ifi204	Interferon-activable protein 204	0.722	0.012	0.108	0.664
Q9JK42	Pdk2	Pyruvate dehydrogenase kinase isozyme 2	0.734	0.005	0.049	0.821
P26645	Marcks	Myristoylated alanine-rich C-kinase substrate	1.486	0.003	1.718	0.001
P05132	Prkaca	cAMP-dependent protein kinase catalytic subunit α			0.049	0.677

P_mod value, moderated P value.

Table 2.

Proteins with substantial changes in abundance in response to PKA-Cβ deletion [P_mod < 0.02 and |log(ratio)| > 0.5]

UniProt ID	Gene Symbol	Annotation	Log2 (PKA-Cα Null/ PKA-Cα Intact)	Pmod Value	Log2 (PKA-Cβ Null/ PKA-Cβ Intact)	Pmod Value
Q8VCT4	Ces1d	Carboxylesterase 1D	−0.370	0.440	−1.358	0.013
Q8CEZ4	Mab21l4	Protein mab-21-like 4	0.180	0.652	−1.351	0.005
P30115	Gsta3	Glutathione S-transferase A3	−0.294	0.202	−1.334	6.23 × 10−5
P08074	Cbr2	Carbonyl reductase [NADPH] 2	−0.178	0.683	−1.186	0.017
Q8JZM8	Muc4	Mucin-4	−0.396	0.353	−1.144	0.017
P00329	Adh1	Alcohol dehydrogenase 1	0.334	0.348	−1.094	0.008
Q7TPW1	Nexn	Nexilin	0.228	0.349	−0.881	0.003
Q9D939	Sult1c2	Sulfotransferase 1C2	−0.535	0.065	−0.869	0.007
Q62469	Itga2	Integrin α2	−0.074	0.743	−0.696	0.009
Q99LB7	Sardh	Sarcosine dehydrogenase, mitochondrial	−0.163	0.457	−0.591	0.017
Q9WUU7	Ctsz	Cathepsin Z	−0.053	0.786	−0.562	0.013
Q8R3G9	Tspan8	Tetraspanin-8	−0.199	0.212	−0.557	0.003
P22935	Crabp2	Cellular retinoic acid-binding protein 2	−0.282	0.088	−0.534	0.004
P63082	Atp6v0c	V-type H+-ATPase 16-kDa proteolipid subunit	−0.101	0.556	−0.526	0.009
Q80V42	Cpm	Carboxypeptidase M	−0.077	0.607	−0.516	0.004
P12265	Gusb	β-Glucuronidase	−0.032	0.819	−0.516	0.003
P17563	Selenbp1	Methanethiol oxidase	−0.122	0.525	−0.506	0.019
Q5FWI3	Cemip2	Cell surface hyaluronidase	0.020	0.899	0.508	0.008
P55264	Adk	Adenosine kinase	−0.297	0.059	0.513	0.004
Q99PT3	Ino80b	INO80 complex subunit B	0.048	0.658	0.551	2.66 × 10−4
Q8K0C4	Cyp51a1	Lanosterol 14-α demethylase	−0.313	0.117	0.563	0.011
P28661	Septin4	Septin-4	0.235	0.106	0.690	2.90 × 10−4
Q80Z25	Ofd1	Oral-facial-digital syndrome 1 protein homolog	0.518	0.020	0.826	0.001
P26645	Marcks	Myristoylated alanine-rich C-kinase substrate	1.486	0.003	1.718	0.001
P68181	Prkacb	cAMP-dependent protein kinase catalytic subunit β	0.717	0.038

P_mod value, moderated P value.

Proteome-wide quantification of protein phosphorylation.

Quantitative data are provided at https://hpcwebapps.cit.nih.gov/ESBL/Database/PKA-singleKO-phospho/ and as Supplemental Data Set S2 for all phosphopeptides. Figure 4 shows changes in phosphopeptide abundances (n = 4635) in both PKA-Cα-null cells and PKA-Cβ-null cells versus their respective PKA-intact (control) cells. Each point shows values for a different phosphopeptide. Phosphopeptides corresponding to previously identified PKA target sites (10) are indicated in red. Most of the PKA sites were found to be decreased in PKA-Cα-null but not PKA-Cβ-null cells relative to controls. A listing of the phosphopeptides most convincingly changed [P_mod < 0.01 and |log(ratio)| > 0.5] in PKA-Cα-null cells is shown in Table 3, whereas changes in PKA-Cβ-null cells are shown in Table 4. Overall, we conclude that PKA-Cα and PKA-Cβ have a substantially different set of phosphorylation targets. This difference could be due either to a difference in target sequence preference or a difference in PKA-Cα and PKA-Cβ interactomes. Figure 5 shows the predicted target sequence preference logos for decreased single-site phosphopeptides in PKA-Cα- and PKA-Cβ-null cells, respectively. Only the PKA-Cα logo was consistent with the familiar (R/K)-(R/K)-X-p(S/T) motif generally accepted to be characteristic of PKA. Among the 82 phosphosites that were decreased in PKA-Cβ-null cells, only four possessed the (R/K)-(R/K)-X-p(S/T) motif. The logos raise the hypothesis that the two PKA catalytic subunits may have different target sequence preferences.

Fig. 4.

Effect of PKA-Cα (left) and PKA-Cβ (right) deletion on phosphopeptide abundances in mouse mpkCCD cells. Red points indicate phosphorylation sites altered in PKA-Cα/PKA-Cβ double knockout cells (10). Arrows to these red points show official gene symbol and amino acid number of phosphorylated site for those sites above the moderated P (P_mod) threshold (−log P_mod > 1.3).

Table 3.

Phosphopeptides most convincingly changed in PKA-Cα-null cells versus PKA-Cα-intact [P_mod < 0.01 and |log(ratio)| > 0.5]

UniProt ID	Gene Symbol†	Amino Acid Number	Annotation	Sequence	Log2 (PKA-Cα Null/ PKA-Cα Intact)	Pmod Value	Log2 (PKA-Cβ Null/ PKA-Cβ Intact)	Pmod Value
P16254	Srp14	44	Signal recognition particle 14-kDa protein	KS*SVEGLEPAENK	−1.41	1.09 × 10−4	0.18	0.472
P31324	Prkar2b	112	cAMP-dependent protein kinase type II-β regulatory subunit	RAS*VCAEAYNPDEEEDDAESR	−1.38	0.001	−0.29	0.402
P05213	Tuba1b; Tuba4a; Tuba1a; Tuba3a; Tuba1c	158	Tubulin α1B chain; tubulin α4A chain; tubulin α1A chain; tubulin α1C chain	LS*VDYGK	−1.36	1.21 × 10−4	0.28	0.266
E9PVZ8	Golgb1	2655	Golgi autoantigen, golgin subfamily b, macrogolgin 1	KVS*EIEDQLK	−1.29	0.001	0.20	0.484
Q68FG2	Sptbn2	2254	Spectrin β-chain	RGS*LGFYK	−1.28	0.003	0.20	0.576
P16254	Srp14	44	Signal recognition particle 14-kDa protein	KS*SVEGLEPAENK	−1.27	1.02 × 10−4	0.25	0.276
P14152	Mdh1	241	Malate dehydrogenase, cytoplasmic	KLS*SAMSAAK	−1.17	1.75 × 10−4	0.07	0.768
Q9ES28	Arhgef7	830	Rho guanine nucleotide exchange factor 7	S*LEEEQR	−1.13	9.54 × 10−5	0.18	0.371
Q9DBC7	Prkar1a	83	cAMP-dependent protein kinase type I-α regulatory subunit	EDEIS*PPPPNPVVK	−1.05	1.34 × 10−4	−0.09	0.642
G3X9K3	Arfgef1	1076	Brefeldin A-inhibited guanine nucleotide-exchange protein 1	EGS*LTGTK	−0.96	0.001	0.09	0.688
Q9CZX9	Emc4	36	Endoplasmic reticulum membrane protein complex subunit 4	SDRGS*GQGDSLYPVGYLDK	−0.93	0.003	0.18	0.485
Q8CI52	Gramd1c	532	Protein aster-C	SS*TDLGFEAK	−0.92	3.38 × 10−4	−0.04	0.832
Q91XA2	Golm1	299	Golgi membrane protein 1	PEEDS*QYPER	−0.91	0.008	0.08	0.794
Q9DBC7	Prkar1a	83	cAMP-dependent protein kinase type I-α regulatory subunit	TDSREDEIS*PPPPNPVVK	−0.90	0.001	−0.11	0.599
Q8CHT1	Ngef	84	Ephexin-1	RAS*DQADLPK	−0.88	0.002	0.32	0.179
Q8CI52	Gramd1c	531	Protein aster-C	S*STDLGFEAK	−0.87	5.31 × 10−5	0.08	0.558
P43277	Hist1h1c; Hist1h1d	37	Histone H1.2; histone H1.3	KAS*GPPVSELITK	−0.84	0.003	0.26	0.263
Q8C341	Suco	1069	SUN domain-containing ossification factor	RTS*FPLIR	−0.80	0.006	0.13	0.597
Q8K4L3	Svil	857	Supervillin	KLS*VDNNTSATDYK	−0.77	0.009	−0.03	0.918
Q67FY2	Bcl9l	118	B cell CLL/lymphoma 9-like protein	SVS*VDSGEQR	−0.75	0.002	−0.86	0.001
F6ZDS4	Tpr	2204	Nucleoprotein TPR	TVPS*TPTLVVPHR	−0.74	0.004	−0.47	0.047
Q8C079	Strip1	59	Striatin-interacting protein 1	KDS*EGYSESPDLEFEYADTDK	−0.73	0.001	0.08	0.642
Q9JKB3	Ybx3	328	Y-box-binding protein 3	S*RPLNAVSQDGK	−0.72	0.004	0.09	0.644
Q6NZF1	Zc3h11a	740	Zinc finger CCCH domain-containing protein 11A	RLS*SASTGKPPLSVEDDFEK	−0.71	0.006	−1.00	0.001
O35609	Scamp3	78	Secretory carrier-associated membrane protein 3	KLS*PTEPR	−0.71	0.004	0.18	0.389
P15066	Jund; Jun	100	Transcription factor jun-D	LAS*PELER	−0.69	0.002	0.07	0.720
Q02248	Ctnnb1	552	Catenin β1	RTS*M^GGTQQQFVEGVR	−0.69	0.003	−0.15	0.455
Q8BUH8	Senp7	12	Sentrin-specific protease 7	RAS*SEIVTEGK	−0.67	8.08 × 10−5	−0.67	8.49 × 10−5
Q6PGG2	Gmip	260	GEM-interacting protein	S*REEAQAK	−0.67	0.001	0.33	0.057
Q04899	Cdk18	75	Cyclin-dependent kinase 18	RLS*LPMDIR	−0.66	0.005	−0.40	0.057
P20357	Map2	1635	Microtubule-associated protein 2	S*GILVPSEK	−0.65	0.005	−0.66	0.004
Q61165	Slc9a1	707	Sodium/hydrogen exchanger 1	IGS*DPLAYEPK	−0.63	0.004	0.17	0.366
Q8K3X4	Irf2bpl; Irf2bp2	13	Probable E3 ubiquitin-protein ligase IRF2BPL; interferon regulatory factor 2-binding protein 2	RQS*CYLCDLPR	−0.63	0.001	0.33	0.047
P70698	Ctps1	575	CTP synthase 1	SGSSS*PDSEITELK	−0.61	0.009	0.36	0.089
A0A087WQ44	Srcap	2771	Snf2-related CREBBP activator protein	TS*ADVEIR	−0.59	0.004	−0.49	0.011
P81122	Irs2	362	Insulin receptor substrate 2	TAS*EGDGGAAGGAGTAGGR	−0.58	0.004	0.19	0.249
Q02248	Ctnnb1	551	Catenin β1	RT*SM^GGTQQQFVEGVR	−0.57	0.008	−0.06	0.754
F7BJB9	Morc3	563	MORC family CW-type zinc finger protein 3	RLS*NPPVENSSYK	−0.57	0.002	−0.29	0.062
Q811L6	Mast4; Mast1; Mast2	1436	Microtubule-associated serine/threonine-protein kinase 4; microtubule-associated serine/threonine-protein kinase 2	SAEPPRS*PLLK	−0.56	0.001	−0.41	0.007
Q8BG05	Hnrnpa3	361	Heterogeneous nuclear ribonucleoprotein A3	SSGSPY*GGGYGSGGGSGGYGSR	−0.54	0.003	−0.23	0.143
P35569	Irs1	3	Insulin receptor substrate 1	AS*PPDTDGFSDVR	−0.54	2.04 × 10−4	−0.09	0.418
P97310	Mcm2	21	DNA replication licensing factor MCM2	RIS*DPLTSSPGR	−0.51	0.001	−0.11	0.378
Q9QXZ0	Macf1	3889	Microtubule-actin cross-linking factor 1	QGS*FSEDVISHK	−0.51	0.008	0.23	0.177
Q8K310	Matr3	188	Matrin-3	RDS*FDDR	−0.50	0.003	0.11	0.423
Q9DBR1	Xrn2	473	5′-3′ exoribonuclease 2	NSSPS*ISPNTSFASDGSPSPLGGIK	0.50	0.003	−0.41	0.010
E1U8D0	Soga1	1300	Protein SOGA1	APS*PTTAAGEESCK	0.52	0.006	0.11	0.494
Q8BTI8	Srrm2	1338	Serine/arginine repetitive matrix protein 2	S*SSELSPEVVEK	0.53	0.006	0.10	0.561
Q8BIA4	Fbxw8	86	F-box/WD repeat-containing protein 8	SRS*PPDRDATEPEPLVDQLIR	0.57	0.003	−0.24	0.126
Q99K30	Eps8l2	217	Epidermal growth factor receptor kinase substrate 8-like protein 2	QPGDS*PQAK	0.57	0.007	−0.02	0.893
P63058	Thra	12	Thyroid hormone receptor α	VECGS*DPEENSAR	0.61	0.006	−0.24	0.208
Q9CQT2	Rbm7	108	RNA-binding protein 7	SGSSHASQDASVSYPQHHVGNLS*PTSTSPNSYER	0.63	0.008	−0.10	0.608
Q9WTQ5	Akap12	584	A-kinase anchor protein 12	GPSEAPQEAEAEEGATS*DGEKKR	0.64	0.008	−0.02	0.908
Q9QXM1	Jmy	704	Junction-mediating and -regulatory protein	STAS*PVPCEEQCHSLPTVLQGQEK	0.67	0.005	−0.29	0.166
P39053	Dnm1	774; 777	Dynamin-1	RS*PTS*SPTPQR	0.67	0.002	−0.31	0.105
Q9Z1T6	Pikfyve	1753	1-Phosphatidylinositol 3-phosphate 5-kinase	GTAGKS*PDLSSQK	0.68	5.23 × 10−5	−0.25	0.039
Q8BTI8	Srrm2	1343	Serine/arginine repetitive matrix protein 2	SSSELS*PEVVEK	0.68	0.002	0.13	0.472
Q8C078	Camkk2	91	Calcium/calmodulin-dependent protein kinase kinase 2	DQPPEADGQELPLEASDPESRS*PLSGR	0.75	9.91 × 10−5	−0.14	0.317
P58802	Tbc1d10a	407	TBC1 domain family member 10A	AILDAEPGPRPALQPS*PSIR	0.78	2.06 × 10−4	−0.34	0.039
P43274	Hist1h1c; Hist1h1t; Hist1h1e; Hist1h1b; Hist1h1d	102	Histone H1.2; histone H1.4; histone H1.5; histone H1.3	GTGAS*GSFK	0.79	0.006	0.10	0.663
Q9ESE1	Lrba	979	Lipopolysaccharide-responsive and beige-like anchor protein	DS*PISPHFTR	0.80	0.003	−0.10	0.659
Q6PDN3	Mylk	355	Myosin light chain kinase, smooth muscle	VPAIGSFS*PGEDRK	0.81	0.001	−0.21	0.261
P56402	Aqp2	256; 261	Aquaporin-2	QS*VELHS*PQSLPR	1.02	4.83 × 10−4	0.44	0.064
P40645	Sox6	454	Transcription factor SOX-6	TS*PVNLPNK	1.19	0.007	−0.10	0.783
P40645	Sox6	439	Transcription factor SOX-6	S*PTSPTQNLFPASK	1.40	0.002	0.13	0.708

PKA-Cβ-null data for the same peptides are given for comparison. P_mod value, moderated P value.

^†Multiple gene symbols are given when a peptide matches sequences in multiple proteins. *Phosphorylated amino acids.

Table 4.

Phosphopeptides changed most convincingly in PKA-Cβ -null cells versus PKA-Cβ intact [P_mod < 0.01 and |log(ratio)| > 0.5]

UniProt ID	Gene Symbol	Amino Acid Number	Annotation	Sequence	Log2 (PKA-Cα Null/ PKA-Cα Intact)	Pmod Value	Log2 (PKA-Cβ Null/ PKA-Cβ Intact)	Pmod Value
Q6NZF1	Zc3h11a	740	Zinc finger CCCH domain-containing protein 11A	RLS*SASTGKPPLSVEDDFEK	−0.71	0.006	−1.00	0.001
Q67FY2	Bcl9l	118	B cell CLL/lymphoma 9-like protein	SVS*VDSGEQR	−0.75	0.002	−0.86	0.001
Q03173	Enah	719	Protein enabled homolog	APST*STPEPTR	0.00	0.988	−0.79	0.002
Q9QZQ1	Mllt4	1201	Afadin	ITSVS*TGNLCTEEQSPPPRPEAYPIPTQTYTR	−0.05	0.793	−0.77	0.002
O08796	Eef2k	66	Eukaryotic elongation factor 2 kinase	T*ECGSTGSPASSFHFK	−0.13	0.552	−0.75	0.004
Q03173	Enah	720	Protein enabled homolog	APSTS*TPEPTR	−0.11	0.633	−0.75	0.005
O55003	Bnip3	60	BCL2/adenovirus E1B 19 kDa protein-interacting protein 3	SSHCDS*PPR	0.27	0.222	−0.70	0.007
Q2VPU4	Mlxip	9	MLX-interacting protein	AADVFM^CS*PR	−0.03	0.897	−0.68	0.010
Q8BUH8	Senp7	12	Sentrin-specific protease 7	RAS*SEIVTEGK	−0.67	8.08 × 10−5	−0.67	8.49 × 10−5
P20357	Map2	1635	Microtubule-associated protein 2	S*GILVPSEK	−0.65	0.005	−0.66	0.004
Q8VDZ4	Zdhhc5	409	Palmitoyltransferase ZDHHC5	SEPSLEPESFRS*PTFGK	−0.27	0.144	−0.64	0.003
Q8CGF1	Arhgap29	1241	Rho GTPase-activating protein 29	ESSEEPALPEGT*PTCQRPR	−0.04	0.743	−0.63	0.001
Q8VI36	Pxn	258	Paxillin	IS*ASSATR	−0.01	0.966	−0.62	0.004
Q8C6B2	Rtkn	518	Rhotekin	T*FSLDAAPADHSLGPSR	−0.01	0.962	−0.60	0.005
Q8C6B2	Rtkn	520	Rhotekin	TFS*LDAAPADHSLGPSR	−0.13	0.423	−0.60	0.002
A1A535	Veph1	422	Ventricular zone-expressed PH domain-containing protein 1	INAESNT*PGSGR	−0.34	0.070	−0.59	0.005
Q9Z2H5	Epb41l1	540	Band 4.1-like protein 1	RLPS*SPASPSPK	−0.17	0.332	−0.57	0.006
P28661	Sept4	68	Septin-4	PQS*PDLCDDDVEFR	−0.02	0.917	−0.55	0.003
Q8C5R2	Proser2	223	Proline and serine-rich protein 2	LAGNEALSPTS*PSK	−0.26	0.119	−0.55	0.004
Q99MR1	Gigyf1	227	GRB10-interacting GYF protein 1	ST*SPDGGPR	−0.05	0.640	−0.51	0.001
Q99MR1	Gigyf1	226	GRB10-interacting GYF protein 1	S*TSPDGGPR	0.05	0.723	−0.50	0.003
Q60864	Stip1	481	Stress-induced-phosphoprotein 1	HDS*PEDVK	−0.12	0.498	0.57	0.005

PKA-Cα-null data for the same peptides are also given for comparison. P_mod value, moderated P value. *Phosphorylated amino acids.

Fig. 5.

Sequence preference logos from sites decreased with PKA-Cα deletion (top) and PKA-Cβ deletion (bottom). Logos were generated using PTM-Logo with a background of all unchanged phosphopeptides.

In vitro phosphorylation by recombinant PKA-Cα versus PKA-Cβ.

PKA-Cα and PKA-Cβ may have different target preferences. To test this hypothesis directly, we carried out in vitro phosphorylation experiments using purified recombinant PKA-Cα and PKA-Cβ to phosphorylate protein extracts obtained from PKA dKO cells. The full data are shown in Supplemental Data Set S3. Figure 6A shows a comparison of the changes in phosphorylation by PKA-Cα versus PKA-Cβ, demonstrating that both PKA catalytic subunits phosphorylate virtually identical substrates. Figue 6B shows that the sequence preferences for PKA-Cα versus PKA-Cβ are virtually identical. This result rules out the hypothesis that the two PKA catalytic subunits have different substrate specificities. The remaining possibility, that the two catalytic subunits have different targets because they are propinquitous to a different set of proteins, remains as the most likely explanation for the differences in targets in the intact cells.

Fig. 6.

Results of in vitro phosphorylation experiments. Protein extracts from PKA double knockout cells (PKA-Cα and PKA-Cβ) were incubated with recombinant PKA-Cα or PKA-Cβ, and phosphorylation was quantified by mass spectrometry. A: there was a marked similarity between responses to the two recombinant kinase proteins. B: sequence preference logos derived from the analysis were almost identical.

Phosphoproteomics as a virtual proximity assay.

In addition to the target sequence, another factor important to the determination of targets for a protein kinase is colocalization because a kinase can only phosphorylate a protein with which it comes into physical contact. In this sense, phosphoproteomic analysis can be viewed as a kind of large-scale proximity assay, identifying kinase/target interactions. To evaluate whether differences in PKA-Cα and PKA-Cβ phosphorylation targets result in part from differences in PKA-Cα and PKA-Cβ localization, we identified Gene Ontology (GO) cellular component terms that are enriched in either set of phosphorylation targets in the intact cell experiments, relative to all proteins detected. Table 5 shows terms enriched in phosphoproteins with altered phosphorylation in PKA-Cα-null cells, whereas Table 6 shows terms enriched in phosphoproteins with altered phosphorylation in PKA-Cβ-null cells. Enriched terms in PKA-Cα-null cells were largely related to cell membranes and membrane vesicles, whereas enriched terms in PKA-Cβ-null cells were related to the actin cytoskeleton and cell junctions, suggesting that in vivo cellular subdomains of PKA-Cα and PKA-Cβ differ.

Table 5.

Gene Ontology cellular component terms enriched in the list of phosphoproteins with altered phosphorylation in PKA-Cα-null cells

Term	Number	Fold Enrichment	Fisher Exact	Proteins
Membrane raft	8	4.3	3.20×10−4	EGFR, EFHD2, PRKAR2B, PRKAR1A, PIKFYVE, CXADR, CTNNB1, HDAC6
Perinuclear region of cytoplasm	12	2.3	3.90×10−3	EGFR, PACS1, EPN3, PRKAR2B, FBXW8, PAK2, SCYL2, PIKFYVE, SPTBN2, RAB3IP, CTNNB1, HDAC6
Membrane part	26	1.6	6.40×10−3	GPRC5C, ACSS2, CXADR, CHCHD6, CTNNB1, EFHD2, PRKAR2B, PIKFYVE, PRKAA1, SYNPO, SREBF1, EGFR, EPN3, OLFR120, KANSL1, SLC33A1, STIM2, CRB3, EMC4, BNIP3L, PRKAR1A, SPTBN2, GRAMD1C, DNM1, HDAC6, SUCO
Cytoplasmic vesicle	11	2	1.90×10−2	EGFR, PACS1, SREBF1, ARHGAP21, EPN3, SCYL2, RAN, PIKFYVE, SPTBN2, CXADR, DNM1
Integral component of membrane	16	1.6	3.60×10−2	SREBF1, EGFR, OLFR120, GPRC5C, KANSL1, SLC33A1, STIM2, CRB3, ACSS2, CXADR, CHCHD6, EMC4, BNIP3L, PIKFYVE, GRAMD1C, SUCO
Membrane-bounded vesicle	21	1.4	5.00×10−2	SREBF1, PACS1, EGFR, SRP14, EPN3, GPRC5C, RAN, CRB3, CXADR, HNRNPA1, CTNNB1, ARHGAP21, FAM65A, PRKAR2B, RPL30, SCYL2, PIKFYVE, SPTBN2, TUBA1B, DNM1, MDH1

Table 6.

Gene Ontology cellular component terms enriched in the list of phosphoproteins with altered phosphorylation in PKA-Cβ-null cells

Term	Number	Fold Enrichment	Fisher Exact	Proteins
Cytoskeleton	23	1.7	2.70 × 10−3	DPF2, SEPT4, ENAH, DYNC1LI2, PDLIM5, AKAP12, IGF2BP2, NEXN, MYO9A, PXN, SMC3, TNKS1BP1, EPB41L1, MACF1, MAP2, MAP4, RANBP2, CDC42EP4, CEP170B, UBXN6, AFAP1, SYNPO, SEPT9
Adherens junction	16	1.9	5.10 × 10−3	EGFR, ENAH, PDLIM5, AKAP12, NEXN, CXADR, PXN, TNKS1BP1, EPB41L1, LIMD1, EEF1D, ERC1, EPS8L2, AFAP1, TJP2, SEPT9
Cell junction	19	1.7	1.30 × 10−2	CLDN8, EGFR, ENAH, PDLIM5, AKAP12, CXADR, NEXN, PXN, TNKS1BP1, EPB41L1, LIMD1, EEF1D, ERC1, CDC42EP4, EPS8L2, AFAP1, TJP2, SYNPO, SEPT9
Plasma membrane	21	1.5	1.90 × 10−2	CLDN8, EGFR, ENAH, IRS2, ZDHHC5, PDLIM5, LRBA, AKAP12, WWC1, ZBTB16, CXADR, PXN, AQP2, TNKS1BP1, EPB41L1, MACF1, MAP4, CDC42EP4, EPS8L2, TJP2, SYNPO
Actin cytoskeleton	9	2.1	2.60 × 10−2	ENAH, MACF1, PDLIM5, CDC42EP4, MYO9A, AFAP1, PXN, SEPT9, SYNPO

PKA-Cα and PKA-Cβ regulate different protein kinases.

Many phosphorylation sites that changed in PKA-Cα-null or PKA-Cβ-null cells did not conform to the classic PKA target motif and presumably undergo changes in phosphorylation as a result of secondary effects on other protein kinases. Table 7 shows protein kinase catalytic proteins that underwent changes in phosphorylation in either PKA-Cα-null or PKA-Cβ-null cells. Many of the altered phosphorylation sites have known effects on enzyme activity, as indicated in the last column. The affected kinases, namely, calcium/calmodulin-dependent protein kinase kinase 2 (Camkk2), epidermal growth factor receptor kinase (Egfr), myosin light chain kinase (Mylk), p21-activated kinase (Pak2), AMP-activated protein kinase-α₁ (Prkaa1), and salt-inducible kinase 2 (Sik2), all underwent changes in phosphorylation in response to PKA-Cα deletion, whereas only Egfr underwent a change in phosphorylation in response to PKA-Cβ deletion. This supports the conclusion that PKA-Cα or PKA-Cβ have different downstream signaling networks.

Table 7.

Protein kinases with altered phosphorylation in response to deletion of PKA-Cα or PKA-Cβ

Gene Symbol	Annotation	Amino Acid Number	Centralized Sequence‡	Log2 (PKA-Cα Null/ PKA-Cα Intact)	Pmod Value	Log2 (PKA-Cβ Null/ PKA-Cβ Intact)	Pmod Value	Kinase Class	Effect of Phosphorylation on Activity†
Camkk2	Calcium/calmodulin-dependent protein kinase kinase 2	85	QELPLEAS*DPESRSP	0.41	0.005	−0.17	0.169	Other	Increases
Camkk2	Calcium/calmodulin-dependent protein kinase kinase 2	91	ASDPESRS*PLSGRKM	0.75	9.91 × 10−5	−0.14	0.317	Other	NI
Cdk18	Cyclin-dependent kinase 18	75	EDLNKRLS*LPMDIRL	−0.61	0.012	−0.39	0.086	CMGC	NI
Cdk18	Cyclin-dependent kinase 18	109	TRMSRRAS*LSDIGFG	−0.65	0.042	−0.18	0.543	CMGC	NI
Eef2k	Eukaryotic elongation factor 2 kinase	66	YYSNLTKT*ECGSTGS	−0.13	0.552	−0.75	0.004	Atypical	NI
Egfr	Epidermal growth factor receptor	695	RELVEPLT*PSGEAPN	−0.55	0.024	−0.63	0.012	Tyr	Altered receptor internalization
Map4k5	Mitogen-activated protein kinase kinase kinase kinase 5	335	SRAERTAS*EINFDKL	−0.70	0.028	−0.48	0.112	STE	NI
Map4k5	Mitogen-activated protein kinase kinase kinase kinase 5	400	PPKPRVNT*YPEDSLP	−0.80	0.027	−0.04	0.894	STE	NI
Mast4	Microtubule-associated serine/threonine-protein kinase 4	1382	SPLARTPS*PTPQPTS	0.04	0.885	−0.55	0.047	AGC	NI
Mast4 or Mast1 or Mast2	Microtubule-associated serine/threonine-protein kinase 4; microtubule-associated serine/threonine-protein kinase 2	1436	KSAEPPRS*PLLKRVQ	−0.559	0.001	−0.411	0.007	AGC	NI
Mylk	Myosin light chain kinase, smooth muscle	355	VPAIGSFS*PGEDRK	0.81	0.001	−0.21	0.261	CAMK	Increases
Nrbp2	Nuclear receptor-binding protein 2	20	EREREDES*EDESDIL	−0.02	0.870	−0.42	0.002	Other	NI
Pak2	Serine/threonine-protein kinase PAK 2	141	TVKQKYLS*FTPPEKD	−2.00	0.043	−1.30	0.167	STE	Increases
Pak2	Serine/threonine-protein kinase PAK 2	197	TKSIYTRS*VIDPIPA	−1.30	0.039	−0.21	0.719	STE	Increases?
Prkaa1	5′-AMP-activated protein kinase catalytic subunit α1	496	ATPQRSGS*ISNYRSC	−0.41	0.009	−0.13	0.347	CAMK	Decreases
Scyl2	SCY1-like protein 2	677	QGKQKRGS*LTLEEKQ	−0.47	0.006	0.09	0.541	Other	NI
Sik2	Serine/threonine-protein kinase SIK2	342	ERLKSHRS*SFPVEQR	−0.42	0.013	−0.14	0.333	CAMK	Increases?
Slk	STE20-like serine/threonine-protein kinase	777	SKAKDSGS*VSLQETR	−0.21	0.293	−0.43	0.041	STE	NI
Wnk1	Serine/threonine-protein kinase WNK1	165	TSKDRPVS*QPSLVGS	0.09	0.578	0.42	0.020	Other	NI

P_mod value, moderated P value; NI, no information available.

^*Phosphorylated amino acids. R or K in positions −3 and −2 are underlined.

^†Based on information at PhospoSitePlus (https://www.phosphosite.org).

Differential regulation of cAMP signaling proteins by PKA-Cα and PKA-Cβ.

One factor involved in the localization of PKA in the cell is its interaction with anchoring proteins such as A-kinase anchoring proteins (AKAPs). Table 8 shows altered phosphopeptides belonging to AKAP proteins or proteins with GO terms containing “cAMP” or “cyclic-AMP” to identify possible other AKAP interactors. Among the phosphoproteins in this list are Akap1, Akap12, cAMP phosphodiesterase 4C (Pde4c), cAMP phosphodiesterase 7A (Pde7a), and Prkaa1 as well as two PKA regulatory subunits, RIα (Prkar1a) and RIIβ (Prkar2b). All of these (except for one site in Akap12) underwent changes in response to PKA-Cα deletion but not PKA-Cβ deletion. Thus, the phosphorylation evidence suggests that PKA-Cα interacts more extensively with components of AKAP complexes than does PKA-Cβ, at least with regard to phosphorylation.

Table 8.

AKAPs and cAMP-associated proteins with altered phosphorylation in either PKA-Cα or PKA-Cβ-null cells

UniProt ID	Gene Symbol	Annotation	Amino Acid Number	Centralized Sequence†	Log2(PKA-Cα Null/ PKA-Cα Intact)	Pmod Value	Log2(PKA-Cβ Null/ PKA-Cβ Intact)	Pmod Value
O08715	Akap1	A-kinase anchor protein 1, mitochondrial	101	TRQVRRRS*ESSGNLP	−0.41	0.003	0.00	0.966
Q9WTQ5	Akap12	A-kinase anchor protein 12	22	PAESDTPS*ELELSGH	0.55	0.035	−0.09	0.716
Q9WTQ5	Akap12	A-kinase anchor protein 12	583	AEAEEGAT*SDGEKKR	0.43	0.018	0.11	0.490
Q9WTQ5	Akap12	A-kinase anchor protein 12	584	EAEEGATS*DGEKKRE	0.64	0.008	−0.02	0.908
Q9WTQ5	Akap12	A-kinase anchor protein 12	683	KRARKASS*SDDEGGP	0.51	0.011	−0.11	0.514
Q9WTQ5	Akap12	A-kinase anchor protein 12	684	RARKASSS*DDEGGPR	0.52	0.019	−0.20	0.305
Q9WTQ5	Akap12	A-kinase anchor protein 12	1292	SNEEQSIS*PEKREMG	0.40	0.061	0.49	0.026
Q3UEI1	Pde4c	cAMP-specific 3′,5′-cyclic phosphodiesterase 4C	301	ELRRSSHT*SLPTAAI	−0.43	0.034	0.26	0.173
P70453	Pde7a	High-affinity cAMP-specific 3′,5′-cyclic phosphodiesterase 7A	58	FETERRGS*HPYIDFR	−0.44	0.020	−0.40	0.031
Q5EG47	Prkaa1	5′-AMP-activated protein kinase catalytic subunit α1	496	ATPQRSGS*ISNYRSC	−0.41	0.009	−0.13	0.347
Q9DBC7	Prkar1a	cAMP-dependent protein kinase type I-α regulatory subunit	83	DSREDEIS*PPPPNPV	−1.01	3.99 × 10−4	0.05	0.816
P31324	Prkar2b	cAMP-dependent protein kinase type II-β regulatory subunit	112	NRFTRRAS*VCAEAYN	−1.38	0.001	−0.29	0.402

P_mod value, moderated P value.

^*Phosphorylated amino acids. R or K in positions −3 and −2 are underlined.

An important role of vasopressin signaling in renal collecting duct cells is transcriptional regulation, particularly regulation of transcription of the Aqp2 gene. Table 9 shows transcription factors with altered phosphorylation in PKA-Cα-null or PKA-Cβ-null cells. Among all transcription factors, 241 phosphopeptides were quantified. There were 14 phosphopeptides that were substantially changed in abundance in 13 different transcription factor proteins. Most of the altered sites had proline in position +1 relative to the phosphorylated S or T, signifying altered phosphorylation by protein kinases in the cyclin-dependent kinase or MAPK families. Only one transcription factor underwent differential phosphorylation at a site with the (R/K)-(R/K)-X-p(S/T) motif, namely, “cAMP-dependent transcription factor ATF-7,” which showed a marked decrease in phosphorylation in PKA-Cα-null but not PKA-Cβ-null cells. Atf7 is a b-ZIP transcription factor that is inactive as a homodimer but can transactivate genes when heterodimerized with members of the activator protein-1 family, including Jund (7), which underwent a decrease in phosphorylation. Beyond the transcription factors, various transcriptional coregulators may participate in vasopressin-mediated regulation of transcription. One such protein is β-catenin, which shows increased phosphorylation at S552, a PKA target site (10), in response to vasopressin. Interestingly, S552 of β-catenin showed a marked decrease in phosphorylation in PKA-Cα-null cells but not PKA-Cβ-null cells (Table 3). In general, as with other categories of proteins, changes in phosphorylation of transcription factors and their coregulators are different in PKA-Cα-null or PKA-Cβ-null cells, again supporting the conclusion that the two PKA catalytic subunits fulfill different regulatory functions.

Table 9.

Phosphopeptides from transcription factor proteins that underwent changes in abundance with PKA-Cα and/or PKA-Cβ deletion

UniProt ID	Gene Symbol	Annotation	Amino Acid Number	Centralized Sequence	Log2 (PKA-Cα Null/ PKA-Cα Intact)	Pmod Value	Log2 (PKA-Cβ Null/ PKA-Cβ Intact)	Pmod Value	Transcription Factor Family
Q3US59	Atf7	cAMP-dependent transcription factor ATF-7	326	TGGRRRRT*VDEDPDE	−1.19	0.039	−0.60	0.266	bZIP
Q61103	Dpf2	Zinc finger protein ubi-d4	142	DPRVDDDS*LGEFPVS	−0.44	0.061	−0.59	0.017	zf-C2H2
Q8K5C0	Grhl2	Grainyhead-like protein 2 homolog	2	______MS*QESDNNK	0.42	0.038	−0.11	0.539	CP2
P15066	Jund (or Jun)	Transcription factor jun-D (or jun)	100	LGLLKLAS*PELERLI	−0.69	0.002	0.07	0.720	bZIP
Q2VPU4	Mlxip	MLX-interacting protein	9	AADVFMCS*PRRPRSR	−0.03	0.897	−0.68	0.010	bHLH
P40645	Sox6	Transcription factor SOX-6	439	KTAEPVKS*PTSPTQN	1.40	0.002	0.13	0.708	HMG
P40645	Sox6	Transcription factor SOX-6	454	LFPASKTS*PVNLPNK	1.19	0.007	−0.10	0.783	HMG
Q9WTN3	Srebf1	Sterol regulatory element-binding protein 1	96	KVTPAPLS*PPPSAPA	−0.80	0.024	0.22	0.500	bHLH
Q9JM73	Srf	Serum response factor	220	TCLNSPDS*PPRSDPT	0.41	0.007	−0.05	0.710	SRF
P63058	Thra	Thyroid hormone receptor α	12	PSKVECGS*DPEENSA	0.61	0.006	−0.24	0.208	TH
P62960	Ybx1	Nuclease-sensitive element-binding protein 1	174	EKNEGSES*APEGQAQ	−0.47	0.020	−0.05	0.792	CSD
Q3UQ17	Zbtb16	MCG3834	307	EESGEQLS*PPVEAGQ	−0.44	0.236	−1.04	0.012	ZBTB
Q8BIQ6	Zfp947	MCG23335	283	FTQKSSLS*IHQMYHT	0.50	0.044	0.07	0.768	zf-C2H2
O35615	Zfpm1	Zinc finger protein ZFPM1	681	RGSEGSQS*PGSSVDD	0.68	0.030	−0.06	0.843	zf-C2H2

^*Phosphorylated amino acids. R or K in positions −3 and −2 are underlined.

DISCUSSION

Based on the observations in this report, we conclude that PKA-Cα and PKA-Cβ do not have redundant regulatory functions. Indeed, while some overlap exists between the two in terms of phosphorylation targets, large differences were seen at a whole phosphoproteome level. PKA-Cα-null cells showed decreased phosphorylation dominated by sites with the classical PKA motif, viz. (R/K)-(R/K)-X-p(S/T), whereas only very few of the decreased phosphorylation sites in PKA-Cβ-null cells contained this motif. However, in vitro incubation with recombinant PKA-Cα and PKA-Cβ resulted in phosphorylation of virtually identical sites, with a predominance of the (R/K)-(R/K)-X-p(S/T) motif. A key question is “If the catalytic regions of PKA-Cα and PKA-Cβ have nearly identical in vitro target specificities, why is there such a difference in phosphorylation targets in the intact cells?”. The question was addressed by the bioinformatics analysis of phosphorylation targets, showing that completely different GO cellular component terms are associated with the two sets of phosphorylation targets. Specifically, PKA-Cα targets were largely related to cell membranes and membrane vesicles, whereas PKA-Cβ targets were related to the actin cytoskeleton and cell junctions, indicating that PKA-Cα and PKA-Cβ interact with different sets of proteins in cells. This difference could arise from differences in the noncatalytic regions of the two kinases that result in a different set of local interactions with potential substrates.

One possibility is that the two catalytic subunits could interact with different AKAPs and/or PKA regulatory subunits (16). Phosphoproteomic results point to a preferential association of PKA-Cα with Prkar2b, which showed a marked decrease at S112 in PKA-Cα null cells but not PKA-Cβ null cells (Table 8). The second possibility is that there could be differential interactions with so-called C-kinase anchoring proteins (26). The third possibility is that there may be a difference in protein-protein interactions with PDZ domain-containing proteins (2). PKA-Cα contains a class 1 COOH-terminal PDZ-ligand motif (–TxF), whereas PKA-Cβ ends with –CxF, which does not conform to a PDZ ligand motif. In a previous study (4), we showed that vasopressin treatment of native inner medullary collecting duct cells results in increased phosphorylation of several PDZ domain-containing proteins [connector enhancer of kinase suppressor of Ras 3 (Cnksr3), PDZ domain-containing protein GIPC1 (Gipc1), multiple-PDZ domain protein (Mpdz), and PDZ and LIM domain protein-5 (Pdlim5)]. That study showed that proteins with class I COOH-terminal three-amino acid motifs are more likely to show increases in phosphorylation in response to vasopressin than proteins without the motif.

In addition to phosphorylation differences, PKA-Cα and PKA-Cβ deletions resulted in many differences in what proteins underwent changes in total protein abundances. Thus, in terms of regulation of protein abundances, PKA-Cα and PKA-Cβ proteins appear to have nonredundant regulatory functions. For example, AQP2 was markedly decreased in PKA-Cα-null cells but not in PKA-Cβ-null cells, despite similar expression levels of the two (10). This result suggests that PKA-Cα is the predominant catalytic subunit responsible for the regulation of AQP2 abundance. Similarly, C3 was markedly decreased in PKA-Cα-null cells but not in PKA-Cβ-null cells. In contrast, the abundance of Muc4 was substantially decreased in PKA-Cβ-null cells but not in PKA-Cα-null cells. A prior study in mpkCCD cells has shown that, in PKA dKO cells, AQP2, C3, and Muc4 mRNA and protein levels were markedly decreased (10).

Data resource.

In addition to the scientific findings highlighted above, this report provides added value in the form of two web resources that allow users to interrogate data from this paper. These resources have been included with other phosphoproteomic data on the Kidney Systems Biology Project website (https://hpcwebapps.cit.nih.gov/ESBL/Database/).

Limitations and future directions.

One limitation of this study is that the use of trypsin, in some cases, produces peptides that are too short to be quantified by mass spectrometry, causing certain important phosphorylation sites to be missed in the analysis. For example, the pKID domain of Creb1 contains a classic PKA phosphorylation site at S133 within the sequence LSRRPS*YRKILNDLS, which, when proteolyzed with trypsin, gives RPS*YR, which is too short to see by mass spectrometry. Future studies are needed using other proteases to allow more comprehensive analysis. Another limitation of the present study is that it was done exclusively in a cultured cell model of the collecting duct. Future studies are needed to assess the specific roles of the two PKA catalytic proteins in the native collecting duct.

GRANTS

This work was funded by National Heart, Lung, and Blood Institute (NHLBI) Projects ZIAHL001285 and ZIAHL006129 (to M.A.K.). K. Salhadar was a member of the Biomedical Engineering Student Internship Program (Robert Lutz, Director) supported by the National Institute for Biomedical Imaging and Bioengineering (June–August 2018). Mass spectrometry used the NHLBI Proteomics Core Facility (M. Gucek, Director).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

V.R. and M.A.K. conceived and designed research; V.R., K.S. and C.Y. performed experiments; V.R., K.S., K.L., C.Y., and M.A.K. analyzed data; V.R., K.S., K.L., E.P., C.Y., and M.A.K. interpreted results of experiments; V.R., K.S., and M.A.K. prepared figures; V.R. and M.A.K. drafted manuscript; V.R., K.S., K.L., E.P., C.Y., and M.A.K. edited and revised manuscript; V.R., K.S., K.L., E.P., C.Y., and M.A.K. approved final version of manuscript.

ENDNOTES

Supplemental data are deposited at https://hpcwebapps.cit.nih.gov/ESBL/Database/Supplemental_Data_PKA_sKO/.

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the data set Accession No. PXD015050 (https://www.ebi.ac.uk/pride/).