Sign in
Systems BiologybioRxiv
Preprint

Kinome-wide CRISPR/Cas9-knockout screening reveals critical protein kinases in vasopressin V2-receptor signaling

Euijung ParkiD, Lihe CheniD, Viswanathan Raghuram, Shaza Khan, Adrián Rafael Murillo‐de‐OzoresiD, Chung‐Lin ChouiD, Chin‐Rang YangiD, Mark A. KnepperiD

Abstract

Identification of signaling networks is an essential goal in systems biology. Here, we use CRISPR/Cas9 knockout screening (employing a whole kinome sgRNA library) to identify functionally critical protein kinases in a well-studied Gs-dependent G-protein coupled receptor (GPCR)-signaling model, namely the vasopressin V2 receptor (V2R) pathway. Screening was done using a specially-designed fluorescence-based reporter cell line with green-fluorescent protein (GFP) co-transcribed with Aqp2, a gene whose transcription is dependent on vasopressin-mediated activation of protein kinase A (PKA). Positive regulators (n=14) included PKA-catalytic subunit (Prkaca) and Dyrk1a (minibrain homolog). Negative regulators (n=12) included PKA-regulatory subunit type I, Stk11 (catalytic subunit of liver kinase B1 [LKB1] complex), and three TGF-{beta} receptor subunits (Tgfbr1, Tgfbr2, Tgfbr3) (see https://esbl.nhlbi.nih.gov/Databases/Kinome-CRISPR-screen/ for full list). Dyrk1a knockout cell lines failed to express AQP2 protein and exhibited a profound decrease in AQP2 mRNA. RNA-sequencing demonstrated widespread increases in cell-cycle transcripts, with a general defect in cell differentiation, accounting for AQP2 loss. TGF-{beta} exposure to un-transformed cells results in a profound decrease in V2R and AQP2 mRNA abundance along with multiple additional transcriptional targets of V2R-PKA signaling, consistent with prior findings in TGF-{beta}-mediated vasopressin escape. Stk11/LKB1 knockout lines displayed marked increases in AQP2 protein and mRNA, even in the absence of vasopressin. RNA-sequencing showed a marked similarity between the responses to Stk11/LKB1 deletion and vasopressin exposure in untransformed cells. Phospho-proteomic data point to opposing roles of Stk11/LKB1 and PKA in the regulation of cAMP-responsive transcriptional coactivator (CRTC) proteins in the transcriptional response to V2R-PKA signaling. Significance StatementCells throughout the body are regulated by extracellular signals like the hormone, vasopressin. Hormonal effects on cellular function are mediated by membrane receptors that trigger biochemical changes, often by inducing chemical modification of the amino acids making up individual proteins, such as addition of function-altering phosphate groups (phosphorylation). Protein phosphorylation is mediated by enzymes known as "protein kinases". Here, we have screened all known protein kinases using modern CRISPR/Cas9 technology to identify those involved in vasopressin action in the kidney. As expected from prior knowledge, the screen identified protein kinase A and one of its regulatory subunits, but also identified several protein kinases not previously implicated in vasopressin action in the kidney.

PDFDOIData
Full-text digest ↓

02 /Specification & Classification

Click any tag to browse matching papers

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Sonnet 4.6 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (6 Jul 2026) · CC0 1.0

A kinome-wide CRISPR knockout screen in a mouse collecting duct Aqp2-GFP reporter line found 26 kinases shaping vasopressin-driven Aqp2 transcription, including Dyrk1a, Stk11 and TGF-β receptors.

Quantitative phosphoproteomics and PKA catalytic-subunit deletion showed phosphorylation changes at many non-PKA-type sites, implying unidentified kinases in vasopressin V2 receptor signaling. This work uses a whole-kinome knockout screen with an Aqp2 transcriptional reporter to find them.

Key results

  1. ShownThe screen recovered 26 kinase hits, with expected PKA biology as an internal control: Prkaca scored as a positive regulator and Prkar1a as a negative regulator, while Prkacb did not score. Hits also clustered in complexes and pathways.

    714 targets with 4 sgRNAs each; GFP-high group compared with the combined GFP-neg and GFP-low groups, using mean values from two independent replicates; 14 positive and 12 negative regulators. Clusters: Stk11/Strada; Tgfbr1, Tgfbr2, Tgfbr3; Map3k1, Map2k4, Map2k7. · Introduction; Results and Discussion; Prkaca and Prkar1a; Figure legends

  2. SuggestedDyrk1a knockout clones lose the vasopressin-induced rise in AQP2 protein. Their transcriptome shows lower Aqp2 and differentiation markers and higher cell-cycle genes, so the effect is likely broad dedifferentiation rather than specific V2R-PKA interference.

    4 independent Dyrk1a KO clones vs 4 WT; Aqp2 mRNA decrease above 90% (value -3.56, Dyrk1a-KO/WT) under 0.1 nM dDAVP; propidium iodide FACS showed a partial shift from G0/G1 to S and G2/M (mean ± SEM, P < 0.05). · Dyrk1a; Overall summary; Figure legends

  3. ShownTGF-β1 strongly suppresses Aqp2 and also lowers Avpr2 (V2 receptor) mRNA. Its transcriptional footprint opposes the vasopressin response and parallels PKA deletion, and it also reaches beyond V2R-PKA signaling (partial epithelial-to-mesenchymal transition-like terms).

    Aqp2 TPM 182±29 vs 1297±53 (n=4 each; P<10^-7), an 86% decrease; Avpr2 TPM 108±6 vs 235±6 (P<10^-6); 1089 transcripts changed with TGF-β vs 278 with dDAVP; Jun TPM 39.8±3.7 vs 129.0±11.2. · Tgfbr1, Tgfbr2, and Tgfbr3; Figure legends

  4. ShownStk11 (LKB1) deletion mimics vasopressin: AQP2 protein and Aqp2 mRNA rise without hormone, and the genome-wide transcript changes closely track the dDAVP response.

    Five independent Stk11 KO clones; RNA-seq 5 KO vs 5 WT without vasopressin; Aqp2 TPM 190+12 (WT) vs 5126+1751 (KO), FDR 0.000004; regression of KO/WT vs dDAVP/vehicle log2 ratios slope 0.957 over 9162 transcripts (P < 10^-8). · Stk11/LKB1; Figure legends

  5. SuggestedStk11 KO attenuates, but does not abolish, the vasopressin response. The authors propose a PKA-SIK-CRTC-CREB model, supported by mined prior phosphoproteomic data but not tested directly here.

    All 24 vasopressin-induced transcripts had smaller log2(dDAVP/vehicle) in Stk11 KO than in WT. Prior SILAC data (3 pairs) show PKA-site phosphorylation rising in SIK1, SIK2 and SIK3 and SIK-site phosphorylation falling in CRTC2 and CRTC3. · Stk11/LKB1; Figure legends

Methods, data and reproducibility

Pooled Brie mouse kinome sgRNA library transduced at MOI 0.1 into Cas9-expressing mpkCCD cells carrying a GFP-Aqp2 reporter; dDAVP-treated cells sorted into four GFP bins and sgRNAs compared with MAGeCK. Hits followed up with clonal KO lines, RNA-seq and TGF-β1 treatment.

  • Pooled kinome CRISPR KO screen in Cas9-expressing GFP-Aqp2 mpkCCD reporter cells (Brie library, 2852 unique gRNAs, 713 genes, four gRNAs per gene), two independent rounds · mouse (mpkCCD cortical collecting duct cell line) · n = 2 independent screening rounds
  • RNA-seq of Dyrk1a KO clones vs WT cells treated with 0.1 nM dDAVP for 48 h · mouse (mpkCCD cells) · n = 4 KO clones vs 4 WT · GSE326336
  • RNA-seq of mpkCCD cells treated with TGF-β1 (10 ng/ml, 48 h) vs vehicle, all in 0.1 nM dDAVP · mouse (mpkCCD cells) · n = n=4 vs n=4 · GSE326336
  • RNA-seq of Stk11 KO vs WT clones without vasopressin, and dDAVP vs vehicle responses in the same lines · mouse (mpkCCD cells) · n = 5 vs 5 · GSE326336
  • Previously published SILAC phosphoproteomics of dDAVP-treated mpkCCD cells (Datta et al.), re-mined for SIK and CRTC phosphosites · mouse (mpkCCD cells) · n = 3 SILAC pairs
Methods & tools
MAGeCK (via Galaxy Europe), FastQC, Cutadapt, Mouse Kinome CRISPR Knockout Library (Brie, Addgene), Illumina NovaSeq 6000, STAR 2.7.10 b, RSEM 1.3.1, edgeR 3.38.4, R, ImageJ, Gene Ontology and KEGG enrichment analysis, propidium iodide flow cytometry
Reagents & environment
STAR 2.7.10 b, RSEM 1.3.1, edgeR 3.38.4 on Ensembl genome 113; MAGeCK, FastQC and Cutadapt run on Galaxy Europe (tool versions not given).
Reproducibility gaps
No analysis code repository is given; statistics were done in Excel and R without shared scripts. Raw screen sgRNA count data and MAGeCK parameters are not clearly deposited.
Limitations (5)
  • CRISPR screens can yield off-target false positives (mitigated by 4 sgRNAs per target) and false negatives where kinases are redundant, since each cell receives only one sgRNA. · stated by the authors
  • Only a subset of the 26 hits was validated in focused studies; the majority is left for future work. · stated by the authors
  • The authors note that whether these findings generalize to other Gαs-coupled GPCRs is a long-term goal, so generalizability is not addressed here. · stated by the authors
  • The SIK-CRTC-CREB mechanism for Stk11 rests on literature and previously published phosphoproteomics; no SIK or CRTC perturbation was performed here. The Dyrk1a cell-cycle mechanism is inferred from RNA-seq and FACS without rescue. · noted in this summary
  • All work is in one mouse cell line (mpkCCD) using clonal knockouts with small replicate numbers; relevance in vivo remains untested. · noted in this summary

Who should read:Read this if you study vasopressin, cAMP-PKA or Aqp2 regulation, or want a worked example of a reporter-based kinome CRISPR screen with RNA-seq follow-up.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Similarity Map
SelectedSame fieldCross-field
THISDatta '20Daly '23Babicz '22Murillo‐de‐ '25Raghuram '20Heydenreich '21
72% semantic matchSystems Biology
Arnab Datta, Chin‐Rang Yang, Karim Salhadar+4 more·British Journal of Pharmacology·Dec 21, 2020· 22 citations