NaCl, but not isosmotic sorbitol, reversibly reduces AGC kinase activation-loop phosphorylation in diverse cultured cells.

Cells were incubated for 10 min in DMEM supplemented with NaCl or sorbitol at the indicated concentrations (mM). Whole-cell lysates were analyzed by immunoblotting with phospho-specific antibodies to activation-loop sites together with total proteins where indicated; membranes were stained with Coomassie Brilliant Blue (CBB) after immunoblotting as a loading control. GAPDH was included as an additional loading control where specified. (A) Dose–response immunoblots showing NaCl-induced decrease in activation-loop phosphorylation in PKN (PKN1/2/3), atypical PKC (PKCζ/λ), and p70 S6 kinase (p70/p85), whereas sorbitol had little or no effect under the same osmolarities. Total PKCλ, total p70/p85, and GAPDH are shown together with CBB. (B) The NaCl-induced decrease in activation-loop phosphorylation in PKN and PKCζ/λ in multiple cell types, including HEK293, HeLa, SH-SY5Y, and primary wild-type mouse embryonic fibroblasts (MEFs). (C) Reversible decrease in phosphorylation levels at the PKN activation loop. COS7 cells were treated with 600 mM NaCl for the indicated durations (upper panel), or incubated for 10 min with 600 mM NaCl or 1200 mM sorbitol followed by washout into DMEM for the indicated recovery times (lower panels). NaCl caused a reversible decrease in activation-loop phosphorylation of PKN, whereas sorbitol did not. Phospho-p38 (P-p38) and total p38 are shown as markers of hypertonic-stress signaling. Representative immunoblots are shown.

Alkali metal cations reduce activation-loop phosphorylation of PKN (and atypical PKC) and the effect is reproduced in cell lysates.

COS7 cells (or lysates, as indicated) were incubated for 10 min with the indicated salts at the indicated concentrations (mM). Whole-cell lysates were analyzed by immunoblotting with phospho-specific antibodies against activation-loop sites together with the corresponding total proteins where shown. Membranes were stained with Coomassie Brilliant Blue (CBB) after immunoblotting as a loading control; PKCλ and PKN1 are included as total-protein references where indicated. Representative immunoblots are shown. (A) Effect of alkali metal ions on the phosphorylation levels in the activation loops of PKN and PKCζ/λ. KCl, RbCl, and CsCl reduced activation-loop phosphorylation of PKN (PKN1/2/3) and PKCζ/λ in intact COS7 cells in a concentration-dependent manner as well as NaCl. (B) Effect of K+ salts on the activation-loop phosphorylation of PKN and PKCζ/λ. K+ reduced activation-loop phosphorylation regardless of the accompanying anion. (C) Decrease in activation-loop phosphorylation of PKN and PKCζ/λ in vitro. Whole-cell lysates were incubated with NaCl, KCl, or sorbitol; soluble cell extracts were incubated with LiCl, NaCl, KCl, CsCl and MgCl₂, or sorbitol, all for 10 min at the indicated mM.

Protein phosphatase PP2A mediates the decrease in PKN activation-loop phosphorylation induced by NaCl or KCl.

COS7 (or HeLa, as indicated) cells or lysates were treated with salts for 10 min at the indicated concentrations (mM). Whole-cell lysates were analyzed by immunoblotting with phospho-specific antibodies to activation-loop sites of PKN (PKN1/2/3) and PKCζ/λ, together with the corresponding total proteins where indicated. Membranes were stained with Coomassie Brilliant Blue (CBB) as a loading control. Representative immunoblots are shown. (A) Effect of okadaic acid on the decrease in activation-loop phosphorylation. COS7 cells were pre-incubated with DMSO or 250 nM okadaic acid for 3 h at 37°C (5% CO2). The cells were then transferred to DMEM containing NaCl at the indicated concentrations while okadaic acid (or DMSO) was kept at the same concentration during the entire 10 min NaCl treatment. (B) Effect of rubratoxin A on the decrease in activation-loop phosphorylation. COS7 soluble extracts were pre-treated with DMSO or 3 µM rubratoxin A for 10 min 30°C, after which they were incubated for an additional 10 min with NaCl or KCl at the indicated concentrations in the continued presence of rubratoxin A (or DMSO). (C) Effect of PP2A knockdown on the activation-loop phosphorylation. HeLa cells were transfected with 10 nM control siRNA or siRNA targeting the catalytic subunit of PP2A, and then cultured for 48 h at 37°C (5% CO2). After incubation, these cells were treated with DMEM supplemented with NaCl or KCl at the indicated concentrations for 10 min. Immunoblot analysis of the HeLa cell lysates was performed using antibodies against the total and phosphorylated forms of PKN and PP2A catalytic subunit. (D) Effects of NaCl and KCl on PP2A phosphatase activity in vitro. PP2A phosphatase activity was measured using radiolabeled phospho-casein as a substrate. Data are mean ± SEM. Data were analyzed using an unpaired t-test (n = 3). **p < 0.01. The input radioactivity of the phospho-casein was 21,324.49 cpm.

Recovery of PKN activation-loop phosphorylation after normalization of K⁺ concentration does not require PDK1, ATP or Mg²⁺.

Whole-cell lysates were analyzed by immunoblotting with antibodies to phospho-PKN (activation loop) and total PKN; membranes were stained with Coomassie Brilliant Blue (CBB) as a loading control. Representative immunoblots are shown. “Con,” “KCl,” and “Re” denote the control, high-K⁺ treatment, and recovery in cell-based experiments, achieved by replacing the high-K⁺ medium with the original medium containing normal K⁺. In in-lysate experiments, “Diluted” indicates a 1:1 dilution into lysis buffer to lower ionic strength. (A) PI3K–PDK1 pathway inhibition with wortmannin. PDK1flox/flox MEFs were pre-incubated with DMSO or 2 μM wortmannin for 3 h at 37°C (5% CO₂), then treated with 150 mM KCl for 10 min, followed by 5 min in normal-K⁺ medium. Wortmannin (or DMSO) was present throughout both phases. (B) PDK1 ablation by Cre and in-lysate reconstitution. PDK1flox/flox MEFs were infected with control adenovirus (PDK1 present) or adenovirus expressing Cre recombinase (PDK1 knockout). After 48 h, cells were treated with 150 mM KCl for 10 min, then incubated in normal-K⁺ medium for 5 min (left panel). For the in-lysate assays (right panel), GST-tagged full-length PKN1 purified from Sf9 cells was incubated with soluble extracts from control virus–infected MEFs (PDK1 present) or from Cre-infected MEFs (PDK1 knockout) in the presence of 150 mM KCl for 10 min, then diluted 1:1 with lysis buffer prior to analysis. “End” indicates endogenous; “Diluted” marks the dilution step. (C) Broad-spectrum kinase inhibition with staurosporine. COS7 cells were treated with 150 mM KCl for 10 min in the presence of DMSO or 1 μM staurosporine, followed by 5 min in normal-K⁺ medium. Staurosporine (or DMSO) was maintained throughout. “STS” indicates staurosporine. (D) ATP- and Mg²⁺ -independent recovery of PKN activation-loop phosphorylation. COS7 soluble extracts were pre-incubated with 1.5 μM apyrase (recombinant, E. coli) with or without 1 mM EDTA for 1 h at 30°C to deplete endogenous ATP, then exposed to 150 mM KCl for 10 min; apyrase (and EDTA, when present) was maintained during the KCl incubation. Samples were then diluted 1:1 with lysis buffer prior to immunoblotting for phospho- and total PKN; CBB staining served as a loading control. Asterisks (*) denote non-specific signals observed in extracts that lacked both EDTA and apyrase.

KCl-induced modulation of activation-loop phosphorylation is preserved in a minimal PKN1 fragment (aa 767–780).

“Re” denotes the recovery phase in cell-based assays, achieved by replacing the high-K⁺ medium with the original medium containing normal K⁺. In lysate assays, “Diluted” indicates a 1:1 dilution into lysis buffer to reduce ionic strength. Blots were probed with antibodies to phospho-PKN (activation loop) and FLAG or GST where indicated; membranes were stained with CBB as a loading control. Representative immunoblots are shown. (A) Schematic of PKN1 and deletion constructs. Domain organization of human PKN1 with the activation loop indicated; the aa 767– 780 segment and Thr774 are marked. Sequence alignment around this region for PKN1/2/3 is shown (conserved Thr in the activation loop). (B) PKN1 deletion mutants expressed in COS7 cells. COS7 cells were transfected with plasmids encoding FLAG-tagged PKN1 mutants (aa 611–942, 637–942, 672–942, 611–928, and 689–942). After 48 h, cells were treated with 150 mM KCl for 10 min, then switched to normal-K⁺ medium for 5 min (“Re”). (C) Minimal fragments retain KCl responsiveness in lysates. GST-PKN1 721–871 and 721–819 were phosphorylated in vitro with PDK1 (3 h). GST-PKN1 767–788 and 767–780 were prepared by co-expression with PDK1. Phosphorylated fragments were mixed with COS7 cell extracts and incubated with or without 150 mM KCl for 10 min; “Diluted” samples were obtained by 1:1 mixing with lysis buffer before analysis.

The phosphate initially used to modify the PKN1 activation loop is reused (recycled) during recovery.

Unless otherwise noted, “Con,” “KCl,” and “Diluted” indicate the control condition, incubation with 150 mM KCl for 10 min, and subsequent 1:1 dilution into 1× lysis buffer to lower ionic strength, respectively. Immunoblots were probed with antibodies against phospho-PKN (activation loop) and GST where indicated; membranes were stained with CBB as a loading control. Representative immunoblots/autoradiographs are shown where applicable. (A) Working models for recovery. Three mechanistic models are considered: (i) intramolecular phosphate transfer within PKN; (ii) transfer from an external phosphate donor (factor D); and (iii) transient transfer to an intermediate (factor X) followed by return to PKN. Yellow “P” indicates the phosphate initially attached to the PKN activation loop; green “P” denotes a phosphate derived from factor D. (B) Detection of phosphorylated PKN1 using Phos-tag SDS-PAGE. COS7 cell extracts were incubated with phosphorylated GST–PKN1 (aa 767–788) in the presence of 150 mM KCl for 10 min; “Diluted” samples were generated by 1:1 mixing with 1× lysis buffer prior to analysis by Phos-tag SDS–PAGE and immunoblotting for phospho-PKN and GST. Bands labeled “P / non-P” indicate phosphorylated/non-phosphorylated species. (C) Detection of radiolabeled phospho-PKN1 by autoradiography. COS7 cell extracts were incubated with 32P-labeled phospho–GST–PKN1 (aa 767–788) under the same KCl and dilution conditions as in (B), followed by autoradiography to detect 32P– PKN1. P/C (positive control): radiolabeled phospho–GST–PKN1 alone; N/C (negative control): COS7 extract alone; PM: protein marker. (D) Recovery requires pre-existing activation-loop phosphorylation. GST–PKN1 (aa 767–788) purified from E. coli was preincubated with or without PDK1, then exposed to COS7 soluble extracts + 150 mM KCl for 10 min, followed by 1:1 dilution as above. Samples were immunoblotted for phospho-PKN and GST. (Only PDK1-pretreated fragments support recovery.) (E) Selectivity of PKN1 fragments undergoing recovery of activation-loop phosphorylation. COS7 extracts containing phospho–GST–PKN1 (aa 767–788) were mixed with non-phosphorylated GST–PKN1 (aa 611–819) or GST–PKN1 (aa 611–800) and processed as in (B). (F) Isoform selectivity. COS7 extracts containing phospho–GST–PKN1 (aa 767–788) were incubated with non-phosphorylated GST–PKN2 (aa 653–949) and analyzed as in (B).

Effect of varying NaCl concentrations on activation-loop phosphorylation of PKN and PKCζ/λ in COS7 cells.

COS7 cells were incubated for 10 min in a HEPES-based medium (4 mM KCl, 2.5 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES) supplemented with NaCl or sorbitol at the indicated concentrations (mM). Cell extracts were immunoblotted with antibodies against phospho-PKN (activation loop), total PKN, phospho-PKCζ/λ, and total PKCλ; membranes were stained with Coomassie Brilliant Blue (CBB) as a loading control. Representative immunoblots are shown.

KCl-induced decrease in PKN phosphorylation in insect-cell extracts (Sf9) but not in E. coli extracts.

Extracts from Sf9 cells and from E. coli strains (BL21 and XL1-Blue) containing phospho-GST–PKN1 (aa 767–788) were incubated with 600 mM NaCl or 150 mM KCl for 10 min. “Diluted” samples were prepared by 1:1 mixing with 1× lysis buffer (50 mM Tris-HCl pH 7.5, 1 mM EGTA, 0.5 mM DTT, 0.1% Triton X-100). Immunoblots were probed for phospho-PKN and GST. Mmembranes were stained with Coomassie Brilliant Blue (CBB) after immunoblotting as a loading control. Representative immunoblots are shown.

Effects of membrane potential and osmotic pressure on PKN activation-loop phosphorylation.

Cell extracts were immunoblotted for phospho-PKN (activation loop) and total PKN; Coomassie Brilliant Blue (CBB) staining of the membrane after immunoblotting was used as a loading control. Representative blots are shown. (A) Ca²⁺ independence. COS7 and HEK293 cells were incubated for 10 min in a HEPES-based medium (10 mM HEPES, 1 mM MgCl₂) supplemented with NaCl, KCl, and CaCl₂ at the indicated mM. (B) Osmolality independence. COS7 and HeLa cells were incubated for 10 min in a HEPES-based medium (10 mM HEPES, 2.5 mM CaCl₂, 1 mM MgCl₂) supplemented with NaCl, KCl, or sorbitol at the indicated mM.

Autophosphorylation of PKN1 and recovery of activation-loop phosphorylation.

Immunoblot analysis was performed using antibodies against the total and phosphorylated forms of PKN, GST and PDK1. Coomassie Brilliant Blue (CBB) staining of the membrane after immunoblotting was used as a loading control. (A) Autophosphorylation in vitro. Full-length GST–PKN1 purified from Sf9 and non-phosphorylated GST–PKN1 (aa 543–942, aa 760–873) purified from E. coli were incubated in 10 mM Tris-HCl (pH 7.5), 5 mM MgCl₂, 30 μM ATP at 30°C or 90 min. (B) Recovery independent of PKN1 catalytic activity in cells. Immortalized PKN1[T778A] MEFs were transfected with empty vector or PKN1 constructs (WT, K644E, S916A), treated with 150 mM KCl for 10 min, then switched to normal-K⁺ medium for 5 min to induce recovery. (C) His-Iα does not block recovery in lysates. Soluble extracts from PDK1flox/flox MEFs ± PDK1 were incubated with full-length GST–PKN1 and His-Iα in 150 mM KCl for 10 min. “Diluted” samples were prepared by 1:1 mixing with 1× lysis buffer (50 mM Tris-HCl pH 7.5, 1 mM EGTA, 0.5 mM DTT, 0.1% Triton X-100).),

Effect of affinity tag on the recovery of PKN1 activation-loop phosphorylation.

COS7 cell extracts were incubated with KCl and the FLAG and (His)6 - tagged proteins; “Diluted” samples were prepared by 1:1 mixing with 1× lysis buffer (50 mM Tris-HCl pH 7.5, 1 mM EGTA, 0.5 mM DTT, 0.1% Triton X-100). Immunoblots were probed with anti–phospho-PKN (activation loop) and anti-FLAG. Membranes were stained with Coomassie Brilliant Blue (CBB) after immunoblotting as a loading control.

Preserved K⁺-dependent loss–recovery dynamics in activation-loop region point mutants.

(A) Representative immunoblots showing activation-loop phosphorylation of PKN1 (P-PKN) and Akt1 (P-Akt) in wild-type (WT) and point-mutant constructs under three conditions: Con (control), KCl (high-K⁺ treatment), and Diluted (post-dilution, i.e., ion reduction). For GST-PKN1 (aa 767–788), mutants include C776A (the conserved Cys at P+2 within the TFCGT motif) and T778A; for GST-Akt1 (aa 301–322), mutants include C310A (the corresponding conserved Cys) and T312A. In all cases, high K⁺ reduces the activation-loop phosphor-signal, which recovers upon ion reduction, similarly to WT. Molecular weight markers (kDa) are indicated. (B) Alignment of activation-loop sequences from PKN1, PKN2, PKN3, and Akt1. The arrow indicates the conserved cysteine (Cys776 in PKN1; Cys310 in Akt1), and the arrowhead marks the activation-loop threonine that, when phosphorylated, is detected by the phospho-specific antibodies used in (A) and (C). (C) Additional GST-PKN1 (aa 767–788) point mutants at activation-loop residues with side chains that could, in principle, form labile phosphate linkages were tested: D770A, R771A (two clones), E780A, alongside WT. As in (A), high K⁺ reduces the activation-loop phospho-signal, which returns after dilution, indicating that these substitutions do not prevent the loss-and-reacquisition behavior.

ATP-independent recovery of PKN1 activation-loop phosphorylation.COS7 extracts containing 32P-labeled phospho–GST–PKN1 (aa 767–788) were incubated ± 1 mM ATP with 150 mM KCl for 10 min.

“Diluted” samples were generated by 1:1 mixing with 1× lysis buffer (50 mM Tris-HCl pH 7.5, 1 mM EGTA, 0.5 mM DTT, 0.1% Triton X-100) prior to autoradiography. Representative autoradiograph is shown.

Absence of recovery with dephosphorylated PKN1 fragment.

Phospho–GST–PKN1 (aa 767–788) was treated with λ-phosphatase at 30° C for 1 h, then precipitated on GST affinity beads and washed three times with 1× lysis buffer (50 mM Tris-HCl pH 7.5, 1 mM EGTA, 0.5 mM DTT, 0.1% Triton X-100) to remove residual phosphatase. COS7 extracts containing either the phospho- or λ-phosphatase–dephosphorylated fragment were incubated with 150 mM KCl for 10 min; “Diluted” samples were prepared by 1:1 mixing with 1× lysis buffer. Immunoblots for phospho-PKN, GST. Membranes were stained with Coomassie Brilliant Blue (CBB) after immunoblotting as a loading control. Representative immunoblot is shown.