Bicarbonate is essential for protein-tyrosine phosphatase 1B (PTP1B) oxidation and cellular signaling through EGF-triggered phosphorylation cascades

Abstract

Protein-tyrosine phosphatases (PTPs) counteract protein tyrosine phosphorylation and cooperate with receptor-tyrosine kinases in the regulation of cell signaling. PTPs need to undergo oxidative inhibition for activation of cellular cascades of protein-tyrosine kinase phosphorylation following growth factor stimulation. It has remained enigmatic how such oxidation can occur in the presence of potent cellular reducing systems. Here, using in vitro biochemical assays with purified, recombinant protein, along with experiments in the adenocarcinoma cell line A431, we discovered that bicarbonate, which reacts with H₂O₂ to form the more reactive peroxymonocarbonate, potently facilitates H₂O₂-mediated PTP1B inactivation in the presence of thioredoxin reductase 1 (TrxR1), thioredoxin 1 (Trx1), and peroxiredoxin 2 (Prx2) together with NADPH. The cellular experiments revealed that intracellular bicarbonate proportionally dictates total protein phosphotyrosine levels obtained after stimulation with epidermal growth factor (EGF) and that bicarbonate levels directly correlate with the extent of PTP1B oxidation. In fact, EGF-induced cellular oxidation of PTP1B was completely dependent on the presence of bicarbonate. These results provide a plausible mechanism for PTP inactivation during cell signaling and explain long-standing observations that growth factor responses and protein phosphorylation cascades are intimately linked to the cellular acid-base balance.

Keywords: bicarbonate; epidermal growth factor receptor (EGFR); oxidative inactivation; peroxiredoxin; peroxymonocarbonate; phosphatase; phosphorylation cascade; protein-tyrosine phosphatase (PTP); redox regulation; redox signaling; thioredoxin reductase.

Figure 1.

Bicarbonate potentiates H₂O₂-dependent inactivation of PTP1B. A, recombinant PTP1B (600 nm) was treated with increasing concentrations of H₂O₂, 0 μm buffer control (●), 3.4 μm (○), 6.3 μm (♦), 12.5 μm (▾) 25 μm (▴), 50 μm (■) (20 mm HEPES, 100 mm NaCl buffer, pH 7.4, containing 0.1 mm diethylenetriaminepentaacetic acid, 0.05% BSA, 1 mm sodium azide) and then assayed for PTP activity at the indicated times. PTP1B activity is given in min⁻¹ (mol of product/mol of enzyme/min). Data points represent means ± S.D. (error bars) (n = 3). B, PTP1B was treated as in A but in the presence of 25 mm bicarbonate (final concentration after adding H₂O₂/bicarbonate and assay buffer). Data points represent means ± S.D. (n = 4). A representative run is presented in Fig. S1.

Figure 2.

Modulation of bicarbonate- and H₂O₂-dependent inactivation of PTP1B in the presence of a Prx2/TrxR1/Trx1 redox system. A, scheme showing relevant reactions in the experimental PTP1B/Prx2/thioredoxin system. S_ox of PTP1B could be either the sulfenic acid or sulfenylamide species, with the former also reducible by TrxR/NADPH in the absence of Trx. B, PTP1B (600 nm) was treated with 50 μm H₂O₂ together with 25 mm bicarbonate in the presence of Trx1 (5 μm), Prx2 (10 μm), NADPH (300 μm), and increasing TrxR1 concentrations of 0 nm (■), 2 nm (▴), 10 nm (▾), 50 nm (♦), 250 nm (○), and 1 μm (□) and with 1 μm TrxR1 using buffer-treated control without H₂O₂ (●). PTP activity was measured after the indicated times. Data points represent means ± S.D. (error bars) (n = 3). C, analyses were performed as in B with TrxR1 (1 μm) and increasing Trx1 concentration, 0 μm (♦), 0.5 μm (▾), 1 μm (▴), 2 μm (■), and 5 μm (●) (n = 1). D, analyses performed as in A with TrxR1 (15 nm) and increasing concentrations of bicarbonate as indicated, 0 mm (■), 5 mm (▴), 10 mm (▾), 15 mm (♦), 25 mm (○), 35 mm (□), and 45 mm (▵) (n = 1).

Figure 3.

Pharmacological inhibition of bicarbonate cotransporters in A431 cells decreases EGF ligand-induced phosphotyrosine formation. Overnight starved A431 cells in regular bicarbonate-containing DMEM 0.1% FCS) were pretreated with the NBC inhibitor S0859 (100 μm) or DMSO control for 1 h prior to EGF stimulation for 2, 4, and 6 min. Cell extracts were separated by SDS-PAGE and Western blotted (IB) for total phosphotyrosine and EGF receptor (A) and specific phosphorylation of the EGF receptor Tyr-992 (pY992) residue (C). Bar diagrams show densitometry ratios for total signal over all phosphotyrosine bands (B) or the specific pTyr-992 signal (D), relative to the immunoblot staining for the EGF receptor. Every ratio is compared with that in unstimulated control, which was set to 1. Results are mean ± S.D. (error bars) for three independent experiments with filled circles showing individual results. Statistically significant differences are indicated (*, p < 0.05). The membranes in the figure stained with Ponceau for total protein loading are shown in Fig. S3.

Figure 4.

Treatment of A431 cells with lactic acid prior to EGF ligand stimulation results in less phosphorylation. A, concentration dependence. Serum-starved A431 cells in regular DMEM containing bicarbonate were pretreated for 2 min with 0, 5, 10, 20, and 40 mm lactic acid and subsequently stimulated with EGF (100 ng/ml) for 5 min. Lysates were analyzed for total phosphotyrosine. B, time course. A431 cells were pretreated for 2 min with 0, 15, and 30 mm lactic acid and subsequently stimulated with EGF (100 ng/ml) for 2, 4, and 6 min. Lysates were analyzed for total phosphotyrosine. The right-hand panels in A and B show densitometry analyses as in Fig. 3 with filled circles showing individual results (n = 3; mean ± S.D. (error bars); *, p < 0.05). The membranes in this figure stained with Ponceau for total protein loading are shown in Fig. S4.

Figure 5.

Bicarbonate increases total phosphorylation in A431 cells after EGF ligand stimulation. A, A431 cells were grown and incubated overnight in low-serum HEPES-buffered (50 mm) DMEM at pH 7.4 with 0, 40, and 60 mm bicarbonate added and equilibrated with 0, 5, and 10% CO₂, respectively, as indicated. The cells were then stimulated with EGF and analyzed for total phosphotyrosine. B, densitometry for three independent experiments, of short- and long-exposed membranes, for high-molecular weight (HMW) phosphorylated proteins, and ∼28 kDa band, respectively (dashed rectangles in A) quantified in relation to Ponceau staining with 2-, 4-, and 6-min time points combined (n = 3; mean ± S.D. (error bars); *, p < 0.05). Symbols indicate individual results. IB, immunoblotting.

Figure 6.

A, bicarbonate is required for EGF-dependent reversible PTP1B oxidation in A431 cells. A431 cells were incubated overnight either in low-serum regular DMEM-containing bicarbonate or in HEPES-buffered DMEM (50 mm, pH 7.4). Pretreatment of the cells in bicarbonate-containing DMEM with the NBC inhibitor S0859 (50 μm) was for 1 h prior to stimulation. Bicarbonate (60 mm) was added to cells in HEPES-buffered DMEM and subsequently stimulated. At the indicated times after EGF stimulation, cells were subjected to the cysteinyl-labeling assay using biotinylated iodoacetyl-PEG2-biotin for analysis of reversible PTP1B oxidation. Biotinylated proteins were purified on streptavidin-Sepharose beads, resolved by SDS-PAGE, and visualized using antibodies against PTP1B. PTP1B control levels were determined from total cell lysate by SDS-PAGE and blotting (IB) against PTP1B. Representative Western blotting of three independent experiments is shown. B, model for regulation of PTP1B activity during growth factor signaling. EGFR activation induces a transient burst of H₂O₂, which reacts with bicarbonate to give PTP1B oxidation via peroxymonocarbonate (red) and activation of phosphorylation pathways (green arrow). Prxs compete for the H₂O₂, and the Trx system decreases the availability of H₂O₂ by supporting the Prx cycle and also acts by reactivating oxidized PTP1B. See “Discussion” for further details.