Saxagliptin but Not Sitagliptin Inhibits CaMKII and PKC via DPP9 Inhibition in Cardiomyocytes

Abstract

Some oral anti-hyperglycemic drugs, including gliptins that inhibit dipeptidyl peptidase 4 (DPP4), have been linked to the increased risk of heart failure (HF) in type-2 diabetic patients. While the cardiovascular safety trial, TECOS, revealed no link between sitagliptin and the risk of HF, a substantial 27% increase in the hospitalization for HF was observed in type-2 diabetic patients treated with saxagliptin within the SAVOR-TIMI 53 trial. A previous in vitro study revealed that saxagliptin impairs the Ca²⁺/calmodulin-dependent protein kinase II (CaMKII)-phospholamban (PLB)-sarcoplasmic reticulum Ca²⁺-ATPase 2a axis and protein kinase C (PKC) activity in cardiomyocytes leading to impaired cardiac contractility and electrophysiological function. However, the link between saxagliptin and its target proteins (CaMKII and PKC) remains to be explored. Since DPP8 and DPP9 (but not DPP4) are expressed by cardiomyocytes and saxagliptin is internalized by cardiomyocytes, we investigated whether DPP8/9 contribute to saxagliptin-mediated inhibition of CaMKII and PKC activity. Structural analysis revealed that the DPP4-saxagliptin interaction motif (S630, Y547) for the cyanopyrrolidine group is conserved in DPP8 (S755, Y669) and DPP9 (S730, Y644). Conversely, F357 that facilitates binding of the anchor lock domain of sitagliptin in the S2 extensive subsite of DPP4 is not conserved in DPP8/9. In parallel, unlike saxagliptin, sitagliptin did not affect phosphorylation of CaMKII/PLB or activity of PKC in HL-1 cardiomyocytes. These findings were recapitulated by pharmacological inhibition (TC-E-5007, a DPP8/9 antagonist) and knock-down of DPP9 (but not DPP8). In primary mouse ventricular cardiomyocytes, saxagliptin (but not sitagliptin) impaired Ca²⁺ transient relaxation and prolonged action potential duration (APD). These results suggest that saxagliptin-DPP9 interaction impairs the CaMKII-PLB and PKC signaling in cardiomyocytes. We reveal a novel and potential role of DPP9 in cardiac signaling. The interaction of saxagliptin with DPP9 may represent an underlying mechanism for the link between saxagliptin and HF. Elucidation of saxagliptin-DPP9 interaction and downstream events may foster a better understanding of the role of gliptins as modulators of cardiac signaling.

Keywords: Ca2+ transient; cardiac electrophysiology; diabetes; gliptins; heart failure.

FIGURE 1

Saxagliptin and sitagliptin internalizes into cardiomyocytes. Representative bright-field (left panels, A–F) and corresponding fluorescent images (right panels, A–F) of HL-1 cardiomyocytes treated with standard external solution (A) alone (control) or containing (B) fluorescamine (2 μM), saxagliptin-fluorescamine adduct [(C) 2–2 μM or (E) 20–2 μM] or sitagliptin-fluorescamine adduct [(D) 2–2 μM or (F) 20–2 μM] for 5 min (n = 6, scale bar: 10 μm). “n” represents the number of experiments.

FIGURE 2

Saxagliptin but not sitagliptin inhibits the CaMKII-PLB axis. Representative Western blots (n = 6) for (A–D) pCaMKII (T286), (E–H) pPLB (T17), and (I) CaMKII and PLB expression in HL-1 cardiomyocytes treated with saxagliptin (A,C,E,G) or sitagliptin (B,D,F,H) for the indicated time points (A,B,E,F,I, 2 μM) and concentrations (C,D, 5 min; G,H, 30 min). GAPDH was used as a loading control. “n” represents the number of experiments.

FIGURE 3

Structural analysis of gliptins-DPP4/8/9 binding. (A) A representative Western blot (n = 3) showing expression of DPP8 and DPP9 in total protein lysate (50 μg) of HL-1 cardiomyocytes and mouse LV. (B) 3D structural alignment of saxagliptin bound to DPP4 (PDB code: 3bjm) and sitagliptin (PDB code: 1×70) aligned with the 3D structure of DPP8 (PDB code: 6eoo) using pymol. The 3D structure of DPP4 and DPP8 are shown as cartoon representation and colored in orange and gray, respectively. The DPP4-bound forms of saxagliptin and sitagliptin are shown in sticks and colored in magenta and green, respectively. The DPP4 and DPP8 amino-acids present in the gliptin-binding surface of DPP4 are shown in sticks and colored in orange and gray, respectively, if conserved in the structural model or, in red and cyan, if not. (C) 3D structural alignment as described in (B) but using DPP9 (PDB code: 6eoq, colored in yellow) in place of DPP8. Representative Western blots (n = 6) for (D) pCaMKII (T286), (E) pPLB (T17), and (F) CaMKII and PLB expression in HL-1 cardiomyocytes treated with TC-E 5007 (2 μM) for the indicated time points. GAPDH was used as a loading control. “n” represents the number of experiments.

FIGURE 4

Saxagliptin inhibits the CaMKII-PLB axis via DPP9 inhibition. (A) mRNA and (B) protein expression of DPP8 and DPP9 in HL-1 cardiomyocytes transfected with si-scr, si-DPP8 or si-DPP9. GAPDH was used as a house keeping gene/protein. A representative Western blot showing (C) pCaMKII (T286), and (D) pPLB (T17) expression in HL-1 cardiomyocytes transfected with si-DPP8 or si-DPP9, and/or treated with saxagliptin or sitagliptin (2 μM each) as indicated for (C) 5 min and (D) 30 min. All values are expressed as mean ± SEM (n = 6). ^∗p < 0.05 vs. si-scr DPP8 and ^#p < 0.05 vs. si-scr DPP9 by one-way ANOVA followed by Tukey’s post hoc test. “n” represents the number of experiments.

FIGURE 5

Saxagliptin inhibits PKC via DPP9 inhibition. Quantification of active PKC levels in HL-1 cardiomyocytes treated with saxagliptin or sitagliptin in a (A) time-dependent (2 μM) or (B) concentration-dependent manner (10 min), and (C) TC-E 5007 (2 μM) for indicated time periods. (D) HL-1 cardiomyocytes were transfected with si-DPP8 or si-DPP9, and/or treated with saxagliptin or sitagliptin (2 μM) for 10 min as indicated to follow measurements of active PKC. All values are expressed as mean ± SEM (n = 6). ^∗p < 0.05 vs. (A,C) 0 min, (B) 0 μM, and (D) si-scr by one-way ANOVA followed by Tukey’s post hoc test. “n” represents the number of experiments.

FIGURE 6

Saxagliptin and TC-E 5007 impairs CaT relaxation and prolongs APD. (A) Representative CaT traces and (B) CaT decay tau (τ_CaT) of control and a saxagliptin-, sitagliptin- or TC-E 5007 (2 μM and 2 h for all)-treated mouse ventricular cardiomyocytes stimulated at 1 Hz frequency (n = 14). (C) Representative action potentials (AP) and (D) AP duration at 90% repolarization of mouse ventricular cardiomyocytes before (control) and after 5 min superfusion with either external solution alone (rundown), saxagliptin or sitagliptin (2 μM for both) and stimulated at 1 Hz frequency (n = 10/5). Controls of rundown, saxagliptin and sitagliptin groups are combined to improve readability of the figure (D). All values are expressed as mean ± SEM. “n” represents the number of (B) experiments and (C) cardiomyocytes/animals. ^∗p < 0.05 vs. control by (B) one-way repeated-measure ANOVA followed by Tukey’s post hoc test or (D) ANCOVA.

FIGURE 7

A schematic presentation of the observed DPP9-mediated effects of saxagliptin on cardiac signaling, contractility and electrophysiology.