Synthesis of the Ca2+-mobilizing messengers NAADP and cADPR by intracellular CD38 enzyme in the mouse heart: Role in β-adrenoceptor signaling

Abstract

Nicotinic acid adenine dinucleotide phosphate (NAADP) and cyclic ADP-ribose (cADPR) are Ca²⁺-mobilizing messengers important for modulating cardiac excitation-contraction coupling and pathophysiology. CD38, which belongs to the ADP-ribosyl cyclase family, catalyzes synthesis of both NAADP and cADPR in vitro However, it remains unclear whether this is the main enzyme for their production under physiological conditions. Here we show that membrane fractions from WT but not CD38^-/- mouse hearts supported NAADP and cADPR synthesis. Membrane permeabilization of cardiac myocytes with saponin and/or Triton X-100 increased NAADP synthesis, indicating that intracellular CD38 contributes to NAADP production. The permeabilization also permitted immunostaining of CD38, with a striated pattern in WT myocytes, whereas CD38^-/- myocytes and nonpermeabilized WT myocytes showed little or no staining, without striation. A component of β-adrenoreceptor signaling in the heart involves NAADP and lysosomes. Accordingly, in the presence of isoproterenol, Ca²⁺ transients and contraction amplitudes were smaller in CD38^-/- myocytes than in the WT. In addition, suppressing lysosomal function with bafilomycin A1 reduced the isoproterenol-induced increase in Ca²⁺ transients in cardiac myocytes from WT but not CD38^-/- mice. Whole hearts isolated from CD38^-/- mice and exposed to isoproterenol showed reduced arrhythmias. SAN4825, an ADP-ribosyl cyclase inhibitor that reduces cADPR and NAADP synthesis in mouse membrane fractions, was shown to bind to CD38 in docking simulations and reduced the isoproterenol-induced arrhythmias in WT hearts. These observations support generation of NAADP and cADPR by intracellular CD38, which contributes to effects of β-adrenoreceptor stimulation to increase both Ca²⁺ transients and the tendency to disturb heart rhythm.

Keywords: CD38; Ca2+; cardiac arrhythmia; cardiac hypertrophy; cyclic ADP-ribose (cADPR); heart; lysosomes; nicotinic acid adenine dinucleotide phosphate (NAADP); sarcoplasmic reticulum (SR); β-adrenoceptor.

Figure 1.

In vitro ADP-ribosyl cyclase activities and NAADP production were not detected in mixed-membrane preparations from CD38^−/− mouse heart. We evaluated the ability of a membrane preparation containing plasmalemma and SR from WT but not CD38^−/− mouse heart to catalyze the synthesis of NAADP and cADPR. A, example traces of the NAADP measurement assay system based on SUEH, in which increasing concentrations of NAADP show a desensitization response to the activating substance NAADP. B, observations reflecting enhanced desensitization at increasing concentrations of NAADP were used to construct the calibration curve. C, using this assay, we show that a membrane preparation from WT mice catalyzed the synthesis of NAADP, whereas a similar preparation from CD38^−/− mouse heart lacked this ability. The time course of NAADP production is shown. D, the average rate of NAADP synthesis from these data (n = 4 for both groups). E and F, the membrane preparation from WT mice catalyzed synthesis of cGDPR (an analogue of cADPR) with NGD as a substrate (the time course is shown in E, and a bar graph showing the average rate of cGDPR synthesis is shown in F), whereas membranes from CD38^−/− mice lacked this ability (n = 6 for both groups). r.f.u. indicates relative florescence unit. Note that WT membranes showed clear synthesis of both messengers, whereas there was no detectable synthesis with CD38^−/− membranes. Data are expressed as mean ± S.E. ***, p ≤ 0.001; n = number of preparations.

Figure 2.

Permeabilization of cardiac myocytes with Triton X-100 and/or saponin enhanced NAADP production and permitted immunolabeling of CD38. A, the rate of NAADP synthesis was higher after permeabilization of the cell membranes with saponin alone, and permeabilization with Triton X-100 in addition to saponin did not further increase the rate of NAADP synthesis (n = 3 in each group). B, this point is further illustrated in the bar graph, which also shows that the ability to synthesize NAADP was lost in myocytes from CD38^−/− mice (n = 4). Omission of NA also abolished synthesis of NAADP in intact cells (n = 5). C, immunolabeling of CD38 using rabbit anti-human CD38 antibody without Triton X-100 showed little staining, although there were surface patches with higher fluorescence intensity (top panels). Following membrane permeabilization with Triton X-100 to allow access of the antibody to the cell interior, there was clear staining with a striated pattern in WT (center panels) but not in permeabilized CD38^−/− cardiac myocytes (bottom panels). D, the representative intensity–distance plot (bottom panel) shows that, in permeabilized cells, the staining observed in the WT myocyte had a much higher intensity than in the CD38^−/− myocyte and showed multiple peaks with a repeating distance interval that resembled the sarcomere spacing. E, similar observations in rabbit ventricular myocytes. The fluorescent images of myocytes permeabilized with Triton X-100 showed clear labeling with CD38 antibody. There was a striated pattern with a similar spacing as that shown by immunolabeling of RyR2. F, no labeling with a striated pattern was observed when Triton X-100 or primary antibodies were omitted. The images show representative staining of the major observation (≥75%) in each group (n ≥ 20). Scale bars = 10 μm, n = number of cells. Data are expressed as mean ± S.E. *, p ≤ 0.05; ***, p ≤ 0.001; n.s., not significant.

Figure 3.

Contribution of NAADP signaling pathway to the β-adrenoreceptor-mediated positive inotropy was lost in the CD38^−/− ventricular myocytes. Ca²⁺ transients were triggered by cardiac action potentials in ventricular myocytes from WT and CD38^−/− mice. A, the amplitudes of Ca²⁺ transients in WT (n = 9) and CD38^−/− myocytes (n = 15) are shown in the presence of isoproterenol. B, example images. Under these conditions, amplitudes of CaTs accompanying action potentials were smaller in cardiac myocytes from CD38^−/− mice than from WT mice. ISO, isoproterenol. C, sarcomere shortening in the presence of isoproterenol accompanying action potentials was also smaller in myocytes from CD38^−/− mice than from WT mice (WT, n = 12; CD38^−/−, n = 11). D, example traces. The observations support a functional role for the CD38 enzyme during β-adrenergic receptor stimulation. E, the effects of isoproterenol on the amplitude of CaTs with and without bafilomycin A1 (10 μm) in ventricular myocytes from WT and CD38^−/− mice (n = 16–25 in each group). F, these observations are replotted to show the magnitude of the isoproterenol-induced changes with and without bafilomycin A1 in myocytes from WT and CD38^−/− mice. G, representative traces. Bafilomycin A1 reduced the effect of isoproterenol on the amplitude of CaTs in myocytes from WT but not CD38^−/− mice. The isoproterenol-induced increase in amplitudes of CaTs was therefore less in the presence of bafilomycin A1 in WT myocytes, whereas, in myocytes from CD38^−/− mice, the isoproterenol-induced increases in CaT amplitude were similar with and without bafilomycin A1. Data are expressed as mean ± S.E. *, p ≤ 0.05; ** p ≤ 0.01; n = number of cells.

Figure 4.

In vitro and in silico evidence for the action of SAN4825 to inhibit NAADP and cADPR production through targeting CD38. A and B, SAN4825 inhibited NAADP (A) and cADPR (B) synthesis in mouse mixed-membrane preparations at both pH 7.2 and pH 4.5. Note that the synthesis of NAADP in the absence of SAN4825 was higher at pH 4.5 than at pH 7.2. C, superimposition of the binding conformation of the NAADP analogue in human CD38 (left, red, PDB code 4F46, chain B) on the most favorable binding conformation of SAN4825 (left, blue) in mouse CD38 (PDB code 2EG9) at pH 7.2 following Glide docking. A superimposition of the binding conformation of ADPRP in human CD38 (right, red, PDB code 4F46, chain A) on the most favorable binding conformation of SAN4825 (right, blue) in mouse CD38 (PDB code 2EG9) is also shown (right). D, the predicted binding sites (orange) for SAN4825 in mouse CD38 following blind docking in the SwissDock server. Inner cavities are shown in yellow. E, the distances between the predicted docking sites and the center of the active site were plotted. Data are expressed as mean ± S.E. *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; n.s., not significant; n ≥ 3 preparations in each group.

Figure 5.

Hearts from CD38^−/− mice or preincubated with SAN4825 showed resistance to arrhythmias during overstimulation of the β-adrenoreceptor pathway. A, the tendency to show disturbances of rhythm was assessed using four different protocols (see “Experimental Procedures”) both before and after exposure to β-adrenoreceptor stimulation with isoproterenol (ISO, 300 nm). In WT hearts, exposure to isoproterenol caused the expected increase in arrhythmias, but this isoproterenol-induced tendency to arrhythmias was suppressed in hearts pretreated with SAN4825 as well as hearts from CD38^−/− mice. B, replotted from A, showing the increase in arrhythmogenicity of hearts in the presence of isoproterenol. C, the observations in A expressed as the proportion of hearts in which isoproterenol increased the tendency to arrhythmias. D, the cumulative percentage of hearts that exhibited arrhythmias in the presence of isoproterenol in the burst pacing experiments with increasing pacing frequency. E, the increase in threshold current required to induce ventricular tachycardia in the presence of isoproterenol. In experiments varying both pacing frequency and current amplitude, hearts from CD38^−/− mice and hearts preincubated with SAN4825 showed reduced tendency to arrhythmias in the presence of isoproterenol. F, the percentages of WT hearts, SAN4825-treated hearts, and CD38^−/− hearts that became arrhythmic (filled columns) during the dynamic pacing and S1S2 pacing experiments, with and without 300 nm isoproterenol. G, representative traces of burst pacing experiments. Data are expressed as mean ± S.E. or percentage. A and B, Mann Whitney Test. C, Fisher exact test. E, Student's t test or one-sample t test. *, p ≤ 0.05; **, p ≤ 0.01. WT, n = 8; SAN4825, n = 8; CD38^−/−, n = 6; n = number of hearts.