Cardiac expression of the CREM repressor isoform CREM-IbΔC-X in mice leads to arrhythmogenic alterations in ventricular cardiomyocytes

Abstract

Chronic β-adrenergic stimulation is regarded as a pivotal step in the progression of heart failure which is associated with a high risk for arrhythmia. The cAMP-dependent transcription factors cAMP-responsive element binding protein (CREB) and cAMP-responsive element modulator (CREM) mediate transcriptional regulation in response to β-adrenergic stimulation and CREM repressor isoforms are induced after stimulation of the β-adrenoceptor. Here, we investigate whether CREM repressors contribute to the arrhythmogenic remodeling in the heart by analyzing arrhythmogenic alterations in ventricular cardiomyocytes (VCMs) from mice with transgenic expression of the CREM repressor isoform CREM-IbΔC-X (TG). Patch clamp analyses, calcium imaging, immunoblotting and real-time quantitative RT-PCR were conducted to study proarrhythmic alterations in TG VCMs vs. wild-type controls. The percentage of VCMs displaying spontaneous supra-threshold transient-like Ca(2+) releases was increased in TG accompanied by an enhanced transduction rate of sub-threshold Ca(2+) waves into these supra-threshold events. As a likely cause we discovered enhanced NCX-mediated Ca(2+) transport and NCX1 protein level in TG. An increase in I NCX and decrease in I to and its accessory channel subunit KChIP2 was associated with action potential prolongation and an increased proportion of TG VCMs showing early afterdepolarizations. Finally, ventricular extrasystoles were augmented in TG mice underlining the in vivo relevance of our findings. Transgenic expression of CREM-IbΔC-X in mouse VCMs leads to distinct arrhythmogenic alterations. Since CREM repressors are inducible by chronic β-adrenergic stimulation our results suggest that the inhibition of CRE-dependent transcription contributes to the formation of an arrhythmogenic substrate in chronic heart disease.

Keywords: Arrhythmia; NCX; Remodeling; Transcription factor CREM.

Fig. 1

a Field stimulation protocol to provoke spontaneous Ca²⁺ events in Indo-1/AM loaded VCMs during a 90 s stimulation pause after a 30 s lasting 1 and 2 Hz stimulation phase. Two event types were detected as displayed in b: sub-threshold Ca²⁺ waves (wCaR) and supra-threshold Ca²⁺ releases (transient-like, tCaR). The ratio tCaR/all measured cells was increased in TG VCMs (c) due to a higher proportion of VCMs showing this event type (d) while the tCaR-frequency in tCaR-positive VCMs was not different between groups (e). The rate of Ca²⁺ waves/VCM was unaltered between groups (f) while the transduction of wCaR into tCaR expressed by the ratio tCaR/(wCaR + tCaR) was increased in TG (g) (white bars CTL, n = 67/11; grey bars TG, n = 66/12; *p < 0.05 vs. CTL)

Fig. 2

a Representative immunoblots and b relative protein levels for total RyR2, Ser2808 and Ser2814 phosphoforms normalized to CSQ. Protein levels were not significantly different between groups, albeit along a noticeable sample to sample variation (white boxes CTL, grey boxes TG, n = 12/group)

Fig. 3

a Diastolic Ca²⁺ _i and b the Ca²⁺ _i transient amplitude were unaltered between groups. c The Ca²⁺ _i transient decay was accelerated in TG myocytes (TTD50 % in ms) pointing to altered Ca²⁺ transport/extrusion mechanisms in TG (white boxes CTL basal, n = 99/10; grey boxes TG basal, n = 97/10; *p < 0.05 vs. CTL). Rapid caffeine application to Indo-1/AM loaded VCMs after prestimulation as displayed in d was performed to assess e SR Ca²⁺ load and f fractional SR Ca²⁺ release which were unaltered between groups (white boxes CTL, n = 39/5; grey boxes TG, n = 33/4)

Fig. 4

a Transport rates for SERCA2a, NCX and the PMCA were determined by exponential fitting to the decay phases of electrically evoked Ca²⁺ transients (r ₁ = SERCA + NCX + PMCA), caffeine induced Ca²⁺ transients (r ₂ = NCX + PMCA) and caffeine induced Ca²⁺ transients with NCX-block by 0Na⁺/0Ca²⁺-solution (r ₃ = PMCA). The SERCA2a transport rate (b) (r_SERCA = r ₁ − r ₂) and the NCX transport rate (c) (r_NCX = r ₂ − r ₃) were increased in TG VCMs while the PMCA transport rate (d) (r _PMCA = r ₃) was unaltered between groups (white boxes CTL, n = 22–53/6; grey boxes TG, n = 30–70/6). e Original registration of a caffeine induced I _NCX in a patch clamp experiment. I _NCX was likewise increased in TG myocytes (f peak I _NCX in pA/pF; white bar CTL, n = 11/3; grey bar TG, n = 11/3) along elevated NCX1 protein levels (g) (n = 12/group) as illustrated by representative immunoblots, whereas Slc8a1 mRNA levels were not different between groups (*p < 0.05 vs. CTL)

Fig. 5

Superimposed representative APs illustrate AP prolongation in TG myocytes (a). The resting membrane potential (RMP) (b) and AP amplitude (c) were unaltered between groups. AP duration until 70 and 90 % repolarization (APD_70,90) was increased in TG VCMs while APD₅₀ was prolonged only in tendency (d) (black CTL, n = 35/13; grey TG, n = 36/13). e Original registration of TG APs with EADs. AP prolongation in TG went along with an increased proportion of VCMs with EADs (f) (*p < 0.05 vs. CTL)

Fig. 6

a I _K,total peak and steady-state amplitudes were determined by a single 25 s voltage pulse to +60 mV (HP −80 mV) for complete transient current inactivation. b I _K,total peak amplitude was reduced by 25 % in TG VCMs whereas the steady-state amplitude was unaltered between groups (black CTL, n = 14/7; grey TG, n = 23/9). c As illustrated by representative immunoblots KChIP2 protein was down-regulated in TG ventricular homogenates (relative protein level, KChIP2 normalized to CSQ, Kv4.2 normalized to Gapdh, n = 5–6). d Relative expression levels of mRNAs encoding major potassium channel subunits in ventricular homogenates from TG and CTL mice. The relative expression level of Kcnd2/Kv4.2 (I _to) was decreased whereas Kcnb1/Kv2.1 (I _Kslow2) was increased in TG [relative mRNA expression level normalized to Hprt (hypoxanthine-guanine phosphoribosyltransferase)]. mRNA levels encoding Kcnip2/KChIP2 (I _to, accessory subunit), Kcna4/Kv1.4 (I _to), Kcna5/Kv1.5 (I _Kslow1), and Kcnd3/Kv4.3 (I _to) were unaltered between groups. e A classic step protocol was used to evoke I _Ca,L (−40 to +65 mV, Δ5 mV, 400 ms duration, HP −80 mV, 1 Hz stimulation frequency). Sodium currents were inactivated by a 200 ms prepulse before each step (truncated in the figure). The calculated I–V-relationship was not different between groups (f) (*p < 0.05 vs. CTL)

Fig. 7

Representative ECG recording of a a WT mice and b a TG mice displaying a VES. c Mean VES/all measured mice, d proportion of mice with VES and e mean VES/positive mouse in older mice (19–21 weeks, n = 20), f–h in young mice (5–7 weeks, n = 12), respectively. Older TG mice had on average more VES/mouse after application of isoproterenol (ISO) because both proportion of mice showing VES and VES frequency in positive mice were tendentially increased. In younger TG mice before the onset of AF the VES rate in event-positive mice (h) was significantly increased after ISO (white bars CTL, grey bars TG; *p < 0.05 vs. CTL, ^# p < 0.05 vs. basal)

Fig. 8

Relative mRNA expression levels of CREM-IbΔC-X on ventricular homogenates from CTL mice after isoproterenol treatment via osmotic minipumps. CREM-IbΔC-X was induced after 10 h isoproterenol treatment in CTL mice (CREM-IbΔC-X mRNA expression level normalized to Hprt, relative to 0 h, n = 11–12, *p < 0.05 vs. 0 h)