Brain-to-heart cholinergic synapse-calcium signaling mediates ischemic stroke-induced atrial fibrillation

Abstract

Background: Stroke-related cardiovascular diseases have attracted considerable attention, with atrial fibrillation (AF) being among the most frequent complications. Despite increasing clinical evidence, experimental models of stroke-induced AF are still lacking, hindering mechanistic discoveries and the development of adequate therapeutics targeting this stroke-heart syndrome (SHS). This study aims to create a rat model of ischemic stroke-induced AF (ISIAF) and to explore the efficacy and mechanism of Wenxin Keli (WK), an antiarrhythmic Chinese medicine. Method: The middle cerebral artery occlusion/reperfusion model was adapted to create subacute brain ischemia in rats with normal cardiac function. Invasive electrophysiologic studies and ex vivo optical mapping were performed to evaluate the altered electrophysiological parameters and Ca²⁺ handling properties. RNA-seq analysis, RT-PCR, and immunohistochemistry (IHC) with immunofluorescence (IF) were employed to assess the SHS model and elucidate the mechanisms of ISIAF and the effects of WK. UPLC/Q-TOF-MS, molecular docking, and whole-cell patch recordings were used to identify the active components of WK for SHS. Results: Ischemic stroke aggravated atrial electrical instability, altered action potential duration (APD), Ca²⁺ transient duration (CaT), conduction heterogeneity, and spatially discordant alternans in SHS rat hearts. These abnormalities were alleviated by WK. RNA-seq analysis revealed that M₃-mediated cholinergic synapse signaling and L-type calcium channel (LTCCs)-mediated Ca²⁺ signaling play prominent roles in ISIAF development and its reversal by WK. UPLC/Q-TOF-MS analysis identified 19 WK components as the main components in plasma after WK treatment. Molecular docking screening identified Dioscin as the major active component of WK. WK and Dioscin reduced I_Ca-L in a concentration-dependent manner with a half-maximal inhibitory concentration of 24.254 ± 2.051 mg/mL and 8.666 ± 0.777 µmol/L, respectively. Conclusion: This study established an experimental model of ISIAF capable of characterizing clinically relevant atrial electrophysiological changes post-cerebral ischemia. Molecular mechanistic studies revealed that the cholinergic-calcium signaling pathway is central to this brain-heart syndrome. Ischemic stroke-induced atrial fibrillation is partially reversible by the Chinese medicine Wenxin Keli, which acts via regulation of the cholinergic-calcium signaling pathway, with its active component Dioscin directly binding to I_KM3 and inhibiting I_Ca-L.

Keywords: Ca2+ signaling; Stroke heart syndrome; Wenxin Keli; atrial fibrillation detected after stroke; cholinergic synapse signaling; ischemic stroke; ischemic stroke-induced AF.

Figure 1

Ischemic stroke caused sinus arrhythmia that was suppressed by WK. (A) Timeline of the MCAO operation, drug treatment, and the following ex vivo cardiac optical mapping. (B) Representative ECG recordings of isolated heart during sinus rhythm. (C) Representative images of conduction in Sham, Model, WK-L, WK-M, WK-H and Met groups under the burst pacing at S1S1. (D) AERP/APD of each group. (E) Statistical results of AF duration in each group. Values were expressed as mean ± SD (n=3-5, ^*P < 0.05 and ^**P < 0.01, Model vs. Sham; ^#P < 0.05 and ^##P < 0.01, Model vs. WK) (F) Incidence of atrial ectopy or fibrillation for each S1 interval. (G) Sinus node recovery duration of each group under different stimulation frequencies (100-50 ms). Values were expressed as mean ± SD (n=3-5, ^*P < 0.05 and ^**P < 0.01, Model vs. Sham; ^#P < 0.05 and ^##P < 0.01, Model vs. WK-L; ^▲P < 0.05 and ^▲▲P < 0.01, Model vs. WK-M; ^★P < 0.05 and ^★★P < 0.01, Model vs. WK-H; ^◆P < 0.05 and ^◆◆P < 0.01, Model vs. Met).

Figure 2

APD and CaT alterations in ISIAF before and after WK treatment via ex vivo optical mapping. (A) Conduction and APD90 maps with four recording sites of six groups under 6Hz. (H) Conduction and CaT90 maps with four recording sites of six groups under 6Hz. (B, I) APD90 and CaT90 traces and quantification in each group. (C, J) Calculation of velocity in AP phase and CaT phase in each right atrium. (D, K) Calculation of rise time in AP phase and CaT phase in each right atrium. (E, L) Calculation of IQR in AP phase and CaT phase in each right atrium. (F) Statistical results of APD 30/80. (M) Calculation diagram of APD/CaT alternans. Values were expressed as mean ± SD (n = 3-5). ^*P < 0.05 and ^**P < 0.01, Model vs. Sham; ^#P < 0.05 and ^##P < 0.01, Model group vs. WK groups. (G, N) The heat maps showed the different conduction rates of the six groups in APD and CaT. Depiction of recovery ratio of APD and CaT at the integral level. Values were expressed as mean ± SD (n=3-5, ^*P < 0.05 and ^**P < 0.01, Model vs. Sham; ^#P < 0.05 and ^##P < 0.01, Model vs. WK-L; ^▲P < 0.05 and ^▲▲P < 0.01, Model vs. WK-M; ^★P < 0.05 and ^★★P < 0.01, Model vs. WK-H; ^◆P < 0.05 and ^◆◆P < 0.01, Model vs. Met).

Figure 3

AF-related changes and Ca²⁺ release restitution in the ISIAF model with or without WK treatment. (A, B) Toff calculation diagram and its values under each condition. (C, D) Typical examples of Tau and calculation of Tau under each condition. Values were expressed as mean ± SD (n=3-5, ^*P < 0.05 and ^**P < 0.01, Model vs. Sham; ^#P < 0.05 and ^##P < 0.01, Model vs. WK-L) (E) Representative recordings of cytosolic CaTs at various S1S2 coupling intervals. (F) The heat maps of different conduction rates of the six groups. (G) Calculation diagram of recovery ratio of CaT. (H) Depiction of recovery ratio of CaT at the S1S2 integral level. Values were expressed as mean ± SD (n=3-5, ^*P < 0.05 and ^**P < 0.01, Model vs. Sham; ^#P < 0.05 and ^##P < 0.01, Model vs. WK-L; ^▲P < 0.05 and ^▲▲P < 0.01, Model vs. WK-M; ^★P < 0.05 and ^★★P < 0.01, Model vs. WK-H; ^◆P < 0.05 and ^◆◆P < 0.01, Model vs. Met). (I) Graphical representation of SR Ca²⁺ fluxes estimation protocol [a: basic Ca²⁺ transient parameters, b: RyR Ca²⁺ leak, c: SR Ca²⁺ content, d: SERCA activity; NT, normal Tyrode solution. (J, K) Quantification of RyR₂ and SERCA functions of rate-constants. [control n=6/18 (animals/cells), WK n=7/20 (animals/cells); ^*P < 0.05 and ^**P < 0.01, Sham vs. WK].

Figure 4

Transcriptome sequencing and IPA analysis of rat atrium in ISIAF and WK treatment. (A) Hierarchical clustering of the Sham, Model, and WK-M groups. The 12 genes for the cholinergic synaptic signaling pathway are labeled. Left panel: Model vs Sham, right panel: WK vs Model. (B) A volcano graph showing altered gene distributions in the Sham, Model, and WK-M groups (|fold change| ≥ 1.25 and P-adj ≤ 0.05). (C) A Venn diagram showing altered genes in ISIAF (Model vs Sham) and after WK treatment (WK vs Model). (D) GO enrichment analysis ranked ion channel activity the highest among the top 10 functions. (E) IPA Core analysis identified the cholinergic synaptic signaling pathway as the top 1 altered pathway associated with disease. The fold change and significance level of the difference in the gene expression were represented by the abscissa and ordinate, respectively. The 12 genes of the cholinergic synaptic signaling pathway were labeled. Genes that have increased or decreased in expression were shown by red or green dots, respectively. There were three independent animals in each group (n = 3). (F) Log₂Fold Change values of the genes in the cholinergic synaptic signaling pathway. (G) Log₂Fold Change values of the genes in the calcium signaling pathway. (H) RT-PCR verification of WK-M regulated genes in cholinergic and calcium signaling pathway. The mRNA expression levels of the CHAT, AChE, CHRM4, CHRM3, SLC5A7, GNG3, ATP2A3, CACNB1 and CACNA1B. Values are given as mean ± SD (n = 3, ^*P < 0.05 and ^**P < 0.01 Model vs. Sham, ^#P < 0.05 and ^##P < 0.01 Model vs. WK).

Figure 5

Verification of WK regulation of the L-type Ca²⁺ channel dynamics. Calcium voltage-gated channel auxiliary subunit beta 1 (CACNB1) constitutes the beta subunit of the L-type calcium channel. (A) Representative trace of the L-type Ca²⁺-current (I_Ca-L) recorded in Control, WK (1, 10, 20, 50 mg/mL), Nif and wash. (B) The concentration-response curve representing the percentage of inhibition of WK (n=3-5). (C) Standardized steady-state activation of I_Ca-L under Control and 20 mg/mL WK. (D) Standardized steady-state deactivation of I_Ca-L under Control and 20 mg/mL WK treatment (n=6). (E, F) Sample trace and pooled data showed the effect of I-V relationship under Control and 20 mg/mL WK treatment. Values are given as mean ± SD (n = 6, ^#P < 0.05 and ^##P < 0.01 WK vs. Control).

Figure 6

Chemical profiling of WK active ingredients in rat plasma. UPLC-MS base peak intensity chromatograms in negative (A) or positive (B) modes. In each panel, WK aqueous solution (Top, 1) Plasma after 1h WK administration (Middle, 2) and Blank plasma (Bottom, 3) are shown.

Figure 7

Virtual screen of WK active compounds targeting muscarinic acetylcholine receptor M3 (CHRM3) and calcium voltage-gated channel auxiliary subunit beta 1 (CACNB1) by molecular docking. (A) The heatmap of docking scores of 19 WK active compounds binding to key targets CACNB1 and CHRM3 (-CDOCKER ENERGY score, kcal/mol). In the bubble plot, the binding energies were represented by bubble colors. A lower stability value indicates a more stable complex. (B) Representative docking complex of key targets and corresponding compounds.

Figure 8

The effect of WK active compounds on I_Ca-L in rat atrium myocytes. (A-H) Whole cell recordings showing (top to bottom) current density, sample trace, combined data and process time of I_Ca-L recorded during exposure to (A) Tanshinone-IIA (100 μmol/L), (B) Nardosinone (100 μmol/L), (C) Cryptotanshinone (100 μmol/L), (D) Ginsenoside Re (100 μmol/L), (E) Ginsenoside Rb1 (100 μmol/L), (F) Ginsenoside Rg1 (100 μmol/L), (G) Atractylenolide (100 μmol/L) and washing under Control conditions. The data are expressed as Mean ± SD (n=3-4). Compared with Control, P > 0.05. (H) The inhibition curve (top), an example trace (middle), time course (bottom) of I_Ca-L recorded under Control conditions of 1, 3, 10, and 30 μmol/L Dioscin. Concentration-response curve representing the percentage of inhibition of Dioscin (n=4).

Figure 9

The effect of WK active compounds on muscarinic acetylcholine receptor M3 (CHRM3) activation in HL-1 cells. (A-E) Representative images and quantitation of HL-1 cells activated by 10 μmol/L ACh (A, D) or 50 μmol/L ACh (B, E) and blocked by 0.3-3 μmol/L Dioscin (Dio). The nuclei were stained in blue by Hoechst and the CHRM3 protein was stained in orange. Scale bar = 100 µm. (C) Bar graph quantitation of Dioscin on the viability of HL-1 cells. Values were expressed as mean ± SD (n = 4-5). (F-G) Representative images (F) and quantitation (G) of ACh-activated HL-1 cells and blocking effects by WK and Dioscin in the presence of 4-DAMP. The nucleus was stained blue, and CHRM3 were labeled green. Scale bar = 100 µm. Values were expressed as mean ± SD (n = 4-5). ^▲▲P < 0.01 Dioscin vs. Ctrl, ^**P < 0.01 vs ACh vs. Ctrl, ^#P < 0.05 and ^##P < 0.01 ACh vs. treat groups, ^●●P < 0.01 Dioscin vs. WK, ^■■P < 0.01 WK vs. 4-DAMP, ^★★P < 0.01 Dioscin vs. 4-DAMP.

Figure 10

An illustration of the proposed role of cholinergic and calcium signaling pathways in ISIAF and WK treatment in neuronal and non-neuronal (cardiac) systems. The genes upregulated by ISIAF (CHAT, SLC5A7, AChE, CACNB1, ATP2A3, and CHRM3) were shown by the red arrow. Genes downregulated by WK (CHAT, SLC5A7, AChE, CACNB1, ATP2A3, and CHRM3) were marked in blue stripes.