心臟纖維母細胞的動態轉變在心臟再生中的關鍵角色

Abstract

心血管疾病是全球死亡的主要原因，其中心肌梗塞（myocardial infarction, MI）常因心肌細胞不可逆的流失而導致心臟衰竭。成年哺乳動物心臟幾乎不具備再生能力，相較之下，某些物種例如斑馬魚、蠑螈以及出生後一週內的新生小鼠，能夠透過現有心肌細胞的增殖再生來修復受損的心肌組織。心臟纖維母細胞（cardiac fibroblasts, CFs）在細胞外基質（extracellular matrix, ECM）生成與組織修復中扮演關鍵角色，並已被證實參與斑馬魚心臟再生。然而，心臟纖維母細胞在哺乳類心肌再生中的功能仍尚未明確。因此，本研究的主要目的在探討心臟纖維母細胞在新生小鼠心臟再生中的角色，以釐清其可能參與的機制。 為了探討心臟纖維母細胞在哺乳類心臟再生中的角色，我們建立了新生小鼠的纖維母細胞特異性去除模型（Col1a2-CreERT2; Rosa26LSL-tdTomato; Rosa26iDTR）。透過注射 tamoxifen 誘導標記表現膠原蛋白第 I 型 α2 鏈（collagen type I alpha 2 chain, Col1a2） 的纖維母細胞，並以白喉毒素 (diphtheria toxin) 進行纖維母細胞特異性剔除後，於出生後第 5 天對小鼠進行心尖切除（apical resection, AR）或心肌梗塞手術。結果顯示，去除 CFs 會嚴重損害心臟再生能力，導致心肌細胞增殖減少並伴隨明顯嚴重纖維化。此外，這些小鼠的死亡率亦顯著上升，AR 組為 42%、MI 組為 55%，對照組則幾乎全數存活。 為了進一步釐清 CFs 調控再生的分子機制，我們對來自 P1、P3 與成年小鼠心臟中分離的 CFs進行整體 RNA 定序（bulk RNA-seq）分析。結果顯示，相較於成年 CFs，新生期 CFs 顯著上調與細胞週期、心臟發育及前驅細胞特徵相關的基因，顯示其具有較高的轉錄可塑性，可能較具支持心臟再生的潛力。鑑於其具有心臟譜系相關基因的表現，我們進一步進行譜系追蹤 (lineage tracing) 的實驗以評估這些纖維母細胞是否可能轉分化為心肌細胞。然而，實驗結果並未支持此假說，顯示 CFs 可能透過其他機制參與再生過程。 為了驗證 bulk RNA-seq結果並進一步解析 CF 的多樣性，我們重新分析了已發表的心臟單細胞 RNA 定序（scRNA-seq）資料集（GSE153480），鑑定出五個具有不同轉錄特徵的 CF 亞群。其中， FB4 於損傷後第 1 天短暫出現，FB3 則於第 3 天出現，顯示再生過程中 CF 亞群在損傷後不同時間點發生動態轉換；然而，FB3及FB4這兩群細胞在缺乏再生能力的 P8 小鼠心臟中則不見表現。偽時間分析（pseudotime analysis）顯示，在排除主要出現在缺乏再生能力之 P8 心臟中的 FB2 後，CF的發展軌跡從 FB1 起始，經由 proliferating FB (Pro.FB) 之後分成兩條路徑：損傷後第 1 天出現 FB4，並於第 3 天轉化為 FB3。此外，我們發現，來自不具再生能力的P8小鼠 CFs 具有較高的膠原纖維排列能力，可能促進形成較為僵硬的疤痕，顯示再生與非再生心臟中的 CFs 具有不同的組織重塑的功能。 綜上所述，我們的研究顯示， Col1a2⁺ fibroblasts 對新生小鼠心臟再生具有重要角色。剔除Col1a2⁺ fibroblasts會抑制心肌細胞增殖、加劇纖維化並降低存活率。新生期 CFs 表現出促進再生的轉錄特徵，並在心臟修復過程中歷經動態的細胞狀態轉換。本研究鑑別出與再生相關的 CF 子群（FB3 與 FB4），並提供了關於促進心肌修復與預防心臟衰竭的嶄新理解與潛在治療靶點。

Cardiovascular disease is the leading cause of death worldwide, with myocardial infarction (MI) frequently resulting in heart failure due to the irreversible loss of cardiomyocytes. Unlike adult mammalian hearts, which have minimal regenerative capacity, certain species—such as zebrafish, newts, and neonatal mice within the first week of life—can regenerate damaged myocardium through the proliferation of existing cardiomyocytes. Cardiac fibroblasts (CFs), which are critical for extracellular matrix (ECM) production and tissue repair, have also been implicated in supporting regeneration in zebrafish. However, whether and how CFs contribute to myocardial regeneration in mammals remains largely unknown. To address this knowledge gap, we aimed to investigate the role of cardiac fibroblasts in neonatal mouse heart regeneration. To investigate the role of CFs in mammalian heart regeneration, we employed a fibroblast-specific ablation model (Col1a2-CreERT2; Rosa26LSL-tdTomato; Rosa26iDTR) in neonatal mice. Following tamoxifen induction and diphtheria toxin–mediated ablation of Col1a2-expressing fibroblasts, we subjected the animals to either apical resection (AR) or myocardial infarction (MI) on postnatal day 5 (P5). Fibroblast depletion resulted in markedly impaired cardiac regeneration, as evidenced by diminished cardiomyocyte proliferation and excessive fibrosis. Notably, this was associated with high mortality rates—42% after AR and 55% after MI—while control animals showed near-complete survival. To uncover the molecular basis underlying CF function in regeneration, we performed bulk RNA sequencing on CFs isolated from postnatal day 1 (P1), day 3 (P3), and adult mouse hearts. Neonatal CFs exhibited upregulation of genes associated with cell cycle activity, cardiac development, and progenitor-like features, including Gata4, Mef2c, and Myh7, compared to adult CFs. These findings highlight the transcriptional plasticity of neonatal cardiac fibroblasts and suggest that they may possess a greater capacity to support cardiac regeneration compared to their adult counterparts. Given the fact that we observed the expression of cardiac lineage-associated genes in neonatal fibroblasts, we next tested whether these cells might transdifferentiate into cardiomyocytes in vivo. However, lineage-tracing experiments did not support this hypothesis, suggesting that fibroblasts may contribute to regeneration through indirect mechanisms. To further characterize cardiac fibroblasts and validate insights from our bulk RNA-seq data, we re-analyzed a publicly available single-cell RNA-seq dataset (GSE153480) and identified five transcriptionally distinct fibroblast clusters. Among these, FB4 and FB3 emerged as regeneration-associated populations in neonatal mouse hearts following MI. FB4 appeared transiently at day 1 post-injury, while FB3 became predominant by day 3, suggesting a dynamic temporal transition of fibroblast states during the regenerative response. In contrast, these populations were scarcely present in non-regenerative P8 hearts. By pseudotime analysis, we found that excluding FB2—which predominantly appears in non-regenerative P8 hearts—the trajectory originates from FB1 and bifurcates after passing through proliferating FB (Pro.FB). One day after MI, the trajectory progresses toward FB4, and by day 3, it transitions into FB3. Functional assays further confirmed that fibroblasts from non-regenerative hearts organize collagen fibers more efficiently, potentially contributing to rigid scar formation. These functional differences suggest that fibroblasts in regenerative and non-regenerative hearts adopt distinct remodeling strategies. Together, our findings demonstrate that Col1a2⁺ fibroblasts are essential for neonatal heart regeneration. Their depletion disrupts cardiomyocyte proliferation, worsens fibrosis, and compromises survival. Neonatal CFs exhibit regenerative transcriptional signature and transition through distinct fibroblast states during repair. These results reveal previously unrecognized fibroblast subpopulations—FB3 and FB4—associated with neonatal cardiac regeneration, offering new insights and potential therapeutic targets for enhancing myocardial repair and restoring cardiac function in heart failure patients.