糖尿病心肌病變之演變中，心臟收縮力之力量依賴性和速度依賴性之指數的時序變化: 氧化壓力和氮化壓力與肌球蛋白重鏈亞型轉換和氧化敏感的轉錄因子之角色

Abstract

糖尿病的盛行率驟增，糖尿病人的心臟血管合併症成為醫療和健康指標的重要課題。糖尿病會導致心肌收縮力受損，但潛在的詳細機制尚不明確。本論文主要的研究目標有三大部分。首先，我們研究的第一部分是以彈性-阻性左心室唧筒模型來探討糖尿病心肌病變的變化過程，從心肌力學的觀點，用力量依賴性(force-dependent)和速度依賴性(velocity-dependent)的心臟收縮指數來探討心臟收縮功能的時序變化。有鑑於心臟收縮功能之探討，其必需同時兼顧心肌載荷條件和心臟主動脈的交互作用，因此，我們研究糖尿病大鼠心臟收縮力的時序演變，並同時探討其中心臟的載荷條件、機械效率和心室動脈偶合(ventriculoarterial coupling)的時序變化。主要目的是以數學模式之心臟力學（即力量依賴性和速度依賴性的心臟收縮指數）來解釋在糖尿病心肌病變過程中，心臟收縮功能惡化之機械原因。另一方面，我們亦嘗試找出在糖尿病心肌病變過程中，造成失代償性心衰竭的心臟機械力學因素。藉以提供臨床醫師有效性的指標，來監控疾病治療的效果或新藥的開發。本博士論文的第二部份，我們的目的是評估抗氧化劑dimethylthiourea (DMTU) 對於糖尿病心肌病變之預防和治療作用。DMTU 是屬於一種強效的羥自由基清除劑。過去少數研究用它來探討糖尿病的動物模式中，抗氧化劑對於血管內皮細胞功能失調的療效。另有少數研究用以探討在心肌梗塞的動物模式中，抗氧化劑對於改善心臟收縮功能之療效。在初步的結果顯示DMTU確實可以改善心肌梗塞後的心衰竭。然而，目前尚未證實DMTU在治療糖尿病心肌病變的角色，我們透過糖尿病心肌病變的動物模式，試圖以彈性-阻性左心室唧筒模型來探討DMTU治療前後，糖尿病心臟收縮功能(即力量依賴性和速度依賴性心肌收縮力指數)的變化。此外，我們針對糖尿病鼠的心臟實驗，研究下游氧化還原敏感的轉錄因子，如心肌細胞增強因子2（myocyte enhancer factor-2, MEF- 2）、心臟自主神經系統、神經嵴衍生物（heart autonomic nervous system and neural crest derivatives, eHAND and dHAND），以及氧化還原壓力指標 (oxidative stress markers) 和肌球蛋白重鏈亞型轉換 (myosin heavy chain isoform switch)。本篇博士論文的第三部份，我們藉由糖尿病動物模式來探討L-NAME(NG-nitro-L-arginine methyl ester:一氧化氮生成酶抑制劑，L-NAME)對糖尿病心肌病變的影響和治療效果，以及其中可能的分子作用機轉。 在本論文的第一部分中，我們在鏈脲佐菌素(streptozotocin, STZ) 引發之糖尿病大鼠模式中，探討其心臟收縮力學、載荷條件和機械效率的時序變化。成年雄性Wistar大鼠隨機分為對照組和STZ誘導的糖尿病組。在注射STZ後的第8週,第16週和第22週，分別執行侵入性血流動力學研究。我們同時測量左心室壓力和主動脈血流，藉由曲線擬合技術，以彈性-阻性左心室唧筒模型來計算最大收縮力（maximal systolic elastance, Emax），最大理論血流量（maximal theoretical flow, Qmax）; 另外，我們藉由單一射血心跳估計技術 (single beat estimation)來計算心室主動脈偶合(ventrioculoarterial coupling) 和機械效率 (mechanical efficiency)。實驗結果顯示，在糖尿病大鼠模式中，其最大收縮力在早期(第8週)就發生顯著下降且持續低瀰。另一方面，最大理論血流量則隨時間逐漸增加，藉以代償最大收縮力的下降。但在第22週後，最大理論血流量開始下降，同時伴隨心輸出量下降的變化。在糖尿病心肌病變過程中，有利的載荷條件(即前負荷增加，後負荷減少)有助於增強心搏量和最大理論血流量，以代償心臟收縮功能。然而心室主動脈偶合失調造成心臟機械效率減弱。我們的研究闡明，後負荷調整的最大理論血流量(Qmaxad）和心臟機械效率的下降，是糖尿病大鼠心臟功能惡化的兩項決定因素。 本論文的第二部分，在確立兩種心肌收縮力之力量依賴性和速度依賴性指數，其對糖尿病心肌病之心肌收縮力評估的有效性之後，我們再深入探討抗氧化劑DMTU對糖尿病心肌病變的動物模式中的預防和治療效果，以及其中可能的作用機轉。文獻報告指出，羥自由基和過氧化氫參與糖尿病之心肌病變的形成有關，但確切的機制仍然不明。我們在STZ誘導的糖尿病大鼠早期和慢性模式中，評估DMTU對於心臟收縮功能力量依賴指數和速度依賴指數的影響。在鏈脲佐菌素(STZ,55毫克/公斤）注射72個小時和8週後，糖尿病大鼠隨機分為 DMTU組（50毫克/公斤/天，腹腔注射）或生理食鹽水組分別治療 6週和12週。所有的老鼠均接受侵入性血流動力學的研究。我們同時測量左心室壓力和主動脈血流，藉由曲線擬合技術，以彈性-阻性左心室唧筒模型來計算最大收縮力和最大理論血流量。在糖尿病大鼠對照組中，標準化的最大收縮力(Emaxn）和後負荷調整的最大理論血流量（Qmaxad）兩個收縮力指標均顯著下降，同時伴隨有肌球蛋白重鏈亞型轉換 (myosin heavy chain isoform switch) 及其上游的轉錄因子 (transcription factor) 的改變，如心肌細胞增強因子-2（MEF-2）和心臟自主神經系統和神經嵴衍生物（dHAND 和eHAND）。在慢性糖尿病大鼠中，DMTU改變了心肌表達的MEF-2和eHAND蛋白量，以及減少了肌球蛋白重鏈亞型轉換，故明顯改善了Qmaxad。我們的研究顯示Qmaxad的改善可能與減少肌球蛋白重鏈亞型的轉換有關。DMTU 對於標準化的最大收縮力(Emaxn)影響效益不顯著。關於預防治療，在早期糖尿病大鼠中，DMTU顯著改善標準化的最大收縮力和後負荷調整的最大理論血流量。我們的研究結果顯示，抗氧化劑可預防糖尿病所造成的心肌病變問題。 在本論文的第三部分，經由先前發展的研究模式，我們利用經過驗證的兩種心肌收縮力指數(即左心室收縮力的力量依賴性和速度依賴性的指數)來闡述糖尿病心肌病的心肌收縮功能。我們探討L-NAME(NG -nitro-L-arginine methyl ester)對糖尿病心肌病變的影響。我們藉由糖尿病動物模式來探討L-NAME的治療效果，及其可能的分子作用機轉。結果顯示使用L-NAME八週後可以顯著地改善糖尿病鼠心臟收縮功能，即左心室收縮之力量依賴性和速度依賴性的指數。我們也發現L-NAME可以減少過氧化物和硝化物引發的自由基離子，改善一氧化氮生成酶之去偶合 (uncoupled nitric oxide synthase)之情形。我們又發現，L-NAME的治療可以減少核仁因子-κB (nuclear factor-κB, NFκB)的活化所引發的一連串發炎反應，和肌球蛋白重鏈亞型轉換。此實驗顯示在糖尿病動物模式下，一氧化氮生成酶之去偶合所引發的自由基離子是造成糖尿病心肌病變的主要原因。此實驗為藥理研究提供了一個未來的研究發展方向。 本博士論文結合心臟力學和分子醫學之技術，研究糖尿病心肌病變的演變過程，其中可能的致病機制。首先，在早期糖尿病心肌病變中，有利的心臟負荷條件促使最大理論血流量和心輸出量的提升，藉以代償在早期糖尿病心肌病變所發生的最大收縮力下降，使心臟收縮功能得以維持。然而，糖尿病心肌病變後期的最大理論血流量開始衰減，同時造成心搏量和心輸出量的下降，導致無法代償而造成顯著的心衰竭 (overt heart failure)。我們的研究結果顯示，最大理論血流量可以作為速度依賴性的心臟收縮功能之指數，可用於預測顯著的心衰竭的發生。此外，心臟機械效率的減少，而不是主動脈液壓能量轉移的減少，可能會引起不利的糖尿病心肌病變作用，而導致促成顯著心衰竭的主因之一。透過最大收縮力、最大理論血流量、心臟負荷條件和心臟機械效率的時序變化的研究結果，我們提供了一個研究方法，以了解糖尿病心肌病變的發病機制。文獻報告指出，肌球蛋白重鏈的異構體轉換可能造成後負荷調整的最大理論血流量的下降。這是值得深入研究，以闡明可能的分子機制，如胞外部分（即膠原蛋白）和幾何因素（即同心心室肥大或是離心心室肥大）。其次，我們證明 DMTU具有預防糖尿病心肌病變的療效。DMTU可使早期糖尿病心肌病變之最大收縮力和後負荷調整之最大理論血流量正常化。然而，在晚期糖尿病心肌病變中，DMTU僅能改善後負荷調整之最大理論血流量，但對於心臟最大收縮力則無顯著的影響。在慢性糖尿病大鼠心肌中，DMTU透過減少脂質過氧化以增強抗氧化能力，以調控心肌對氧化壓力反應的影響。至於對胰島素的分泌，DMTU並無增加的作用。DMTU明顯改善速度依賴性的心臟收縮力指數，而對力量依賴性的心臟收縮力指數則無顯著差異。在慢性糖尿病大鼠模式中，DMTU治療的優勢可能會涉及改善心肌特異性轉錄因子，如MEF-2和eHAND，以及逆轉肌球蛋白重鏈亞型轉換 (mysin heavy chain switch)。這些研究結果將提供另一個改善糖尿病心肌病變的治療方法。最後，本論文證實，過度的氧化壓力和氮化壓力，均會產生大量的自由基離子，引發實驗性糖尿病之心臟內收縮蛋白和酵素的功能失調，並引發與發炎反應相關的轉錄因子活化，因此，肌球蛋白重鏈亞型轉換和過氧亞硝酸鹽(peroxynitrite)成為糖尿病心肌病變的可能主因。

Diabetes mellitus may result in impaired cardiac contractility, but the underlying mechanical and molecular mechanisms remain unclear. There are three major aims of the present doctoral thesis. First, we aimed to investigate the temporal alternations in cardiac mechanics (i.e., force-dependent and velocity-dependent indices of cardiac contractility) in the evolution of diabetic cardiomyopathy (DCM) in terms of the elastance-resistance left ventricle (LV) pump model. In addition, we investigated the temporal changes in loading conditions, mechanical efficiency and ventriculoarterial coupling to elucidate the detailed cardiac mechanics attributable to decompensated heart failure in streptozotocin (STZ)-induced diabetic rats. Second, we aimed to evaluate the preventive and therapeutic effects of dimethylthiourea (DMTU), a potent hydroxyl radical scavenger, on force-dependent and velocity-dependent indices of cardiac contractility in both early and chronic stages of STZ-diabetic rats. We sought to do this in terms of force-dependent and velocity-dependent indices of myocardial contractility by using the elastance-resistance LV model. Furthermore, we planned to carry out experiments to examine downstream transcription factors, such as myocyte enhancer factor-2 (MEF-2) and heart autonomic nervous system and neural crest derivatives (dHAND and eHAND), activated by oxidative stress and markers of oxidative stress and the expression of isoforms of myosin heavy chain (MHC) in STZ-diabetic rat hearts. Third, we also evaluated the therapeutic efficacy of NG-nitro-L-arginine methyl ester (L-NAME), a non-specific inhibitor of nitric oxide synthases (NOS), in STZ-diabetic rats in terms of the elastance and resistance of left ventricle (LV). Furthermore, we planned to carry out experiments to investigate the underlying molecular mechanisms by examining the levels of nitrosative stress and oxidative stress markers, and its downstream signaling, i.e., nuclear factor-κB (NFκB), and coupling of NOS, as well as modulation of MHC isoform in STZ-diabetic rats.

In the first part, we aimed to investigate the temporal alterations in cardiac mechanics, loading conditions as well as mechanical efficiency in the evolution of systolic dysfunction in STZ-diabetic rats. Adult male Wistar rats were randomized into control and STZ-induced diabetic groups. Invasive hemodynamic studies were done at 8, 16, and 22 weeks post STZ injection. Maximal systolic elastance (Emax) and maximum theoretical flow (Qmax) were assessed by curve fitting techniques; ventriculoarterial coupling and mechanical efficiency by a single beat estimation technique. In contrast to early occurring and persistently depressed Emax, Qmax progressively increased with time, but was decreased at 22 wks post STZ injection, which temporally correlated with the changes in cardiac output. The favorable loading conditions enhanced stroke volume and Qmax, while ventriculoarterial uncoupling attenuated the cardiac mechanical efficiency in diabetic animals. The changes in Emax and Qmax are discordant during the progression of contractile dysfunction in the diabetic heart. Our present study showed that attenuated afterload-adjusted Qmax (Qmaxad) and afterload-adjusted LV weight normalized Qmax (Qmaxadn) and cardiac mechanical efficiency, occurring preceding overt systolic heart failure, are two major determinants of deteriorating cardiac performance in diabetic rats. After validation of the usefulness of both force-dependent and velocity-dependent indices of cardiac contractility in diabetic cardiomyopathy, we then aimed to investigate the preventive and therapeutic effects of DMTU on cardiac mechanics in STZ-diabetic rat hearts. It has been reported that hydroxyl radicals and hydrogen peroxide are involved in the pathogenesis of systolic dysfunction in diabetic rats, but the precise mechanisms and the effect of antioxidant therapy in diabetic subjects have not been elucidated. We aimed to evaluate the effects of DMTU on both force-dependent and velocity-dependent indices of cardiac contractility in STZ-induced early and chronic diabetic rats. Seventy-two hours and eight weeks after STZ (60 mg/kg) injection, diabetic rats were randomized to either DMTU (50 mg/kg/day, IP) or vehicle treatment for 6 and 12 weeks, respectively. All rats were then subjected to invasive hemodynamic studies. Again, Emax and Qmax were assessed by curve fitting techniques in terms of the elastance-resistance model. Both LV weight normalized Emax (Emaxn) and Qmaxad were depressed in diabetic rats, concomitant with altered MHC isoform composition and its upstream regulators, such as MEF-2 and heart autonomic nervous system and neural crest derivatives (eHAND and dHAND). In chronic diabetic rats, DMTU markedly attenuated the impairment in Qmaxad, and normalized the expression of MEF-2 and eHAND, and MHC isoform composition, but exerted an insignificant benefit on Emaxn. Regarding preventive treatment, DMTU significantly ameliorated both Emaxn and Qmaxad in early diabetic rats. We also checked blood glucose and insulin concentrations in controls and diabetic rats. The results showed that plasma insulin levels were significantly decreased after STZ injection and blood glucose levels were significantly increased in diabetic rats, but there were no significant differences after DMTU treatment in diabetic groups. In the second part, our current study shows that the advantage of DMTU in chronic diabetic rats might involve normalization of MEF-2 and eHAND, as well as reversal of MHC isoform switch. Finally, we evaluated the therapeutic efficacy of L-NAME, a non-specific inhibitor of NOS, in STZ-diabetic rats in terms of the elastance and resistance of left ventricle. Several reports have shown that nitrosative stress (NS) plays an essential role in diabetic cardiomyopathy. However, the precise mechanism by which the NS leads to compromised cardiac contractility has not been elucidated. We aimed to test the hypothesis that uncoupled endothelial NOS (eNOS) under excessive nitrosative stress and oxidative stress may enhance the translocation of NFκB and the resultant MHC proteolysis and isoform switch. Four weeks after STZ (60 mg/kg) or vehicle injection, male Wistar rats were randomized to receive treatment with either L-NAME (30 mg/kg/day in drinking water) or vehicle for another 8 weeks and then followed by invasive hemodynamic studies. Similarly, both Emax and Qmax were assessed by curve fitting techniques; the Ea by a single beat estimation technique. Analysis of low temperature sodium dodecyl sulfate polyacrylamide gel eletrophoresis (SDS-PAGE) of eNOS and Western blotting of NFκB-p65 in nuclear extract of the LV were done in the control, STZ and STZ+L-NAME groups. In parallel, catalase and oxidized to reduced glutathione ratio (GSSG/GSH ratio), and 3-nitrotyrosine (3-NT) in the LV were also measured. Both Emaxn and Qmaxadn were significantly depressed in 12-week diabetic rats, accompanied with nuclear translocation of NFκB-p65 and MHC isoform switch. L-NAME treatment not only attenuated the reductions in both Emaxn and Qmaxadn in diabetic rats, but also significantly modulated NFκB translocation, and MHC isoform switch analyzed by real time polymerase chain reaction (RT-PCR). We demonstrated that chronic inhibition of NOS reduced uncoupled eNOS and levels of NS and oxidative stress markers on one hand, and improved both in vivo force-dependent and velocity-dependent indices of cardiac contractility on the other hand. Our results suggested that NS might play an essential role in the pathogenesis of diabetic cardiomyopathy by mediating the activation of NFκB, and subsequent MHC isoform switch. In conclusion, the present doctoral thesis combined cardiac mechanics studies and molecular studies to demonstrate the possible mechanisms attributable to overt heart failure in the evolution of diabetic cardiomyopathy. First, at early stage of DCM, the enhanced Qmax and favorable loading conditions play complementary roles to persistently depressed Emax, and cardiac performance is then well-preserved in early period of diabetic rats. However, Qmax is attenuated at later stages, which is temporally correlated with the declines in stroke volume and cardiac output. The compensatory offset between Emax and Qmax would be lost and depressed cardiac performance would then ensue. Like systolic elastance, Qmax can serve as a velocity-dependent dimension of cardiac contractile function and can predict the occurrence of overt systolic dysfunction. In addition, cardiac mechanical efficiency, rather than aortic hydraulic energy transfer, is diminished preceding overt systolic heart failure, and may play a detrimental role in the evolution of contractile dysfunction in STZ-induced diabetic rats. By unraveling the temporal changes in Emax and Qmax, and loading conditions as well as cardiac mechanical efficiency, our present study provides a comprehensive understanding of the pathogenesis of contractile dysfunction in diabetic rat heart. It has been reported that isoform switch of MHC might play a role in depressed Qmaxad, it warrants further study to elucidate the possible molecular mechanisms attributable to attenuated Qmaxad, such as extracellular component (i.e., collagen), and geometric factors (i.e., concentric ventricular hypertrophy). Second, we demonstrated that DMTU normalizes Emaxn and Qmaxad in early stage of DCM, but normalizes only Qmaxad in late stage of diabetic cardiomyopathy. DMTU, by reducing lipid peroxidation and enhancing antioxidant capacity, has disparate effects on modulation of the myocardial response to increased oxidative stress in chronic diabetic rat hearts, with a dramatic restoration of velocity-dependent index of cardiac contractility but an insignificant benefit on force-dependent index. The advantages of DMTU treatment might involve normalization of cardiac-specific transcription factors, such as MEF-2 and eHAND, as well as the reversal of MHC isoform switch in chronic diabetic rats. These observations suggest a way towards an additional therapeutic approach to systolic dysfunction in diabetes. Finally, we showed that chronic inhibition of NOS may ameliorate both force-dependent and velocity-dependent indices of cardiac performance in diabetic rats. The underlying molecular mechanisms of L-NAME treatment might be mediated through attenuation of uncoupled eNOS and nitrosative/oxidative stress, and NFκB translocation from the cytosol to the nucleus, and subsequent reversal of MHC isoform switch.