Nitroxyl-mediated disulfide bond formation between cardiac myofilament cysteines enhances contractile function

Abstract

Rationale: In the myocardium, redox/cysteine modification of proteins regulating Ca(2+) cycling can affect contraction and may have therapeutic value. Nitroxyl (HNO), the one-electron-reduced form of nitric oxide, enhances cardiac function in a manner that suggests reversible cysteine modifications of the contractile machinery.

Objective: To determine the effects of HNO modification in cardiac myofilament proteins.

Methods and results: The HNO-donor, 1-nitrosocyclohexyl acetate, was found to act directly on the myofilament proteins, increasing maximum force (F(max)) and reducing the concentration of Ca(2+) for 50% activation (Ca(50)) in intact and skinned cardiac muscles. The effects of 1-nitrosocyclohexyl acetate are reversible by reducing agents and distinct from those of another HNO donor, Angeli salt, which was previously reported to increase F(max) without affecting Ca50. Using a new mass spectrometry capture technique based on the biotin switch assay, we identified and characterized the formation by HNO of a disulfide-linked actin-tropomyosin and myosin heavy chain-myosin light chain 1. Comparison of the 1-nitrosocyclohexyl acetate and Angeli salt effects with the modifications induced by each donor indicated the actin-tropomyosin and myosin heavy chain-myosin light chain 1 interactions independently correlated with increased Ca(2+) sensitivity and force generation, respectively.

Conclusions: HNO exerts a direct effect on cardiac myofilament proteins increasing myofilament Ca(2+) responsiveness by promoting disulfide bond formation between critical cysteine residues. These findings indicate a novel, redox-based modulation of the contractile apparatus, which positively impacts myocardial function, providing further mechanistic insight for HNO as a therapeutic agent.

Figure 1. HNO reacts with free thiols via a reactive intermediate

Following the formation of N-hydroxysulfinamide, the reaction proceeds by one of two pathways; A, isomerization to form a sulfinamide group or B, in the presence of an additional thiol formation of a disulfide bond and hydroxylamine.

Figure 2. Effect of NCA on rat cardiac muscle

A, Raw tracings of intracellular Ca²⁺ transient (left) and force (right) at varied NCA concentrations. B, Pooled data of the dose-response of [Ca²⁺]_i (left) and force development (right) to NCA (0–20 µmol/L). Note that twitch force increased significantly without increases in resting force at varied NCA concentrations (n=7–8/group). C, Effect of NCA on [Ca²⁺]_i transient (left) and force development (right) at varied external Ca²⁺. At any given [Ca²⁺]_o, twitch force increased significantly after NCA treatment while [Ca²⁺]_i transient was not affected. * p<0.05 versus no drug (n=5 in each group). D, Effect of NCA on force-frequency relation. NCA treatment did not affect [Ca²⁺]_i transient at any given frequencies of stimulation but increased force development at higher stimulation frequencies (* p<0.05 versus no drug, ** p<0.01 versus no drug, n=6 in each group).

Figure 3. NCA acts directly on the myofilament proteins increasing F_max and decreasing Ca₅₀

A, Steady-state force-[Ca²⁺]_i relationship in intact trabeculae before and after NCA (2.5 µmol/L) (n=5). B, Force-[Ca²⁺] relation in skinned trabeculae before and after NCA treatment (n=6). C, Reversal of NCA treatment’s effect on force-[Ca²⁺] in skinned muscles. The muscles were treated with DTT (5 mmol/L) for 10 min after first force-[Ca²⁺] was obtained in the presence of NCA treatment alone, and a second force-[Ca²⁺] relation was obtained in the presence of NCA+DTT treatment (n=3). D, 1-nitrosocyclohexyl pivalate (NCP), a pro-compound of NCA with similar structure that does not release HNO, did not affect force-[Ca²⁺] relation in skinned muscles (n=3). See Table 1.

Figure 4. Detection, capture and site identification of HNO modified proteins

A, Modified biotin switch assay schema outlining thiol blocking, reduction and biotin labelling steps, as well as capture of intact proteins or digested peptides for MS/MS analysis. B, Silver stained gel of rat cardiac myofibrils treated with HNO/NO-donors or control compounds and subjected to the biotin switch assay (n=3). HNO modifications were reduced by 5 mmol/L DTT but were resistant to treatment with 1 mmol/L ascorbate (black arrowheads) while NO modifications were reversed with ascorbate (outline arrowhead). C, Summary of subtractive proteomic site mapping study comparing AS and NCA treatments (n=3). Sites specific to NCA treatment were candidates for the Ca²⁺ sensitization (Ca₅₀) effect while sites in common were attributed to conferring the maximum force (F_max) increase. See Table 2 for identified sites and Online Table I for MS data.

Figure 5. HNO treatment induces the formation of disulfide linked dimers

A, 1 µg of rat cardiac myofibrils treated with HNO/NO donors or control compounds was separated under reducing or non-reducing conditions, western blots probed for candidate proteins identified in Table 2 (n=4). In each case the change in mobility was reversed with treatment of 5 mmol/L DTT. B, MS analysis of silver stained gel confirmed presence of monomeric actin in the NCA treatment and also revealed a similar, but less abundant, species of actin in AS treated samples (n=2) (Online Table II). C, Evaluation of the interaction between actin and TM with NCA treatment. Purified rabbit skeletal TM (0.03 µg) and isolated rat cardiac myofibrils (1 µg) were treated with NCA or AS and evaluated by 1D non-reducing western blot probing for TM (left) and actin (right) indicating a co-migrating species reversed by DTT (n=3). D, Fluorescent DIGE gel of the same samples were independently labelled; purified TM (Cy2-blue), NCA treated myofibrils (Cy3-green) and AS treated myofibrils (Cy5-red) (n=3). E, MS analysis (lower right) of the same gel region identified both actin and TM only in the non-reduced lane (n=2) (Online Table II). This analysis demonstrates a NCA specific actin-TM heterodimer linked by the formation of a disulfide bond.

Figure 6. MLC1 Cys81 is necessary for increase in F_max induced by treatment with HNO donors

A, Sequence alignment comparing isoforms of rat cardiac and skeletal MLC1 in the region surrounding cardiac Cys81. B, Steady-state force–[Ca²⁺]_i relations in before (open symbols) versus after (filled symbols) NCA treatment (2 µmol/L) from cardiac or skeletal muscles indicating loss of force increase with loss of Cys81 in skeletal isoform with NCA (n=5, each group). C, Diagonal gel shift assay (non-reduced (NR) and reduced (DTT, 100 mmol/L)) using 10 µg of skeletal or cardiac myofibrils indicating loss of higher molecular weight forms of MLC1 in skeletal HNO treated preparations while maintaining the higher molecular weight form of TM (n=3).