Phosphorylation of Lamin A/C at serine 22 modulates Nav1.5 function

Abstract Variants in the LMNA gene, which encodes for Lamin A/C, are associated with cardiac conduction disease (CCD). We previously reported that Lamin A/C variants p.R545H and p.A287Lfs*193, which were identified in CCD patients, decreased peak I Na in HEK‐293 cells expressing Nav1.5. Decreased peak I Na in the cardiac conduction system could account for patients’ atrioventricular block. We found that serine 22 (Ser 22) phosphorylation of Lamin A/C was decreased in the p.R545H variant and hypothesized that lamin phosphorylation modulated Nav1.5 activity. To test this hypothesis, we assessed Nav1.5 function in HEK‐293 cells co‐transfected with LMNA variants or treated with the small molecule LBL1 (lamin‐binding ligand 1). LBL1 decreased Ser 22 phosphorylation by 65% but did not affect Nav1.5 function. To test the complete loss of phosphorylation, we generated a version of LMNA with serine 22 converted to alanine 22 (S22A‐LMNA); and a version of mutant R545H‐LMNA that mimics phosphorylation via serine 22 to aspartic acid 22 substitution (S22D‐R545H‐LMNA). We found that S22A‐LMNA inhibited Lamin‐mediated activation of peak I Na by 63% and shifted voltage‐dependency of steady‐state inactivation of Nav1.5. Conversely, S22D‐R545H‐LMNA abolished the effects of mutant R545H‐LMNA on voltage‐dependency but not peak I Na. We conclude that Lamin A/C Ser 22 phosphorylation can modulate Nav1.5 function and contributes to the mechanism by which R545H‐LMNA alters Nav1.5 function. The differential impact of complete versus partial loss of Ser 22 phosphorylation suggests a threshold of phosphorylation that is required for full Nav1.5 modulation. This is the first study to link Lamin A/C phosphorylation to Nav1.5 function.


| INTRODUCTION
Laminopathies are a broad spectrum of diseases involving cardiac and skeletal anomalies attributed to mutations in the LMNA gene, which encodes for Lamin A/C. These disorders comprise over 12 clinically heterogeneous syndromes (Bertrand et al., 2011). Lamin A/C is a type-V intermediate filament that forms the nuclear lamina underlying the inner membrane of the nucleus (Houben et al., 2013). LMNA mutations are highly prevalent in patients with cardiac conduction disease (CCD), often associated with eventual dilated cardiomyopathy (DCM; Anselme et al., 2013;Barra et al., 2012;Hermida-Prieto et al., 2004;Keller et al., 2012;Malek et al., 2011;Mounkes et al., 2005;Olaopa et al., 2018;Saj et al., 2010;Shaw et al., 2002;Zaragoza et al., 2016). A clinical study of familial autosomal DCM found that approximately 33% of patients were positive for LMNA mutations (Arbustini et al., 2002). LMNA patients require implantable cardioverterdefibrillator or pacemaker therapy to prevent cardiac arrest and sudden death (Anselme et al., 2013;Kumar et al., 2016;Olaopa et al., 2018). The clinical presentations of these patients, including atrioventricular (AV) block and progressive CCD, are similar to those observed in disorders caused by mutations in the SCN5A gene (Olaopa et al., 2018;Wang et al., 2002), which encodes for the cardiac sodium channel (Na v 1.5).
Na v 1.5 is primarily localized within the sarcolemma, in contrast to the nuclear localization of Lamin A/C. Mutations in SCN5A are a major cause of AV block, sick sinus syndrome, progressive CCD, and eventual DCM leading to sudden death (Amin et al., 2010;Chockalingam et al., 2012;Holst et al., 2010;Lee et al., 2016;Makita, 2009;Shuraih et al., 2007). Due to the similarities between clinical presentations of disorders associated with both LMNA and SCN5A mutations, studies by our group and others have sought to functionally link variants in Lamin A/C to Na v 1.5 activity (Liu et al., 2016;Markandeya et al., 2016;Olaopa et al., 2018). Our group reported that two Lamin A/C variants (p.R545H and p.A287Lfs*193), found in patients with CCD, significantly decrease peak sodium current (I Na ) and shift the voltage-dependency of steady-state inactivation of Na v 1.5 (Olaopa et al., 2018). The p.R545H variant is due to a missense point mutation (c.1634G>A); while the p.A287Lfs*193 variant is caused by a single nucleotide deletion (c.859delG) and subsequent frame shift, leading to a premature termination codon (Olaopa et al., 2018).
Although the transcriptional and post-translational regulation of Lamin A/C within the heart is not fully understood, its phosphorylation has functional importance (Buxboim et al., 2014;Haas & Jost, 1993;Kochin et al., 2014;Mitsuhashi et al., 2010;Torvaldson et al., 2015;Wu et al., 2011). Phosphorylation at serine 22 (Ser 22) plays an important role in cell cycle regulation, nuclear stability, and signaling between the nucleoskeleton and cytoskeletal structures--the LINC (Linker of Nucleoskeleton and Cytoskeleton) complex (Buxboim et al., 2014;Kochin et al., 2014;Osmanagic-Myers et al., 2015;Torvaldson et al., 2015). Additionally, a p.S22L Lamin A/C variant was identified in a genetic screen of cardiac transplant patients with DCM (Pethig et al., 2005), indicating a clinically relevant role for Ser 22 in cardiac disease. Thus, we sought to determine if Ser 22 is involved in the mechanism by which the p.R545H variant affects Na v 1.5 function, with the aim of identifying a potential therapeutic target for CCD patients with similar LMNA mutations. We used a novel small molecule (lamin-binding ligand, LBL1) that binds Lamin A/C in the N-terminal region encompassing Ser 22 (Chao et al., 2017;Li et al., 2018), as well as genetic (Ser 22 modification) approaches to modulate Ser 22 phosphorylation and determine its role in Na v 1.5 function.

| Cell culture and transfection
HEK-293 cells were grown and maintained in DMEM media (Thermo Fisher), supplemented with 10% fetal bovine serum (ATCC). Cells were serum-starved overnight and transiently transfected with identical amounts of plasmid DNA using Effectene (Qiagen). Cells were treated 24 hrs post-transfection with vehicle dimethyl sulfoxide (1% DMSO), 5 or 10 µM LBL1. Cells were harvested 24 hrs post-treatment or 48 hrs post-transfection.

| Patch clamp experiments
Cells were harvested using trypsin-EDTA (Thermo Fisher) for 2 min and transferred to the patch chamber for wholecell recording of GFP-labeled cells, which was performed as previously described (Olaopa et al., 2018;Yu et al., 2014). Whole-cell configuration was made in bath solution (in mM): NaCl 140, KCl 5, CaCl 2 1.8, MgCl 2 1, 4-(2-hydro xyethyl)-1-piperazineethanesulfonic acid (HEPES) 5, and Glucose 10 (pH 7.4 adjusted with NaOH). Pipette resistances were 1.5-3 MΩ; solution (in mM): NaF 10, CsF 110, CsCl 20, ethylene glycol tetraacetic acid (EGTA) 10, and HEPES 10 (pH 7.35 adjusted with CsOH). After achieving a giga-seal, the test pulse current was nulled by adjusting the pipette capacitance compensator with both fast and slow components. After break-in, the whole-cell charging transient was nulled by adjusting whole-cell capacitance and series resistance. Voltage control protocols were generated with Axopatch 200B amplifier/Digidata 1440A or MultiClamp 700A acquisition system using pCLAMP-10 software (Molecular Devices). Whole-cell recording was analyzed using Clampfit 10.2 (Molecular Devices). All experiments were carried out at room temperature. Conductance G (V) was calculated by the equation: where I is the peak current, E rev is the measured reversal potential, and V m is the membrane potential. The normalized peak conductance was plotted as a function of membrane potentials. Steady-state inactivation was estimated by pre-pulse protocols (500 ms) from a holding potential of −140 mV. Steady-state activation and inactivation were fitted with the Boltzmann equation: where y represents variables; V h , midpoint; k, slope factor; and V m , membrane potential.

| Statistical analyses and data availability
The Mann-Whitney-Wilcoxon rank test was performed for patch clamp analyses. One-way ANOVA test was performed for western blot densitometry. For statistical significance p < 0.05 was used. Data are presented as mean ± SE. The data associated with this manuscript will be made available.

| Mutagenesis of Ser 22 residue
We generated LMNA phosphorylation plasmids that either mimic loss of phosphorylation in wild type (S22A-LMNA) or constitutive phosphorylation in the R545H-LMNA mutant (S22D-R545H-LMNA). We selected the R545H-LMNA mutant for our study due to the previously described decrease in Ser 22 phosphorylation. To mimic loss of phosphorylation, we genetically substituted alanine (GCG codon) for serine (TCG codon) residue at position 22 (p.S22A); conversely, to mimic phosphorylation, we genetically substituted aspartic acid (GAC codon) for serine (TCG codon) residue at position 22 (p.S22D). We confirmed successful mutagenesis by sequencing and western blot (Figure 1). Cells that expressed either the S22A-LMNA or S22D-R545H-LMNA plasmids appeared negative for Ser 22 phosphorylation, indicating the serine residue had been successfully mutated.

| Ser 22 Lamin phosphorylation is reduced in R545H-LMNA mutant
We found that Ser 22 Lamin A/C phosphorylation was reduced by 60% in cells expressing R545H-LMNA compared to wild type (Figure 2a,c). The A287Lfs-LMNA mutation decreased total Lamin A/C levels but did not impact Ser 22 phosphorylation (Figure 2a,d) and was therefore not a subject of this study. R545H-LMNA did not alter Na v 1.5 whole-cell protein content (Figure 2b), indicating that the decrease in peak I Na was not due to a change in total channel expression. We did not assess localization of Na v 1.5 to the membrane, and cannot rule out a role for decreased cell surface expression of sodium channels contributing to the decrease in peak I Na . Alpha-actinin 2, which has been shown to modulate Na v 1.5 function via direct interaction (Ziane et al., 2010), was likewise unchanged by R545H-LMNA (Figure 2b).

| Mimicking Ser 22 phosphorylation in R545H-LMNA partially restores Na v 1.5 function
To test if Ser 22 played a role in the mechanism by which R545H-LMNA decreased peak I Na (Olaopa et al., 2018), we generated a version of R545H-LMNA that substituted aspartic acid for serine at position 22 (S22D-R545H-LMNA) to mimic phosphorylation at that site. We then measured peak I Na in cells transfected with S22D-R545H-LMNA. We . Thus, mimicking Ser 22 phosphorylation partially rescued Na v 1.5 function disrupted by the R545H-LMNA mutation, as normal voltage-dependency was restored but peak I Na levels were not (Table 1).   reduced Ser 22 phosphorylation by 65% while 5 µM LBL1 had no effect (Figure 5a-c). To determine the effect of LBL1-suppressed Lamin A/C phosphorylation on Na v 1.5 function, we assessed peak I Na and voltage-dependency in cells treated with 10 µM LBL1 compared to vehicle. We anticipated that inhibition of Lamin Ser 22 phosphorylation would lead to changes in Na v 1.5 function similar to those seen with the R545H-LMNA or S22A-LMNA mutations. However, decreasing Ser 22 phosphorylation by LBL1 in wild-type Lamin did not change peak I Na or steady-state inactivation or activation (Figure 5d-f; Table 1).

| DISCUSSION
The goal of this study was to understand the effect of Lamin A/C Ser 22 phosphorylation on Na v 1.5 function, with the aim of identifying a novel therapeutic target for CCD patients with LMNA mutations. Our findings indicate that Ser 22 phosphorylation plays a role in Laminmediated Na v 1.5 modulation and can alter peak I Na and voltage-dependency of the steady-state inactivation.
However, there appears to be a threshold level of phosphorylation necessary in wild-type Lamin A/C. Partial loss of Ser 22 phosphorylation via the lamin-binding small molecule LBL1 did not affect Na v 1.5 function, but complete loss of Ser 22 phosphorylation via genetic substitution had significant effects on channel function that were consistent with the R545H-LMNA variant (Table 1). Another possible explanation for the discrepancy between the effect of partial and complete loss of Ser 22 phosphorylation is that other Lamin A/C phosphorylation sites, in addition to Ser 22, may play a cumulative role in modulating Na v 1.5. This alternative interpretation is supported by our findings, which show that mimicking constitutive Ser 22 phosphorylation in the R545H-LMNA mutant only partially restored Nav1.5 function. This suggests that there are other factors and/or phosphorylation sites that also play a role in governing the effect of R545H-LMNA or other LMNA mutations on sodium channel function or cell surface expression. To explore this possibility, future studies will be focused on assessing the role of additional Lamin A/C phosphorylation sites and developing chemical and genetic tools that can specifically identify and modulate these sites.
Lamin A/C is an intermediate filament organized into a tripartite structure: a non-helical N-terminal head domain encompassing the Ser 22 phosphorylation site; an αhelical coiled-coil rod domain encompassing the 287 residue; and an immunoglobulin fold (Ig-fold) domain at the C-terminal end encompassing the 545 residue (Ho & Lammerding, 2012;Osmanagic-Myers et al., 2015). The Ig-fold domain, which spans residues 430-545, is believed to be involved in Lamin A/C's protein-protein functional interactions (Dhe-Paganon et al., 2002;Krimm et al., 2002;Shumaker et al., 2008), including with cytoskeletal linkers that provide the machinery by which sarcolemma proteins like Na v 1.5 might be modulated (Markandeya et al., 2016;Olaopa et al., 2018;Stroud et al., 2014). Consistent with this, both the R545H and A287Lfs mutations alter the Igfold domain. R545H is a point mutation at the end of the Ig-fold, a location that has been implicated in other laminopathies including DCM and CCD with loss of peak I Na (Chan et al., 2016;Liu et al., 2016;Malek et al., 2011;Saj et al., 2010). The A287Lfs frame shift mutation alters the sequence from A287 in the rod domain to the premature stop codon in the Ig-fold. In contrast, LBI1 binds Lamin A/C in the N-terminal region encompassing Ser 22 (Chao et al., 2017;Li et al., 2018).
Dimerization of Lamin A/C is driven by coiled-coil formation of its central rod domains (Ho & Lammerding, 2012) (Dhe-Paganon et al., 2002Krimm et al., 2002;Stuurman et al., 1998). Lamin A/C dimers assemble head-to-tail into polar polymers, which require an overlapping interaction between the head and tail domains (Heitlinger et al., 1992;Sasse et al., 1998). These polymers then laterally assemble in an anti-parallel fashion into nonpolar filaments (Ben-Harush et al., 2009). These reports, in addition to our findings, lead us to propose a structural paradigm by which the R545H mutation could affect Ser 22 phosphorylation via antiparallel head-to-tail interaction with neighboring dimers (graphical abstract). In the case of the A287Lfs mutant, which does not result in loss of Ser 22 phosphorylation levels, this head-totail interaction is likely abolished due to the frame shift and subsequent truncation of this region of the protein. This truncation could explain why modulation of Ser 22 phosphorylation level does not occur in the A287Lfs mutant. The complex structure of lamin polymers may also explain why a twofold increase in LBL1 concentration induced a significant loss of Ser 22 phosphorylation. Experimental approaches aimed at exploring these possibilities, while important, are outside the scope and focus of this study.
From a clinical perspective, our results suggest that Lamin A/C phosphorylation may be a potential therapeutic target for patients with specific LMNA mutations and CCD. Small molecules that enhance, rather than block, Ser 22 phosphorylation might partially restore Na v 1.5 function in disease caused by R545H and similar LMNA mutations. This would mimic the partial rescue of Na v 1.5 function we observed in cells expressing the S22D-R545H-LMNA variant. It could also help to improve diagnosis and prognosis in patients with similar LMNA mutations by developing tools that could assess their Lamin A/C phosphorylation state.
In summary, our data indicate that Lamin A/C Ser 22 phosphorylation modulates Na v 1.5 function. This phosphorylation appears to be part of the mechanism by which the R545H-LMNA mutation affects Na v 1.5 function. To our knowledge, this is the first study to link Lamin A/C phosphorylation and Na v 1.5 function.