Enhancing atrial‐specific gene expression using a calsequestrin cis‐regulatory module 4 with a sarcolipin promoter

Abstract Background Cardiac gene therapy using the adeno‐associated virus serotype 9 vector is widely used because of its efficient transduction. However, the promoters used to drive expression often cause off‐target localization. To overcome this, studies have applied cardiac‐specific promoters, although expression is debilitated compared to that of ubiquitous promoters. To address these issues in the context of atrial‐specific gene expression, an enhancer calsequestrin cis‐regulatory module 4 (CRM4) and the highly atrial‐specific promoter sarcolipin were combined to enhance expression and minimize off tissue expression. Methods To observe expression and bio‐distribution, constructs were generated using two different reporter genes: luciferase and enhanced green fluorescent protein (EGFP). The ubiquitous cytomegalovirus (CMV), sarcolipin (SLN) and CRM4 combined with sarcolipin (CRM4.SLN) were compared and analyzed using the luciferase assay, western blotting, a quantitative polymerase chain reaction and fluorescence imaging. Results The CMV promoter containing vectors showed the strongest expression in vitro and in vivo. However, the module SLN combination showed enhanced atrial expression and a minimized off‐target effect even when compared with the individual SLN promoter. Conclusions For gene therapy involving atrial gene transfer, the CRM4.SLN combination is a promising alternative to the use of the CMV promoter. CRM4.SLN had significant atrial expression and minimized extra‐atrial expression.

been managed previously through transcriptional targeting. 8 Because the virus regulates transgene production through its promoter, promoter alterations have been used to control gene expression in various organs to avoid off-target effects. To overcome this barrier, cardiac-specific promoters have been applied.
Current cardiac-specific promoters investigated include α-myosin heavy chain, myosin light chain, enhanced myosin light chain and cardiac troponin T promoters. [9][10][11][12][13][14] Among the promoters investigated, some have shown chamber specificity. Because the atrium and ventricle differ morphologically, functionally and are molecularly distinct, pathophysiology also varies in cardiac diseases. Thus, using a cardiac-specific promoter for universal transgene expression in the heart in certain disease states can also affect healthy tissues negatively. Accordingly, the use of a specific promoter is critical in chamber-specific diseases.
Currently, several promoters have been observed to regulate cardiac chamber-specific gene expression. Among these are the atrial myosin light chain-2a, slow myosin heavy chain-3, atrial natriuretic factor and sarcolipin (SLN) for the atrium and the myosin light chain-2V for the ventricle. 15 The uniqueness of these promoters has been exploited in other methods in addition to improving specificity, such as allowing for cardiomyocytes generated from human-induced pluripotent stem cells to be separated by subtype-specific promoter-driven action potentials. 16 However, the application of these promoters to drive chamber-specific transgene expression has been limited as a result of the compromised efficiency of these promoters. To apply these promoters for use in gene therapy, the efficient transduction of these promoters is essential.
To overcome this, two sequences established previously were used to make an atrial-specific promoter with enhanced expression. We approached this issue by taking advantage the atrial-specific promoter of SLN. The SLN promoter was chosen because the SLN protein is characteristically expressed in the atrial chamber of the heart. Furthermore, it has been observed to change dependent on disease states. Sarcolipin was found to be up-regulated in rodent models of congenital heart disease and in patients with preserved left ventricular ejection fraction and chronic isolated mitral regurgitation. 17,18 By contrast, SLN was found to be down-regulated in patients with chronic atrial fibrillation. 19 In Duchenne muscular dystrophy mice, a reduction in SLN expression was found to mitigate associated cardiomyopathy. 20 Alone, the SLN promoter activity is unsubstantial compared to that of ubiquitous promoters. Thus, the addition of calsequestrin 2 cis-regulatory module 4 (CRM4), a cardiomyocyte-specific enhancer, which showed superior activity in the heart, was used to heighten transgene expression of SLN within the heart. 21 In addition to cardiac tissue, high SLN levels are found in the skeletal muscle and diaphragm. 22

| pTR.SLN.Luc
The shortened SLN promoter used was 1029 bp. This was identified as the primary core promoter sequence by removal of polyadenylation sequence from a 1253-bp human SLN promoter pur-

| Generation of AAV
The following self-complementary AAV (serotype 9) constructs were

| Transverse aortic constriction (TAC)
Mice were anesthetized with a solution mixture of 95 mg kg -1 ketamine and 5 mg kg -1 xylazine administered via intraperitoneal injection.
Mice were ventilated with a tidal volume of 0.2 mL and a respiratory rate of 110 breaths per minute (Harvard Apparatus). A longitudinal incision of 2-3 mm was made in the proximal sternum to allow visualization of the aortic arch. The transverse aortic arch was ligated between the innominate and left common carotid arteries with an overlaid 27-gauge needle. The needle was then immediately removed, leaving a discrete region of constriction.

| Fluorescence imaging
Mouse heart tissues were cut in a sagittal plane and fixed in 4% paraformaldehyde for 24 hours and moved to 20% sucrose for cryoprotection for 24 hours before being cryopreserved to OCT compound (Tissue-Tek, Sakura Finetek, Torrance, CA, USA) at -80°C. Tissues were sectioned into 6-μm thick slices (CM 3050S; Leica, Wetzlar, Germany). The slides were dehydrated in 100% ethanol for 10 minutes and dipped in two changes for xylene. Slides were sealed with glass slides using mounting medium with 4′,6-diamidino-2-phenylindole (DAPI) (Vectashield, Burlingame, CA, USA) and analyzed using a

| Western blot analysis
Tissue homogenates were lysed using RIPA buffer (Thermo Scientific) with protease inhibitor cocktail (Sigma-Aldrich). Protein quantification was conducted using Pierce BCA Protein Assay Kit (Thermo Scientific).
Proteins were resolved on 10% SDS-PAGE gels followed by transfer to nitrocellulose membrane (Bio-Rad, Munchen, Germany). Antibodies

| Statistical analysis
Statistical analyses were performed using a two-tailed Student's t-test, with significant differences demonstrated as appropriate. Data are reported as the mean ± SD. sequence significantly improved SLN promoter activity (p < 0.005).  A luciferase assay was conducted to determine bio-distribution. Luciferase activity was normalized to background values. Statistical significance was measured with a two-tailed Student's t-test and significant differences are indicated: *p < 0.05, **p < 0.01, ***p < 0.005. Bar graphs represent the mean ± SD (n = 4) (p < 0.005), skeletal muscle (p < 0.01), diaphragm (p < 0.005), lung (p < 0.05), liver (p < 0.005) and kidney (p < 0.005) compared to control.

| EGFP
The luciferase was exchanged with an EGFP reporter gene: AAV9-

Protein analysis
Bio-distribution was observed in the atrium, ventricle, skeletal muscle, diaphragm, brain, lung, liver and kidney with western blotting using
Quantified luciferase values are shown in the table in supporting information (Table S1).   25 However, the promoter used to drive expression of the gene of interest often causes undesired activity in off-target tissues, especially in hepatic tissue. 7 Thus, cardiac-specific promoters have been identified to improve cardiac specificity. [9][10][11][12][13][14][15][16] Even within these promoters, chamber specificity has been noted. 14,15,23 Because the atrium and ventricle have individual functions, mechanisms and molecular signatures, pathophysiology varies in cardiac diseases as well.

Fluorescence imaging
Thus, using a chamber-specific promoter may be a better option than a global cardiac transgene expressing promoter. However, the transduction efficiency of chamber-specific promoters is often compromised as a result of poor activity. 10 With the current research on transcriptional regulation, we designed a combined a novel combination with synergistic effects to obtain enhanced atrial specificity. 21 Various enhancer sequences have been reviewed, 26-30 although FIGURE 5 mRNA analysis validated EGFP gene expression profiles. In vivo gene expression was observed 3 weeks post tail vein delivery of control and EGFP containing CMV, SLN and CRM4.SLN vectors. qRT-PCR analysis was conducted to determine bio-distribution with EGFP and normalized with 18S. Statistical significance was measured with a two-tailed Student's t-test and significant differences are indicated: *p < 0.05, **p < 0.01, ***p < 0.005. Bar graphs represent the mean ± SD (n = 4)

FIGURE 6
Visualization of biodistribution. In vivo gene expression in the atrium and ventricle of control and EGFP containing CMV, SLN and CRM4.SLN vectors was observed by fluorescence imaging using a confocal microscope at 200× magnification. Bright field, DAPI, EGFP and merged images were taken under same exposure and normalization settings CRM4 was the most promising candidate as a result of its superior cardiac specificity. Thus, we have manipulated the promoter by adding CRM4, a cardiomyocyte-specific enhancer, to improve transcriptional regulation of SLN. 21 In the present study, we evaluated an atrial-specific promoter sequence enhanced by a cis-regulatory module in vitro and in vivo. Initially, to evaluate promoter efficiency, we observed expression in HL1 cells and confirmed that the CRM4 within the gene cassette significantly improved luciferase activity driven by the SLN promoter in atrial cells (p < 0.005) ( Figure 3A). Thus, we followed with in vivo bio-distribution studies in wild-type mice to quantify potential offtarget effects in six other tissues, including the skeletal muscle, diaphragm, brain, lung, liver and kidney. Skeletal muscle and diaphragm were investigated because high SLN expression was reported. 22,23 Other tissues were also measured to evaluate AAV9 off-target effects. [5][6][7] To compare gene expression levels under SLN and CRM4. SLN promoters, two reporter genes were used for assessment: luciferase and EGFP. The luciferase assay showed that SLN.Luc was not only expressed in the atrium, but also detected in the ventricle and liver.
On the other hand, the CRM4.SLN.Luc showed highly selective luciferase activity in the atrium compared to the ventricle and other tissues.
Strikingly, CRM4 significantly minimized off-target effects even compared to SLN promoter used alone.
For alternative verification for protein and mRNA analysis, EGFP was used as a supplementary reporter gene. Protein evaluation with western blotting supported the luciferase assay data, which showed that the CRM4.SLN.EGFP had exceptional atrial specificity and basal expression level in other tissues (Figure 4). The EGFP reporter gene offered additional insights by showing expression in skeletal muscle as well. Because SLN is also significantly expressed in skeletal muscle, expression was not unanticipated. However, in conjunction with the luciferase assay data, which showed expression in the atrium to be significantly stronger when all samples were compared in reference, skeletal expression was concluded to be lower than atrial expression. qRT-PCR data of the EGFP transferred mice also supported atrial-specific expression with CRM4.SLN.EGFP ( Figure 5). Inconsistencies between RNA and protein levels in certain tissues such as the lung, kidney and brain reflect that our acquired data for qRT-PCR showed relative values instead of actual mRNA. Furthermore, visualization of EGFP fluorescence was observed with sectioned tissue of the atrium and ventricle ( Figure 6). The strong correlation between the luciferase and EGFP reporter genes both supported that the CRM4.SLN combination enhanced atrial specificity. Furthermore, we tested whether specificity to the atrium was preserved at higher doses of CRM4.SLN.
Although, at lower dosages, EGFP was previously not observed in the liver, at higher dosages, expression in the liver occurred. Nevertheless, CRM4.SLN confers exclusive atrial specificity among these three promoters.
For future application of this construct, the CRM4.SLN can be used for atrial-specific gene therapy in disease models. Because sarcolipin expression is increased in heart failure states (Figure 1), the sarcolipin promoter in diseased states may be advantageous, similar to that of the ANF promoter. 11 Furthermore, current therapeutic gene transfer for atrial fibrillation is highly limited, such as direct injection to the right and left atrium or to the atrioventricular node for rate control. 31,32 However, these approaches can lead to tissue damage and inflammatory response. 33,34 The most effective reported method is the gene painting method, which uses a poloxamer gel, dilute trypsin and vector mixture to increase contact time and penetration. 35 However, this must be performed in open heart settings, which increases mortality risk. 36 The use of this construct can offer a non-invasive and technically simple approach for atrial-specific gene therapy. The CRM4 and SLN combination caused robust and highlyspecific atrial activity and is a promising method of targeting transcriptional mechanisms to improve atrial transgene transduction.