Modulation of ADARs mRNA expression in congenital heart defect patients

Adenosine (A) to inosine (I) RNA editing, is a hydrolytic deamination reaction catalyzed by adenosine deaminase (ADAR) acting on RNA enzymes. RNA editing is a molecular process that involves the post-transcriptional modification of RNA transcripts. Interestingly, few studies have been carried out to determine the role of RNA editing in vascular disease. The current study found that in blood samples positive for congenital heart disease (CHD) ADAR1 and ADAR2 expression change at RNA level was opposite to each other. That is, an increase of ADAR1 mRNA was noticed in human CHD cases, whereas ADAR2 mRNA was vastly down-regulated. The increase in ADAR1 may be explained by the stress induced by CHD. The dramatic decrease in ADAR2 in CHD cases was unexpected and prompted further investigation into its effects on the heart. Therefore we performed expression analysis on a microarray data encompassing ischemic and non-Ischemic cardiomyopathy patient myocardial tissues. A strong down-regulation of ADAR2 was observed in both ischemic and especially non-ischemic cases. However, ADAR1 showed a mild increase in the case of non-ischemic myocardial tissues. To further explore the role of ADAR2 with respect to heart physiology. We selected a protein coding gene filamin B (FLNB). FLNB is known to play an important role in heart development. Although there were no observable changes in its expression, the editing levels of FLNB dropped dramatically in ADAR2-/- mice. We also performed miRNA profiling from ADAR2 -/- mice heart tissue revealed a decrease in expression of miRNAs. It is established that aberrant expression of these miRNAs is often associated with cardiac defects. This study proposes that sufficient amounts of ADAR2 might play a vital role in preventing cardiovascular defects.


Abstract 27
Adenosine (A) to inosine (I) RNA editing, is a hydrolytic deamination reaction catalyzed by 28 adenosine deaminase (ADAR) acting on RNA enzymes. RNA editing is a molecular process that 29 involves the post-transcriptional modification of RNA transcripts. Interestingly, few studies have 30 been carried out to determine the role of RNA editing in vascular disease. The current study 31 found that in blood samples positive for congenital heart disease (CHD) ADAR1 and ADAR2 32 expression change at RNA level was opposite to each other. That is, an increase of ADAR1 33 mRNA was noticed in human CHD cases, whereas ADAR2 mRNA was vastly down-regulated. 34 The increase in ADAR1 may be explained by the stress induced by CHD. The dramatic decrease 35 in ADAR2 in CHD cases was unexpected and prompted further investigation into its effects on 36 the heart. Therefore we performed expression analysis on a microarray data encompassing 37 ischemic and non-Ischemic cardiomyopathy patient myocardial tissues. A strong down-38 regulation of ADAR2 was observed in both ischemic and especially non-ischemic cases. 39 However, ADAR1 showed a mild increase in the case of non-ischemic myocardial tissues. To 40 further explore the role of ADAR2 with respect to heart physiology. We selected a protein 41 coding gene filamin B (FLNB). FLNB is known to play an important role in heart development.

55
Congenital Heart Disease (CHD) is defined as structural or functional heart defect. It belongs to 56 a heterogeneous group of diseases and can be classified anatomically, clinically, environmental factors such as maternal diabetes or rubella are identified in some cases but for 63 most babies born with a heart defect the cause remain unknown(6).

64
The multi-lineage differentiation during cardiogenesis is orchestrated by a precise spatial and 65 temporal regulation of gene expression. Genetic studies in humans and knockout embryos have 66 identified various genes, such as TBX5, GATA4,CX43,NOTCH1 and VEGF 67 responsible for sporadic and inherited CHD cases (7).

68
In humans, the most prevalent type of RNA editing is adenosine (A) to inosine (I) (8). This 69 complex post-transcriptional hydrolytic deamination reaction is carried out by adenosine 70 deaminase (ADAR) family of enzymes. This family acts on double stranded RNA and comprise 71 of three members ADAR1, ADAR2 and ADAR3. ADAR1 ad ADAR2 are actively involved in 72 adenosine deamination however, ADAR3 is non-functional (9). Different studies have shown 73 that the extent of RNA editing not only varies among individuals but also show high tissue 74 specificity. Approximately 2.5 million sites in human transcriptome undergo editing however a 75 vast majority of them lie in the Alu elements located mostly in the introns and UTR (untranslated 76 region) (10). However, the functional consequence of majority of RNA editing events still 77 remain elusive. RNA editing is known to modulate splicing, coding potential, transcript stability 78 and even alters the processing and targeting of the microRNAs (miRNA) (8,11,12 shown an increase in expression of a lysosomal cysteine protease encoded by cathepsin S RNA 87 (CTSS) via ADAR1 mediated RNA editing followed by HuR recruitment . Cathepsin S has a 88 role in vascular inflammatory processes and the CTSS mRNA editing is increased in hypoxic or 89 pro-inflammatory conditions as wells as in patients suffering from clinical or subclinical vascular 90 damage (14).

91
In the current study we have determined the RNA level of ADAR1 and ADAR2 in congenital 92 heart disease patients. We also checked the relative gene expression of FoxP1 which is an 93 important transcription factor crucial for angiogenesis. We found a strong down-regulation of 94 ADAR2 and an up-regulation of ADAR1. Interestingly, microarray data analysis of human non-95 ischemic myocardial tissues showed similar trend. Interestingly the ischemic myocardial tissues 96 showed completely opposite trend. To further explore the role of ADAR2 in heart physiology, 97 we used ADAR2 knockout mouse. Although no strong anomaly in heart physiology was 98 observed as documented previously (15) however, we found down-regulation of different 99 microRNAs. reports were consulted to confirm the presence of congenital heart defects and all sample 107 collection was done pre-operatively. Whereas 13 control samples were collected from healthy 108 individuals using same parameters. Interviews were conducted personally using the specified 109 questionnaires. Information on age, gender, medication and family history was recorded.

110
Perspective study was initiated after getting approval from ethical committees of both CIIT and 111 collaborating hospitals.   Female mouse whole heart as dissected at the age of post natal day 6 (P6), homogenized and 152 total RNA was extracted using TriFast reagent according to manufacturer's instructions 153 (PEQLAB Biotechnologie GmbH, Erlangen, Germany). miRNA library preparation was 154 performed as previously described (21). Agilent QPCR NGS library quantification kit. Cluster generation and sequencing was carried out 159 using the Illumina Genome Analyzer IIx system according to the manufacturer's guidelines.

160
Illumina sequencing was performed at the CSF NGS Unit (csf.ac.at). After sequencing at a read 161 length of 36 base pairs, adaptor sequences were removed using Cutadapt (22).

162
Mapping to mature miRNA sequences 163 Mapping of clipped reads to mature miRNA sequences was performed as described previously 164 Mapping was performed using NextGenMap, restricting the mapped reads to have at least 90% 165 identity (# differences/alignment length) (23).

167
ADAR2 has the lowest expression in CHD patients 168 RNA was extracted from 35 affected congenital heart disease patients and 13 normal individuals 169 were taken as control. Most of the patients had VSD (Ventricular septal defects). All CHD 170 samples were pre-operative cases. Since recent reports indicated, that RNA editing might be 171 involved in cardiogenisis (13,14). We checked the expression of the functional RNA editing 172 enzymes using quantitative realtime PCR (qPCR). A significant decline in ADAR2 expression 173 was observed. However, on the contrary, ADAR1 showed a significant increase (Figure 1a). The Heart defect specific function 191 We further investigated whether this increase in ADAR1 is specific for a heart defect. As in our 192 study the patients were suffering from different forms of congenital heart disease. ADAR1 was 193 strongly up-regulated in ASD followed by VSD. However, it shows approximately 3fold 194 increase-in TOF and CAVSD (Figure 2a and d). On the contrary, ADAR2 shows a strong 195 significant decline in CAVSD, TOF and VSD. (Figure 2b and d). This specificity of gene 196 expression of ADAR1 and ADAR2 with heart defects can be answered by the differences in the    Since increase in ADAR1 has been found in CHD patients. We determined whether this 260 observed increase in ADAR1 is due to deregulation of ADAR2. Therefore, we determined 261 ADAR1 level in the absence of ADAR2. We did not find any significant change in ADAR1 level 262 ( Figure 6). Therefore, we can conclude that the observed increase in ADAR1 is soley because of 263 the CHD defect.

272
A previous study focusing on ADAR2 -/mice showed a statistically significant decrease in heart 273 rate (15). We performed RNA sequencing of ADAR2 -/mice heart samples in triplicates and 274 observed approximately ~2-fold decrease in miR-29b level consistently at P6 stage.

275
Approximately 1.5 fold down-regulation has been observed for miR451-b, miR451-a, miR19b, 276 (Table 1). To our surprise, we did not observe any up-regulated miRNAs. Quantitative trait loci 277 (QTLs) associated with miR-29 a and b show their potential involvement in cardiac diseases 278 (29). miR-29 family shows strong expression in lung, kidney and the heart. It expresses 279 predominantly approximately 5-12 folds in cardiac fibroblasts as compared to cardiomyocytes.

280
Moreover, the miR-29 family is down-regulated in fibrotic scars after myocardial infarction and 281 can lead to cardiac fibrosis by boosting collagen expression. miR29-b has an antifibrosis role as 282 it targets promoters of several extracellular matrix genes (30). Recent reports have documented a 283 cardioprotective role of miR29-b. miR 29b inhibits angiotensin II induced cardiac fibrosis by 284 targeting TGF-ß /Smad3 pathway (31).

285
After miR-29b, miR-451(a and b), miR-19b1and different members of let-7, family also showed 286 down-regulation in ADAR2 knockout mice heart However, they showed significant but small 287 down-regulation of only about 1-1.5 fold ( Adenosine deamination by ADARs is a post-transcriptional event that can diversify the 298 transcripts both at sequence as well as structure level. The deregulation of editing has been 299 associated with number of diseases (8,13,33).

300
ADARs play a significant role in development. Moreover the tissue and site specific editing 301 largely affects the differential expression of substrate transcripts (38,39). In our study, we found 302 a strong down-regulation of ADAR2 and a strong increase in ADAR1 in ASD and VSD patient 303 samples. This observation is in line with previous finding demonstrating increased CTSS mRNA 304 editing due to up-regulation of ADAR1 in human atherosclerotic plaques (14). Since the 305 expression analysis was performed only on the PBMCs from normal and CHD patients we 306 further extended our study to myocardial tissues. Expression analysis of myocardial tissues from 307 ischemic and non-ischemic patients showed a significant decline in ADAR2 expression level 308 (Figure 3). This result supported our finding that ADAR2 not only down-regulates in PBMCS 309 but also showed decreased expression in human cardiomyopathy tissues.

310
The stong downregulation of ADAR2 with respect to heart disease urged us to further explore 311 what heart related processes might be ADAR2 regulating? To address this query, we used 312 ADAR2 knock out mouse. We chose FLNB which plays an essential role in the heart however 313 lack of ADAR2 strongly decreased its editing. We did not observe any change in FLNB  ADARs can modulate microRNA processing and also are capable of retargeting the microRNA 321 to different substrate (35). Like ADAR1, ADAR2 also can modulate microRNA processing.

322
Since a number of micro RNAs like miR-1, miR-423 are associated with heart disease we 323 thought of investigating the microRNA profile in ADAR2 knock out mouse heart (40).
324 Surprisingly, we did not observe any up-regulated micro RNAs in the ADAR2 -/mice heart.

325
However, we observed a decline of ~1.5-2 fold in miR-29b, miR-451, miR19 and members of let 326 -7 family (Table 1). 327 miR-29b family regulates a plethora of proteins at RNA level that are involved in cardiac 328 fibrosis. This family has highest expression in the heart fibroblast population and comprises of 329 three members miR-29a, miR-29b and miR-29c. miR-29b differs only by one base from miR-