Alternative splicing is not a key source of chemerin isoforms diversity

Background Chemerin is a chemoattractant protein with adipokine and antimicrobial properties encoded by the retinoic acid receptor responder 2 (RARRES2) gene. Chemerin bioactivity largely depends on carboxyl-terminal proteolytic processing that generates chemerin isoforms with different chemotactic, regulatory, and antimicrobial potentials. While these mechanisms are relatively well known, the role of alternative splicing in generating isoform diversity remains obscure. Methods and results Using rapid amplification of cDNA ends (RACE) PCR, we determined RARRES2 transcript variants present in mouse and human tissues and identified novel transcript variant 4 of mouse Rarres2 encoding mChem153K. Moreover, analyses of real-time quantitative PCR (RT-qPCR) and publicly-available next-generation RNA sequencing (RNA-seq) datasets showed that different alternatively spliced variants of mouse Rarres2 are present in mouse tissues and their expression patterns were unaffected by inflammatory and infectious stimuli except brown adipose tissue. However, only one transcript variant of human RARRES2 was present in liver and adipose tissue. Conclusion Our findings indicate a limited role for alternative splicing in generating chemerin isoform diversity under all tested conditions. Supplementary Information The online version contains supplementary material available at 10.1007/s11033-022-08174-7.


Introduction
Protein isoforms can play important roles in various biological processes, such as growth, differentiation, and signal transduction. They can originate from separate genes, or a single gene can code for multiple proteins through alternative mRNA splicing. Alternative polyadenylation, RNA editing, and posttranslational modification can also create functionally distinct proteins. However, the alternative splicing of transcripts is one of the main sources of proteomic diversity mChem162K (mouse), where the number and capital letter indicate the terminal amino acid position and code, respectively [14]. Pro-chemerin is converted to chemotactically active isoforms through posttranslational carboxyl-terminal processing by proteases belonging to the coagulation, fibrinolytic, and inflammatory cascades. The most active form of human chemerin, hChem157S, is produced by direct cleavage of six C-terminal amino acids by neutrophil elastase or cathepsin G. [15]. Different proteolytic activities can generate isoforms with low or no activity, including 152G, 153Q, 154 F, 155 A, 156 F, and 158 K [15][16][17][18]. Several murine chemerin isoforms have been characterized in a mouse model of obesity, with mChem156S and mChem155F exhibiting the highest biological activity. Mouse chemerin undergoes tissue-specific proteolytic cleavage similar to human chemerin [19].
While mechanisms of proteolytic processing in generating chemerin isoforms are relatively well described, the role of alternative splicing remains obscure. Both mouse and human RARRES2 genes are comprised of six exons and five introns [20,21]. While in humans, only a single transcript (NM_002889.4) encoding a 163 amino acid (aa) protein has been described, [20] in the mouse, three alternatively spliced transcripts (NM_001347168.1, NM_027852.3, and NM_001347167.1) encoding 162 or 163 aa proteins have been reported [21]. The mChem162K and hChem163S proteins are the major chemerin forms present in mouse and human plasma, respectively [19].
Generating multiple chemerin isoforms is critical for controlling its local and context-specific bioactivity. Therefore, understanding the mechanisms underlying the diversity of chemerin isoforms is particularly important. Here, we show that alternatively spliced variants of mouse Rarres2 are present across different tissues and organs. Moreover, in addition to the variants encoding mChem163K and mChem162K, we have identified a novel transcript variant encoding mChem153K. We demonstrate that inflammatory and infectious conditions do not affect the expression pattern of Rarres2 splice variants. In contrast to murine chemerin, only one transcript variant was found in human liver and adipose tissue. We provide novel insights into the mechanisms that may contribute to chemerin isoform diversity and activity.

Materials
All chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise stated. Phosphate-buffered saline (PBS) buffer was obtained from PAN Biotech (Aidenbach, Germany). Mouse recombinant IL-1β and OSM were obtained from R&D Systems (Minneapolis, MN, USA).

Clinical material
Visceral WAT and liver samples were collected during bariatric surgery. All human studies were approved by the Jagiellonian University Institutional Bioethics Committee (protocol number KBET/87/B/2014) and adhered to the Declaration of Helsinki. All participants provided their written informed consent to participate in these studies.

Animal studies
This study used male 8-to 12-week-old C57BL/6 mice. The mice were maintained under specific pathogen-free conditions in the animal care facility in the Faculty of Biochemistry, Biophysics, and Biotechnology at Jagiellonian University. IL-1β and OSM were injected intraperitoneally at doses of 10 µg/kg body weight (BW) and 160 µg/kg BW, respectively, as previously described [22]. After 48 h (h), different tissues were isolated and subjected to RT-qPCR analysis. All animal procedures were approved by the First Local Ethical Committee on Animal Testing at the Jagiellonian University (Krakow, Poland; permit number 41/2014) in accordance with the ARRIVE guidelines and the Guidelines for Animal Care and Treatment of the European Community. The mice were sacrificed by an overdose of anesthesia (a mixture of ketamine and xylazine), followed by cervical dislocation.

RACE PCR
Total RNA was extracted for all tissues as described by Chomczynski and Sacchi [23], and 3' and 5' RACE PCR was performed with the 3' and 5' RACE System Kits (Invitrogen; Carlsbad, CA, USA) according to the manufacturer's recommended protocol. The following Rarres2 specific primers were used: 5'-GTGTGGACAGAGCT-GAAGAAGTGCTCTTC-3' (3' RACE) and 5'-CTG-GAGAAGGCAAACTGTCCAGGTAGGAAGTAG-3' (5' RACE). RACE PCR products were separated by agarose gel electrophoresis, with bands of interest excised from the gel and purified using the Gel-Out Concentrator kit (A&A Biotechnology; Gdynia, Poland), and ligated into the pTZ57/ RT vector using InsTAclone PCR Cloning Kit (Thermo Scientific; Waltham, MA, USA), followed by heat shock transformation of the plasmid into chemically competent Top10 E.coli (Invitrogen; Carlsbad, CA, USA). Selected bacterial colonies were subjected to colony PCR using standard M13 primers. Plasmid DNA was recovered from positive clones using GeneJET Plasmid Miniprep Kit (Thermo Scientific; Waltham, MA, USA) and sequenced at Genomed (Warsaw, Poland). All results were analyzed using SnapGene Viewer (GSL Biotech LLC; San Diego, CA, USA).

RT-QPCR and quantification of RARRES2 transcript variants
Total RNA was extracted with the Total RNA Zol-Out Kit (A&A Biotechnology; Gdynia, Poland) and converted to complementary DNA (cDNA) using NxGen M-MulV reverse transcriptase (Lucigen Corporation; Middleton, WI, USA) with random primers (Promega Corporation; Madison, WI, USA) and oligo dT (Genomed; Warsaw, Poland). RT-PCR was performed with a CFX96 thermocycler (Bio-Rad Laboratories; Hercules, CA, USA) using SYBR Green I and a universal PCR master mix (A&A Biotechnology; Gdynia, Poland) with the following mouse-specific primers: chemerin_all_variants The expression stabilities of commonly used reference genes were assessed as previously described [22]. Relative gene expression normalized to the geometric mean of these housekeeping genes was calculated using the 2 −ΔΔCT method [24]. RIV were obtained using the method of Londoño et al. [25]. The PCR efficiency of each primer set was calculated using CFX Maestro Software (Bio-Rad; Hercules, CA, USA) using pcDNA3.1 plasmids encoding mChem162K and mChem153K as a template.

Alternative splicing analyses of RNA-seq datasets
Rarres2 expression levels in different tissues and cell lines and isoform quantities were obtained from VastDB [26]. To assess isoform ratios in publicly available RNA-seq datasets, we calculated PSI scores with vast-tools [26]. We also analyzed datasets from the gene expression omnibus (GEO) database maintained by The National Center for Biotechnology Information (NCBI; Bethesda, MA, USA) that investigated the molecular effects of a high-fat diet (accessions GSE76133, GSE75984, and GSE117249) and transcriptional changes after infection with Staphylococcus aureus (GSE108718), Toxoplasma gondii (GSE119855), and influenza virus (GSE114232). Differential splicing analyses were performed with the diff module of vast-tool.

Statistical analysis
Differential splicing quantification in RNA-seq datasets was performed using vast-tools with flags -r 0.95 and -m 0.1. All other data were analyzed using STATISTICA 13 (Stat-Soft; Tulsa, OK, USA). Results were visualized using Prism (GraphPad Software; San Diego, CA, USA) and presented as mean ± standard deviation (SD). Comparison between groups used the Student's t-test. For multiple group comparisons, analysis of variance (ANOVA) with Tukey's posthoc test was used. Differences were considered statistically significant if they had a p-value < 0.05.

Characterization of alternatively spliced RARRES2 transcript variants
To identify transcript variants of mouse and human chemerin present in tissues, including liver and white adipose tissue (WAT), 3' and 5' rapid amplification of cDNA ends (RACE) PCR was performed. We detected one RARRES2 transcript variant in human tissues and three variants in mouse tissues ( Fig. 1 A-B and Fig. S1 A-B). Mouse Rarres2 variant 1 is the longest transcript and encodes the longer isoform 1 (mChem163K). Rarres2 variant 2 uses an alternate in-frame splice site in the 3' coding region and encodes the shorter protein isoform 2 (mChem162K). Rarres2 variant 3 differs in the 5' UTR and uses the same alternate in-frame splice site in the 3' coding region as variant 2, and therefore also encodes the shorter protein isoform 2 (mChem162K). In addition to the previously reported variants 1, 2, and 3, we have identified a novel variant 4, created by an alternate inframe splice site in the 3' coding region ( Fig. 1 A-B). While it contains exons 1 to 6, a 30 bp fragment is missing from exon 5 (Fig. 1 A). This novel transcript variant 4 of mouse Rarres2 was not predicted nor annotated by Ensembl [27] and RefSeq [28].
We performed a multiple sequence alignment of the four murine chemerin protein isoforms (Fig. 2). Isoform mChem162K, encoded by Rarres2 transcript variants 2 and 3, is the major form of chemerin in plasma [19]. Isoform mChem163K, encoded by Rarres2 transcript variant 1, has one extra glutamine at position 128. Interestingly, the newly discovered isoform mChem153K, encoded by Rarres2 transcript variant 4, is missing 10 amino acids (128-137) relative to mChem163K, consistent with its 27 bp deletion in mChem162K and mChem163K, but not mChem153K because transcript variant 4 is not present in the VastDB. We found transcript variants 2 and 3 (mChem162K) to be the dominant forms expressed in all investigated tissues, with an average percent spliced in (PSI) score of ~ 68.5 (Fig. 3 A). However, transcript variant 1 (mChem163K) accounted for up to 42% of transcripts in the cerebellum and pancreas. Our findings with the RNA-seq data were consistent with these patterns (Fig. 3B). However, Rarres2 transcript variant 4 (mChem153K) was rare, with a PSI score of < 1.5 (Fig. 3 A). Notably, the expression patterns of the Rarres2 transcript variants were unaffected by a high-fat diet or viral, bacterial, and parasite infections. In addition, there were no statistically significant differences between the control and treatment groups. However, levels of the newly discovered Rarres2 transcript variant 4 tended to increase in the kidney exon 5. Notably, all amino acid changes found in murine pro-chemerin isoforms are due to exon 5.

Expression pattern of mouse Rarres2 splice variants across different tissues and experimental conditions
Because only one RARRES2 transcript variant was found in human liver and adipose tissue, we focused on the role of Rarres2 alternative splicing in mouse tissues. We first determined the tissue expression patterns of the four alternatively-spliced transcripts using publicly-available nextgeneration RNA sequencing (RNA-seq) data and standard real-time quantitative PCR (RT-qPCR).
Using VastDB, [26] an atlas of alternative splicing profiles and functional associations in vertebrate cell and tissue types, we quantified transcript variants encoding chemerin expression in mouse adipocytes and human 3D skin cultures [7,22]. Therefore, we explored whether these cytokines affected the balance between newly discovered Rarres2 transcript variant 4 and the other transcript variants in mouse tissues. We found diminished levels of transcript variant 4 in brown adipose tissue (BAT) of IL-1β and OSM treated animals (Fig. 4 C). This was associated with up-regulation of total Rarres2 mRNA levels (Fig. S2). The transcript ratio remained stable in other tissues, and there and skin after a high-fat diet and S. aureus infection. Further studies are needed to investigate this finding.
Consistent with our analysis of publicly available RNAseq data, Rarres2 transcript variant 4 was found to be rare in our RT-qPCR data, with its highest expression level found in the heart (Fig. 4 A). The median of the relative incidence values (RIV) [25] of transcript variant 4 varied from ~ 0.14% in the liver to ~ 2.13% in the large intestine (Fig. 4B).
We have previously shown that acute-phase cytokines, interleukin 1β (IL-1β), and oncostatin M (OSM) regulate In this study, we have described for the first time a novel transcript variant 4 of mouse Rarres2 that encodes a 153 aa chemerin isoform 3 (mChem153K). Compared to isoform 1 (mChem163K), mChem153K is missing 10 aa at positions 128-137. This deletion may significantly affect protein structure because it removes a cysteine residue involved in forming one of three intrachain disulfide bonds [14]. Indeed, conformational changes due to this deletion may underlie our inability to purify mChem153K expressed in E.coli to determine the physiological role of this isoform (data not were no statistically significant differences between control and cytokine-treated mice.

Discussion
Our understanding of the posttranslational modifications of chemerin that generate a variety of protein isoforms has increased significantly over the last two decades. However, these studies focused mainly on the proteolytic processing of human (hChem163S) or mouse pro-chemerin (mChem162K) by extracellular proteases [15,16,19,29,30]. Alternative splicing is a key factor in increasing Fig. 4 Acute-phase cytokines do not affect the relative incidence ofRarres2transcript variant 4 levels across distinct mouse tissues.Rarres2 transcript variant 4 expression in different organs and tissues of control animals presented as relative expression values (A) or RIV (B). The relative incidence of Rarres2 transcript variant 4 in selected tissues of acute-phase cytokine or PBS treated animals (C). Data are presented as the mean ± standard deviation (A and C) or as the median (B) of at least three independent replicates. Statistical significance between the control (PBS) and the cytokine-treated animals is indicated by an asterisk; *p < 0.05 by the Student's t-test. Key: V. WAT -visceral white adipose tissue, S. WAT -subcutaneous white adipose tissue, INTEST. -intestine chemerin isoform mChem157S exhibits slightly increased antibacterial activity than mChem156S (Fig. S3A), but no change in chemotactic activity was observed (Fig. S3B). However, further studies are required to determine whether mChem157S encoded by the mouse Rarres2 variant 1 has any physiological role.
While four transcript variants of mouse Rarres2 encoding three protein isoforms are known, there is only one confirmed transcript variant for human RARRES2, which is translated into the hChem163S precursor protein [42]. Liver has one of the highest levels of alternative splicing among human tissues [43]. However, we did not find any additional transcript variants of human RARRES2 in the liver or adipose tissue based on 3' and 5' RACE PCR.
Our study has provided new insight into the mechanisms accounting for chemerin isoforms diversity. We have reported for the first time the identification of rare transcript variant 4 for mouse Rarres2 that encodes mChem153K (protein isoform 3). Rarres2 transcript variants 1 to 4 were present in all investigated mouse tissues, and the most abundant transcript variants encode chemerin isoform mChem162K. Our findings showed that the splicing pattern of RARRES2 mRNA was unaltered by a high-fat diet and bacterial, viral, or parasitic infection, nor by proinflammatory cytokine treatment. We found only one transcript variant of RARRES2 in human tissues. Altogether, our findings indicate that alternative splicing of RARRES2 in human and mouse tissues has a limited role in generating chemerin isoforms diversity under the tested conditions.

Declarations
Animal studies All animal procedures were approved by the First Local Ethical Committee on Animal Testing at the Jagiellonian University (Krakow, Poland; permit number 41/2014) in accordance with the ARRIVE guidelines and the Guidelines for Animal Care and Treatment of the European Community.
Human studies All human studies were approved by the Jagiellonian University Institutional Bioethics Committee (protocol number KBET/87/B/2014) and adhered to the Declaration of 1964 Helsinki. All participants provided their written informed consent to participate in these studies. shown). Our in silico and in vivo studies have revealed that transcript variant 4 accounts for only a small fraction of Rarres2 transcripts under physiological conditions. The average percentage across all mouse tissues investigated using RNA-seq and RT-qPCR was 0.55% and 1.31%, respectively. Therefore, transcript variant 4 might reflect inaccurate or inappropriate splicing creating abnormal transcripts of no functional significance. This phenomenon has been observed with almost all genes, and its frequency has been estimated to be at least 0.1% for each intron [31].
A frequent outcome of alternative splicing is decreased gene function due to the production of non-functional instead of functional isoforms which can be caused by alterations in protein functional domains [32]. Nonetheless, all Rarres2 transcript variants are generated by an alternate in-frame splice site in the 3' coding region of exon 5 or differences in the 5' UTR (variant 3 only). These modifications do not affect the C-terminal region of chemerin, which is crucial for its bioactivity [5].
Alternative transcripts are often differentially expressed between cells and tissues and possess different functions [33][34][35]. Moreover, changes in alternative splicing events can be associated with exposure to different stimuli [ 36]. Indeed, altered chemerin expression may be relevant in pathological conditions such as obesity, cancer, and inflammation [10,29,[37][38][39]. Various inflammatory and metabolic mediators regulate chemerin expression in a cell-typedependent manner [22,40]. We have previously shown that IL-1β and OSM upregulate chemerin expression in human skin cultures [7] and mouse adipocytes [22]. Moreover, bacteria such as S. aureus upregulate chemerin levels in models of the human epidermis and mouse skin [7]. Indeed, skin transcriptome analyses of antimicrobial peptides differentially regulated after skin infection with C. acnes or Leishmania braziliensis revealed elevated RARRES2 transcript levels [41]. In this study, we have shown that the splicing pattern of Rarres2 mRNA is not altered under most tested conditions including a high-fat diet or bacterial, viral, or parasitic infection, nor by cytokine treatment in different mouse organs except BAT, where down-regulation of transcript variant 4 was observed. This may suggest its limited physiological role since total chemerin mRNA levels were elevated. Therefore, these factors are not major determining factors in Rarres2 splice site selection.
We also explored differences in antimicrobial and chemotactic activity between biologically active chemerin isoforms mChem157S and mChem156S since they differ by only a single amino acid, the glutamine at position 128. This change does not directly affect chemerin's antimicrobial region (p4), which is located in the middle of the protein (positions 66-85 and 68-87 for human and mouse chemerin, respectively) [6]. Our preliminary findings indicate that