TMEM244 Is a Long Non-Coding RNA Necessary for CTCL Cell Growth

Transmembrane protein 244 (TMEM244) was annotated to be a member of the TMEM family, which are is a component of cell membranes and is involved in many cellular processes. To date, the expression of the TMEM244 protein has not been experimentally confirmed, and its function has not been clarified. Recently, the expression of the TMEM244 gene was acknowledged to be a diagnostic marker for Sézary syndrome, a rare cutaneous T-cell lymphoma (CTCL). In this study, we aimed to determine the role of the TMEM244 gene in CTCL cells. Two CTCL cell lines were transfected with shRNAs targeting the TMEM244 transcript. The phenotypic effect of TMEM244 knockdown was validated using green fluorescent protein (GFP) growth competition assays and AnnexinV/7AAD staining. Western blot analysis was performed to identify the TMEM244 protein. Our results indicate that TMEM244 is not a protein-coding gene but a long non-coding RNA (lncRNA) that is necessary for the growth of CTCL cells.


Introduction
The transmembrane protein (TMEM) family comprises proteins that are embedded in the cell membrane and span both intracellular and extracellular environments. TMEMs are components of various cell membranes, such as mitochondrial membranes, Golgi membranes, lysosomes, and the endoplasmic reticulum [1]. They are involved in many cellular processes, such as the transport of ions and molecules across impermeable membranes, membrane trafficking, and signaling transduction pathways [2]. Studies showed that TMEMs' expression can be down-or upregulated in several cancers [3,4] and is associated with tumor progression, disease stage, and patient survival [5,6]. Because membrane proteins are involved in essential cellular pathways, they are often targets of pharmaceutical agents [7]. Yet, for most TMEMs, the mechanism of their involvement in carcinogenesis is still unknown.
The TMEM244 gene is located in chromosome 6q22.33 and comprises five exons. Under physiological conditions, TMEM244 is expressed at a low level in the brain and the pituitary glands. To date, the existence of the TMEM244 protein has not been experimentally demonstrated, but the gene is supposed to encode a protein of 128 amino acids and a molecular mass of 14,657 Da. Izykowska et al., were the first to pay attention to the TMEM244 gene, as it was identified among four other genes (EHD1, MTMR2, RNF123, and TOX) to be involved in the rearrangements affecting gene expression in Sézary syndrome (SS) patients compared to controls [8]. Further studies showed that TMEM244 is expressed in T-cell lymphomas as a result of specific hypomethylation of its promoter, and this expression is associated with poor overall survival in T-cell lymphoma patients [9,10]. A significantly higher expression of TMEM244 was identified in Sézary syndrome patients, not only compared to healthy individuals but also to SS clinical mimickers, such as mycosis fungoides and erythrodermic manifestations of non-malignant diseases, therefore indicating its diagnostic potential. Moreover, higher expressions of TMEM244 in CD4+ and CD8+ subsets of memory cells (CD45RO+) were identified, which is in line with the immunophenotype of Sézary cells [11]. The purpose of this study was to establish the function of the TMEM244 gene, which has not been investigated yet.

TMEM244 Has a Low Protein-Coding Potential and Cannot Be Detected at the Protein Level
For the detection of the TMEM244 protein, two CTCL cell lines were examined: SeAx and HH, as well as other non-CTCL cell lines, HDLM2, D341med, and COLO684, with predicted high levels of the TMEM244 transcript.
A high level of TMEM244 expression in the Jurkat cell line was induced using two lentiviral systems with different promoters (human cytomegalovirus; CMV or 3phosphoglycerate kinase; PGK). FLAG tags were introduced either at the N-or C-terminus of the TMEM244 gene. TMEM244 expression on the mRNA level, as determined using RT-qPCR, was the highest in Jurkat-CMV (mean = 390 × 10 3 ± 102 × 10 3 ) and Jurkat-PGK (mean = 190 × 10 3 ± 69 × 10 3 ). In cell lines with endogenous TMEM244 expression, the highest level was detected in D341med (mean = 180 × 10 3 ± 67 × 10 3 ), followed by COLO684 (mean = 59 × 10 3 ± 16 × 10 3 ), SeAx (mean = 2.5 × 10 3 ± 0.82 × 10 3 ), HDLM2 (mean = 2.4 × 10 3 ± 0.88 × 10 3 ), and HH (mean = 0.7 × 10 3 ± 0.2 × 10 3 ) ( Figure 1). Upon confirmation of the expression of TMEM244 in selected cell lines, Western blot analysis using a custom-made anti-TMEM244 antibody and an anti-FLAG antibody, with a wild-type cell line used as a control, was performed. and this expression is associated with poor overall survival in T-cell lymphoma patients [9,10]. A significantly higher expression of TMEM244 was identified in Sézary syndrome patients, not only compared to healthy individuals but also to SS clinical mimickers, such as mycosis fungoides and erythrodermic manifestations of non-malignant diseases, therefore indicating its diagnostic potential. Moreover, higher expressions of TMEM244 in CD4+ and CD8+ subsets of memory cells (CD45RO+) were identified, which is in line with the immunophenotype of Sézary cells [11]. The purpose of this study was to establish the function of the TMEM244 gene, which has not been investigated yet.

TMEM244 Has a Low Protein-Coding Potential and Cannot Be Detected at the Protein Level
For the detection of the TMEM244 protein, two CTCL cell lines were examined: SeAx and HH, as well as other non-CTCL cell lines, HDLM2, D341med, and COLO684, with predicted high levels of the TMEM244 transcript.
A high level of TMEM244 expression in the Jurkat cell line was induced using two lentiviral systems with different promoters (human cytomegalovirus; CMV or 3-phosphoglycerate kinase; PGK). FLAG tags were introduced either at the N-or C-terminus of the TMEM244 gene. TMEM244 expression on the mRNA level, as determined using RT-qPCR, was the highest in Jurkat-CMV (mean = 390 × 10 3 ± 102 × 10 3 ) and Jurkat-PGK (mean = 190 × 10 3 ± 69 × 10 3 ). In cell lines with endogenous TMEM244 expression, the highest level was detected in D341med (mean = 180 × 10 3 ± 67 × 10 3 ), followed by COLO684 (mean = 59 × 10 3 ± 16 × 10 3 ), SeAx (mean = 2.5 × 10 3 ± 0.82 × 10 3 ), HDLM2 (mean = 2.4 × 10 3 ± 0.88 × 10 3 ), and HH (mean = 0.7 × 10 3 ± 0.2 × 10 3 ) ( Figure 1). Upon confirmation of the expression of TMEM244 in selected cell lines, Western blot analysis using a custom-made anti-TMEM244 antibody and an anti-FLAG antibody, with a wild-type cell line used as a control, was performed. Moreover, commercially available brain lysate was analyzed for TMEM244 protein expression, as according to the Human Protein Atlas database TMEM244 is expressed on the mRNA level in the brain (https://www.proteinatlas.org/ENSG00000203756-TMEM244/tissue; accessed on 30 January 2020). All Western blot analyses showed signals Moreover, commercially available brain lysate was analyzed for TMEM244 protein expression, as according to the Human Protein Atlas database TMEM244 is expressed on the mRNA level in the brain (https://www.proteinatlas.org/ENSG00000203756-TMEM2 44/tissue; accessed on 30 January 2020). All Western blot analyses showed signals from positive controls, but they failed to detect the TMEM244 protein ( Figure S1). The TMEM244 protein was neither detected with the anti-FLAG antibody in cell lines with induced TMEM244 overexpression ( Figure S1A) nor with the specific anti-TMEM244 antibody in the same cell lines, and it was not detected in cell lines or brain tissue with high endogenous TMEM244 expression ( Figure S1B,C). Furthermore, in silico analysis showed a very low coding probability of TMEM244 [0.097], with a cutoff <0.364 indicating a noncoding sequence [12]. Taken together, these results indicate that TMEM244 is not expressed at the protein level.

TMEM244 Transcript Is Primarily Localized in the Cytoplasm
Using the long non-coding RNA subcellular localization predictor (lncLocator; http:// www.csbio.sjtu.edu.cn/bioinf/lncLocator/; accessed on 2 September 2021), the TMEM244 transcript was predicted to be present, mainly in the cytoplasm and partially in the nucleus (Figure 2A). To confirm this prediction, subcellular fractionation was performed in SeAx and HDLM2 cell lines and FISH RNA analysis in SeAx and HDLM2 cells. In both cell lines, the level of the TMEM244 transcript in the cytoplasm (SeAx = 67%; HDLM2 = 55%) was higher than that observed in the nucleus (SeAx = 29%; HDLM2 = 39%) ( Figure 2B). FISH results further confirmed that in SeAx and HDLM2 cells, the prominent distribution of TMEM244 was in the cytoplasm ( Figure 2C).

Inhibition of TMEM244 Results in Decreased Cell Growth in CTCL Cell Lines
To establish the function of TMEM244 in cancer cells, the effect of TMEM244 knockdown was analyzed in SeAx and HH cell lines with endogenous TMEM244 expression. The effectiveness of the shRNAs targeting the TMEM244 transcript was confirmed for both cell lines ( Figure S2). TMEM244 silencing resulted in a strong negative effect on cell growth. On day 22 after transduction, the ratio of GFP-positive cells decreased by more than 50% in the cell lines treated with TMEM244-specific shRNAs compared to non-targeting and scrambled controls ( Figure 3).
The knockdown of TMEM244 decreased in the GFP+ cell population for all constructs in both CTCL cell lines. In HH, all shRNAs showed a strong effect, with a reduction of 57%, 55%, and 68% for shRNA 1, shRNA 2, and shRNA 3, respectively, compared to SCR and NT. The effects on the growth of SeAx cells were stronger for shRNA 2 and shRNA 3, with a reduction of 78% and 74% compared to SCR, and only a mild effect was observed for shRNA 1 (36%).
To further investigate the mechanism of growth inhibition upon TMEM244 knockdown on cancer cells, Annexin V/7AAD staining was conducted ( Figure 4). from positive controls, but they failed to detect the TMEM244 protein ( Figure S1). The TMEM244 protein was neither detected with the anti-FLAG antibody in cell lines with induced TMEM244 overexpression ( Figure S1A) nor with the specific anti-TMEM244 antibody in the same cell lines, and it was not detected in cell lines or brain tissue with high endogenous TMEM244 expression ( Figure S1B,C). Furthermore, in silico analysis showed a very low coding probability of TMEM244 [0.097], with a cutoff <0.364 indicating a noncoding sequence [12]. Taken together, these results indicate that TMEM244 is not expressed at the protein level.

TMEM244 Transcript Is Primarily Localized in the Cytoplasm
Using the long non-coding RNA subcellular localization predictor (lncLocator; http://www.csbio.sjtu.edu.cn/bioinf/lncLocator/; accessed on 2 September 2021), the TMEM244 transcript was predicted to be present, mainly in the cytoplasm and partially in the nucleus (Figure 2A). To confirm this prediction, subcellular fractionation was performed in SeAx and HDLM2 cell lines and FISH RNA analysis in SeAx and HDLM2 cells. In both cell lines, the level of the TMEM244 transcript in the cytoplasm (SeAx = 67%; HDLM2 = 55%) was higher than that observed in the nucleus (SeAx = 29%; HDLM2 = 39%) ( Figure 2B). FISH results further confirmed that in SeAx and HDLM2 cells, the prominent distribution of TMEM244 was in the cytoplasm ( Figure 2C).

Inhibition of TMEM244 Results in Decreased Cell Growth in CTCL Cell Lines
To establish the function of TMEM244 in cancer cells, the effect of TMEM244 knockdown was analyzed in SeAx and HH cell lines with endogenous TMEM244 expression. The effectiveness of the shRNAs targeting the TMEM244 transcript was confirmed for both cell lines ( Figure S2). TMEM244 silencing resulted in a strong negative effect on cell growth. On day 22 after transduction, the ratio of GFP-positive cells decreased by more than 50% in the cell lines treated with TMEM244-specific shRNAs compared to non-targeting and scrambled controls ( Figure 3).
The knockdown of TMEM244 decreased in the GFP+ cell population for all constructs in both CTCL cell lines. In HH, all shRNAs showed a strong effect, with a reduction of 57%, 55%, and 68% for shRNA 1, shRNA 2, and shRNA 3, respectively, compared to SCR and NT. The effects on the growth of SeAx cells were stronger for shRNA 2 and shRNA 3, with a reduction of 78% and 74% compared to SCR, and only a mild effect was observed for shRNA 1 (36%).

Inhibition of TMEM244 Results in Decreased Cell Growth in CTCL Cell Lines
To establish the function of TMEM244 in cancer cells, the effect of TMEM244 knockdown was analyzed in SeAx and HH cell lines with endogenous TMEM244 expression. The effectiveness of the shRNAs targeting the TMEM244 transcript was confirmed for both cell lines ( Figure S2). TMEM244 silencing resulted in a strong negative effect on cell growth. On day 22 after transduction, the ratio of GFP-positive cells decreased by more than 50% in the cell lines treated with TMEM244-specific shRNAs compared to non-targeting and scrambled controls ( Figure 3).
The knockdown of TMEM244 decreased in the GFP+ cell population for all constructs in both CTCL cell lines. In HH, all shRNAs showed a strong effect, with a reduction of 57%, 55%, and 68% for shRNA 1, shRNA 2, and shRNA 3, respectively, compared to SCR and NT. The effects on the growth of SeAx cells were stronger for shRNA 2 and shRNA 3, with a reduction of 78% and 74% compared to SCR, and only a mild effect was observed for shRNA 1 (36%). Only a slight, statistically non-significant difference in the number of apoptotic cells in the HH cell line was detected. In SeAx, although also non-significant, the effect was stronger, with an increase of 7% and 3% using shRNA 1; 21%, and 18% using shRNA 2; and 14% and 11% using shRNA 3, compared to SCR and NT, respectively.

Identification of Novel Alternative Transcripts of TMEM244
This study led to the identification of two novel alternative TMEM244 transcripts. Besides the two known transcripts-variant one with 5 exons (RefSeq NM_001010876; ENST00000368143.6) and variant two with an extra 5 exon (ENST00000438392.2), available in the Genome Browser-two additional variants were identified: variant three without exon 4 and variant four without exons 2 and 3 ( Figure 5A and Figure S4). The expression profile of each splice variant in cell lines with TMEM244 expression (SeAx, HH, Hut78, HDML2, and D341med) was performed using RT-qPCR with variant-specific primers. In most cell lines, the expression level was as follows; variant1 > variant2 > variant3 > variant4, except for D341med, where the expression of variant three was higher than that of variant two ( Figure 5B).  To further investigate the mechanism of growth inhibition upon TMEM244 knockdown on cancer cells, Annexin V/7AAD staining was conducted (Figure 4). Only a slight, statistically non-significant difference in the number of apoptotic cells in the HH cell line was detected. In SeAx, although also non-significant, the effect was stronger, with an increase of 7% and 3% using shRNA 1; 21%, and 18% using shRNA 2; and 14% and 11% using shRNA 3, compared to SCR and NT, respectively.

Identification of Novel Alternative Transcripts of TMEM244
This study led to the identification of two novel alternative TMEM244 transcripts. Besides the two known transcripts-variant one with 5 exons (RefSeq NM_001010876; ENST00000368143.6) and variant two with an extra 5′ exon (ENST00000438392.2), available in the Genome Browser-two additional variants were identified: variant three without exon 4 and variant four without exons 2 and 3 ( Figures 5A and S4). The expression profile of each splice variant in cell lines with TMEM244 expression (SeAx, HH, Hut78, HDML2, and D341med) was performed using RT-qPCR with variant-specific primers. In most cell lines, the expression level was as follows; variant1 > variant2 > variant3 > vari-ant4, except for D341med, where the expression of variant three was higher than that of variant two ( Figure 5B).   Only a slight, statistically non-significant difference in the number of apoptotic cells in the HH cell line was detected. In SeAx, although also non-significant, the effect was stronger, with an increase of 7% and 3% using shRNA 1; 21%, and 18% using shRNA 2; and 14% and 11% using shRNA 3, compared to SCR and NT, respectively.

Identification of Novel Alternative Transcripts of TMEM244
This study led to the identification of two novel alternative TMEM244 transcripts. Besides the two known transcripts-variant one with 5 exons (RefSeq NM_001010876; ENST00000368143.6) and variant two with an extra 5′ exon (ENST00000438392.2), available in the Genome Browser-two additional variants were identified: variant three without exon 4 and variant four without exons 2 and 3 ( Figures 5A and S4). The expression profile of each splice variant in cell lines with TMEM244 expression (SeAx, HH, Hut78, HDML2, and D341med) was performed using RT-qPCR with variant-specific primers. In most cell lines, the expression level was as follows; variant1 > variant2 > variant3 > vari-ant4, except for D341med, where the expression of variant three was higher than that of variant two ( Figure 5B).  To assess the coding potential of each transcript variant, the in silico analysis was performed using CPAT. The analysis confirmed that neither of the detected variants had the potential to encode a protein. The coding potential was 0.0976, 0.0003, 0.0359, and 0.0043 for variants one, two, three, and four, respectively.

Discussion
TMEMs are a very heterogeneous group of more than 300 genes, which have been included based on the in silico analysis of their DNA sequence. To be annotated as a TMEM gene, the predicted protein structure must contain at least one putative transmembrane segment that spans completely or partially through biological membranes [13]. Some TMEMs have been experimentally shown to encode a protein and, upon functional characterization, have been renamed and reclassified [14]. Still, for many of them, including TMEM244, neither the protein nor the function have been experimentally demonstrated.
Our study demonstrated that despite its annotation, based only on the in silico analysis of the predicted protein structure, the TMEM244 gene does not seem to encode a protein but, rather, belongs to the long-non-coding RNA (lncRNAs) family. LncRNAs are defined as ≥200 nucleotides long RNAs that are spliced and polyadenylated like mRNAs; however, they lack protein-coding activity. In silico analysis revealed minor coding potential of the 545 nucleotides long TMEM244 transcript. TMEM244 has an open reading frame (ORF) but, as shown for other lncRNAs, such as LINC00116 or LINC00948, possessing an ORF does not determine protein production [15,16]. TMEM244, like lncRNAs, is poorly conserved and its expression level is lower compared to protein-coding genes [17]. In addition, the TMEM244 mRNA level is highest in the brain and pituitary glands, which is typical for lncRNAs.
While mRNAs are very specifically located on the ribosomes in the cytoplasm, lncR-NAs may occupy diverse sites, including chromatin, subnuclear domains, nucleoplasm, and cytoplasm [18]. Furthermore, in tumors, the cellular localization of lncRNAs is related to their functions. LncRNAs located in the nuclear compartment usually control transcription and post-transcriptional processing. Since TMEM244 is mainly located in the cytoplasm, this suggests its involvement in the regulation of translation, mRNA turnover, protein stability, sponging of cytosolic factors, and the modulation of signaling pathways [15,19].
Recently, emerging evidence showed that lncRNAs could promote cell proliferation and, therefore, be engaged in carcinogenesis. For instance, the lncRNA HOXD cluster antisense RNA 1 (HOXD-AS1) was upregulated and promoted cell proliferation in cervical cancer, while lncRNA EPIC1 promoted proliferation and inhibited apoptosis of gallbladder cancer cells [16,20]. LncRNAs can also affect apoptosis by acting as a competitive endogenous RNA (ceRNA) for miRNA and binding to the sequence at the 5 end of the miRNA, therefore reducing target mRNA expression and ultimately affecting cell apoptosis. Moreover, lncRNAs can act directly or indirectly on death receptors [16]. Previous studies have demonstrated that in some cases, a non-protein-coding locus can give rise to functionally distinct transcript isoforms [21][22][23][24]. Recently, it was shown that the switch in the lncRNA HOTAIR start site after the induction of differentiation promotes the inclusion of HOTAIR exon 3, containing a protein-binding domain, which likely changes its function [25]. We showed that TMEM244 is necessary for the growth of cells where its expression is at a relatively high level, such as in CTCL cell lines. However, the mechanism behind this observation is still unknown.
It is, however, worth mentioning that the experiments were focused on the sense strand of TMEM244 based on the GenBank (NCBI) transcription annotation (Gene ID: 253582), and it is not known if the anti-sense strand of TMEM244 is expressed or whether it plays any role in cell proliferation.

Fluorescence In Situ Hybridization (FISH) Assay
Fluorescence-labeled probes for TMEM244 and GAPDH RNA were designed and synthesized, and FISH experiments were performed according to the manufacturer's protocol, using the Stellaris™ FISH technology kit (Biosearch Technologies, Hoddesdon, UK). Twenty Quasar ® 570-labeled probes for the TMEM244 transcript were designed using Stellaris ® Probe Designer version 4.2 (LGC Biosearch Technologies, Berlin, Germany). The nuclei were stained with DAPI, and Human GAPDH with the Quasar ® 570 Dye Stellaris ® FISH Probe was used as a cytoplasmic marker ( Figure S3). Images were acquired using a Leica DMI8 laser-scanning confocal microscope (Leica Microsystems, Wetzlar, Germany). Cells were imaged with an HC PL APO CS2 100×/1.40 oil objective lens and processed using Leica Application Suite X software (Leica Microsystems, Wetzlar, Germany). All samples were imaged under the same optical conditions.

Western Blot
Whole cell lysates were prepared from 5-10 × 10 6 cells. Cells were washed with phosphate-buffered saline (PBS) and lysed in RIPA buffer (Merck KgaA, Darmstadt, Germany) containing 1X protease inhibitor cocktail (Bioshop Canada Inc., Burlington, ON, Canada) for 30 min on ice. Samples were centrifuged at 14,000× g for 30 min to remove DNA or debris. As an additional control, MG132 proteosome inhibitor (Merck KgaA, Darmstadt, Germany) was added to the cell cultures to prevent possible TMEM244 degradation. Total protein concentrations of the cell extracts were measured using the Pierce BCA Protein Quantitation kit (Thermo Fisher Scientific™, Waltham, MA, USA), and the samples were stored at −80 • C until assayed. Prior to loading on gel, samples were heated at 95 • C for 5 min in a heating block. A synthetic peptide-fragment of the putative TMEM244 protein used for mice immunization was used as a positive control. Not centrifuged and nonheated proteins were used as controls for the sample preparation procedure. Human brain whole tissue lysate was commercially available (Novus Biologicals LLC a Bio-Techne Brand, Centennial, CO, USA). Western blotting was performed as previously described [9]. Primary antibodies (anti-FLAG (F1804, 1:1000, Merck KGaA, Darmstadt, Germany), anti-TMEM244 (custom-made; 1:1000, Proteogenix, Schiltigheim, France)) were used, as well as HRP-labeled secondary antibodies (sc-2005, 1:10,000, Santa Cruz Biotechnology, Dallas, TX, USA). The signal was detected by chemiluminescence with Clarity Western ECL Substrate (Bio-Rad, Hercules, CA, USA) using ChemiDoc™ Imaging Systems (Bio-Rad, Hercules, CA, USA). Quantitative analysis was performed using ImageLabTM Software. The WB results were normalized using a stain-free technique, by measuring total protein directly on the WB membrane.

GFP Competition Assay
SeAx and HH cells were infected with miRZIP lentivirus, aiming at the infection percentage of 50%. The percentages of GFP-positive cells were measured using the flow cytometry (CytoFLEX S Flow Cytometer, Beckman Coulter, Indianapolis, IN, USA) on day 4 post-transduction and monitored tri-weekly for three weeks. Data were analyzed using Kaluza Analysis Software (Beckman Coulter, Indianapolis, IN USA). To determine the effect on cell growth, the percentage of GFP-positive cells on day 6 was set to 100%, and the fold difference relative to this starting point was calculated for each time point.
To determine significant differences in the GFP assays, we used mixed model analysis as described previously [27].

Apoptosis Assay
The percentages of apoptotic cells were determined in SeAx and HH cells harvested on day 8 after transduction with the lentiviral miRZIP vectors aiming at an infection percentage of >95%. Briefly, cells were washed twice with cold phosphate-buffered saline and resuspended at a concentration of 1 × 10 6 cells/mL in 1X Binding Buffer. Cells were stained with Annexin V APC and 7AAD according to the manufacturer's protocol (BD Biosciences, San Jose, CA, USA) and analyzed via flow cytometry (CytoFLEX S Flow Cytometer, Beckman Coulter, Indianapolis, IN, USA).

RACE-PCR
Both 5 -and 3 -rapid amplification of cDNA ends (RACE) were performed using the SMARTer ® RACE 5 /3 kit (Takara Bio Inc., San Jose, CA, USA), according to the manufacturer's instructions. Briefly, 1 µg of total RNA isolated from the SeAx cell line was converted into the RACE-Ready first-strand cDNA. For the preparation of 5 -RACE-Ready cDNA/3 -RACE-Ready cDNA, the 5 -CDS Primer A/3 -CDS Primer A, respectively, were mixed with RNA. Two rounds of PCR amplification were performed, the first one with a specific primer 10XUPM (universal primer mix) and a gene-specific primer: TMEM244r primer for the 3 RACE and TMEM244f primer for the 5 RACE. To perform nested PCR, Universal Primer Short (UPM short) was added, as well as TMEM244r2 inner primer for the 3 RACE and TMEM244f2 inner primer for the 5 RACE. PCR products were analyzed using 1% agarose gel. Prior to qRT-PCR, the PCR products were purified using the QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany). Subsequently, colony PCR was performed using RedTaq Polymerase (Merck KGaA, Darmstadt, Germany) and M13f, M13r primers. Different size bands were sequenced to identify the possible isoforms. Isoforms detected using RACE were confirmed using RT-PCR and primers designed to identify isoforms. The GAPDH gene was used as a positive control. All primer sequences are listed in Table S1.

RNA Extraction and Real-Time Quantitative PCR (RT-qPCR)
RNA was extracted using TRI Reagent (Merck KGaA, Darmstadt, Germany) according to the manufacturer's protocol. The quantity of RNA was measured using the NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific™, Waltham, MA, USA), and the quality was determined by 0.8% agarose gel electrophoresis with ethidium bromide staining. cDNA was synthesized from 0.3 µg or 0.5 µg of RNA using SuperScript™ IV Reverse Transcriptase with random hexamer primers (Invitrogen™, Waltham, MA, USA). TMEM244 expression was analyzed using TaqMan Gene Expression Assays (Applied Biosystems, Foster City, CA, USA) (Hs02340633_m1) with intron-spanning primers located in the second and third exons. Beta-2 microglobulin (B2M) (Hs00984230_m1), with intron-spanning primers located in the first and second exons, was used as a reference gene for sample normalization. Relative gene expression was calculated using the median ct value method (2 −∆CT ). Expression levels of different isoforms were measured using 5× HOT FIREPol ® EvaGreen ® qPCR Supermix (Solis Biodyne, Tartu, Estonia) and primers specific to each isoform. The results were normalized to the GAPDH reference gene. Standard curves for each isoform were prepared as follows: PCR products for each TMEM244 isoform and GAPDH gene were cloned into the pGEM ® -T Easy Vector and transformed into bacteria. Vectors were isolated using the GeneJET Plasmid Miniprep Kit (Thermo Fisher Scientific™, Waltham, MA, USA), sequenced, and digested with the BstXI enzyme (NEB, Ipswich, MA, USA). Serial dilutions were prepared to obtain the concentration from 10 9 to 10 1 copy numbers. RT-qPCR was performed using the CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA).

Isolation of Nuclear and Cytoplasmic RNA
Cytoplasmic, nuclear, and chromatin RNA were isolated using an adaptation of the CD4+ T-cell nuclei extraction by Danko et al. [28,29]. RNA from the isolated fractions was reverse transcribed and used for qRT-PCR as described above. All samples were tested for TMEM244 as well as tRNA lys, RPPH1, DANCER (cytoplasmic controls), U3SNORNA, ANRIL (nuclear controlss), and KTN1_AS1, KTN1_AS1_intron (chromatin controls). The sum of the cytoplasmic, nuclear, and chromatin expression levels of each transcript was set to 100%, and the percentage of each transcript localized to each compartment was determined. All controls showed the expected localization in each experiment, confirming successful fractionation. The primers used for RT-qPCR are listed in Table S1.

Conclusions
Our study is the first to experimentally verify the presence of TMEM244 protein. Different TMEM244 transcript variants were identified; however, none of them had significant coding potential and they were all expressed at a lower level compared to the main transcript variant. Although TMEM244 transcripts are localized in the cytoplasm, it appears that they do not encode a protein but are, rather, lncRNAs. Obtained results demonstrate that TMEM244 mRNA is necessary for cellular growth of CTCL cells; therefore, it might be considered a new therapeutic target for the treatment of CTCL. Further study is needed to elucidate the in vivo effect and the downstream signaling pathway through which TMEM244 functions in CTCL cells, as well as the function of its novel transcript variants.