Expression, Regulation, and Functions of the Galectin-16 Gene in Human Cells and Tissues

Galectins comprise a family of soluble β-galactoside-binding proteins, which regulate a variety of key biological processes including cell growth, differentiation, survival, and death. This paper aims to address the current knowledge on the unique properties, regulation, and expression of the galectin-16 gene (LGALS16) in human cells and tissues. To date, there are limited studies on this galectin, with most focusing on its tissue specificity to the placenta. Here, we report the expression and 8-Br-cAMP-induced upregulation of LGALS16 in two placental cell lines (BeWo and JEG-3) in the context of trophoblastic differentiation. In addition, we provide the results of a bioinformatics search for LGALS16 using datasets available at GEO, Human Protein Atlas, and prediction tools for relevant transcription factors and miRNAs. Our findings indicate that LGALS16 is detected by microarrays in diverse human cells/tissues and alters expression in association with cancer, diabetes, and brain diseases. Molecular mechanisms of the transcriptional and post-transcriptional regulation of LGALS16 are also discussed based on the available bioinformatics resources.

Galectin expression profiles vary significantly between different cells and tissues. Some galectins are commonly expressed with low tissue specificity, e.g., galectin-1 and galectin-3, while others are highly-tissue specific [3].
LGALS16 was characterized in placental tissue by Than and co-authors [11] and together with two other galectins (LGALS13 and LGALS14) was found to be upregulated in differentiated trophoblast cells to confer immunotolerance at the maternal-fetal interface [12]. These three galectin genes are located in a cluster of four human protein-coding galectin genes on chromosome 19 and

Statistical Analysis
Statistical analysis was performed using GraphPad Prism 9 for Windows, version 9.1.2 (GraphPad Software, San Diego, CA, USA) and the data were presented as mean ± SD. One-way analysis of variance (ANOVA) was used to determine statistical significance across treatments followed by Tukey's honestly significant difference test to detect which means were statistically significant at a value of p < 0.05. LGALS16 gene structure and the mRNA sequence. (a) LGALS16 (4735 bp) is located on chromosomal band 19q13.2 and contains 4 exons (ENSG00000249861). (b) NCBI reference sequence of LGASL16 mRNA (NM_001190441.3). Each exon is highlighted with red, orange, blue, and black representing exons 1, 2, 3, and 4, respectively. The protein coding sequence (CDS) is indicated in capitals while UTRs in small characters. The oligonucleotide sequences for PCR amplification are boxed.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism 9 for Windows, version 9.1.2 (GraphPad Software, San Diego, CA, USA) and the data were presented as mean ± SD. One-way analysis of variance (ANOVA) was used to determine statistical significance across treatments followed by Tukey's honestly significant difference test to detect which means were statistically significant at a value of p < 0.05.

Molecular Characteristics of Galectin-16 Gene and Recombinant Protein
The LGALS16 gene structure and molecular details were described by Than and coauthors [11].
LGALS16 is found only in primates and is part of the chromosome 19 gene cluster containing four proteincoding genes (LGALS10, LGALS13, LGALS14, LGALS16) [11,12,14,29]. The diversification and evolutionary origin of this cluster, including LGALS16, is thought to be related to placenta development and mediated by transposable long interspersed nuclear elements (LINEs), which are commonly found at the boundaries of large inversions and gene duplication units [11,30,31]. The relevant rearrangements and subsequent gains and losses of duplicated genes and pseudogenes are proposed to have enabled anthropoids to sustain highly invasive placentation and placental phenotypes, such as longer gestation for larger offspring and an increased body to brain size ratio [11].
To the best of our knowledge, no studies are available on native galectin-16 at the protein level whereas recombinant protein has been produced and tested. The crystal structure of recombinant galectin-16 and its mutants was solved by Si and co-authors [32]. Recombinant galectin-16 is a monomeric protein, which is composed of 142 amino acids and has a typical galectin structure of the CRD β-sandwich with two sheets formed by six β-strands on the concave side (S1-S6) and five β-strands on the convex side (F1-F5) ( Figure 2). This group also showed that galectin-16 lacks lactose-binding ability unless arginine (Arg55) is replaced with asparagine in S4 β-strand. In comparison, an earlier report showed that recombinant galectin-16 and two other human galectins (galectin-13 and galectin-14) can bind lactose-agarose beads and are efficiently and competitively eluted by lactose [11]. More insights into this discrepancy are required considering multiple interfering factors, mutations/replacements of amino acids within the CRD, and different study designs. Regardless, both glycan-dependent and glycan-independent interactions might be essential for galectin-16 similar to other galectins [3].
LGALS16 is found only in primates and is part of the chromosome 19 gene cluster containing four protein-coding genes (LGALS10, LGALS13, LGALS14, LGALS16) [11,12,14,29]. The diversification and evolutionary origin of this cluster, including LGALS16, is thought to be related to placenta development and mediated by transposable long interspersed nuclear elements (LINEs), which are commonly found at the boundaries of large inversions and gene duplication units [11,30,31]. The relevant rearrangements and subsequent gains and losses of duplicated genes and pseudogenes are proposed to have enabled anthropoids to sustain highly invasive placentation and placental phenotypes, such as longer gestation for larger offspring and an increased body to brain size ratio [11].
To the best of our knowledge, no studies are available on native galectin-16 at the protein level whereas recombinant protein has been produced and tested. The crystal structure of recombinant galectin-16 and its mutants was solved by Si and co-authors [32]. Recombinant galectin-16 is a monomeric protein, which is composed of 142 amino acids and has a typical galectin structure of the CRD β-sandwich with two sheets formed by six β-strands on the concave side (S1-S6) and five β-strands on the convex side (F1-F5) ( Figure  2). This group also showed that galectin-16 lacks lactose-binding ability unless arginine (Arg55) is replaced with asparagine in S4 β-strand. In comparison, an earlier report showed that recombinant galectin-16 and two other human galectins (galectin-13 and galectin-14) can bind lactose-agarose beads and are efficiently and competitively eluted by lactose [11]. More insights into this discrepancy are required considering multiple interfering factors, mutations/replacements of amino acids within the CRD, and different study designs. Regardless, both glycan-dependent and glycan-independent interactions might be essential for galectin-16 similar to other galectins [3].

Expression Patterns and Functions of LGALS16 in Cells and Tissues
Experimental studies focusing on LGALS16 are limited and an essential source of relevant information about this gene is Gene Expression Omnibus (GEO), a data repository for microarray and RNA-sequencing data [33]. Overall, 287 datasets are available on GEO (November 2021 search) reporting LGALS16 expression in 52 types of tissues and various cell lines based on the following platforms: Affymetrix Human Genome (n = 27), Affymetrix Human Gene (n = 151), Agilent (n = 31), Human Unigene (n = 1), Illumina Human (n = Each color corresponds to the exon from which the amino acids were encoded with red, orange, blue, and black representing exons 1, 2, 3, and 4, respectively. The anti-parallel β-sheets of F-face (F1-F5) and S-face (S1-S6) strands as well as a short α-helix are showed. (b) The crystal structure was extracted from Protein Data Bank (available online: rcsb.org, accessed on 6 September 2021), PDB ID: 6LJP.

Expression Patterns and Functions of LGALS16 in Cells and Tissues
Experimental studies focusing on LGALS16 are limited and an essential source of relevant information about this gene is Gene Expression Omnibus (GEO), a data repository for microarray and RNA-sequencing data [33]. Overall, 287 datasets are available on GEO (November 2021 search) reporting LGALS16 expression in 52 types of tissues and various cell lines based on the following platforms: Affymetrix Human Genome (n = 27), Affymetrix Human Gene (n = 151), Agilent (n = 31), Human Unigene (n = 1), Illumina Human (n = 82), MCI Human (n = 1), NuGO (n = 1), and Sentrix Human (n = 15). Quantification of differences in LGALS16 expression between different platforms is challenging. However, evaluation of gene expression values within the same GEO datasets demonstrates that LGALS16 can be classified as a gene with relatively low expression in comparison with LGALS1 (a widely expressed galectin with a low tissue specificity) and ACTB (a common housekeeping gene) (Table 1). Indeed, regardless of the platform, average GEO percentile rank of expression for LGALS16 measured with different arrays ranged 4-32% on a scale of 1-100% while the range was 63-100% for LGALS1 and 94-100% for ACTB. Available GEO profiles do not contain relevant datasets with LGALS16 for placenta for comparison, however, the Human Protein Atlas (HPA) reports tissue-specific overexpression of LGALS16 in placenta followed by brain tissues and retina ( Figure 3a). The biological meaning and reasons of overexpression of LGALS16 in these diverse tissues is unknown and requires further investigations in the context of developmental biology. For instance, the complex mechanisms of the placenta-brain axis of cell development [34] could be addressed in terms of the unique association of LGALS16 with these tissues.  In comparison with tissues, HPA reports the expression of two human cell lines including placental choriocarcinoma cell teratoma cell line SuSa, which probably can be used as appropria biological role of LGALS16 gene (Figure 3b). Human syncytiotro minally differentiated placental cells, can also serve as a str LGALS16 overexpression [10][11][12].
To develop experimental models for studying LGALS16 fun examined the gene expression in BeWo cells and an additional p the context of trophoblastic differentiation. The expression of L nificantly increased in both cell lines after 36 h (JEG-3 cells) and ment with a potent cell-permeable and metabolically stable activ protein kinase 8-Br-cAMP (250 µM), which coincided with upreg encoding chorionic gonadotropin subunits 3 and 5 ( Figure 4). As In comparison with tissues, HPA reports the expression of LGALS16 mRNA only in two human cell lines including placental choriocarcinoma cell line BeWo and testicular teratoma cell line SuSa, which probably can be used as appropriate systems to explore the biological role of LGALS16 gene (Figure 3b). Human syncytiotrophoblasts, which are terminally differentiated placental cells, can also serve as a strong positive control for LGALS16 overexpression [10][11][12].
To develop experimental models for studying LGALS16 functions and regulation, we examined the gene expression in BeWo cells and an additional placental cell line JEG-3 in the context of trophoblastic differentiation. The expression of LGALS16 mRNA was significantly increased in both cell lines after 36 h (JEG-3 cells) and 48 h (BeWo cells) treatment with a potent cell-permeable and metabolically stable activator of cAMPdependent protein kinase 8-Br-cAMP (250 µM), which coincided with upregulation of CGB3/5, genes encoding chorionic gonadotropin subunits 3 and 5 ( Figure 4). As chorionic gonadotropin is one of the biomarkers of placenta and trophoblastic differentiation, our results suggest classifying LGALS16 to the same category of biological molecules. Other studies also reported significant upregulation of LGALS16 in association with processes of cellular differentiation, even if the basal levels were relatively low. Thus, treatment of BeWo cells with forskolin, an inducer of cyclic adenosine 3 ,5 -monophosphate (cAMP), stimulated trophoblastic differentiation and simultaneous LGALS16 overexpression [12]. An interesting example of LGALS16 upregulation was reported in a model of intestinal differentiation of Caco-2 cells induced by a combined treatment with dexamethasone and p44/42 MAPK inhibitor PD98059 [35]. Therefore, LGALS16 may deserve further attention as a factor associated with processes of cellular differentiation and tissue development.
stronger than in the cytoplasm suggesting that the transport of galectin-16 into the nucleus might play a role in regulating intranuclear processes. These authors showed that the binding partner of galectin-16 is c-Rel, a member of the NF-κB family of transcription factors (TFs), which is involved in the regulation of multiple processes such as apoptosis, inflammation, immune responses, tumorigenesis, cell growth and differentiation [32,36]. All NF-κB family members, including c-Rel, have a conserved N-terminal DNA-binding/dimerization domain, known as the Rel homology domain (RHD) [37]. Recombinant galectin-16 strongly binds to the RHD which might inhibit c-Rel and prevent activation of anti-apoptotic genes, such as Bcl-2 and Bcl-xL, promoting T-cell apoptosis during pregnancy [32]. An additional aspect of LGALS16 functions may contribute to the rescue of glucose restriction-induced cell death in a model of a whole genome gain-of-function CRISPR activation using human mitochondrial disease complex I mutant cells [38].  An important function of galectin-16 as well as placental galectin-13 and galectin-14 is the ability to induce apoptosis of CD3+ T cells, which was detected by flow cytometry of cells double-stained with annexin-V-FITC and propidium iodide [11]. Considering the high expression of galectin-16 in differentiated trophoblasts, the apoptotic mechanism might contribute to the immune tolerance at the maternal-fetal interface reducing the danger of maternal immune attacks on the fetus and enabling anthropoid primates to evolve long gestation periods while retaining highly invasive placentation. The details of this regulation are obscure since there are no studies addressing the secretion of galectin-16 from trophoblasts. However, intracellular EGFP-tagged recombinant galectin-16 was readily localized in the nucleus and cytoplasm of transfected cells including, HeLa, 293T, HCT-116, SMMC-7721 and Jurkat cells [32]. In fact, the nuclear staining was much stronger than in the cytoplasm suggesting that the transport of galectin-16 into the nucleus might play a role in regulating intranuclear processes. These authors showed that the binding partner of galectin-16 is c-Rel, a member of the NF-κB family of transcription factors (TFs), which is involved in the regulation of multiple processes such as apoptosis, inflammation, immune responses, tumorigenesis, cell growth and differentiation [32,36]. All NF-κB family members, including c-Rel, have a conserved N-terminal DNA-binding/dimerization domain, known as the Rel homology domain (RHD) [37]. Recombinant galectin-16 strongly binds to the RHD which might inhibit c-Rel and prevent activation of anti-apoptotic genes, such as Bcl-2 and Bcl-xL, promoting T-cell apoptosis during pregnancy [32]. An additional aspect of LGALS16 functions may contribute to the rescue of glucose restriction-induced cell death in a model of a whole genome gain-of-function CRISPR activation using human mitochondrial disease complex I mutant cells [38].

Transcription Factors
Multiple TFs can be involved in the regulation of LGALS16 expression based on the presence of specific response elements in the promoter regions of the gene. Original analysis of retrotransposons within the 10 kb 5 UTR by Than and co-authors demonstrated that the LGALS16 promoter has binding sites for GATA2, TEF5, and ESRRG, which are also involved in the regulation of important trophoblast-specific genes such as ERVWE1 (marker of cell fusion), CGA, and CGB3 (markers of chorionic gonadotropin production, a hormone released by differentiated trophoblasts to maintain pregnancy) [12]. The contribution of these TFs in regulating LGALS16 expression was claimed to vary, especially with decreased regulation from GATA2, due to the specific layout and properties of transposable elements (L1PA6 and L1PREC2) within the 5 UTR of this gene as compared to two other placental genes, LGALS13 and LGALS14. Additional shared TFs for the placental galectin gene cluster include TFAP2A and GCM1, which have binding sites within ALU transposable elements next to L1PREC2. Experimental evidence of this regulation was confirmed in a model of forskolin-induced differentiation of primary trophoblasts, which revealed timedependent upregulation of LGALS16 in parallel with the expression of TEAD3, ESRRG, GCM1, and ERVWE1 [12]. It is interesting to note that this study did not reveal the effect of 5-azacytidin on LGALS16 expression in BeWo trophoblast cells as compared to other upregulated placental galectins, which suggested a minor role of DNA methylation in the context of LGALS16 regulation.
To enrich this analysis, we used human choriocarcinoma cell line JEG-3 and Qiagen RT 2 Profiler™ PCR Array to test changes in the mRNA transcript levels of 84 TFs during trophoblastic differentiation induced by 8-Br-cAMP. Overall, 60 TFs were upregulated in this assay including three top genes encoding Jun B proto-oncogene (JUNB), SMAD family member 9 (SMAD9), and activating transcription factor 3 (ATF3) ( Figure 5). Since all of these three genes are expressed in placenta and brain tissues [21,[39][40][41][42][43], which are LGALS16-positive, this observation provides a new insight into possible transcriptional regulation of this gene. JUNB and ATF3 belong to a family of TFs with a basic leucine zipper DNA binding domain, with JUNB preferentially binding to the 12-O-tetradecanoylphorbol-13-acetate response element sequence and ATF3 binding to the cAMP response element in promoters with the consensus sequence, TGACGTCA [44]. They are subunits of activating protein 1 (AP-1) TFs, which function as homodimers or heterodimers in association with other members of JUN, FOS, ATF, and MAF protein families [44]. JUNB was reported to be directly involved in processes of trophoblastic cell syncytialization [45,46], while upregulation of ATF3 was associated with cellular stress responses [41,47,48], decidualization [47], and preeclampsia [49]. In comparison, SMAD9 is activated by bone morphogenic proteins (BMPs), a subfamily of the transforming growth factor-β (TGF-β) family [40,50]. Although some BMPs such as BMP-4 can be regulated in a downstream manner from the cAMP pathway [51], the connection between SMAD9 and cAMP is still unclear. GATA2 was found to be slightly upregulated, which may suggest that the enrichment of L1PREC2 in the 5 UTR still plays a role in regulating LGALS16 despite the insertion of L1PA6 [12]. Interestingly, CREB1, a major regulator downstream of the cAMP pathway was not upregulated in this RT-qPCR array analysis suggesting that post-translational modification and transcriptional activation might be essential for this TF. We further analyzed the 2 kb region upstream from the transcription start LGALS16 by extracting the sequence from Ensembl (Release 104) and performing in analysis of TF-binding sites using PROMO virtual laboratory with a 0% dissimilar dex. Putative binding sites for seventeen TFs were identified, which may represen cific response elements, enhancers, or silencers ( Figure 6). To reveal common patte the expression of the predicted TFs, we watched for their protein levels in two LGA positive tissues, the cerebellum and placenta, using the expression scores (high, me low) available at HPA. Within this set of data, two TFs (CEBPβ and TFII-I) were c terized by high protein expression levels, seven TFs had variable levels (GR, NFAT STAT4, TCF-4E, TFIID, and YY1), four TFs (ERα, FOXP3, Pax5, and PR A) showe expression, and no HPA data were available for GRα, GRβ, PR B, and XBP-1 in the sues (Figure 7). Thus, the role of the predicted TFs in tissue-specific transcriptional lation of LGALS16 can be different and remains to be studied. We further analyzed the 2 kb region upstream from the transcription start site of LGALS16 by extracting the sequence from Ensembl (Release 104) and performing in silico analysis of TF-binding sites using PROMO virtual laboratory with a 0% dissimilarity index. Putative binding sites for seventeen TFs were identified, which may represent specific response elements, enhancers, or silencers ( Figure 6). To reveal common patterns in the expression of the predicted TFs, we watched for their protein levels in two LGALS16-positive tissues, the cerebellum and placenta, using the expression scores (high, medium, low) available at HPA. Within this set of data, two TFs (CEBPβ and TFII-I) were characterized by high protein expression levels, seven TFs had variable levels (GR, NFAT1, p53, STAT4, TCF-4E, TFIID, and YY1), four TFs (ERα, FOXP3, Pax5, and PR A) showed low expression, and no HPA data were available for GRα, GRβ, PR B, and XBP-1 in these tissues (Figure 7). Thus, the role of the predicted TFs in tissue-specific transcriptional regulation of LGALS16 can be different and remains to be studied. We further analyzed the 2 kb region upstream from the transcription start site of LGALS16 by extracting the sequence from Ensembl (Release 104) and performing in silico analysis of TF-binding sites using PROMO virtual laboratory with a 0% dissimilarity index. Putative binding sites for seventeen TFs were identified, which may represent specific response elements, enhancers, or silencers ( Figure 6). To reveal common patterns in the expression of the predicted TFs, we watched for their protein levels in two LGALS16positive tissues, the cerebellum and placenta, using the expression scores (high, medium, low) available at HPA. Within this set of data, two TFs (CEBPβ and TFII-I) were characterized by high protein expression levels, seven TFs had variable levels (GR, NFAT1, p53, STAT4, TCF-4E, TFIID, and YY1), four TFs (ERα, FOXP3, Pax5, and PR A) showed low expression, and no HPA data were available for GRα, GRβ, PR B, and XBP-1 in these tissues ( Figure 7). Thus, the role of the predicted TFs in tissue-specific transcriptional regulation of LGALS16 can be different and remains to be studied.

miRNAs
Post-transcriptional control of mRNA availability for protein miRNAs which can hybridize to complementary sequences in prot the 3′ untranslated region and either block protein translation or in tion [52]. Multiple miRNAs were predicted to target the LGALS16 matics tools, such as Diana Tools [17], miRabel [18], miRDB [19], which use different algorithms and methods. A robust applicatio default options shows that five miRNAs (hsa-miR-3155a, hsa-miR hsa-miR-4778-5p, hsa-miR-6783-5p) are predicted by all four of t ( Figure 8). These miRNAs among others can be considered as pers regulating the stability and/or translational potential of the LGAL cially in relevant tissues such as placenta and brain. Indeed, a signi miR-4778-5p expression during gestation in exosomes from mater

miRNAs
Post-transcriptional control of mRNA availability for protein synthesis depends on miRNAs which can hybridize to complementary sequences in protein-coding mRNAs at the 3 untranslated region and either block protein translation or induce mRNA degradation [52]. Multiple miRNAs were predicted to target the LGALS16 transcript by bioinformatics tools, such as Diana Tools [17], miRabel [18], miRDB [19], and TargetScan [20], which use different algorithms and methods. A robust application of these tools using default options shows that five miRNAs (hsa-miR-3155a, hsa-miR-3155b, hsa-miR-4689, hsa-miR-4778-5p, hsa-miR-6783-5p) are predicted by all four of these online platforms (Figure 8). These miR-NAs among others can be considered as perspective candidates for regulating the stability and/or translational potential of the LGALS16 transcripts, especially in relevant tissues such as placenta and brain. Indeed, a significant decrease in hsa-miR-4778-5p expression during gestation in exosomes from maternal blood was associated with preterm birth pregnancies [53]. Expression of hsa-miR-3155a was significantly upregulated in the anterior cingulate cortex of deceased patients with major depressive disorder [54]. In comparison, the expression of hsa-miR-4689 was downregulated in exosomes isolated from the plasma of patients with mesial temporal lobe epilepsy with hippocampal sclerosis compared to controls [55]. Differential expression of exosomal hsa-miR-4689 and hsa-miR-6783-5p was reported in patients with intracranial aneurysms [56]. Human miRNA tissue atlas confirms expression of hsa-miR-3155a, hsa-miR-3155b, hsa-miR-4689, and hsa-miR-4778-5p in brain among other tissues at variable levels [57]. Unraveling possible mechanisms of miRNA-mediated regulation of galectin-16 in these tissues awaits future research.

LGALS16 and Human Diseases
Dysregulation of the placenta-specific gene cluster containing LGALS16 is with a pregnancy complication known as preeclampsia, which can be highly fa the mother and fetus. As such, LGALS16 together with LGALS13 and LGALS14 firmed to satisfy the criteria of placenta enriched genes in a comprehensive stud Seq datasets from 302 placental biopsies [58]. However, although increasing ex LGALS13, LGALS14, and LGALS16 was observed during forskolin-induced sy tion and differentiation of primary trophoblasts and BeWo cells in culture, only and LGALS14 were downregulated in preeclampsia with no significant c LGALS16 [12]. Remarkably, LGALS16 does not show sex-biased expression dep the chromosomal sex of the fetus while LGALS13 and LGALS14 are notably e fetal male placentas based on the chorionic villus transcriptome [59]. These asp lectin network regulation remain unclear in the context of placental disorders opment.
Alterations in the expression or mutations of LGALS16 have been also re several other diseases and based mostly on microarray and RNA-Seq analysis the application of this gene as a biomarker is still unknown. Gene expressio with RNA sequencing data revealed that LGALS16 was detected as an upregu in fusiform gyrus tissue sections of 219 autopsy-confirmed Alzheimer's cases neurologically normal age-matched controls [60].
LGALS16 was also recognized tissue-specific gene within genome-wide associations with several neuroimagi atric traits [61]. Further, LGALS16 was expressed two-fold higher in chronic m kemia granulocytes compared to controls [62]. Copy number variations were id chromosome 19 for multiple genes including LGALS16 in association with c tures, such as histological type, ethnicity, disease stage, and familial history, of cer using tumor samples from a Brazilian cohort [63]. In addition, LGALS16 mined to be a moderate impact variant associated with autism spectrum diso sisting of a missense single nucleotide variant (SNV), which was reported as d by bioinformatic tools SIFT and PolyPhen-2 [64].
LGALS16 also had greater SN the 3′ flank region with one or more mutations in patients with diffuse large phoma [65]. Moreover, a LGALS16 SNP was revealed to be associated with ins

LGALS16 and Human Diseases
Dysregulation of the placenta-specific gene cluster containing LGALS16 is associated with a pregnancy complication known as preeclampsia, which can be highly fatal for both the mother and fetus. As such, LGALS16 together with LGALS13 and LGALS14 were confirmed to satisfy the criteria of placenta enriched genes in a comprehensive study of RNA-Seq datasets from 302 placental biopsies [58]. However, although increasing expression of LGALS13, LGALS14, and LGALS16 was observed during forskolin-induced syncytialization and differentiation of primary trophoblasts and BeWo cells in culture, only LGALS13 and LGALS14 were downregulated in preeclampsia with no significant changes of LGALS16 [12]. Remarkably, LGALS16 does not show sex-biased expression depending on the chromosomal sex of the fetus while LGALS13 and LGALS14 are notably elevated in fetal male placentas based on the chorionic villus transcriptome [59]. These aspects of galectin network regulation remain unclear in the context of placental disorders and development.
Alterations in the expression or mutations of LGALS16 have been also reported for several other diseases and based mostly on microarray and RNA-Seq analysis, although the application of this gene as a biomarker is still unknown. Gene expression profiling with RNA sequencing data revealed that LGALS16 was detected as an upregulated gene in fusiform gyrus tissue sections of 219 autopsy-confirmed Alzheimer's cases versus 70 neurologically normal age-matched controls [60].
LGALS16 was also recognized as a brain tissue-specific gene within genome-wide associations with several neuroimaging psychiatric traits [61]. Further, LGALS16 was expressed two-fold higher in chronic myeloid leukemia granulocytes compared to controls [62]. Copy number variations were identified in chromosome 19 for multiple genes including LGALS16 in association with clinical features, such as histological type, ethnicity, disease stage, and familial history, of breast cancer using tumor samples from a Brazilian cohort [63]. In addition, LGALS16 was determined to be a moderate impact variant associated with autism spectrum disorder, consisting of a missense single nucleotide variant (SNV), which was reported as detrimental by bioinformatic tools SIFT and PolyPhen-2 [64].
LGALS16 also had greater SNVs within the 3 flank region with one or more mutations in patients with diffuse large B-cell lymphoma [65]. Moreover, a LGALS16 SNP was revealed to be associated with insulin secretion in a cohort of African Americans [66]. This study also showed that interactions between this LGALS16 SNP and others, such as an intergenic SNP upstream of the LYPLAL1 gene, have also been associated with type 2 diabetes risk. The LGALS16 transcript was one of the top 50 down-regulated mRNA present in the exosomes isolated from the cerebrospinal fluid in patients with meningeal carcinomatosis in comparison with healthy controls [67].

Conclusions
Although the galectin-16 gene was described more than 10 years ago [11], the regulation, functions, and clinical aspects of this tissue-specific molecule are largely unexplored. Primary association of LGALS16 with placental tissue has been challenged by its detection in brain tissues and several cancer cell lines as followed from available microarray and RNA-seq databases. There are bioinformatics indications that the expression of LGALS16 changes in association with Alzheimer's disease, chronic myeloid leukemia, breast cancer, B-cell lymphoma, and type 2 diabetes. Although LGALS16 was not significantly impacted at the gene level in preeclampsia, there remain questions regarding regulation at the protein level, which cannot be properly addressed at this time due to the absence of commercially available specific galectin-16 antibodies. The results obtained with recombinant galectin-16 are promising, but there is still a gap in our understanding of why the expression of endogenous galectin-16 protein has not been reported. Nevertheless, among the possible functions of galectin-16 in these and other tissues, its contribution to the regulation of cellular differentiation and programmed cell death (apoptosis) warrants special attention. Lastly, the use of proper cell culture models and the examination of multiple factors (transcription regulators and miRNA) is evidently the first line of study to position galectin-16 within a complex galectin network in cells. The generation of galectin-16-specific antibody and LGALS16 knockout cell lines using CRISPR/Cas9 technology might be required steps to unravel the role and significance of this molecule in the context of cell biology.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.