1 INTRODUCTION

Over the last decade, targeted therapies and immune checkpoint inhibitors have significantly improved the management of patients with melanomas. The choice of treatment is based both on the detection of driver mutations and on anatomopathological analysis that explores histopathological criteria and the immunohistochemical characteristics of the melanoma (Ascierto et al., 2020). These investigations define the percentage of tumor cells per sample and are based on the detection of highly expressed markers that are found specifically in melanocytes, including human melanoma black-45 (HMB-45), Melan-A, tyrosinase (TYR), S100, and TYRP1 (Gogas et al., 2009). These biomarkers can also be used to characterize tumor heterogeneity during relapse (Bai et al., 2019; Rambow et al., 2018). Recently, we and other researchers showed that TYRP1 mRNA and proteins are more than just humble markers (El Hajj et al., 2015; El Hajj et al., 2013; Ghanem & Journe, 2011; Gilot et al., 2017). Therefore, an in-depth knowledge of the regulation of TYRP1 promoter, RNAs, and proteins is important to elucidate the “indirect oncogenic activity” of TYRP1 mRNA, which is based on miR-16 sequestration. Moreover, the restricted expression of TYRP1 in melanocytes and the retinal pigment epithelium (RPE) suggests that TYRP1 mRNA might be a remarkable target for cancer therapy, as there is no fear of harmful effects on the other cells in the body, which are all TYRP1-negative. We also discuss here an antisense oligonucleotide strategy aimed at avoiding miRNA sequestration on TYRP1, thereby restoring the mRNA activity of the tumor suppressor miR-16.

2 TYRP1 GENE REGULATION 2.1 Gene

Human TYRP1 cDNA was isolated from melanoma cells in 1990. This gene is located on chromosome 9 (9p23) at base pairs 12,693,385–12,710,266 in the NCBI GRCh38/hg38 assembly. It encodes the human homolog of the mouse brown (b) locus gene product (Bennett et al., 1990; Jackson, 1988). The human TYRP1 gene is spread over 24 kbp of genomic DNA, as compared to 18 kbp for the mouse version (Shibahara et al., 1992; Sturm et al., 1995), and the human and mouse TYRP1 proteins are encoded by 8 exons (Figure 1).

Human TYRP1 gene and its products. Human TYRP1 gene, transcript (a) and protein (a, b) details according to Ensembl

The TYRP1 GeneID at NCBI is 7306 (https://www.ncbi.nlm.nih.gov/gene/7306). Additional information is available on other websites, including Ensembl:ENSG00000107165, MIM:115501, and Vega:OTTHUMG00000021034. Synonyms for TYRP1 include OCA3, TYRRP, GP75, CATB, TRP, b-PROTEIN, TYRP, CAS2, and TRP1.

2.2 Expression

TYRP1 is involved in the production of melanin pigment, so it is mostly expressed in cell types that produce melanin, including melanocytes and the RPE (Murisier & Beerman, 2006). The RPE originates from the optic neuroepithelium, is located close to the retina, and is crucial for eye organogenesis and vision (Bharti et al., 2006). Melanocytes, which are derived from the neural crest, can be classified into two groups: cutaneous/classical melanocytes which can be found in the skin (epidermis and dermis); and non-cutaneous/non-classical melanocytes which colonize the eye, inner ear, meninges, heart, and adipose tissues (Petit & Larue, 2016; Randhawa et al., 2009; Yajima & Larue, 2008). While TYRP1 can be detected in tissues colonized by non-cutaneous melanocytes, this RNA is mainly detected in the skin (Figure 2a).

(a) RNAseq analysis of human TYRP1 expression in different tissues from the Human Protein Atlas dataset. Results are expressed in transcripts per million kilobases). (b) TYRP1 mRNA expression in 10,967 tumor samples from the TCGA Pan-Cancer Atlas (www.cbioportal.org). (c) TYRP1 expression in TCGA melanoma biopsies (n = 459 melanoma samples). To illustrate the expression heterogeneity of pigmentation-related mRNAs, the genes TYR, DCT, MITF, and MLANA were also analyzed. These results are based upon data generated by the TCGA research network (http://cancergenome.nih.gov/)

Most cutaneous melanocytes are located in the epidermis, and are follicular and interfollicular (often called “epidermal”) melanocytes. These are both involved in hair and skin pigmentation as well as in protecting skin against DNA damage or oxidative stress. Melanocytes in the hair follicles also contribute to the elimination of the toxic by-products that result from melanin synthesis, while epidermal melanocytes are involved in inflammatory response by acting as phagocytic cells (Colombo, Berlin, Delmas, & Larue, 2011). TYRP1 expression in both cell types is dependent on melanocyte differentiation, and seems to be necessary for the maturation of the melanosome, the organelle that synthesizes, stores, and transports melanin. Only melanocytes with mature melanosomes (types III and IV) seem to express TYRP1 (Cichorek, Wachulska, Stasiewicz, & Tymińska, 2013; Jimbow et al., 2000; Raposo & Marks, 2007).

Follicular melanocyte maturation follows the hair development cycle, starting with the anagen growth phase of active melanogenesis, then a regressive phase where mature melanocytes undergo apoptosis, and finally the telogen quiescent phase (Qiu, Chuong, & Lei, 2019; Schneider, Schmidt-Ullrich, & Paus, 2009). During the hair cycle, TYRP1 is only expressed in the anagenic follicular melanocytes localized in the hair matrix which are responsible for hair pigmentation (Slominski et al., 2005). Both hair cycles and pigmentation are regulated by the circadian clock, and TYRP1 levels have been shown to depend on the expression of clock genes. Indeed, silencing the core clock genes BMAL1 and PER1 extends the anagen phase and increases TYRP1 expression levels (Hardman et al., 2015; Plikus et al., 2015).

TYRP1 is also highly expressed in tumors derived from melanocytes, cutaneous, and uveal melanomas. In benign nevus and melanoma tumors, TYRP1 mRNA expression levels are variable (Figure 2b,c). The three members of the tyrosinase family, tyrosinase (TYR), the dopachrome tautomerase (DCT/TYRP2), and TYRP1, are not strictly correlated even if their expression levels are all at least in part governed by the same transcription factor, MITF (melanocyte/microphthalmia-associated transcription factor). In melanoma biopsies from the cutaneous skin cancer cohort in the Cancer Genome Atlas (TCGA), the expression pattern of TYRP1 is well correlated with those of TYR and MLANA (Figure 2c). TYRP1 can also be detected in other types of cancers arising from non-melanocytic lineage tissues such as colon and breast cancers (Hsu et al., 2018; Montel et al., 2009). Taken together, these studies highlight the specific TYRP1 expression profile in tissues, as well as its involvement in the pigmentation process.

2.3 TYRP1 gene promoter and enhancer

TYRP1 expression is tightly associated with melanocyte differentiation and pigmentation. Even if MITF plays a predominant role in TYRP1 expression by targeting the proximal promoter, TYRP1 expression also depends on other transcriptional factors as well as on a distal enhancer.

2.3.1 Role of the distal enhancer of Tyrp1

Numerous positive and negative transcription regulators have been identified for the TYRP1 promoter (Table 1 and Figure 3). Data have been mainly produced from murine models (Tyrp1). MITF’s ability to transactivate TYRP1 and Tyrp1 promoters has been clearly demonstrated in numerous studies including those in knockout mice (Hemesath et al., 1994; Hodgkinson et al., 1993). MITF belongs to the MiTF/TFE group in the basic helix-loop-helix—leucine-zipper (bHLH-LZ) transcription factor family (Hemesath et al., 1994; Hodgkinson et al., 1993; Nakayama et al., 1998). The bHLH-LZ transcription factors bind an E-box consensus sequence (CANNTG), while MITF binds the M-box, a sequence of 11 bp that contains an E-box core (AGTCATGTGCT). In mice, this box is located upstream of the TATA box in the Tyrp1 promoter and transactivates it (Bertolotto et al., 1998). Furthermore, even if it is to a lesser extent, Mitf induces Tyrp1 expression by binding an E-box (CAAGTG) located at −238/−233 (Bertolotto et al., 1998). In human cells, data from chromatin immunoprecipitation (ChIP) experiments followed by DNA sequencing have demonstrated a physical interaction between MITF and TYRP1 (Strub et al., 2011). Moreover, MITF cDNA overexpression promotes TYRP1 expression by causing MITF to bind to the M-box (Yasumoto et al., 1997). The human TYRP1 M-box (AATCATGTGCT) occurs from −197 to −186 on the GCRh38/hg38 assembly, and it is a strong activator of pigment cells (Shibata et al., 1992; Yasumoto et al., 1997).

TABLE 1. Direct activators and repressors of TYRP1 expression Factor Role Model Reference Mouse Human MITF Activator X X Bertolotto et al. (1998), Shibata et al. (1992) and Yasumoto et al. (1997) TFEB Activator X Verastegui et al. (2000) TFE3 Activator X Verastegui et al. (2000) Pax3 Activator X Galibert et al. (1999) Tbx2 Repressor X Carreira et al. (1998) p53 Activator X Nylander et al. (2000) p63α Activator X Nylander et al. (2000) p73α Activator X Nylander et al. (2000) Sox10 Activator X X Murisier et al. (2006) OTX2 Activator X Martínez-Morales et al. (2003) (a) Transcription factors bind the TYRP1 enhancer and promoter at specific binding sites. Some have been studied in mouse or human models or in both. Notably, the M-box is green. Tbx2 and Pax3 binding sites in the human TYRP1 promoter were predicted according to their binding sites from a murine model, and these predicted sites are in orange italics. OTX2 binding sites are depicted for retinal pigment epithelium cells. (b) A schema showing the role of MITF, SOX10, and BRG1 in the TYRP1 promoter. In differentiated cells, this promoter becomes accessible to MITF. SOX10 binds to the enhancer and recruits BRG1 to form a chromatin loop. Consequently, these three proteins are able to transactivate the TYRP1 (Marathe et al., 2017)

While both MITF and the M-box are required for TYRP1 expression in melanocytes, a conserved transcriptional enhancer also seems to play a crucial role (Marathe et al., 2017; Murisier et al., 2006). This element is located at approximately −15 kb and contains Sox10 binding sites, which are perfectly conserved between the mouse and human genomes. By combining ChIP experiments and formaldehyde-assisted isolation of regulatory elements (FAIRE) analysis, the authors showed that the distal enhancers of Tyrp1 are in an inherently open chromatin state prior to differentiation. Upon differentiation of mouse melanoblast Melb-a cells, the chromatin becomes accessible at the proximal promoter of Tyrp1 (Marathe et al., 2017). This study strongly suggests that SOX10 binds to the distal enhancer and recruits brahma-related gene-1 (BRG1) to form a loop with the proximal promoter, which contains MITF-binding sites. In Figure 3b, we present a schema illustrating this hypothetical mechanism. A chromatin loop promoting the connection between enhancer and proximal promoter of Tyrp1 likely occurs in the presence of the three main actors, SOX10, BRG1, and MITF. MITF transcription factor expression increases during differentiation, and its rising expression allows for the transactivation of pigmentation genes such as Tyrp1 and Tyr.

2.3.2 Role of the first intron in Tyrp1 mRNA expression

The mouse Tyrp1 promoter contains MSEu and MSEi, two related cis-acting elements that are upstream and initiator melanocyte-specific elements, respectively. These two GTGTGA sequences are binding sites for Pax3 and Tbx2 (Carreira et al., 1998; Galibert et al., 1999). Pax3 binding on MSEu (−241 to −235 in the GRCm38/mm10 assembly) and on MSEi leads to Tyrp1 transcription activation, whereas Tbx2 acts as a repressor (Carreira et al., 1998). These two sites and their respective positions are not conserved in the human TYRP1 promoter. Two GTGTGA sequences separated by 21 nucleotides are found in the TYRP1 intron 1 (+334 to +340 and +361 to +367 in the GRCh38/hg38 assembly). Moreover, it has been proposed that TYRP1 is regulated by PAX3 through a TGTCACACTT sequence (Corry & Underhill, 2005). In human TYRP1, this sequence is localized in intron 1 between the two GTGTGA sequences just discussed. This finding is of particular interest because it is possible that the TYRP1 exon 1 and intron 1 are important for TYRP1 regulation (Murisier et al., 2006).

2.3.3 Transcription factors E3 and EB (Tfe3 and Tfeb)

MITF is not the only bHLH-LZ transcription factor able to promote TYRP1 expression. Tfeb and Tfe3 share a high homology with Mitf, and in the B16 mouse melanoma cell line, their ectopic expression transactivates the Tyrp1 promoter by binding the M-box (Verastegui et al., 2000). On the other hand, endogenous Tfe3 cannot bind to the M-box as a homodimer, and the authors did not detect any Tfe3/Mitf heterodimers. It is therefore difficult to be sure of the physiological roles of Tfe3 and Tfeb in basal Tyrp1 expression when Mitf proteins are present, and further investigations are required to prove their ability to promote Tyrp1 without ectopic expression. As observed in Mitf-knockout mice, there is a strong decrease in Tyrp1 levels when Mitf is depleted (Hemesath et al., 1994; Hodgkinson et al., 1993). This suggests that Mitf is required for Tyrp1 induction, and Tfe3 and/or Tfeb are not able to compensate for the lack of Mitf and induce Tyrp1.

2.3.4 Inducibility of TYRP1 gene by ultraviolet radiation

Interestingly, p53, which is activated in response to ultraviolet (UV) rays, can bind and transactivate the TYRP1 promoter via two clusters, each of which is composed of three binding sites (Nylander et al., 2000). These clusters are separated by 14 nucleotides, and their binding sites are all located in the −121 to −42 fragment. The p53-homolog proteins p63α and p73α also transactivate the TYRP1 promoter.

2.3.5 Tyrp1 expression repressors

In normal cells, TYRP1 mRNA is mainly expressed in melanocytes and in the RPE (Figure 2). The orthodenticle homeobox 2 (OTX2) protein transactivates the TYRP1 promoter via three binding sites (Martínez-Morales et al., 2003). OTX2 is an RPE-specific factor, while SOX10 and PAX3 are specifically expressed in melanocytes (Murisier & Beermann, 2006) and in testes. These transcription factors are also at least partly responsible for TYRP1 cell-specific expression.

In conclusion, the transcription factors governing murine Tyrp1 expression are better known than those driving human TYRP1. In order to map the regulatory elements of the human TYRP1 gene, it might be interesting to perform in situ tiling screens, using clustered regularly interspaced short palindromic repeats (CRISPR) for saturation mutagenesis experiments along the TYRP1 promoter (Canver et al., 2015).

2.4 RNA

The reference sequences (RefSeqs) for TYRP1 mRNA are NM_000550.3 in NCBI and ENST00000388918.9 in Ensembl (Figure 1). The mRNA length is 2,896 bp. The gene has seven transcripts (splice variants), 346 orthologs, and two paralogs. It is a member of one Ensembl protein family and is associated with one phenotype (Sarangarajan & Boissy, 2001): Type 3 oculocutaneous albinism (OCA3), described in more detail below.

The NM_000550.3 transcript contains eight exons and is associated with 526 variations. Importantly, NM_000550.3 replaced NM_000550.2 in 2019 as the official RefSeq (Figure 4). NM_000550.3 and NM_000550.2 differ by 1 bp in the 5′ untranslated region (UTR) and 19 bp in the 3′-UTR. In contrast, NM_000550.1 has two deletions of 4 nucleotides and four single nucleotide polymorphisms (SNPs) in the 3′-UTR, as well as being shorter (99 bp) than NM_000550.3. To date, the consequence of the two deletions observed in NM_000550.1 (AAGT, 2090–2094 and ATTA, 2264–2268) remains elusive. Interestingly, the chimpanzee genome also contains these two deletions, which suggests that the differences between NM_000550.1 and NM_000550.3 are probably not the result of sequencing experiment artifacts.

3′-UTR variability of TYRP1 mRNA. (a) Three mRNA variants have been successively used as reference sequences in the NCBI RefSeq database, with the NM_000550.3 isoform being used currently. This variant is distinguished from NM_000550.1 by the existence of four SNPs and two short deletions. (b) Distribution of the four SNPs in the CEU cohorts CEPH (Utah Residents with Northern and Western European Ancestry), JPT (Japanese in Tokyo, Japan), and YRI (Yoruba in Ibadan, Nigeria), as per https://www.ncbi.nlm.nih.gov/projects/SNP and ensembl.org/Homo_sapiens/Variation/Population

The four SNPs that differ between the NM_000550.3 and NM_000550.1 transcripts are rs683, rs2762464, rs910, and rs1063380 (Figure 4). Due to their physical proximity, these SNPs are tightly linked and inherited together (Li et al., 2012). These SNP alleles are associated with individuals from various backgrounds, including Caucasians (CEU CEPH, Utah Residents with Northern and Western European Ancestry), Asians (JPT, Japanese in Tokyo, Japan), and Africans (YRI, Yoruba in Ibadan, Nigeria). In 2012, Jinjing Li and colleagues demonstrated that rs683 and rs910 determine TYRP1 mRNA levels (Li et al., 2012), and they also showed that these SNPs are located in the microRNA response elements (MREs) for miR-155 (Figure 5).

Impact of SNPs on TYRP1 mRNA decay. (a) Relative positions of microRNA response elements (MREs) and the SNPs rs683 and rs910 on the TYRP1 3′-UTR. (b) These SNPs modify the base pairing between miR-155 and TYRP1 alleles. (c) Table showing the consequences of these SNPs on TYRP1 decay and expression as a function of TYRP1 allele and patient origin

MicroRNAs (miRNAs) are small (~22 nt) non-coding RNAs that mediate mRNA translation arrest and/or RNA decay through a perfect base pairing between miRNA seed sequences (nt 2–7) and the MRE of their RNA targets (Bartel, 2018). Three MRE-155s have been identified on the TYRP1 3′-UTR (Figure 5a). We recently confirmed that the first one seems to play a more minor role in TYRP1 RNA decay than to the other two (Gilot et al., 2017), and this is probably due to the proximity of the stop codon and the ribosome passage on this region. We also showed that the rs910 SNP in the third MRE-155 exerts a moderately increased effect on TYRP1 expression over that of rs683 in the second MRE. Indeed, the rs683-A allele, TYRP1-A from NM_000550.1, reduces miR-155-induced TYRP1 mRNA decay much more strongly than the C allele, TYRP1-C from NM_000550.3 (El Hajj et al., 2015; Li et al., 2012). Our results confirm that miR-155 affects TYRP1-A mRNA less than it affects TYRP1-C mRNA. In other words, the TYRP1 mRNAs expressed by two-thirds of CEU individuals carry the ancestral allele which disrupts the interaction between miR-155 and TYRP1, thereby contributing to elevated levels of TYRP1 mRNA in CEU individuals. In contrast, miR-155 strongly reduces the half-lives of the TYRP1-C alleles in the JPT and YRI cohorts (Figure 5b,c).

Interestingly, Li et al. analyzed the expression levels of TYRP1 as a function of population localization across the globe. They observed a strong negative correlation between the population's latitude of residence and the miR-155-mediated repression of TYRP1 (Li et al., 2012). Thus, people with high levels of sun exposure may display a faster turnover of TYRP1 than other populations.

Besides SNPs, other factors such as RNA-binding protein (RBPs) have been described as modulating the binding of miRNA on target sequences (Soemedi et al., 2014). Reciprocally, SNPs located on miRNA genes may affect mature sequences of the miRNAs and consequently modify their MRE-binding affinity. Huang et al. showed that miR-196a-2 carrying the rs11614913 C allele downregulates TYRP1 expression more strongly than the miRNA with the T allele. They proposed that this model could explain the relatively low TYRP1 expression in normal human melanocytes that contain that rs11614913 C allele (Huang et al., 2013). The authors also showed that this miR-196a-2 C-variant genotype is associated with a decreased risk of vitiligo, a depigmentation disorder characterized by the destruction of melanocytes because of an inherent sensitivity to oxidative stress (Cui et al., 2015; Jimbow et al., 2001; Picardo et al., 2015). The researchers hypothesized that this miR-196a-2 variant downregulates its targets, including TYRP1, thereby exerting potentially protective effects on human melanocytes under oxidative stress and reducing the risk of vitiligo (Huang et al., 2013). It has been proposed that normal TYRP1 protein expression protects vitiligo melanocytes from early death (Jimbow et al., 2001; Mansuri et al., 2016), indicating that miRNA regulation in the expression of TYRP1 in vitiligo should be clarified.

We recently identified three miR-16 binding sites on the TYRP1 3′-UTR (Figure 5a). These sites are common on the two alleles (NM_000550.3 and NM_000550.1). These MRE-16s have a non-coding role in TYRP1 mRNA function, and this is discussed at the end of this review.

2.5 TYRP1 protein and its functions

TYRP1 mRNA encodes TYRP1, a protein containing 537 amino acids (NP_000541.1 or ENSP00000373570). TYRP1 is synthesized as a core 55 kDa polypeptide (Figures 1 and 6), and the mature glycoprotein is ~75 kDa (Vijayasaradhi et al., 1991). Mature TYRP1 is a transmembrane protein that contains two zinc ions in its active sites instead of the copper ions found in tyrosinase (Lai et al., 2017, 2018). According to Ensembl (Figure 1), different TYRP1 proteins can be translated from three transcripts (splice variants). Despite using three different anti-TYRP1 antibodies that target different epitopes (G-17, Ab23, and PEP1), we were not able to identify the other putative TYRP1 proteins in cutaneous melanoma biopsies (Gilot et al., 2017).

TYRP1 protein domains and post-translational modifications. TYRP1 contains a signal-peptide region, cysteine-rich regions, two metal-binding domains, and a unique transmembrane region. Phosphorylation and glycosylation sites were obtained from www.phosphosite.org and www.uniprot.org

Human and mouse TYRP1/Tyrp1 have six potential N-glycosylation sites (Figure 6; Vijayasaradhi et al., 1991; Xu et al., 2001). The mean N-glycan composition of Tyrp1 is 16% high mannose, 16% biantennary, and more than 65% tri- and tetra-antennary structures (Branza-Nichita et al., 2000; Negroiu et al., 1999). This pattern of glycosylation seems important for TYRP1 processing, trafficking, and stability (Xu et al., 2001). The role of the N-glycosylation of TYR family members was recently discussed by Lai et al. (2020), who identified Asn371 on TYR as the most important N-glycosylation site for enzyme functioning. It might improve TYR stability and/or activity, and it corresponds to amino acid 385 in the human TYRP1 protein. Based on the crystal structure of glycosylated human TYRP1, the authors suggested that Asn385 glycosylation on TYRP1 may also impact TYRP1 activity and/or stability.

Interestingly, the classical glycosylation pathway can be bypassed (Branza-Nichita et al., 2000). In mice, this process occurs when the endomannosidase pathway directly in the Golgi compartment is used, and this leads to a different glycan pattern (Negroiu et al., 1999).

In addition, TYRP1 is phosphorylated on five sites in human proteins (S46, S137, S207, T222, S270) and on one site in murine proteins (S535; PhosphoSitePlus website [phosphosite.org]). These sites were assigned by mass spectrometry. Notably, only the serine in position 46 is not conserved between the human and mouse proteins. Their biological consequences of these sites remain to be elucidated.

After its synthesis in the endoplasmic reticulum and passage through the Golgi apparatus, the TYRP1 protein is transported through the endosomal recycling pathway, reaching the melanosomes through tubular–vesicular endosomal domains (Jani et al., 2015). The melanocyte's ultimate goal is to deliver melanin pigment to the keratinocytes, where it acts as an “umbrella” in the keratinocyte nucleus, preserving or limiting DNA alterations in response to UV irradiation. The path from the endoplasmic reticulum to the keratinocyte is a multistep process involving many proteins, and this “TYRP1 route” is not yet completely clear (Raposo & Marks, 2007).

In most animal models, TYRP1 is involved in the synthesis of eumelanin, a photoprotective pigment (Boissy et al., 1996; Zhao et al., 1996). However, its exact role in humans is not yet well-defined because of differing experimental results. Mouse Tyrp1 catalyzes the oxidation of 5,6-dihydroxyindole-2-carboxylic acid (DHICA), a major eumelanogenic intermediate (Jiménez-Cervantes et al., 1994; Kobayashi et al., 1994). A contrario, human TYRP1 does not express any DHICA oxidase activity (Boissy et al., 1998). Winder et al. showed that mouse Tyrp1 could be a dopachrome tautomerase that catalyzes the conversion of l-dopachrome in DHICA, probably in cooperation with Dct/Tyrp2 (Winder et al., 1994). Another study reported that human and mouse TYRP1 have catalase activity that could play a role in decomposing hydroperoxide to avoid disrupting melanin and its precursors (Halaban & Moellmann, 1990). Other potential enzymatic functions in pigmentation have been ascribed to human and/or murine TYRP1, such as the hydroxylation of tyrosine and the oxidation of DOPA or 5,6-dihydroxyindole (DHI; Sarangarajan & Boissy, 2001; Sarangarajan et al., 2000; Zhao et al., 1994). Because the TYRP1 structure was recently solved (Lai et al., 2017), these different TYRP1 activities should be revisited. The authors provided evidence for the presence of two zinc ions in the active site, which makes it unlikely that human TYRP1 could be a redox enzyme.

TYRP1 has been shown to play a non-enzymatic role in pigmentation by affecting the stability of TYR, the main pigmentation enzyme. Murine Tyrp1 forms a dimer with Tyr and thereby stabilizes it (Kobayashi & Hearing, 2007; Kobayashi et al., 1999). Human TYRP1 also forms a dimer with human TYR (Dolinska et al., 2019; Wu & Park, 2003). Tyrp1’s enhancement of Tyr stability is lost with the Tyrp1 brown mutation form (guanine to adenine at position 503; Manga et al., 2000). A previous report convincingly demonstrated that ectopic expression of human TYRP1 cDNA enhances TYR activity (Zhao et al., 1996). TYRP1 also contributes to melanosomal structure and maturation (Sarangarajan & Boissy, 2001). Mutation of Tyrp1 that leads to the mouse “light” phenotype (hairs pigmented only at their tips) has been associated with disrupted melanosomal structures (Johnson & Jackson, 1992). Moreover, expression of C110Y (a mutated brown TYRP1) by cDNA overexpression causes abnormal melanosome structures (Sarangarajan & Boissy, 2001; Sarangarajan et al., 2000).

TYRP1 mutations can lead to the OCA3 phenotype, a form of albinism characterized by rufous or brown skin coloring as well as visual anomalies. Several TYRP1 mutations have been identified and linked to this disease, which mainly occurs in the African population (Mártinez-García & Montoliu, 2013; Sarangarajan & Boissy, 2001; Simeonov et al., 2013) although OCA3-like cases have also been identified in Caucasian populations (Rooryck et al., 2006). In “brown” mutant mouse melanocytes (melan-b), Tyrp1 is not correctly processed, so the endoplasmic reticulum (ER) is retained (Toyofuku et al., 2001). Indeed, Toyofuku and colleagues hypothesized that OCA3 could be a Tyrp1 ER retention disease. In the Solomon Islands, over-representation of blond-haired people has also been associated with TYRP1 mutation (Kenny et al., 2012). Abnormal expression and cellular localizations of TYRP1 proteins have also been described in Hermansky–Pudlak syndrome, a multisystemic disease that includes oculocutaneous albinism (Boissy et al., 2005; Helip-Wooley et al., 2007; Richmond et al., 2005).

In conclusion, TYRP1 is clearly involved in the pigmentation cascade, although the exact role of human TYRP1 proteins remains a matter of debate. Further investigations are also needed to better understand the role of TYRP1 in other cellular processes.

2.6 Non-coding functions of TYRP1 mRNAs

There are few studies concentrating on TYRP1 or DCT rather than the more popular TYR. In light of this, we recently demonstrated a non-coding function of TYRP1 mRNA: promotion of melanoma growth (Gilot et al., 2017). This result is consistent with our previous publications, which show that high levels of TYRP1 mRNAs in patients with metastatic melanomas are associated with poor clinical outcomes (El Hajj et al., 2015; El Hajj et al., 2013; Ghanem & Journe, 2011; Journe et al., 2011). The “oncogenic role” of TYRP1 mRNA has long been ignored, as authors almost exclusively associated TYRP1 mRNA with the pigmentation process. However, two publications have suggested that TYRP1 mRNA could be involved in processes other than pigmentation (Fang et al., 2002; Vijayasaradhi et al., 1995). Indeed, Tyrp1 was implicated in melanocyte proliferation and survival (Li et al., 2004). In addition, an RNAi screen showed that the TYRP1 defect phenotype is characterized by G2 arrest (Kittler et al., 2007), although the TYRP1 protein or mRNA responsible for this phenotype has not yet been identified. In the same way, transfection of antisense oligonucleotides targeting TYRP1 in melanocytes and melanoma cell lines leads to G1 arrest, with increased apoptosis in vitro and decreased tumor growth in vivo (Li et al., 2004). Collectively, these data suggest that the expression levels of TYRP1 mRNA and/or TYRP1 proteins are associated with melanoma cell and/or melanocyte proliferation and survival.

In 2017, we showed that TYRP1 mR