The official HUGO gene nomenclature page says that GPR172A (Gamma-hydroxybutyrate receptor) and SLC52A2 (riboflavin transporter, member 2) are the same. The sequence reported by Andriamampandry seems to be the same, but I find no other evidence. Can anyone enlighten me?
This isn't a case of gene splicing causing different protein variants. In the studies that identified these two functions (GHB sensitivity and riboflavin transport), they were using DNA derived from mRNA (cDNA), which means what was being expressed in their experiments did not have introns, so there was no chance for alternative splicing.
This gene has a weird history, here is the summary:
Takeda et al. 2002: Identification of G protein-coupled receptor genes from the human genome sequence
Takeda and co. download a copy of the human genome and translate it looking for GPCRs. They look at all the ORF that DIDNT have introns: "We collected intronless open reading frames (ORFs), which were long enough to cover GPCRs from the human genome". They also only look at things with 6-8 transmembrane helices as determined by the prediction program SOSUI. One of the GPCRs they found was named hGCRP41 (hypothetical protein similar to GPCR) accession number: AB083623. They report it probably has 8 transmembrane helices, and 418aa. This protein probably doesn't exist because they read straight through two introns in their analysis. The protein also terminates in the middle of a frame-shifted exon.
Ericsson et al. 2003: Identification of receptors for pig endogenous retrovirus
Next, Ericsson et al come along looking for receptors for the pig endogenous retrovirus and find this gene. They used cDNA to make a bunch of transduced cell lines that had difference human cDNAs expressed. They found two related proteins that could serve as receptors to the pig endogenous retrovirus. They named the first one HuPAR-1, which is 445aa long and has 11 transmembrane helices. This is the length of one of the proper splice variants of this gene.
Andriamampandry et al. 2007: Cloning and functional characterization of a gammahydroxybutyrate receptor identiﬁed in the human brain
Andriamampandry is looking for gammahydroxtbutyrate (GHB) receptors by using cDNA and seeing which cDNA-transduced cells become sensitive to GHB. They find two nice proteins that are sensitive to GHB, GHBh1 and C12K32. GHBh1 is identical to HuPAR-1, it has 445aa and 11 transmembrane helices. C12K32 is a frameshifted version of GHBh1 that is slightly longer, but has the same number of helices. They had previously found a 512aa ortholog in rats, in 2013.
Yao et al. 2010: Identification and Comparative Functional Characterization of a New Human Riboflavin Transporter hRFT3 Expressed in the Brain.
Finally, we have Yao et al, who searched for homologs to their Riboflavin transporters. They found one in a cDNA library, and it was 445aa long. But they noticed upon BLASTing it that people had been calling it a GPCR. They point out that this is weird since it has 11 transmembrane helices, and go on to show that it is a Riboflavin transporter, concluding: "The molecular function of GPR172A has yet to be determined. We designated it hRFT3 (GenBank accession no. AB522904) based on its functional characterization as shown below." This paper is in pretty poor form, because somehow they don't address the Andriamampandry paper which presented data supporting its function as a GPCR.
So this protein, in this specific spliced formation, has been shown to be a virus receptor, a GHB G protein-coupled receptor, and a riboflavin transporter. It is very weird that it is a GPCR with 11 transmembrane domains... they very rarely have more or less than 7. I'd wait for more studies to confirm both of these findings before I'd accept that this protein is an XXL GPCR moonlighting as a riboflavin transporter.
Its fairly common for a gene to have multiple names. I cant say for certain, but Jax and GeneCards also lists them as being synonymous so I would say its safe to say they are the same
Edit in reply to comment:
In higher level Eukaryotes the vast majority of genes have more than one isoform as a result of alternative splicing, and this can drastically change the protein sequence and function of that protein. So even if it comes from the gene, two different isoforms can have completely unrelated functions.
According to ensembl, this gene has 14 isoforms.
Of course there could also be some type of post translational modification that could also change the function, but Id say mostly likely if the same gene produces two different functioning proteins its because of alternative splicing.