UniProt: Q9HAB3 (S52A2_HUMAN). Gene: SLC52A2 (HGNC:30224). Synonyms: GPR172A, PAR1, RFT3, hRFT3.
445 aa, 11 predicted transmembrane helices, cryo-EM structure PDB 8XSM.
This is the single most important thing to get right for this gene. Two independent
naming schemes exist for the three human riboflavin transporters and they are
cross-wired:
The UniProt entry for Q9HAB3 confirms the identity: RecName "Solute carrier family 52,
riboflavin transporter, member 2"; AltName "Riboflavin transporter 3; Short=hRFT3"
[file:human/SLC52A2/SLC52A2-uniprot.txt "AltName: Full=Riboflavin transporter 3;"].
PMID:24253200 states it explicitly: "RFVT2 (formerly RFT3)" and "RFVT3 (formerly RFT2)"
PMID:24253200.
Consequence for annotation review: in PMID:21854757 (Subramanian 2011), the entity called
hRFT-3 is this gene (SLC52A2/RFVT2), whereas the "hRFT-2 apical" result in that paper is
about SLC52A3. So the GO:0016323 (basolateral plasma membrane) IDA attributed to Q9HAB3
from PMID:21854757 corresponds to that paper's hRFT-3, which was described as
"localized predominantly within intracellular vesicles, although expression was evident at the
BLM of some cells" PMID:21854757.
The curator (who read the full text and knew hRFT-3 = SLC52A2) correctly captured the partial
BLM signal. Note this is NOT the dominant/canonical localization; the mainstream picture is
plasma-membrane / cell-surface expression.
Biallelic SLC52A2 loss-of-function → riboflavin transporter deficiency: childhood-onset
axonal sensorimotor neuronopathy with sensorineural deafness, optic atrophy, bulbar palsy,
respiratory insufficiency; treatable with high-dose oral riboflavin.
[file:human/SLC52A2/SLC52A2-uniprot.txt "autosomal recessive progressive neurologic disorder characterized by"];
PMID:24253200 abstract: "SLC52A2 mutations cause reduced riboflavin uptake and reduced riboflavin
transporter protein expression" and high-dose oral riboflavin gives "significant and sustained
clinical and biochemical improvements"
PMID:24253200.
Two IntAct IPI annotations to GO:0005515:
- PMID:25416956 (Rolland 2014, human interactome map) — interactor CDC23 (Q9UJX2).
- PMID:32296183 (Luck 2020, HuRI binary interactome) — interactor FAM209A (Q5JX71).
Both interactors match UniProt's INTERACTION section
[file:human/SLC52A2/SLC52A2-uniprot.txt "Q9HAB3; Q9UJX2: CDC23;"],
[file:human/SLC52A2/SLC52A2-uniprot.txt "Q9HAB3; Q5JX71: FAM209A;"].
These are systematic Y2H/high-throughput hits with no functional interpretation; bare
"protein binding" is uninformative → MARK_AS_OVER_ANNOTATED (not removed; experimental IPI).
PMID:17197387 (Andriamampandry 2007) cloned "GHBh1"/GPR172A from human frontal cortex as a
putative gamma-hydroxybutyrate (GHB) receptor with GTP-gamma-S-sensitive signalling. This is the
basis of the GO:0062124 (4-hydroxybutyrate receptor activity) IDA and one of the plasma-membrane
IDAs. UniProt retains this only tentatively: "May also act as a receptor for 4-hydroxybutyrate
(Probable)" with ECO:0000305 [file:human/SLC52A2/SLC52A2-uniprot.txt "May also act as a receptor for 4-"].
The protein is now firmly established as a riboflavin transporter (SLC/TCDB 2.A.125 e-RFT family;
no GPCR fold). The GHB-receptor characterization is best treated as an over-annotation /
superseded secondary claim; per curation policy an experimental IDA is marked
MARK_AS_OVER_ANNOTATED rather than removed. The plasma-membrane IDA from the same paper is
retained (the localization observation is consistent with the transporter).
GO:0006771 (riboflavin metabolic process) TAS from Reactome R-HSA-196843 "Vitamin B2 (riboflavin)
metabolism". SLC52A2 is a transporter, not a metabolic enzyme; it participates in the pathway only
by importing the vitamin. The specific and correct BP is GO:0032218 riboflavin transport (already
annotated). GO:0006771 is an over-annotation of a transporter as a metabolic-process participant →
MARK_AS_OVER_ANNOTATED.