Focused Curation Report — GLT-1 (NEUCR, Q1K4S3): "glucose transport and proton coupling"

Gene: glt-1 / NCU01633 / Q1K4S3 (Neurospora crassa)
Focus type: function_assignment
Hypothesis slug: glucose-transport-and-proton-coupling
Disputed function: carbohydrate:proton symporter activity (GO:0005351) + proton transmembrane transport, in addition to established low-affinity glucose uptake.

Executive Judgment

Verdict: Weakly supported / over-annotated as a specific mechanistic claim.

The seed hypothesis — that GLT-1 retains carbohydrate:proton symport in addition to low-affinity glucose uptake — rests entirely on ancestry (PANTHER IBD/IBA) inference, not on any protein-specific measurement. Three independent lines of evidence argue that the proton-symporter molecular function is not the gene product's actual activity:

1. Annotation provenance. Every GO term on Q1K4S3 is evidence code IBA:GO_Central (PANTHER PAINT), including GO:0005351. There is no experimental (EXP/IDA/IMP) support for proton coupling.
2. Family logic. GLT-1 belongs to the MFS Sugar Porter family (PTHR48022; Pfam PF00083), which contains both proton symporters (XylE, plant STP) and facilitated-diffusion uniporters (mammalian GLUT/SLC2, yeast HXT). Mechanism is not conserved across this family, so a family-level "sugar:proton symporter activity" cannot be assumed for a given member (PMID 24847886).
3. Direct sequence discriminator, with an intra-organism control. GLT-1 lacks the acidic TM1 protonation-site residue that is obligatory for sugar:H⁺ symport. At the position that is Asp27 in the symporter XylE and Asp40 in the symporter STP1 (and Asn29 in the uniporter GLUT1), GLT-1 has a non-protonatable Val48 — even less symport-compatible than GLUT. Critically, the same scan run on the Neurospora high-affinity paralogs shows they RETAIN the acidic residue — HGT-1 (Q7S0I5) Asp52, HGT-2 (Q7S3Q4) Asp35 — while only the low-affinity GLT-1 has lost it. Within one genome, proton-coupling capacity thus segregates to the high-affinity (system-II-like) transporters, exactly matching the classical physiology, and pinpoints GLT-1 as the member that specifically lost the coupling determinant. A valine cannot act as proton donor/acceptor (diagnostic per PMID 28617850).

The seed's fair caveat holds: the identification paper (PMID 28115989) infers system-I assignment from affinity/expression and did not run a GLT-1-specific proton-flux/electrophysiology assay, so it does not experimentally refute symport at the protein level. Hence "refuted" is too strong; but the convergent structural/evolutionary evidence makes proton-coupled symport the less likely mechanism. Curation recommendation: do not assert GO:0005351 as a core/experimental function; treat it as IBA-only and preferably generalize the MF to glucose/carbohydrate transmembrane transporter activity (uniporter). This is a lead requiring curator verification.

Evidence Matrix

GO Curation Implications

• GO:0005351 (MF, carbohydrate:proton symporter activity) — recommend: do NOT retain as an asserted/core function. It is IBA-only, family-level, and is contradicted by the missing TM1 protonation-site residue. Options for the curator, in order of preference:
• Generalize the MF to an experimentally-safe, mechanism-agnostic term supported by the primary data — e.g., D-glucose transmembrane transporter activity (GO:0055056) / glucose:proton symporter activity? NO → prefer the non-coupled glucose transmembrane transporter activity (GO:0005355). These capture the demonstrated low-affinity glucose uptake without the unsupported proton-coupling claim.
• If house style requires keeping the PANTHER IBA row, retain GO:0005351 strictly as IBA (do not promote to experimental, do not treat as core), and add the more precise experimentally-based glucose transporter MF above it.
• Proton transmembrane transport (BP, GO:1902600 / GO:0015992) — recommend: remove / do not add. No evidence; depends on the symport assumption that the sequence contradicts.
• BP: carbohydrate/glucose transmembrane transport (GO:0008643 / GO:1904659) — retain (well supported experimentally).
• CC: membrane / plasma membrane (GO:0016020 / GO:0005886) — retain (MFS integral membrane; plasma-membrane glucose uptake demonstrated).

Note on the seed's PANTHER argument: it is factually correct that PTN002362095 descends from the PTN000627328 IBD for GO:0005351 with no lineage-loss row, so the IBA annotation is procedurally valid. But procedural ancestry validity ≠ demonstrated mechanism; the curator should weight the protein-specific sequence and physiology evidence over an un-lost family default.

Mechanistic Scope

• Immediate molecular function (what is actually tested): transmembrane movement of D-glucose across the plasma membrane, low affinity (system I). Established by heterologous radioglucose kinetics (PMID 28115989/24581151).
• The disputed add-on (coupling mechanism): whether that glucose movement is coupled to inward H⁺ flux (symport, concentrative, energized by the PM H⁺-ATPase gradient) versus facilitated diffusion (uniport, down-gradient). This is the crux — glucose specificity is not in question; coupling is.
• Downstream / not the tested function: GLT-1's roles in carbon catabolite signaling, cellulase regulation, and glucose-gradient sensing (PMID 33509260) are pleiotropic pathway consequences, not the transport mechanism, and should not be used to justify a symporter MF.

Conflicts and Alternatives

• Uniporter vs symporter (primary conflict). Sequence (Val48 at the protonation site) and Sugar-Porter-family logic favor a facilitated-diffusion uniporter, aligning GLT-1 with low-affinity yeast HXT and mammalian GLUT rather than with H⁺-symporters. This competes directly with the IBA symporter default.
• Whole-cell energy dependence artifact. Schneider & Wiley (1971) found system I is also inhibited by azide/DNP. This could be misread as proton coupling, but uncouplers collapse membrane potential and cellular ATP and perturb all PM transport indirectly; it is not evidence of stoichiometric H⁺:glucose symport by GLT-1. Only system II concentrates 3-O-methylglucose.
• System-to-gene mapping caveat. The historical "system I/II" physiology was measured on whole mycelia (mixtures of transporters). Equating GLT-1 exactly with all of "system I" is itself an inference; the modern gene-level data (PMID 28115989) support low-affinity glucose uptake but did not isolate coupling.
• Paralog / subfamily confusion (now with positive evidence). PANTHER subfamily label "low affinity glucose transporter MstE" (Aspergillus) reinforces low-affinity glucose uniport, not symport. The high-affinity, proton-coupling-capable activity in N. crassa segregates to the HGT-1/HGT-2 paralogs, which retain the TM1 acidic residue (Asp52 / Asp35) that GLT-1 has lost (Val48). This is a lineage-controlled contrast within one genome: the family-default symporter GO is appropriate (at the sequence level) for the high-affinity paralogs but not for GLT-1. Care is needed not to carry the paralogs' coupling determinant onto GLT-1 via family-level IBA.

Knowledge Gaps

1. No direct GLT-1 energetics assay. Checked: no EXP GO, no electrophysiology in the retrieved literature. Matters because it is the only way to prove uniport vs symport. Resolve with: proteoliposome glucose flux ± ΔpH/Δψ, or two-electrode voltage clamp in Xenopus oocytes measuring glucose-evoked H⁺ current.
2. Single-residue diagnostic is asymmetric. Checked: an expanded 11-protein panel. The signal is directional: a non-protonatable residue (GLT-1 Val48) is strong evidence against symport (a proton donor/acceptor is physically required), whereas an acidic residue is necessary but not sufficient — human GLUT2 (P11168) is a bona fide uniporter yet carries Asp27 at this position (a false-positive for "symport"). GLT-1 sits on the robust side of this asymmetry (it lacks even the neutral amide of GLUT1/3/5). PMID 28617850 further cautions that multiple residues tune thermodynamics. Resolve with: full structure-based alignment / AlphaFold model + inspection of the complete proton-relay network (TM1/TM4/TM7/TM10) against XylE and GLUT1.
3. Membrane-potential dependence of GLT-1 specifically. Historical energy dependence is whole-cell. Resolve with: GLT-1-only heterologous system + protonophore controls.

Discriminating Tests

• Proteoliposome reconstitution of purified GLT-1: measure glucose accumulation against a gradient with vs without an imposed ΔpH (inside-alkaline). Concentrative uptake requiring ΔpH ⇒ symport; equilibrative only ⇒ uniport. Most decisive.
• Oocyte electrophysiology: glucose-induced inward current that is H⁺-dependent ⇒ symport; electroneutral/no current ⇒ uniport.
• Structure-guided mutagenesis: if a symport-like acidic residue can be introduced at the Val48 position and confers H⁺ coupling, that would delimit the native mechanism.
• Comparative bioinformatics (fast): extend the TM1 protonation-site scan across all N. crassa Sugar Porter members and characterized fungal symporters/uniporters to confirm GLT-1 clusters with uniporters (partially done here for GLT-1 vs XylE/STP1/GLUT).

Curation Leads (require curator verification)

• Action change: Downgrade/《do not assert》 GO:0005351 on GLT-1 to IBA-only or replace with mechanism-agnostic glucose transmembrane transporter activity (GO:0005355) / D-glucose transmembrane transporter activity (GO:0055056); do not add proton transmembrane transport BP.
• Retain: carbohydrate/glucose transmembrane transport BP (GO:0008643 / GO:1904659) and membrane CC (GO:0016020), with experimental support from PMID 28115989 / 24581151.
• Candidate references + snippets to verify:
• PMID 28617850 — "the variation between an acidic residue (Asp27 in XylE) and a neutral one (Asn29 in GLUT1)…key element for their mechanistic divergence."
• PMID 24847886 — "Structural comparison of the uniporter GLUT1 with its bacterial homologue XylE, a proton-coupled xylose symporter…"
• PMID 5573732 — "The apparent K(m) for glucose uptake by system I and system II are 25 and 0.04 mm, respectively."
• PMID 6444943 — "a second galactose transport activity…required energy, had a high affinity…was shown to be the same as glucose transport system II."
• Suggested curator question: Is there any GLT-1-specific proton-flux/voltage-clamp datum? If not, the symporter MF should not be a core assertion.
• Suggested experiment: proteoliposome ΔpH-dependence assay (see Discriminating Tests).

Provenance artifacts:
- provenance/glt1_TM1_protonation_site_analysis.py (executable) + ...analysis.csv — core symporter/uniporter TM1 scan incl. Neurospora HGT-1/-2.
- provenance/glt1_TM1_expanded_panel.csv — 11-protein control panel (documents the GLUT2 acidic-but-uniporter caveat and the diagnostic's asymmetry).
- provenance/glt1_GO_decision_table.csv — per-term recommended curation actions.

GO decision table (summary; leads requiring curator verification):

Artifacts

• OpenScientist final report
• OpenScientist final report
• OpenScientist glt1 GO decision table
• OpenScientist glt1 TM1 expanded panel
• OpenScientist glt1 TM1 protonation site analysis