AIGR Gene Hypothesis Deep Research — Candida albicans PDE1 (Q5AGE4)

Hypothesis slug: pde1-glucose-gpcr-pathway
Focus type: function_assignment
Term under evaluation: GO:0110034 — negative regulation of adenylate cyclase-activating glucose-activated G protein-coupled receptor signaling pathway
Target: PDE1 / Q5AGE4, Candida albicans SC5314 (NCBITaxon:237561)

Summary

Verdict: Partially supported but over-annotated — the specific GO:0110034 annotation should be removed (or not propagated) for C. albicans Pde1.

The seed hypothesis bundles two separable claims whose evidence diverges sharply. The first — that Pde1 negatively regulates the glucose-triggered cAMP rise — is well supported. Pde1 is a bona fide low-affinity (class II) cyclic-AMP phosphodiesterase, and it is specifically the enzyme that terminates the cAMP spike produced after glucose addition or intracellular acidification (PMID: 20558315; ortholog PMID: 9880329). There is no dispute about Pde1's role as a negative regulator of cAMP signalling downstream of glucose.

The second claim — that the cAMP rise is generated through a glucose-activated G protein-coupled receptor (Gpr1) and the Gα protein Gpa2 — is refuted in C. albicans. The study that directly asked where glucose-induced cAMP comes from found that deleting GPR1 or GPA2 does not abolish glucose-induced cAMP, whereas deleting CDC25 or RAS1 does (PMID: 15673611). The C. albicans Gpr1 ligand is moreover unresolved and may be amino acids such as methionine rather than glucose (PMID: 15667329).

GO:0110034 is a highly specific Biological Process term that asserts three mechanistic commitments at once — a GPCR-mediated, glucose-activated, adenylate-cyclase-activating pathway — exactly the architecture the C. albicans data contradict. Critically, the annotation sits on Q5AGE4 by phylogenetic inference only (IBA; GO_REF:0000033), with no C. albicans experimental support. It is a classic over-propagation from the Saccharomyces cerevisiae glucose-sensing paradigm into a species where the wiring differs. The correct, evidence-backed annotations to retain are GO:0004115 (MF, IDA), GO:0141162 (BP, IMP), and the broader GO:1902660 (BP).

Key Findings

Finding 1 — Pde1 terminates the glucose-induced cAMP rise (the negative-regulation activity is real)

Pde1 is a class II 3′,5′-cyclic-AMP phosphodiesterase whose physiological role is to down-regulate the cAMP transient that follows glucose addition. Wilson et al. (2010) provide direct biochemical and genetic evidence in C. albicans: "Pde1, but not Pde2, is responsible for down-regulation of cAMP signalling induced by glucose addition or intracellular acidification" (PMID: 20558315). This cleanly separates Pde1, the low-affinity enzyme that shapes the acute, agonist-induced spike, from Pde2, the high-affinity enzyme that sets basal cAMP tone.

The specificity is conserved from the S. cerevisiae ortholog, where "deletion of PDE1, but not PDE2, results in a much higher cAMP accumulation upon addition of glucose or upon intracellular acidification" (PMID: 9880329). UniProt Q5AGE4 records the molecular function directly: GO:0004115 (3′,5′-cyclic-AMP phosphodiesterase activity), IDA (PMID:8075796), with the class II PDEase domain (Pfam PF02112, PDEase_II; metallo-hydrolase fold).

The portion of the hypothesis asserting that Pde1 negatively regulates glucose-induced cAMP signalling is therefore correct. It maps onto the experimentally supported term GO:0141162 (negative regulation of cAMP/PKA signal transduction, IMP, PMID:20558315) and the molecular-function term GO:0004115. Importantly, this negative-regulation role is an action on the second messenger, not on the receptor.

Finding 2 — The glucose-induced cAMP rise is Cdc25/Ras1-dependent, NOT Gpr1/Gpa2-dependent (the GPCR specificity is refuted)

This is the decisive finding. Maidan et al. (2005) performed the exact discriminating experiment — which upstream module is required for glucose-induced cAMP in C. albicans: "deletion of neither CaGpr1 nor CaGpa2 affects glucose-induced cAMP signaling. In contrast, the latter is abolished in strains lacking CaCdc25 or CaRas1, suggesting that the CaCdc25-CaRas1 rather than the CaGpr1-CaGpa2 module mediates glucose-induced cAMP signaling in C. albicans" (PMID: 15673611).

This directly contradicts the mechanistic claim embedded in GO:0110034 — that the pathway being negatively regulated is a glucose-activated GPCR pathway. In C. albicans, the glucose→cAMP route runs through the Ras guanine-nucleotide-exchange factor Cdc25 and the small GTPase Ras1 to adenylyl cyclase (Cyr1/Cdc35), not through Gpr1→Gpa2→Cyr1.

A competing earlier claim exists. Miwa et al. (2004) reported that "GPR1 and GPA2 are required for a glucose-dependent increase in cellular cAMP" (PMID: 15302825). This is a genuine conflict in the primary literature. However, the Maidan study specifically dissected the glucose→cAMP requirement with clean epistasis against both branches, and Wilson et al. (2010) further reframed the Gpa2 contribution as linked to intracellular acidification and found Gpa2's other morphogenetic effects to be "cAMP pathway-independent." The weight of mechanistically targeted evidence favors a Cdc25/Ras1 source for glucose-induced cAMP.

The second, more specific half of the hypothesis — that Pde1 terminates a cAMP rise generated through Gpr1/Gpa2 — is therefore not supported and is actively contradicted in this organism.

Finding 3 — C. albicans Gpr1 is likely an amino-acid (methionine) sensor, not a confirmed glucose receptor

Even the "glucose-activated" qualifier of the receptor is questionable in C. albicans. Maidan, Thevelein & Van Dijck (2005) found that Gpr1 internalization is induced by specific amino acids such as methionine and concluded: "it remains unclear whether Gpr1 senses sugars, as in Saccharomyces cerevisiae, or specific amino acids like methionine" (PMID: 15667329). The carbon-source-induced yeast-to-hypha transition was found to depend on both the presence of amino acids and on Gpr1.

The ligand specificity that would justify calling the Gpr1 pathway "glucose-activated" is thus unresolved in C. albicans and may in fact be amino-acid-based. This further undermines the exact wording of GO:0110034, which bundles "glucose-activated" into the term definition.

Finding 4 — GO:0110034 on Q5AGE4 is IBA-only and contradicted by species-specific data (over-propagation)

A QuickGO annotation retrieval for UniProtKB:Q5AGE4 (performed this run) shows that GO:0110034 is assigned solely by IBA (Inferred from Biological ancestor; GO_REF:0000033, GO Central phylogenetic inference). There is no IDA/IMP (experimental) support for this specific term in C. albicans. Across organisms the term is dominated by inference (in the first 100 annotations: IBA = 18, IEA = 10, IMP = 4, IC = 1, IGI = 1), i.e., it is largely propagated across fungal/amoebal phosphodiesterase orthologs (including Dictyostelium pdsA, S. pombe cgs2, Ustilago).

By contrast, Q5AGE4's experimentally anchored terms are GO:0004115 (MF, IDA; PMID:8075796), GO:0141162 (BP, IMP; PMID:20558315), and filamentous-growth terms (BP, IMP; PMID:20558315). GO:0110034 is thus a phylogenetic carry-over from the S. cerevisiae glucose-GPCR paradigm, applied to a species where the glucose→cAMP route is experimentally shown to be Cdc25/Ras1-dependent — precisely the kind of over-propagation curators should correct.

Mechanistic Model / Interpretation

The hypothesis conflates what Pde1 does (correct) with how the upstream cAMP is generated (incorrect for this organism). The diagram below summarizes the C. albicans wiring as supported by primary data.

  ========= S. cerevisiae paradigm (the source of GO:0110034) =========
  glucose ──> Gpr1 (GPCR) ──> Gpa2 (Gα) ──┐
                                    ├──> Cyr1 (adenylyl cyclase) ──> cAMP ──> PKA
  intracellular acidification ─────────────┘                               │
                                                        Pde1 ◄──────┘ (terminates spike)

  ============== C. albicans reality (PMID:15673611; 20558315) ==============
  glucose ──> Cdc25 (RasGEF) ──> Ras1 ─────┐
                                    ├──> Cyr1/Cdc35 ──> cAMP ──> PKA ──> morphogenesis
  (Gpr1/Gpa2: amino-acid sensing;           │
   Gpa2 linked to acidification) ───────────┘
                                                        Pde1 ◄──── terminates glucose/
                                                                   acidification-induced cAMP

Component-by-component comparison:

Pde1's negative regulation operates at the level of cAMP hydrolysis, two steps downstream of any receptor. Even if a Gpr1/Gpa2 input contributed to cAMP under some conditions, Pde1 does not "negatively regulate the GPCR pathway" in a receptor- or Gα-specific way — it degrades the common output (cAMP) that integrates signals from Cdc25/Ras1, acidification, and potentially Gpr1/Gpa2. The specificity named in GO:0110034 is therefore both mechanistically mislocated (Pde1 acts on the messenger, not the pathway head) and factually contradicted (the glucose cAMP source is Ras-dependent in this species).

Evidence Base / Evidence Matrix

GO Curation Implications (leads — require curator verification)

The supported BP level is "negative regulation of cAMP-mediated / glucose-mediated signaling"; the MF is cAMP phosphodiesterase activity. The GPCR-specific BP term (GO:0110034) is not justified by primary C. albicans data. "Protein binding" is not invoked — a more informative set of terms is supported. The recommendation concerns a BP term only; the MF (GO:0004115) is unaffected, and no CC change is implied.

Mechanistic Scope

Pde1's direct molecular function is hydrolysis of the 3′,5′-cyclic phosphodiester bond of cAMP, i.e., it acts on the diffusible second messenger downstream of adenylyl cyclase (Cyr1). Any "negative regulation of a receptor signaling pathway" is therefore indirect: Pde1 lowers the shared cAMP pool irrespective of which upstream input (glucose→Cdc25/Ras1, intracellular acidification, or amino-acid→Gpr1/Gpa2) generated it. Pde1 does not bind or modify Gpr1, Gpa2, or Cyr1.

• Direct activity: cAMP hydrolysis (GO:0004115, IDA) — well supported.
• Direct regulatory consequence: lowering cAMP after glucose/acidification spikes → negative regulation of cAMP/PKA signalling (GO:0141162, IMP) — well supported.
• Downstream phenotypes (not core function): filamentous growth / virulence contributions are loss-of-function phenotypes mediated through altered cAMP/PKA tone, not receptor-level activity of Pde1.
• Mis-scoped assertion: ascribing GPCR-branch specificity to a second-messenger-degrading enzyme conflates the enzyme's activity with the particular stimulus whose output it happens to buffer — a scope error compounded by the factual refutation of the Gpr1/Gpa2 glucose source.

Conflicts and Alternatives

1. Internal reference conflict (Miwa 2004 vs. Maidan 2005). PMID:15302825 reports Gpr1/Gpa2 are required for glucose-dependent cAMP; PMID:15673611 reports they are not, assigning glucose-induced cAMP to Cdc25/Ras1 and reassigning Gpr1/Gpa2 to amino-acid sensing. Curation should weight the more directly targeted, later study; Wilson 2010 additionally attributes Gpa2's cAMP contribution to acidification.

2. Organism carry-over + ligand ambiguity. The glucose-sensing GPCR role of Gpr1/Gpa2 is well established in S. cerevisiae, but in C. albicans the Gpr1 ligand is unresolved — internalization is induced by amino acids such as methionine (PMID:15667329). The "glucose-activated GPCR" premise is a probable S. cerevisiae→C. albicans carry-over; the term may fit S. cerevisiae Pde1 but not C. albicans Pde1.

3. Alternative interpretation. Pde1's phenotypes (filamentous growth, virulence contribution) are downstream, pleiotropic consequences of global cAMP/PKA buffering, not evidence of GPCR-branch-specific action. Even granting a minor Gpr1/Gpa2 contribution to cAMP, Pde1 degrades the common pool; GO:0110034's receptor-pathway specificity remains misplaced for a phosphodiesterase.

Limitations and Knowledge Gaps

Provenance note: the annotation-evidence breakdown (IBA-only for GO:0110034; experimental support for GO:0004115/GO:0141162) is derived from a QuickGO/UniProt retrieval conducted during the investigation. This is a database-level check, not a wet-lab result, and is reported as such.

Discriminating Tests

1. Unified-condition epistasis panel. Measure glucose-induced cAMP kinetics in isogenic gpr1Δ, gpa2Δ, cdc25Δ, ras1Δ, and WT (plus pde1Δ in each background) under tightly controlled glucose concentration and extracellular pH. Directly adjudicates Miwa (2004) vs. Maidan (2005) and identifies which input Pde1 buffers.
2. Ligand specificity of Gpr1. Dose-response cAMP and receptor-internalization assays comparing glucose, sucrose, and methionine/other amino acids to establish the physiological C. albicans Gpr1 ligand.
3. Interaction screen (AP-MS / two-hybrid) for Pde1 against Gpr1/Gpa2/Cyr1 to test any direct branch coupling.
4. Comparative annotation audit. Confirm whether S. cerevisiae Pde1 (not C. albicans) is the proper home for a glucose-GPCR-linked term, and review other fungal IBA propagations of GO:0110034.

Curation Leads (require curator verification)

• Action: Remove / do-not-accept the existing IBA GO:0110034 annotation on C. albicans Pde1 (Q5AGE4). Keep the experimentally grounded GO:0004115 (MF, IDA) and GO:0141162 (BP, IMP), and the broader GO:1902660 (BP). If a stimulus-linked BP is desired, prefer GO:1902660 (glucose-mediated signaling) without the GPCR qualifier. The term is currently present only through phylogenetic (IBA/GO_REF:0000033) propagation from orthologs, not through C. albicans experimental evidence.
• Key references to cite in the review:
• PMID:15673611 — "deletion of neither CaGpr1 nor CaGpa2 affects glucose-induced cAMP signaling. In contrast, the latter is abolished in strains lacking CaCdc25 or CaRas1..." (refutes GPCR source).
• PMID:20558315 — "Pde1, but not Pde2, is responsible for down-regulation of cAMP signalling induced by glucose addition or intracellular acidification." (supports Pde1→glucose-cAMP termination).
• PMID:9880329 — "deletion of PDE1, but not PDE2, results in a much higher cAMP accumulation upon addition of glucose or upon intracellular acidification." (ortholog).
• PMID:15302825 — "GPR1 and GPA2 are required for a glucose-dependent increase in cellular cAMP." (competing claim; note it is superseded by 15673611).
• PMID:15667329 — "it remains unclear whether Gpr1 senses sugars... or specific amino acids like methionine." (ligand ambiguity).
• Suggested curator question: Does GO:0110034's definition require action at the receptor/G-protein step (which Pde1 does not perform) or merely regulation of the pathway's cAMP output?
• Suggested experiments: the unified-condition epistasis panel and Gpr1 ligand dose-response assays under Discriminating Tests.

Artifacts (provenance): /tmp/evidence_matrix_PDE1_GPCR.csv, /tmp/go_decision_table_PDE1.csv, /tmp/GO0110034_evidence_codes.png (computed from QuickGO: ~82% of GO:0110034 annotations are IBA/IEA inferred), /tmp/PDE1_verdict_summary.png, plus executed UniProt (Q5AGE4) and QuickGO fetches confirming domain/GO content and the IBA-only evidence basis of the GO:0110034 annotation.

Final Verdict

The enzymatic and negative-regulatory core of the hypothesis is correct — Pde1 terminates the glucose-induced cAMP spike — but the specific claim that it regulates a glucose-activated GPCR (Gpr1/Gpa2) pathway is refuted in C. albicans, where glucose-induced cAMP is Cdc25/Ras1-dependent. GO:0110034 is attached to Q5AGE4 only by phylogenetic inference and should be removed or down-weighted in favor of the experimentally grounded terms GO:0004115 (MF, IDA) and GO:0141162 (BP, IMP). Overall: partially supported, over-annotated.

Artifacts

• OpenScientist final report
• OpenScientist final report