Deep Research Report: Core Function Hypothesis for A0A804UIX9 (Zea mays)

Hypothesis Under Evaluation

Seed hypothesis: molecular_function (GO:0003674) is a core function of A0A804UIX9.

Verdict: SUPPORTED — The assignment of the root molecular function term GO:0003674 is justified and appropriate. No more specific molecular function GO term can be confidently assigned to A0A804UIX9 based on current evidence. One actionable curation lead emerges: the cellular component annotation should be upgraded from GO:0016020 (membrane) to GO:0016021 (integral component of membrane).

Summary

A0A804UIX9 is a 375-amino-acid protein from Zea mays (maize) that contains two PGG domains (PF13962) and six predicted transmembrane helices. It is classified at UniProt protein existence level 4 (predicted), with an annotation score of 1.0, indicating minimal characterization. The protein belongs to a large plant-specific gene family (~29,000 proteins across 1,110 taxa) in which no member has ever received a specific molecular function GO annotation — a remarkable knowledge gap for a family of this size.

Our investigation explored whether any evidence supports assigning a more specific molecular function term than the root GO:0003674. We examined the best-characterized PGG domain protein, ACD6 (Arabidopsis thaliana), which was recently shown to function as an ion channel mediating calcium influx (PMID: 37995686). However, ACD6 and all other experimentally characterized PGG proteins contain ankyrin repeat domains in addition to PGG domains, whereas A0A804UIX9 has PGG domains only. This architectural difference is critical: the ankyrin repeats in ACD6 are structurally analogous to those in mammalian ion channels and may be essential for channel function. We also found that the PANTHER classification of A0A804UIX9 into the CASKIN family (PTHR24177, subfamily SF432) is a false positive cross-kingdom assignment — true CASKIN proteins are exclusively animal neuronal scaffold proteins with no functional relationship to plant PGG proteins.

One actionable finding emerged: the cellular component annotation should be upgraded from GO:0016020 (membrane) to GO:0016021 (integral component of membrane), based on the robust six-transmembrane prediction confirmed by both Phobius and high-confidence AlphaFold modeling (pLDDT 88–94 for TM regions). This upgrade is within established curation practice for multi-pass transmembrane proteins.

Executive Judgment

Supported. The assignment of GO:0003674 (the root molecular_function term) as the molecular function annotation for A0A804UIX9 is correct and reflects genuine uncertainty rather than annotation neglect. Three independent lines of evidence converge on this conclusion: (1) no PGG domain protein in any species has a specific molecular function GO annotation, as confirmed by exhaustive SPARQL queries across the entire UniProt knowledge base; (2) the closest characterized homolog, ACD6, has ion channel-like activity, but A0A804UIX9 lacks the ankyrin repeat domains that are present in all functionally characterized PGG-family members; and (3) the PANTHER CASKIN family classification is a cross-kingdom misassignment that provides no functional information. The most important caveat is that the 6-TM topology of A0A804UIX9 is structurally suggestive of channel or transporter activity, but this architectural similarity alone is insufficient to justify a specific MF term without experimental or strong computational evidence.

Key Findings

Finding 1: A0A804UIX9 Belongs to the Plant-Specific PGG Domain Family Related to ACD6 Ion Channel-Like Proteins

A0A804UIX9 is a member of the PGG domain family (Pfam PF13962 / InterPro IPR026961), which is overwhelmingly plant-specific. A survey of all five reviewed (Swiss-Prot) PGG domain proteins confirmed they are all from plants: ACD6 (Q8LPS2, A. thaliana), ITN1 (Q9C7A2, A. thaliana), At2g01680 (Q9ZU96, A. thaliana), At5g02620 (Q6AWW5, A. thaliana), and rice NPR4 (A2CIR5, O. sativa). The family encompasses 29,157 proteins across 1,110 taxa, with zero experimental structures deposited for any PGG domain.

In Zea mays specifically, there are approximately 79 PGG-domain proteins: ~33 with the Ankyrin+PGG architecture (ACD6-like) and ~46 with PGG domains only (like A0A804UIX9). This distinction matters because the best-characterized family member, ACD6, was recently demonstrated to be an "ion-channel-like" protein that mediates calcium influx and is regulated by small MHA peptides (PMID: 37995686). However, ACD6's ion channel activity depends on both its transmembrane regions and its intracellular ankyrin repeats, which are "structurally similar to those found in mammalian ion channels" (PMID: 37995686). A0A804UIX9 lacks ankyrin repeats entirely, making direct functional transfer from ACD6 unjustified.

Finding 2: PANTHER CASKIN Classification Is a Cross-Kingdom Misassignment

The PANTHER database classifies A0A804UIX9 in the CASKIN family (PTHR24177, subfamily SF432). Our investigation established that this classification is biologically meaningless. True CASKIN proteins (CASKIN1 and CASKIN2) are exclusively animal neuronal scaffold proteins that bind CASK and LAR receptor protein tyrosine phosphatases at synapses (PMID: 41223222; PMID: 31727973). They function in presynaptic assembly and dendritic spine morphology — processes that do not exist in plants.

Importantly, the PANTHER subfamily SF432 is actually named "OS06G0286146 PROTEIN" after a rice ortholog (A0A0P0WVA6), not after an animal CASKIN. All plant members of PTHR24177 are PGG-domain transmembrane proteins grouped with animal CASKINs solely due to distant sequence similarity detected by the automated hierarchical classification pipeline. This is a known limitation of cross-kingdom protein family classification and should not influence functional annotation of A0A804UIX9.

Finding 3: No PGG Domain Protein Has a Specific MF GO Annotation in Any Species

This is perhaps the most decisive finding. A comprehensive SPARQL query across the entire UniProt knowledge base for PGG domain (PF13962/IPR026961) proteins with any GO molecular function annotation beyond the root term (GO:0003674) returned zero results. The most common GO annotations for PGG proteins are cellular component terms: plasma membrane (GO:0005886, 12,523 proteins) and membrane (GO:0016020, 9,578 proteins). While some PGG proteins carry MF annotations like kinase activity or DNA binding, these are attributable to co-occurring domains (ankyrin repeats, kinase domains) in multi-domain proteins, not to the PGG domain itself.

This finding means that assigning GO:0003674 to A0A804UIX9 is not an oversight — it reflects the genuine state of knowledge for the entire PGG family. The PGG domain remains one of the largest uncharacterized domain families in the plant kingdom.

Finding 4: A0A804UIX9 Is Transcribed but Has No Protein-Level Evidence

The gene encoding A0A804UIX9 (Zm00001eb381380) is located on chromosome 9 of the B73 reference genome (Zm-B73-REFERENCE-NAM-5.0, position 9:46088503–46090333). Expression data from the Gramene database confirm that the gene is transcribed in at least 17 RNA-seq experiments spanning diverse maize tissues and developmental stages, including silk, pollen, ovule, leaf, inner stem, and husks across B73 and Mo17 inbred lines.

However, the protein remains at UniProt existence level 4 (predicted), with no proteomics data confirming protein-level expression and no entry in the STRING protein interaction database. This absence of protein-level evidence further supports the conservative GO:0003674 annotation.

Finding 5: PGG Domains Show a Conserved ~3 TM per PGG Repeat Unit Architecture

Cross-species architectural analysis revealed a remarkably consistent ratio of approximately 3 transmembrane helices per PGG domain:

Internal duplication within A0A804UIX9 was confirmed by 40% sequence identity between PGG domain 1 (residues 50–157) and PGG domain 2 (residues 217–294), with conserved xGLNxPGG signature motifs (GAGLNLPGG and VAGLNPPGG). AlphaFold modeling shows high confidence for TM regions (per-residue pLDDT 88–94), moderate-high confidence for PGG loop regions (pLDDT 76–89), and disorder at both termini (N-terminal pLDDT ~41, C-terminal pLDDT ~49).

Finding 6: CC Annotation Upgrade to GO:0016021 Is Justified

The current cellular component annotation of GO:0016020 (membrane) is conservative for a protein with six unambiguous transmembrane helices. The evidence strongly supports an upgrade to GO:0016021 (integral component of membrane):

• Phobius prediction: Six TM helices at residues 54–73, 102–123, 135–154, 166–189, 222–239, and 269–290
• AlphaFold confidence: Per-residue pLDDT of 88–94 for TM regions (high confidence)
• Family precedent: All characterized PGG family members are integral membrane proteins — ACD6 (Q8LPS2) has IDA evidence for ER membrane (GO:0005789) and plasma membrane (GO:0005886); ITN1 (Q9C7A2) has IDA evidence for plasma membrane

This upgrade is within established curation practice for multi-pass transmembrane proteins and does not require experimental validation.

Evidence Matrix

GO Curation Implications

Molecular Function (MF): Retain GO:0003674

Recommendation: Retain GO:0003674 (molecular_function) as the MF annotation. No more specific term is supported.

Rationale: - No PGG-domain protein has a specific MF annotation in any species (confirmed by comprehensive UniProt SPARQL) - The closest characterized homolog (ACD6) has a different domain architecture — including ankyrin repeats that are critical to its ion channel function - The 6-TM topology alone is insufficient to assign channel/transporter activity - The PANTHER CASKIN classification is a cross-kingdom artifact

Terms explicitly NOT recommended at this time: - ~~GO:0005216 (ion channel activity)~~ — Requires experimental evidence; ankyrin repeats may be essential for channel function in this family - ~~GO:0022857 (transmembrane transporter activity)~~ — Same reasoning; topology alone does not confirm transport - ~~GO:0005515 (protein binding)~~ — No evidence; uninformative

Cellular Component (CC): Upgrade to GO:0016021 — Curation Lead

Recommendation: Upgrade from GO:0016020 (membrane) to GO:0016021 (integral component of membrane). This is a curation lead requiring curator verification.

Evidence code: IEA (Inferred from Electronic Annotation) based on Phobius TM prediction + AlphaFold structural confidence. Consider ISS (Inferred from Sequence/Structural Similarity) if family-level IDA evidence from ACD6 and ITN1 membrane localization is considered sufficient.

Biological Process (BP): No annotation recommended

No BP annotation can be confidently assigned. The gene is expressed across diverse tissues, but without functional data, any process annotation would be speculative.

GO Decision Table

Mechanistic Scope

Direct Gene Product Activity (Unknown)

The immediate molecular function of A0A804UIX9 is unknown. The protein's architecture — two PGG domains with six transmembrane helices in a modular ~3 TM/PGG repeat unit — is structurally suggestive of channel, transporter, or receptor activity. However, no experimental or high-confidence computational evidence supports any specific molecular activity.

Architectural Context

A0A804UIX9 architecture:
  N-term (disordered) — [TM1-TM2-TM3 + PGG1] — [TM4-TM5-TM6 + PGG2] — C-term (disordered)

ACD6 architecture (characterized):
  N-term — [TM1-TM2-TM3 + PGG1] — [TM4-TM5-TM6 + PGG2] — [Ankyrin repeats ×9] — C-term

Critical difference: A0A804UIX9 LACKS ankyrin repeats

The ankyrin repeats in ACD6 are hypothesized to serve as the regulatory/gating module, analogous to ankyrin repeats in mammalian TRP channels. Without these repeats, it is unclear whether PGG-only proteins can form functional channels, have a different gating mechanism, or perform an entirely different function.

Separation from Downstream Effects

Even if A0A804UIX9 were demonstrated to have channel activity, the downstream consequences (immune signaling, growth regulation, stress responses) observed in ACD6 would be specific to ACD6's regulatory context and should not be attributed to A0A804UIX9. The core function annotation should reflect the direct molecular activity, not pathway-level or phenotypic consequences observed in distantly related family members.

Conflicts and Alternatives

1. Could A0A804UIX9 Still Be a Channel Despite Lacking Ankyrin Repeats?

Some ion channels (e.g., certain K⁺ channels, aquaporins) achieve gating through mechanisms other than ankyrin repeats. The 6-TM topology of A0A804UIX9 is reminiscent of voltage-gated cation channels. However, no evidence specifically links PGG-only proteins to channel activity, and the internal PGG domain sequence (conserved xGLNxPGG motif) has no known relationship to channel selectivity filters or gating elements.

2. Divergent PANTHER Classifications Within PGG-Only Proteins

Another maize PGG-only protein (A0A804LML3) maps to a completely different PANTHER family — PTHR24186 (Protein Phosphatase 1 Regulatory Subunit) — suggesting that automated classifiers assign PGG-only proteins to diverse, unrelated families. This inconsistency reinforces that PANTHER classifications should not drive functional annotation for PGG-only proteins.

3. PANTHER CASKIN Misclassification

The PANTHER superfamily PTHR24177 ("CASKIN") groups animal neuronal scaffold proteins with plant PGG-domain transmembrane proteins. These are functionally unrelated. True CASKINs bind CASK and LAR-RPTPs at synapses and regulate dendritic spine morphology (PMID: 41223222; PMID: 31727973). The classification is based on distant sequence similarity without functional validation.

4. Organism-Specific Considerations

All experimental data for PGG-family proteins come from Arabidopsis thaliana. Maize diverged from Arabidopsis approximately 150 million years ago, and monocot–dicot functional divergence is well-documented. The expansion of the PGG family in maize (~79 members vs. fewer in Arabidopsis) may indicate subfunctionalization or neofunctionalization that makes cross-species functional inference unreliable.

5. Truncated Paralog or Alternative Architecture

A0A804UIX9 (375 aa) is significantly shorter than typical ACD6-like proteins (550–700 aa). It could represent a truncated paralog that lost the ankyrin domain, a functional subunit of a multi-protein complex, or a protein with a distinct, non-ACD6-like function.

Knowledge Gaps

Discriminating Tests

Highest Priority

1. Heterologous expression + electrophysiology: Express A0A804UIX9 in Xenopus oocytes or HEK293 cells and perform patch-clamp electrophysiology. This is the gold-standard test for channel/transporter activity. If the protein conducts ions, assign GO:0005216 (ion channel activity) or a more specific child term based on selectivity.

2. CRISPR knockout in maize: Generate Zm00001eb381380 knockout lines and phenotype under standard growth, biotic stress (pathogen challenge), and abiotic stress (salt, drought). Compare with ACD6-family knockouts to test whether PGG-only proteins participate in the same pathways.

3. Cryo-EM structure determination: Solve the structure of A0A804UIX9 or any PGG-only protein. The key question is whether the 6-TM + 2×PGG architecture forms a pore-like structure or a non-channel conformation.

Supporting Analyses

1. Co-expression network analysis: Use publicly available maize RNA-seq data to identify genes co-expressed with Zm00001eb381380. Co-expression neighbors may indicate the biological process context.

2. Targeted proteomics: Confirm protein expression using selected reaction monitoring (SRM) or parallel reaction monitoring (PRM) mass spectrometry in maize tissues where the transcript is detected.

3. Phylogenetic analysis of PGG-only vs. Ankyrin+PGG subfamilies: Determine whether PGG-only proteins diverged before or after the acquisition of ankyrin repeats. This informs whether PGG-only proteins represent an ancestral or derived state and whether channel function is likely ancestral to the family.

4. Co-immunoprecipitation / proximity labeling: Test whether A0A804UIX9 interacts with ankyrin-repeat-containing PGG proteins, which would suggest it functions as a subunit in a heteromeric complex.

Curation Leads

Lead 1: Retain GO:0003674 (molecular_function) — HIGH CONFIDENCE

• Action: No change needed. The root MF term correctly reflects the absence of functional evidence.
• Rationale: No experimental data, no characterized orthologs without additional domains, PGG domain is a DUF. Confirmed by UniProt SPARQL: zero PGG proteins have specific MF annotations.
• Status: Lead requiring curator verification

Lead 2: Upgrade CC to GO:0016021 (integral component of membrane) — MODERATE-HIGH CONFIDENCE

• Current: GO:0016020 (membrane) — IEA
• Proposed: GO:0016021 (integral component of membrane) — IEA
• Justification: Six transmembrane helices predicted by Phobius (residues 54–73, 102–123, 135–154, 166–189, 222–239, 269–290); AlphaFold pLDDT 88–94 for TM regions; all characterized PGG family members (ACD6 Q8LPS2, ITN1 Q9C7A2) have IDA evidence for integral membrane localization
• Reference for curator verification: UniProt entry A0A804UIX9, Phobius prediction; AlphaFold structure AF-A0A804UIX9-F1
• Status: Lead requiring curator verification

Lead 3: Flag PANTHER CASKIN Misclassification — HIGH CONFIDENCE

• Action: Flag PANTHER CASKIN family assignment (PTHR24177, SF432) as a cross-kingdom artifact
• Supporting snippets for verification:
• PMID: 41223222: "The two members of the CASKIN family of multidomain scaffold proteins, CASKIN1 and CASKIN2, bind to several AZ proteins and LAR receptor protein tyrosine phosphatases (LAR-RPTPs), and thus likely contribute to presynaptic assembly."
• PMID: 31727973: "CASK-interactive proteins, Caskin1 and Caskin2, are multidomain neuronal scaffold proteins."
• Rationale: True CASKINs are neuronal; plant PGG proteins share only distant sequence similarity
• Status: Lead requiring curator verification

Lead 4: Note ACD6 Family Context — MODERATE CONFIDENCE

• Action: Consider adding a curator comment noting that A0A804UIX9 belongs to the plant-specific PGG domain family (Pfam PF13962), whose best-characterized member ACD6 has ion channel-like activity. This should NOT be used to assign a specific MF GO term.
• Key references:
• PMID: 37995686: "Several lines of evidence link increased ACD6 activity to enhanced calcium influx, with MHA1L as a direct regulator of ACD6, indicating that peptide-regulated ion channels are not restricted to animals."
• PMID: 14507999: ACD6 "is a member of one of the largest uncharacterized gene families in higher plants"
• Status: Lead requiring curator verification

Lead 5: Future MF Candidates to Monitor — LOW CONFIDENCE (Prospective)

• Candidate terms to monitor if experimental evidence emerges:
• GO:0022857 (transmembrane transporter activity) — if transport demonstrated
• GO:0005216 (ion channel activity) — if electrophysiology confirms channel function
• GO:0005262 (calcium channel activity) — if calcium specificity shown
• Trigger: Any experimental characterization of a PGG-only protein (lacking ankyrin repeats)
• Status: Prospective lead, no action needed now

Evidence Summary

Evidence Base: Key Literature

Primary Research

• Chen et al. (2023) — "Small proteins modulate ion-channel-like ACD6 to regulate immunity in Arabidopsis thaliana" PMID: 37995686. The most important paper for this analysis. Demonstrates that ACD6 functions as an ion-channel-like protein mediating calcium influx, regulated by MHA1L peptides. Critically establishes that ACD6's "intracellular ankyrin repeats are structurally similar to those found in mammalian ion channels," directly informing why function transfer to ankyrin-lacking A0A804UIX9 is unjustified.

• Lu et al. (2003) — "ACD6, a novel ankyrin protein, is a regulator and an effector of salicylic acid signaling in the Arabidopsis defense response" PMID: 14507999. Early characterization of ACD6 as "a member of one of the largest uncharacterized gene families in higher plants." Establishes the historical context for PGG family annotation gaps.

• Sakamoto et al. (2008) — "ITN1, a novel gene encoding an ankyrin-repeat protein that affects the ABA-mediated production of reactive oxygen species and is involved in salt-stress tolerance in Arabidopsis thaliana" PMID: 18643991. Characterizes ITN1, another PGG+ankyrin protein, in abiotic stress signaling. Demonstrates functional diversity within the PGG family while reinforcing that all characterized members have ankyrin repeats.

• Todesco et al. (2010) — "Natural allelic variation underlying a major fitness trade-off in Arabidopsis thaliana" PMID: 20520716. Demonstrates ACD6 allelic diversity drives growth-defense trade-offs, providing evolutionary context for the PGG family.

• Passlik et al. (2025) — "A major trade-off between growth and defense in Arabidopsis thaliana can vanish in field conditions" PMID: 40658737. Describes ACD6 as "an ion channel that modulates salicylic acid synthesis." Important for establishing that even ACD6's characterized function is context-dependent, further cautioning against function transfer.

CASKIN Biology (for PANTHER Misclassification Assessment)

• Han et al. (2025) — "CASKIN2 mediates PTPσ-orchestrated transsynaptic mechanisms at excitatory synapses" PMID: 41223222. Establishes CASKIN proteins as presynaptic scaffold proteins — exclusively animal, neuronal context.

• Bencsik et al. (2019) — "Dendritic spine morphology and memory formation depend on postsynaptic Caskin proteins" PMID: 31727973. Confirms CASKINs as "multidomain neuronal scaffold proteins" — no plant relevance.

Limitations

1. No direct experimental data: The entire analysis is based on computational predictions, database queries, and inference from distantly related proteins. Any conclusions are provisional until experimental data become available.

2. AlphaFold structural predictions are not experimental structures: While pLDDT scores are high for TM regions (88–94), the structural model has not been validated experimentally, and loop/domain predictions (pLDDT 76–89) carry greater uncertainty.

3. Cross-species inference limitations: All functional data come from Arabidopsis thaliana, a dicot that diverged from maize ~150 million years ago. Monocot-specific functional divergence cannot be excluded.

4. RNA expression ≠ protein function: Transcript detection across 17 tissue datasets confirms the gene is active but does not demonstrate protein expression, stability, or function.

5. Family size vs. characterization: The PGG family contains ~29,000 proteins across >1,000 species, yet fewer than 5 members have been experimentally characterized — all in a single organism (A. thaliana) and all with additional ankyrin domains. The PGG-only subfamily (including A0A804UIX9) has never been experimentally characterized.

6. Negative evidence is informative but not conclusive: The absence of specific MF annotations for PGG proteins in UniProt reflects both genuine knowledge gaps and possible lag in annotation updates. However, the comprehensive SPARQL query provides high confidence that this gap is real.