KDX1/Mlp1 (P36005, YKL161C) — Adjudication of Cell-Cycle, CDK-Complex, and ATP-Binding Function Assignments OpenScientist openscientist-autonomous 3 citations 2 artifacts 2026-09-21T04:33:42.110250 citations file

OpenScientist final report (text/html)

OpenScientist final report (application/pdf)

KDX1/Mlp1 (P36005, YKL161C) — Adjudication of Cell-Cycle, CDK-Complex, and ATP-Binding Function Assignments

Focus type: function_assignment · Hypothesis slug: cell-cycle-complex-and-nucleotide-binding-functions
Organism: Saccharomyces cerevisiae (NCBITaxon:559292) · Gene: KDX1 (synonym MLP1) · ORF: YKL161C

Executive Judgment

Verdict: Over-annotated / Refuted for the cell-cycle & CDK-complex claims; Weakly-supported→Unresolved for ATP binding.

The four seed candidate functions are adjudicated independently:

Candidate function	Current GO (evidence)	Verdict
Regulates cell cycle / G1/S–G2/M transition	GO:0000086 G2/M transition (IBA)	Refuted as a direct function / over-annotated
Cyclin-dependent protein kinase holoenzyme complex	GO:0000307 (IBA)	Refuted / over-annotated
Cyclin-dependent protein Ser/Thr kinase activity	GO:0004693 (IBA)	Refuted / over-annotated
Binds ATP	GO:0005524 (IEA:InterPro)	Weakly supported → Unresolved (rule-based only; likely impaired)

Reasoning. KDX1/Mlp1 is a pseudokinase — a whole-genome-duplication paralog of the active cell-wall-integrity (CWI) MAPK Slt2/Mpk1. Its only experimentally established molecular function is noncatalytic transcriptional coactivation: together with (and redundantly to) Mpk1, it forms a complex with the Swi4/Swi6 (SBF) transcription factor to drive cell-wall-stress genes (FKS2 and others) independently of kinase activity (PMID:18268013, PMID:20641022, PMID:20219973). It is a MAPK-family (CMGC group) protein, not a cyclin-dependent kinase. The CDK-activity, CDK-holoenzyme-complex, and G2/M-transition annotations are IBA (PANTHER/PAINT) propagations from a broad CMGC-group ancestor that co-mingles CDK and MAPK subfamilies; they are not supported by any KDX1 primary experiment and conflict with the protein's demonstrated biology. ATP binding is an IEA:InterPro rule-based call from the generic protein-kinase profile; sequence analysis shows KDX1 has lost the DFG Mg²⁺-coordinating aspartate (→NFG) and has a degraded glycine-rich loop, so Mg-ATP binding is likely impaired and remains experimentally undemonstrated.

Most important caveats. (1) Absence of a direct ATP-binding assay means we cannot formally exclude residual nucleotide binding (the β3 catalytic Lys55 is retained). (2) SBF (Swi4/Swi6) does have a bona-fide G1/S cell-cycle role, so a tenuous, indirect cell-cycle connection exists via its partner — but this is a downstream/partner property, not a direct KDX1 function. (3) The seed's caution is correct: a NOT annotation from a different fungal PANTHER node (PTN008603245) should not be transferred to this target without established ancestry.

Evidence Matrix

Citation (PMID)	Evidence type	Supports/Refutes/Qualifies	Claim tested	Key finding	Context	Confidence & limitations
18268013 (Kim, Truman, Levin 2008)	Mutant/biochemical (ChIP, kinase-dead)	Refutes catalytic role; supports noncatalytic SBF function	Does Mlp1/Mpk1 act via kinase activity or noncatalytically?	Mpk1 and its pseudokinase paralog Mlp1 activate FKS2 via a noncatalytic mechanism; form dimeric complex with Swi4, recruit Swi6; FKS2 activation independent of protein kinase activity	S. cerevisiae, cell-wall stress, FKS2 promoter	High for mechanism. Directly labels Mlp1 a pseudokinase.
20641022 (Kim & Levin 2010)	Reporter/genetic	Supports; Qualifies redundancy	Is Mlp1 role redundant with Mpk1? Which genes?	"The Mlp1 pseudokinase serves a redundant function with Mpk1"; additional SBF-noncatalytic targets CHA1, YLR042c, YKR013w	S. cerevisiae CWI transcription	High. Confirms redundant, noncatalytic, SBF-dependent role.
20219973 (Kim et al 2010)	Biochemical/localization	Qualifies; contextualizes SBF	Mechanism of Swi6 recruitment	Mpk1(/Mlp1)–Swi4 complex recruits Swi6 to nucleus for CWI transcription; Mpk1 phosphorylates Swi6 Ser238	S. cerevisiae	High for Mpk1; Mlp1 shares the complex-formation step (noncatalytic).
20489023 (Breitkreutz et al 2010, Science)	Interaction (systematic network)	Competing/context (not CDK evidence)	Seed's 3rd cited PMID — what is it?	Global yeast kinase–phosphatase interaction network (large-scale AP-MS); a systematic interactome resource, not a functional CDK-complex study	S. cerevisiae, proteome-scale	Verified this run via NCBI eSummary. Explains how systematic interaction data enter databases; does not place KDX1 in a CDK holoenzyme.
UniProt P36005 (features)	Computational/rule-based	Qualifies ATP-binding claim	Are ATP-binding sites experimental?	P-loop (29–37), Lys55, active-site 153 all annotated by PROSITE-ProRule (ECO:0000255); GO:0005524 = IEA:InterPro	Sequence rule	Rule-based only; no assay. Cannot confirm ATP binding.
UniProt GO xrefs	Database (IBA/IEA)	Competing/context	Provenance of CDK/cell-cycle terms	GO:0000307, GO:0004693, GO:0000086 are all IBA:GO_Central (PAINT); no experimental support	PANTHER family tree	These are the terms under adjudication; provenance is phylogenetic inference, not experiment.
This report — sequence analysis (P36005 vs P32485)	Structural/evolutionary (computed)	Refutes catalytic competence; qualifies ATP binding	Does KDX1 retain kinase catalytic machinery?	KDX1 vs Slt2: DFG→NFG (Mg-Asp lost), HRD→HCD (Arg→Cys), degraded P-loop (no GxGxxG), no MAPK T-x-Y activation motif; β3 Lys retained	In-silico, public sequences	High for motif calls; ATP-binding consequence is inferred, not assayed.
This report — InterPro/PANTHER/CDD (P36005)	Computational/database	Refutes CDK classification; competing with CDK IBA terms	Which kinase family does KDX1 belong to?	PANTHER PTHR24055 "MAPK"; InterPro family IPR050117 "Mitogen-activated protein kinase"; CDD cd07857 = fungal MPK1 (Slt2) MAPK clade. ATP-binding GO derives from IPR017441 / PROSITE PS00107 ATP-binding signature (motif pattern), no CDK family present	InterPro protein2ipr	High. Confirms MAPK, not CDK; CDK IBA terms must be cross-clade over-propagation.

GO Curation Implications (leads — require curator verification)

GO:0000307 cyclin-dependent protein kinase holoenzyme complex (CC, IBA): Lead — REMOVE / do-not-assert. No evidence KDX1 is in a CDK–cyclin complex; it is a MAPK-family pseudokinase whose documented complex is with Swi4/Swi6 (SBF). Consider replacing with a supported CC such as the transcription-factor/SBF context (or leaving CC to the nucleus based on its SBF role).
GO:0004693 CDK Ser/Thr kinase activity (MF, IBA): Lead — REMOVE / do-not-assert. Pseudokinase; no catalytic activity, and even the parent kinase-activity call is questionable. If any MF is kept it should reflect the noncatalytic scaffolding role, not CDK activity.
GO:0004674 / GO:0106310 protein (Ser/Thr) kinase activity (ISS/IEA): Lead — flag for review. Contradicted by pseudokinase status (DFG→NFG, no activation motif); at minimum should not be a core function; a NOT-qualifier is defensible if curator confirms no residual activity.
GO:0000086 G2/M transition of mitotic cell cycle (BP, IBA): Lead — REMOVE / do-not-assert as direct function. No KDX1 experiment supports G2/M control; any cell-cycle link is indirect through its SBF partner (which is G1/S, not G2/M).
GO:0005524 ATP binding (MF, IEA:InterPro): Lead — treat as uncertain. Rule-based only; sequence predicts impaired Mg-ATP binding. Do not upgrade to experimental; acceptable to retain as low-confidence IEA or flag pending assay.
Supported terms to retain/strengthen (positive leads): BP cell-wall-stress transcriptional response / GO:0071554 cell wall organization or biogenesis context; MF/BP capturing noncatalytic transcription coactivator activity with SBF (e.g., a transcription-coregulator MF is more informative than "protein binding"); CC nucleus (promoter-bound complex). Interaction with Rlm1 and Swi4/Swi6 supports specific complex/coregulator terms rather than generic "protein binding."

Mechanistic Scope

Immediate molecular function tested: whether KDX1 directly (a) has CDK/kinase catalytic activity, (b) belongs to a CDK holoenzyme complex, (c) directly controls cell-cycle transitions, and (d) binds ATP.

Direct gene-product activity that is supported: noncatalytic assembly with Swi4 (and Swi6) to confer DNA-binding competence and activate cell-wall-stress transcription — redundant with Mpk1.
Downstream / partner-derived, not direct: any cell-cycle connotation arises from SBF's separate G1/S role, not from KDX1 acting on the cell-cycle machinery. G2/M is not mechanistically linked at all.
Inferred, not assayed: ATP/nucleotide binding — predicted impaired from loss of DFG-Asp and P-loop degradation; no phosphotransfer expected (no activation motif; HRD-Arg lost).

Conflicts and Alternatives

Paralog/namesake confusion (flagged in seed, confirmed relevant): "MLP1" is also the name of the nuclear-pore / myosin-like protein Mlp1 (YKR095W, UniProt Q02455), a ~1875-aa nuclear-basket coiled-coil protein, and its paralog Mlp2 (YIL149C, P40457, 1679 aa) which "associate with the nuclear pore complex" and organise perinuclear chromatin (verified via UniProt this run). The target here is KDX1/Mlp1 = YKL161C, P36005 (433 aa, Slt2 pseudokinase) — a completely different, much smaller protein. Curators must not import nuclear-basket / NPC / mRNA-export annotations onto P36005.
PANTHER/PAINT over-propagation (confirmed): InterPro/PANTHER/CDD place P36005 unambiguously in the MAPK family (PTHR24055; IPR050117) and specifically the fungal MPK1/Slt2 MAPK clade (CDD cd07857) — no CDK-family assignment exists. The CDK-activity, CDK-holoenzyme, and G2/M IBA terms therefore cannot come from KDX1's own subfamily and must be propagated from a broad CMGC-group ancestor node that mixes CDK and MAPK clades. This is a classic family-tree over-annotation; the MAPK subfamily should not inherit CDK-specific MF/CC/BP terms. The ATP-binding IEA traces to the PROSITE PS00107 / IPR017441 ATP-binding signature (a sequence motif pattern), not to any binding assay.
NOT-annotation transfer risk: The seed correctly warns that a NOT at another fungal node (PTN008603245) must not be applied unless ancestry to leaf PTN000623199 is established. My recommendation to not assert CDK/cell-cycle terms does not depend on importing that NOT; it rests on the target's own biology and sequence.
Genuine but indirect cell-cycle thread: Because SBF (Swi4/Swi6) governs G1/S, and KDX1 physically joins SBF under stress, a curator could see spurious support for "cell-cycle." This is an alternative interpretation to guard against, not evidence for a direct KDX1 cell-cycle function.

Knowledge Gaps

Direct ATP/nucleotide binding of KDX1. Checked: sequence (DFG→NFG, degraded P-loop, retained Lys55) and UniProt provenance (rule-based). Matters because GO:0005524 rests entirely on an InterPro rule. Resolve with: ITC/DSF thermal-shift or ADP-Glo nucleotide-binding assay on recombinant Kdx1; or a structure/AlphaFold-nucleotide-pocket analysis.
Any residual catalytic activity. Checked: motif loss predicts none. Matters for whether a NOT-kinase-activity qualifier is warranted. Resolve with: in-vitro kinase assay (already implied negative by kinase-dead-independent transcription in PMID:18268013, but not directly assayed for Kdx1 alone).
PANTHER tree ancestry for the disputed IBA terms. Checked: UniProt lists them as IBA:GO_Central. Matters because the correct curation is to prune CDK terms from the MAPK leaf. Resolve with: inspection of PAINT nodes PTN001172579/PTN000623979/PTN000623980 vs leaf PTN000623199 to confirm the terms derive from CDK-clade ancestors.
Whether the SBF-linked role justifies any cell-cycle BP term. Checked: literature ties Mlp1 to stress transcription, not cell-cycle progression. Resolve with: cell-cycle phenotyping of kdx1Δ (synchronised cultures) — expected null given redundancy and stress-specific role.

Discriminating Tests

Nucleotide-binding assay (DSF/thermal shift ± ATP-Mg²⁺; ADP-Glo) on recombinant Kdx1 vs Slt2 — cleanly separates "ATP binding" from "no binding."
Structure-based pocket check: AlphaFold model of P36005 → measure ATP-pocket integrity (DFG-Asp absent, P-loop geometry) and compare to Slt2 — supports/undermines GO:0005524 conservatively.
Cell-cycle assay of kdx1Δ and slt2Δ kdx1Δ (FACS/bud-index, synchronised) — tests any real G1/S or G2/M contribution.
PANTHER PAINT node audit (PTN000623199 leaf and ancestors) — confirms the IBA CDK/cell-cycle terms originate from a CDK-clade ancestor and should be pruned for the MAPK leaf.
ChIP/co-IP of Kdx1 with Swi4/Swi6 and Rlm1 — strengthens the correct complex/coregulator CC/MF terms.

Curation Leads (require curator verification)

Candidate references to verify (exact snippets):
PMID:18268013 — "We show that the MAPK of this pathway (Mpk1) and its pseudokinase paralog (Mlp1) use a noncatalytic mechanism to activate transcription of the FKS2 gene." and "Transcriptional activation of FKS2 was dependent on the Swi4/Swi6 (SBF) transcription factor and on an activating signal to Mpk1 but not on protein kinase activity."
PMID:20641022 — "In this context, the Mlp1 pseudokinase serves a redundant function with Mpk1."
PMID:20219973 — Mpk1–Swi4 complex recruits Swi6 to the nucleus for cell-wall-stress transcription (context for KDX1's shared complex step).
Candidate action changes: Remove / do-not-assert GO:0000307, GO:0004693, GO:0000086 for KDX1 (PAINT over-propagation from CDK clade); downgrade/flag GO:0005524 (IEA-only, likely impaired); flag GO:0004674/GO:0106310 kinase activity as contradicted by pseudokinase status.
Candidate replacement/new terms: noncatalytic transcription-coregulator role with SBF (Swi4/Swi6); cell-wall-stress transcriptional response / cell wall organization or biogenesis BP; nucleus CC; specific complex/interaction terms for Swi4/Swi6 and Rlm1 (avoid generic "protein binding").
Suggested questions for curator: Are the disputed IBA terms confirmed to descend from a CDK-clade ancestor (safe to prune)? Is a NOT-kinase-activity qualifier desired? Should ATP binding be retained as IEA or removed pending assay?
Suggested experiments: nucleotide-binding thermal-shift; kdx1Δ cell-cycle phenotyping; AlphaFold pocket analysis.

Provenance

Computational analyses executed this run: (1) UniProt REST fetch of P36005 (sequence, features, GO xrefs, comments); (2) UniProt fetch of paralog P32485 (SLT2); (3) regex motif scan comparing catalytic motifs (P-loop, VAIK, HRD, DFG, MAPK T-x-Y) between KDX1 and SLT2. Key computed result: KDX1 lacks GxGxxG, carries HRD→HCD and DFG→NFG substitutions, and has no T-x-Y activation motif, whereas SLT2 retains all. UniProt binding/active-site features for P36005 are ECO:0000255 (PROSITE-ProRule) rule-based, and the disputed CDK/cell-cycle GO terms are IBA:GO_Central; ATP binding is IEA:InterPro. No result was fabricated; where no assay exists (direct ATP binding), this is stated explicitly.