AIGR Gene Hypothesis Deep Research — SSQ1 (Q05931): Folding/Refolding Activity and the Nop1 Interaction

Target gene: SSQ1 / YLR369W / Ssc2p — Saccharomyces cerevisiae mitochondrial Hsp70 (UniProt Q05931)
Focus type: function_assignment
Hypothesis slug: folding-refolding-and-secondary-client-interactions

Executive Judgment

Verdict: Partially supported, with important precision caveats — and one clearly over-annotated sub-claim.

The seed hypothesis makes two independent claims that must be judged separately.

Claim 1 — Ssq1 retains protein folding/refolding or unfolded-protein holdase activity alongside its Fe–S role. This is partially supported but must be narrowed. Direct in vitro assays establish that Ssq1 retains generic Hsp70 behavior: ATP-regulated binding to unfolded substrate proteins (PMID: 11601843) and aggregation prevention of guanidine-denatured rhodanese by light scattering, strongest without nucleotide or with ADP (PMID: 16431909, Fig 4C). These are holdase / unfolded-protein-binding activities. Critically, no published assay demonstrates productive refolding or reactivation (i.e., restored enzymatic activity of a denatured substrate) by Ssq1. The rhodanese experiment measured aggregation suppression by light scattering, not recovered rhodanese activity; the Nfs1 experiments measured protection of already-folded enzyme activity, not reactivation of a misfolded species. Furthermore, Ssq1 lacks the general folding co-chaperone Mdj1, the matrix J-protein that powers Ssc1-driven refolding, and it is a specialist duplicate dedicated to Fe–S cluster transfer. Therefore a general "protein refolding" (GO:0042026) annotation over-reaches, whereas a holdase (GO:0140309) or substrate-binding characterization is defensible.

Claim 2 — Ssq1 has a biologically meaningful interaction with Nop1 (P15646). This is weakly supported / likely a high-throughput artifact. The interaction rests entirely on two whole-cell TAP-MS datasets (PMID: 16554755 and PMID: 19536198) that are not independent — they share authors, methods, and data lineage, and the second uses the alias "Ssc2." IntAct records both as moderate-score (0.56) spoke-expanded co-complex "associations" detected by TAP, with no direct binary-binding or functional assay. The two proteins occupy incompatible compartments (mitochondrial matrix vs. nucleolus), making a direct physical interaction topologically implausible. This interaction is not among Ssq1's current live GO annotations and should not be added.

The most important caveat for curation: the folding/refolding GO terms currently on Ssq1 are phylogenetic inferences (IBA) propagated from PANTHER family PTHR19375, not Ssq1-specific experimental findings. They are legitimate ancestral inferences (no experimental NOT/IRD exists on the path), but they are less precise than what direct Ssq1 assays actually show.

Key Findings

Finding 1 — Ssq1 retains ATP-regulated unfolded-substrate binding / holdase activity, but no refolding (foldase) activity is demonstrated

Two independent lines of direct in vitro evidence converge on a holdase, not foldase, characterization.

First, Schmidt et al. 2001 (PMID: 11601843) report that "Ssq1 showed typical chaperone properties by binding to unfolded substrate proteins in an ATP-regulated manner." This is the canonical behavior of the Hsp70 substrate-binding domain — nucleotide-modulated capture and release of exposed hydrophobic segments on unfolded clients. It establishes that Ssq1 has not lost the generic Hsp70 substrate-binding capacity.

Second, Dutkiewicz et al. 2006 (PMID: 16431909) provide the crucial biochemical detail the seed hypothesis flags. Their Fig 4C shows Ssq1 protects guanidine-denatured rhodanese against aggregation, measured by light scattering, with protection strongest in the absence of nucleotide or with ADP. Their Fig 4A/B show ATP/Jac1-independent protection of purified Nfs1 activity, which the authors attribute to "nonspecific binding of Ssq1p to Nfs1p [that] helped to prevent its unfolding." Fig 4D shows that Ssq1-depleted mitochondrial lysate retains unchanged Nfs1 activity. Every one of these readouts is aggregation prevention or unfolding prevention (holdase) — none measures the recovery of catalytic activity from a misfolded starting state (foldase/refolding).

The mechanistic distinction matters for GO curation. The obsolete term GO:0051082 ("unfolded protein binding") captured the older framing; the precise modern molecular-function term for the demonstrated action is GO:0140309 "unfolded protein holdase activity" (aggregation prevention plus escort to an acceptor), which also fits the ATP/ADP-modulated Isu1→Grx5 cluster-delivery cycle (PMID: 23615440). By contrast, the foldase/refolding terms — GO:0140662 "ATP-dependent protein folding chaperone," GO:0044183 "protein folding chaperone (in unfolded protein binding)," and BP GO:0042026 "protein refolding" — assert an activity (productive folding/reactivation) that no Ssq1 assay demonstrates.

Bottom line: the seed's holdase framing is supported; the folding/refolding framing is not directly supported and would over-annotate if asserted as Ssq1 experimental fact.

Finding 2 — Ssq1 lacks the general folding co-chaperone Mdj1 and is a specialized Fe–S factor, arguing against a general refolding annotation

The Hsp70 folding cycle is not autonomous; productive refolding requires a J-protein to trigger ATP hydrolysis and lock substrate, plus a nucleotide exchange factor to reset the cycle. In the yeast matrix, the multifunctional Hsp70 Ssc1 performs general protein folding/refolding using the J-protein Mdj1. Schmidt et al. 2001 (PMID: 11601843) explicitly report that "no interaction of Ssq1 with the two other mitochondrial Hsp70-cochaperones, Tim44 and Mdj1, was observed." Without Mdj1, Ssq1 lacks the machinery to run a productive folding cycle on general clients — consistent with its retained but unproductive-for-folding holdase behavior.

Dutkiewicz et al. 2006 (PMID: 16431909) reinforce this: they conclude "Ssq1p/Jac1p/Mge1p are not important for Fe/S cluster synthesis on Isu1p," and the only in vitro stimulatory effect they observed (on Nfs1) was ATP- and Jac1-independent — i.e., not a productive ATP-driven folding cycle but nonspecific unfolding prevention.

Two further papers frame the specialization. Pukszta et al. 2010 (PMID: 20224575) describe Ssq1 as an mtHsp70 duplicate that "specializes in iron-sulphur cluster biogenesis"; importantly, the inference of folding loss in that paper is drawn from partner (Jac1) specialization, not from a direct negative refolding assay — a nuance the seed correctly demands be preserved. Voisine et al. 2000 (PMID: 10779357) report only a Yfh1 maturation/processing delay in Δssq1 (mature Yfh1 ~75% of WT) with no aggregation or protease-sensitivity difference — a specific negative on one client that neither proves nor disproves generic holdase capacity on other clients.

The synthesis: Ssq1's established core is a specialized ATP-dependent Isu/Grx5 Fe–S cluster transfer factor (holdase/escort plus scaffold remodeling), sitting within the biological process GO:0016226 "iron-sulfur cluster assembly." A broad "protein folding chaperone" / "protein refolding" annotation is too strong as a statement about Ssq1's demonstrated activity.

Finding 3 — The Ssq1–Nop1 interaction rests on two non-independent whole-cell AP-MS datasets across incompatible compartments; likely a high-throughput artifact

The Ssq1(Ssc2p)–Nop1 interaction derives from Krogan et al. 2006 (PMID: 16554755), a genome-scale study that "used tandem affinity purification to process 4,562 different tagged proteins," and Gong et al. 2009 (PMID: 19536198), in which "systematic analysis of physical TAP-tag based protein–protein interactions of all known 63 chaperones in Saccharomyces cerevisiae has been carried out."

These two sources are not independent evidence. Gong 2009 and Krogan 2006 share authors (Krogan, Greenblatt, Emili) and TAP-MS data lineage; both IPI rows therefore trace to essentially one high-throughput, spoke-inferred dataset. Neither reports a direct binary-binding assay or a functional assay for Ssq1–Nop1. Compartmentally, Ssq1 is a mitochondrial-matrix Hsp70 (GO:0005759), whereas Nop1/fibrillarin (P15646) is a nucleolar rRNA 2′-O-methyltransferase; a direct physical interaction between them is topologically implausible. Abundant nucleolar proteins and "sticky" Hsp70s are classic AP-MS false-positive contaminants.

An IPI "protein binding" annotation founded on two non-independent AP-MS studies, with no orthogonal validation and a cross-compartment mismatch, is over-annotation. Native matrix residence alone does not formally disprove the interaction, but the burden of proof (direct binding, co-localization, a functional consequence) is unmet.

Finding 4 — The folding/refolding GO annotations are phylogenetic (IBA from PANTHER PTHR19375), not experimental; the Nop1 interaction is NOT a current GO annotation

A QuickGO annotation pull for Q05931 (28 annotations) clarifies the provenance of every term at issue. The folding claims are both IBA (ECO:0000318, GO_REF:0000033 = PANTHER): GO:0044183 "protein folding chaperone" (MF, IBA) and GO:0042026 "protein refolding" (BP, IBA), plus GO:0031072 "heat shock protein binding" (MF, IBA). PANTHER geneinfo confirms Q05931 is in family PTHR19375 (subfamily SF197), protein class "Hsp70 family chaperone" (PC00027), which propagates GO:0044183/GO:0031072 by descent — matching the seed's described PTHR19375 folding/refolding IBD path with no NOT/IRD assertion on the path.

By contrast, the experimentally grounded MF annotations are only GO:0016887 "ATP hydrolysis activity" (PMID: 12756240, PMID: 16431909, PMID 26545917) and GO:0005515 "protein binding" (IPI) — whose current references are PMID 12756240, 12947415, and 37968396 (Isu/Jac1/structural partners), not the Nop1 datasets. The Nop1 IPI (PMID 16554755 / 19536198) does not appear among Q05931's current live GO annotations; if it exists at all, it is an interaction-database/AP-MS entry, not a live GO curation. Localization is experimentally mitochondrial matrix (GO:0005759; PMID 10779357, 11273703, 8707841), with only an IBA cytoplasm term (GO:0005737).

Curation consequence: the folding/refolding terms under review are conserved-ancestor inferences, legitimately retainable as IBA (no experimental NOT exists), but less precise than the direct evidence, which supports holdase/substrate-binding (best captured by GO:0140309). GO:0042026 "protein refolding" is the weakest/most over-reaching term (no Ssq1 reactivation assay exists). The Nop1 interaction should not be added to GO on AP-MS-only, non-independent, cross-compartment evidence.

Finding 5 — IntAct confirms Ssq1–Nop1 rests only on two TAP co-complex records (score 0.56); Ssq1's other "protein binding" IPI is also HTP AP-MS

A programmatic IntAct query (findInteractions/Q05931) returned 123 interaction records for Ssq1. Exactly two involve Nop1/P15646, both with detectionMethod = 'tap' (tandem affinity purification), type "association"/"physical association," intactMiscore = 0.56, expansionMethod = spoke/co-complex, from publicationPubmedIdentifier 19536198 and 16554755 — i.e., the same two non-independent AP-MS datasets, with no direct binary method. Ssq1's IntAct partner list is dominated by abundant nuclear/chromatin/cytosolic proteins (e.g., ASF1, HHT1, HIR3, INO80, EAF6/EAF7, IES1, ELF1) alongside the genuine ISU1/ISU2 — a textbook signature of sticky-Hsp70 AP-MS promiscuity.

Separately, one of Ssq1's current GO "protein binding" (GO:0005515) IPI references, PMID 37968396 (Michaelis et al., Nature 2023, "The social and structural architecture of the yeast protein interactome"), is itself a genome-scale affinity-enrichment MS study rather than a targeted binary assay. So even Ssq1's non-Isu/Jac1 "protein binding" evidence is largely high-throughput. This confirms that the Nop1 record should be treated as a likely high-throughput artifact / non-core and not curated as biologically meaningful.

Mechanistic Model / Interpretation

Ssq1 is best understood as a specialized, "de-generalized" Hsp70: it kept the physical hardware of a chaperone (a nucleotide-binding domain with ATP hydrolysis, and a substrate-binding domain that captures unfolded/exposed segments) but shed the co-chaperone connections (Mdj1, Tim44) and acquired a dedicated J-protein (Jac1) and a specific client (Isu) that redirect it toward one job: Fe–S cluster transfer.

   GENERAL matrix Hsp70 (Ssc1)                SPECIALIST duplicate (Ssq1)
   ---------------------------                ---------------------------
   Co-chaperones: Mdj1, Tim44, Mge1           Co-chaperones: Jac1, Mge1 (NO Mdj1/Tim44)
   Clients: broad (import, folding)           Client: Isu (Fe-S scaffold), + Grx5 acceptor
   Activity: import + productive REFOLDING     Activity: holdase/escort + scaffold remodeling
                              Retained generic property: ATP-regulated
                              unfolded-substrate BINDING / aggregation
                              prevention (HOLDASE) -- but NOT refolding

The Fe–S transfer cycle Ssq1 executes (PMID: 23615440; PMID: 12756240):

  Jac1 (J-protein) targets Isu1 --> Ssq1 (ATP) binds Isu1 PVK motif
--> ATP hydrolysis (stimulated by Jac1+Isu1) --> ADP-Ssq1 grips Isu1 tightly
--> Grx5 binds a DISTINCT site on Ssq1 --> Fe-S cluster passed Isu1 --> Grx5
--> Mge1 exchanges ADP for ATP --> release/reset

Where does "folding/refolding" fit? The holdase step — nucleotide-modulated grip on an exposed, unfolded region — is real and generic, and it plausibly doubles as aggregation prevention for at-risk clients (rhodanese in vitro; Nfs1 protection). But productive refolding — the ATP-driven conformational work that returns a misfolded protein to a native, catalytically active state — has never been demonstrated for Ssq1, and the loss of Mdj1 removes the machine's ability to do it on general clients. The correct GO granularity therefore separates:

Evidence Matrix

GO Curation Implications

Lead requiring curator verification — treat all recommendations below as candidate actions, not final edits.

1. BP GO:0042026 "protein refolding" (currently IBA): Most over-reaching term. No Ssq1 assay demonstrates reactivation/refolding of a misfolded substrate. Lead: flag as non-core; consider whether the IBA should be down-weighted or annotated with a caveat, since direct experimental evidence contradicts a productive foldase interpretation for Ssq1 even though the ancestral inference is formally valid.

2. MF GO:0044183 "protein folding chaperone" / GO:0031072 "heat shock protein binding" (currently IBA): Retainable as ancestral inference (legitimate IBA; no experimental NOT on the PTHR19375 path). Lead: retain but recognize these are less precise than the direct evidence.

3. Add a more precise, experimentally grounded MF: the demonstrated activity is aggregation prevention + escort/cargo delivery. Lead: consider GO:0140309 "unfolded protein holdase activity" (with experimental evidence from PMID 11601843 and 16431909) as the term that best matches direct assays — a positive, informative alternative to the generic folding terms. Do not invent a new foldase claim.

4. Retain experimental core: MF GO:0016887 "ATP hydrolysis activity" and BP GO:0016226 "iron-sulfur cluster assembly" (and cluster transfer) are well supported and should remain the primary functional characterization. CC GO:0005759 "mitochondrial matrix" is experimentally supported.

5. Nop1 interaction: Do NOT add an IPI "protein binding" annotation to Nop1. It is not a current live GO annotation, and the underlying evidence is two non-independent AP-MS datasets with a compartment mismatch. If any interaction annotation to Nop1 exists downstream in an interaction database, curators should mark it as non-core / likely artifact.

Avoid "protein binding" (GO:0005515) as a final recommendation for the retained functional claim — the holdase term (GO:0140309) or ATP hydrolysis (GO:0016887) is more informative.

GO Decision Table

Mechanistic Scope

The immediate molecular function under test is substrate-binding-domain activity of an Hsp70 — nucleotide-regulated capture of unfolded/exposed polypeptide segments — and whether that extends to productive folding/refolding (net conformational work restoring native state) or is limited to holdase behavior (binding + aggregation prevention).

• Direct gene-product activity (supported): ATP-regulated unfolded-substrate binding; aggregation prevention; ATP hydrolysis stimulated by Jac1+Isu1; Isu1/Grx5 binding and Fe–S cluster transfer.
• Downstream / loss-of-function phenotypes (must not be conflated with direct activity): Yfh1 processing delay in Δssq1; reduced Fe–S enzyme activities; elevated iron; mtDNA maintenance defects. These are consequences of impaired cluster transfer, not evidence of a foldase.
• Inference-only (not direct assay): "folding loss" argued from partner (Jac1) specialization (PMID 20224575) is an inference, not a negative refolding experiment; the seed correctly insists this be labeled as such.

Conflicts and Alternatives

• Paralog/alias confusion: Ssq1 is also called Ssc2p, and its paralog Ssc1 is the genuine general folding Hsp70. AP-MS datasets using the "Ssc2" alias (PMID 19536198) risk carry-over confusion, and the folding-competent behavior belongs to Ssc1, not Ssq1. Curators should ensure folding annotations are not migrating from the paralog.
• In vitro–only activity: the holdase/aggregation-prevention effects (rhodanese, Nfs1) are in vitro and explicitly described as nonspecific (PMID: 16431909); their physiological weight is uncertain. This qualifies — but does not eliminate — the holdase claim.
• Compartment mismatch (Nop1): mitochondrial matrix vs. nucleolus is the strongest single argument against a biologically meaningful direct Ssq1–Nop1 interaction; the most parsimonious alternative is an AP-MS contaminant/co-purification artifact.
• Database carry-over: the folding GO terms are PANTHER-family IBA propagation, an alternative explanation to any assumption that they reflect Ssq1-specific experiments.

Limitations and Knowledge Gaps

1. No direct Ssq1 refolding assay exists. Checked: PMID 11601843, 16431909, 23615440, 12756240 — all measure binding, aggregation prevention, ATPase, or cluster transfer. Why it matters: a true foldase claim requires a reactivation assay. Resolver: an in vitro luciferase/rhodanese reactivation assay (recovered enzymatic activity, not just light scattering) ± Jac1/Mge1.
2. Physiological relevance of the holdase activity. Checked: rhodanese/Nfs1 protection framed as nonspecific. Why it matters: determines whether GO:0140309 should be a core or non-core annotation. Resolver: client-capture proteomics in vivo under stress; test whether Ssq1 protects endogenous matrix clients beyond Isu.
3. Full-text access to the AP-MS supplements for Ssq1–Nop1. Checked: abstracts and IntAct provenance; not the raw supplement rows. Why it matters: to confirm exact Nop1–YLR369W entries and formally establish non-independence. Resolver: inspect Krogan 2006 and Gong 2009 supplementary tables directly. This is an unresolved source-access limit and is reported plainly.
4. Whether any minor extra-mitochondrial pool of Ssq1 exists. Checked: experimental localization is matrix (GO:0005759); cytoplasm term is IBA only. Why it matters: a nuclear/cytosolic pool could (barely) rationalize a Nop1 contact. Resolver: high-sensitivity fractionation / imaging — but prior data strongly favor matrix-exclusive residence.

Discriminating Tests

1. Refolding vs. holdase discrimination: in vitro reactivation assay of chemically denatured firefly luciferase or rhodanese — measure recovered enzymatic activity (foldase readout) vs. suppressed light scattering (holdase readout) — for Ssq1 ± Jac1 ± Mge1 ± ATP, side-by-side with Ssc1+Mdj1 as a positive foldase control. Predicted outcome: Ssq1 suppresses aggregation but does not restore activity.
2. Mdj1-dependence test: attempt to reconstitute Ssq1-driven refolding by supplying Mdj1; if refolding does not appear, it confirms specialization removed foldase capacity.
3. Nop1 direct-binding test: purified Ssq1 + purified Nop1 by ITC/SPR/pulldown; and reciprocal endogenous co-IP with fractionation controls. Predicted outcome: no specific direct interaction.
4. Provenance audit: extract the exact Ssq1(YLR369W/Ssc2)–Nop1 rows from Krogan 2006 and Gong 2009 supplements to formally document non-independence and spoke inference.

Proposed Follow-up Actions / Curation Leads

All leads below require curator verification.

• Lead A (folding terms): Retain IBA GO:0044183 / GO:0031072 as ancestral inference; flag BP GO:0042026 "protein refolding" as non-core / over-reaching for Ssq1 (no reactivation assay). Verify snippet: "Ssq1p/Jac1p/Mge1p are not important for Fe/S cluster synthesis on Isu1p" (PMID: 16431909).
• Lead B (positive replacement MF): Consider adding experimentally supported GO:0140309 "unfolded protein holdase activity." Verify snippets: "binding to unfolded substrate proteins in an ATP-regulated manner" (PMID: 11601843); "nonspecific binding of Ssq1p to Nfs1p helped to prevent its unfolding" (PMID: 16431909).
• Lead C (Nop1 interaction): Do not add an IPI "protein binding" annotation to Nop1 (P15646). Verify snippets: "We used tandem affinity purification to process 4,562 different tagged proteins" (PMID: 16554755); "Systematic analysis of physical TAP-tag based protein–protein interactions of all known 63 chaperones" (PMID: 19536198). Rationale: non-independent AP-MS, spoke-inferred, IntAct score 0.56, matrix-vs-nucleolus mismatch.
• Lead D (retain core): Keep MF GO:0016887 (ATP hydrolysis), BP GO:0016226 (Fe–S cluster assembly/transfer), CC GO:0005759 (mitochondrial matrix) as the primary, experimentally grounded characterization.
• Suggested curator questions: (1) Should IBA folding terms carry a caveat when direct assays show only holdase? (2) Is GO:0140309 preferred over generic folding terms for the experimental MF? (3) Should the Nop1 interaction be explicitly marked non-core/artifact in any interaction annotation?
• Suggested experiments: luciferase/rhodanese reactivation assay (foldase vs. holdase); Mdj1-reconstitution test; Ssq1–Nop1 direct-binding assay + fractionation.

Source-Access Limitations (reported plainly)

• Full-text figures cited in the seed (PMID 12756240 Fig 8; PMID 16431909 Fig 4A–D; PMID 23615440 Fig 3D; PMID 10779357 Discussion) were evaluated via abstracts, the seed's detailed figure descriptions, and cross-referenced literature — not by rendering the author PDFs programmatically. The conclusions rest on the abstract-level statements and the seed's figure summaries, which are internally consistent with the abstracts.
• The exact Krogan 2006 / Gong 2009 supplementary rows for Ssq1(YLR369W/Ssc2)–Nop1 were not opened directly; non-independence is inferred from shared authorship, methods, and data lineage plus IntAct provenance. This should be confirmed against the raw supplements before finalizing.
• No fabricated results are reported; computed provenance (QuickGO annotation pull, PANTHER geneinfo, IntAct findInteractions/Q05931) is described as executed database queries.

Conclusion

The seed hypothesis is partially supported with precision caveats. Ssq1 retains generic Hsp70 unfolded-substrate binding and aggregation-prevention (holdase) activity, but no evidence supports a productive folding/refolding (foldase) activity, and it lacks the Mdj1 co-chaperone needed to run one. The existing folding annotations (GO:0044183, GO:0042026) are IBA carry-over from PANTHER PTHR19375, retainable as ancestral inference but less precise than the experimentally grounded holdase (GO:0140309) — with "protein refolding" the most over-reaching (treat as non-core). The Nop1 interaction is weakly supported / likely artifactual, is not a current GO annotation, and should not be added.

Artifacts

• OpenScientist final report
• OpenScientist final report
• OpenScientist provenance ssq1 curation