Where RPA actually is: nucleoplasmic and diffuse until there is single-stranded DNA, then punctate prereplication foci at the G1/S border, damage-induced foci with RAD51, and a PML-body pool - and the subunits differ in how well any of this is measured
also called RPA subnuclear localisation · RPA replication foci · RPA PML bodies
The RPA clamp has no fixed address in the nucleus. It floats diffusely until single-stranded DNA appears, and then gathers into bright spots - at sites where DNA copying is about to start, and at damage. It is also found in PML bodies, a distinct nuclear compartment.
The same thing, for a scientist
**RPA's localisation is conditional, which is the point.** All three subunits are annotated to **Nucleus** with experimental evidence codes; RPA1 and RPA2 additionally carry **Nucleus, PML body** [E1]. HPA immunofluorescence gives **RPA1 Nucleoplasm** and **RPA2 Nucleoplasm and Nuclear bodies**, both reliability 'Supported', and **for RPA3 HPA has NO subcellular location at all** (record present, reliability 'Approved', location field null) [E2]. The condition-dependent picture: **punctate RPA2 staining indicative of early or prereplication foci appears when cells are synchronised at the G1/S border** [E3]; **ATR kinase activity regulates the intranuclear translocation of ATR and RPA after ionising radiation** [E4]; and damage induces RAD51 accumulation into multiple nuclear foci **colocalising with RPA2** [E5]. **Three things this entry exists to prevent.** (i) **Do not upgrade the complex's nuclear call to RPA3** - RPA3's compartment rests on two UniProt attributions and no imaging call. (ii) **Do not read RPA2 foci in this corpus as a localisation measurement of RPA2** - in every corpus appearance they are the *instrument* of a resection assay. (iii) **The one negative that constrains the menin axis lives here**: at the G1/S border, with RPA2 prereplication foci visible, **menin did not mobilise into them** [E3]. **No RPA localisation of any kind has been measured in a KMT2A-rearranged cell** [E6].
Per-subunit, because they are not equally measured
| subunit | UniProt Nucleus | UniProt PML body | HPA call | HPA reliability |
|---|---|---|---|---|
| RPA1 (P27694) | 2 ECO:0000269 (17596542, 17959650) | yes (17959650) | Nucleoplasm | Supported |
| RPA2 (P15927) | 5 ECO:0000269 (10982866, 12814551, 17596542, 20154705, 21504906) | yes (12814551) | Nucleoplasm, Nuclear bodies | Supported |
| RPA3 (P35244) | 2 ECO:0000269 (17596542, 9430682) | no | none - field is null | Approved |
RPA2 is the best-measured subunit by a clear margin - five coded attributions against two - and it is also the one with the most antibodies, which is not independent of that fact. RPA3 is the least measured, and at 121 aa and 14 kDa a missing HPA call may reflect how hard it is to raise a specific antibody rather than a lack of interest.
'Nuclear bodies' (HPA) and 'PML body' (UniProt) are not the same call and should not be merged: HPA's term covers any discrete nuclear focus, UniProt's names a specific compartment. They are consistent, not identical.
The conditional addresses, and the negative that matters most
Prereplication foci at G1/S. When HeLa cells were synchronised at the G1/S border, punctate RPA2 staining indicative of early or prereplication foci was observed 3. This is the classic RPA localisation and it is cell-cycle-gated.
And in the same experiment, menin did not go there. "Although punctate RPA2 staining indicative of early or prereplication foci was observed when cells were synchronized at the G1/S border... menin localization appeared to be similar to that in untreated cells" 3. That negative is the strongest constraint on any replication-fork model of the menin-RPA2 contact, and it is recorded here as well as in interface-men1-rpa2 because a localisation entry is where a later beat will look for it.
Damage-induced translocation. ATR kinase activity regulates the intranuclear translocation of ATR and RPA following ionising radiation 4 - which is also one of UniProt's coded citations for RPA2's PML-body annotation, so the database row and the localisation study are the same paper in this instance (unusually; rule 91 warns that they normally are not).
Damage-induced foci with RAD51. RAD51 accumulates into multiple nuclear foci colocalising with RPA2 after damage 5. This corpus's gene-rad51 holds that record.
And the locus-level address. The one place RPA has been put at a genomic position relevant to this project is the KMT2A 11q23 breakpoint cluster region after etoposide - which is an occupancy measurement, not a compartment, and it is held in protein-rpa-heterotrimer and in the analysis entry 6.
| subject | protein-rpa-heterotrimer |
|---|---|
| subcellular | Nucleus, all three subunits, experimentally coded. |
| subnuclear localization | FOUR distinct calls, all condition-dependent, none constitutive: (1) diffuse NUCLEOPLASM in the unperturbed cell (HPA, RPA1 and RPA2); (2) PUNCTATE PREREPLICATION FOCI at the G1/S border (RPA2, immunofluorescence in HeLa, PMID 12509449); (3) DAMAGE-INDUCED FOCI colocalising with RAD51 and RAD51AP1, with an ATR-kinase-dependent intranuclear translocation after ionising radiation (PMID 12814551, and gene-rad51's UniProt record); (4) NUCLEAR PML BODY (RPA1 and RPA2, one ECO:0000269 each). RPA3 has none of its own. |
| localization is conditional | YES and this is the defining feature. RPA has no genomic or subnuclear address without single-stranded DNA. A compartment call for RPA is therefore nearly uninformative and the focus pattern is the measurement. |
| localization shown for fusion | NO - no RPA localisation of any kind has been measured in a KMT2A-rearranged cell, in any lineage. This is the plain-NO case rule 86 reserves for 'nobody has looked in any fusion'. The nearest thing is a DESIGNED but UNRUN experiment in this corpus using RPA2 foci as a resection readout. |
| measured in a blood cell | false |
| epigenetic readout | NONE. No histone mark is involved in RPA's localisation as recorded here. gamma-H2A.X marks the damage sites where RPA foci form but is not a determinant of RPA's own address in any source read by this beat. |
| absence class | never-measured |
| absence class note | For RPA localisation in a KMT2A-rearranged cell. Cheap to close - it is an immunofluorescence experiment in a cell line this project already works with, and the corpus has already designed one that would produce the image as a by-product. |
| fetched | 2026-09-24 |
subject subunits
[ "gene-rpa1", "gene-rpa2", "gene-rpa3" ]
per subunit — the numbers behind the chart
{
"RPA1": {
"uniprot": "P27694",
"uniprot_nucleus_eco269": 2,
"uniprot_pml_body": true,
"hpa_location": [
"Nucleoplasm"
],
"hpa_reliability": "Supported"
},
"RPA2": {
"uniprot": "P15927",
"uniprot_nucleus_eco269": 5,
"uniprot_pml_body": true,
"hpa_location": [
"Nucleoplasm",
"Nuclear bodies"
],
"hpa_reliability": "Supported"
},
"RPA3": {
"uniprot": "P35244",
"uniprot_nucleus_eco269": 2,
"uniprot_pml_body": false,
"hpa_location": null,
"hpa_reliability": "Approved",
"note": "NO HPA subcellular-location value exists. Do not upgrade the complex's nuclear call to this subunit's imaging record - there is none."
}
}the menin negative
{
"observation": "menin showed no detectable mobilisation into replication foci; at the G1/S border with punctate RPA2 prereplication foci visible, menin's localisation appeared similar to untreated cells",
"pmid": "12509449",
"why_recorded_here": "it is the strongest constraint on any replication-fork model of the menin-RPA2 contact, and a localisation entry is where a later beat will look for it. Also held in interface-men1-rpa2.",
"strength": "a 2003 immunofluorescence negative - it bounds the FRACTION of menin that relocalises, not the existence of the contact"
}locus level address
{
"site": "MLL/KMT2A 11q23 breakpoint cluster region after etoposide",
"pmid": "21048951",
"fusion_present": false,
"note": "an occupancy measurement, not a compartment. Held in full in protein-rpa-heterotrimer and in the beat's analysis entry."
}cell type
{
"answer": "NOT LINEAGE-SPECIFIC. Every imaging observation cited here is in HeLa or an HPA cell line (U-2 OS, A-431, U-251 MG by default). No haematopoietic cell appears anywhere."
}cell state
{
"state": "G1/S BORDER for prereplication foci; POST-DAMAGE (ionising radiation, camptothecin, etoposide) for damage foci. Both are the states in which RPA has an address at all."
}measurements — the numbers behind the chart
[
{
"quantity": "count",
"subject": "ECO:0000269 nuclear attributions per RPA subunit",
"value": 5,
"units": "attributions",
"n": 3,
"universe": "UniProtKB P27694, P15927, P35244 subcellular-location comments, fetched 2026-09-24",
"null": null,
"note": "RPA2 has 5, RPA1 has 2, RPA3 has 2. The value field holds RPA2's, the best-measured subunit; the spread is the point, and it tracks antibody availability rather than biological importance."
},
{
"quantity": "count",
"subject": "RPA subunits with an HPA subcellular-location call",
"value": 2,
"units": "subunits",
"n": 3,
"universe": "HPA search_download API responses for RPA1, RPA2, RPA3, 2026-09-24",
"null": 3,
"note": "RPA3's location field is null. null column is what a fully imaged complex would give."
},
{
"quantity": "count",
"subject": "RPA localisation measurements in a KMT2A-rearranged cell",
"value": 0,
"units": "measurements",
"n": 1,
"universe": "15 PubMed queries with every querytranslation read back plus a corpus read, 2026-09-24",
"null": null,
"note": "zero of any kind - compartment, subnuclear, focus or occupancy. The one locus-level RPA measurement in the KMT2A literature used cells with no fusion."
}
]- RPA2 foci are the field's standard readout for replication stress and for resection, so any intervention that increases them will look like it is engaging a specific mechanism when it may only be increasing ssDNA. A localisation readout alone cannot attribute a mechanism, and this corpus's own proposed experiments lean on RPA2 foci.
- Chromatin-bound RPA INTENSITY (the exhaustion readout) and RPA2 FOCI (the resection readout) are different measurements that look alike on a microscope. Conflating them would let an RPA-exhaustion phenotype be reported as increased resection, or the reverse.
- What does RPA2 focus formation look like in a KMT2A-rearranged blast, at baseline and under a menin inhibitor? The corpus has designed this experiment for a different purpose (resection competence) and the localisation image is a free by-product of running it.
- Is the RPA PML-body pool functionally distinct, and is it present in haematopoietic cells? Both RPA1 and RPA2 carry the annotation on a single attribution each, and PML bodies are a compartment this corpus touches elsewhere.
- Does menin relocalise to a SMALL fraction of RPA2 foci? The 2003 negative is a bulk-imaging result and a few-per-cent pool would be invisible to it. Proximity ligation would answer it at a sensitivity the original experiment did not have.
- Why does RPA3 have no HPA localisation call? Antibody difficulty at 14 kDa is the obvious explanation and it is a guess; the HPA antibody records were not examined.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt annotates all three RPA subunits to Nucleus with experimental evidence codes and two of them to nuclear PML body: RPA1 (P27694) Nucleus on ECO:0000269 PubMed 17596542 and 17959650 plus 'Nucleus, PML body' on 17959650; RPA2 (P15927) Nucleus on five ECO:0000269 attributions (PubMed 10982866, 12814551, 17596542, 20154705, 21504906) plus 'Nucleus, PML body' on 12814551; RPA3 (P35244) Nucleus on ECO:0000269 PubMed 17596542 and 9430682, with no PML-body annotation.
UniProtKB P27694, P15927 and P35244 subcellular-location comments with their evidence codes and PubMed attributions, all three fetched live 2026-09-24 · UniProt Consortium, EMBL-EBI/SIB/PIR 2026What this evidence cannot show
Rule 91 in full force: an ECO:0000269 means a curator saw an experiment in that paper, not that the paper is about localisation, is the founding observation, or is the best study available. Of the nine attributions here, exactly ONE was opened and title-checked by this beat - PubMed 12814551, 'ATR kinase activity regulates the intranuclear translocation of ATR and RPA following ionizing radiation', which IS a localisation study [E4]. The other eight were not opened. Note also that these are UNION annotations across papers and conditions: nothing here says what fraction of RPA is in a PML body, or under what conditions, and RPA's whole behaviour is conditional.
Evidence that argues against this (1)
- A compartment annotation is close to uninformative for RPA specifically, because RPA is diffusely nucleoplasmic in an unperturbed cell and its interesting localisation is a CONDITIONAL focus pattern that a static compartment call cannot express. 'Nucleus' is true and carries almost no information; the PML-body call is the only part that discriminates, and it rests on one attribution per subunit. This corpus has already measured that UniProt localisation is not a measurement in a quarter of its entries, and this is a case where the annotation is correct and still nearly useless. This repository 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Human Protein Atlas immunofluorescence gives RPA1 (ENSG00000132383) Nucleoplasm at reliability 'Supported', RPA2 (ENSG00000117748) Nucleoplasm and Nuclear bodies at reliability 'Supported', and for RPA3 (ENSG00000106399) a record at reliability 'Approved' with NO subcellular-location value.
Human Protein Atlas search_download API, three queries ('RPA1', 'RPA2', 'RPA3') with columns gene/ensembl/synonym/subcellular-location/reliability, all fetched 2026-09-24 · Human Protein Atlas, proteinatlas.org 2026What this evidence cannot show
Antibody-based immunofluorescence in fixed cultured cell lines - not a haematopoietic cell and not a primary cell. 'Supported' is HPA's middle reliability tier, below 'Enhanced'. The antibody identifiers are not returned by this column set (rule 74). The null location for RPA3 is a statement about the API response and could mean the experiment was never run, was not released, or the column is unpopulated - it is NOT a report that HPA looked and saw nothing. Ensembl ids are HPA's own joins and were not independently resolved. Same-day, single fetch.
Evidence that argues against this (1)
- HPA's 'Nuclear bodies' for RPA2 and UniProt's 'PML body' are being reported side by side and a reader will merge them - they should not be. HPA's vocabulary covers any discrete nuclear focus, so an RPA2 damage focus, a prereplication focus and a PML body would all score as 'Nuclear bodies' in a fixed cell. The two annotations are consistent and neither confirms the other. UniProt Consortium 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In HeLa cells synchronised at the G1/S border, punctate RPA2 staining indicative of early or prereplication foci was observed, while menin's localisation appeared similar to that in untreated cells - menin showed no detectable mobilisation into replication foci.
The 32-kilodalton subunit of replication protein A interacts with menin, the product of the MEN1 tumor suppressor gene. · Sukhodolets KE, Hickman AB, Agarwal SK, Sukhodolets MV, Obungu VH, Novotny EA, Crabtree JS, Chandrasekharappa SC, Collins FS, Spiegel AM, Burns AL, Marx SJ · Molecular and Cellular Biology 23(2):493-509 2003What this evidence cannot show
PMC full text read for this statement on 2026-09-24; the immunofluorescence images were not viewed and no quantitation (fraction of cells with foci, focus counts) was extracted - 'punctate staining indicative of early or prereplication foci' and 'appeared to be similar' are the paper's own qualitative language. HeLa, not haematopoietic, no fusion. 2003 imaging, so the resolution and the focus-calling standards are those of the era. A negative from a colocalisation experiment is sensitive to antibody, fixation and the fraction of protein involved: a small menin pool at foci could have been missed.
Evidence that argues against this (1)
- The menin negative is a NEGATIVE from a 2003 immunofluorescence experiment and should not be over-read. It bounds the fraction of menin that relocalises, not the existence of a contact - a contact involving a few per cent of nuclear menin would look exactly like this. The correct reading is that the contact is not a bulk replication-fork event, which is weaker than 'menin is not at forks'. Searched for a later study of menin at replication foci and found none. NCBI PubMed E-utilities 2026
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
ATR kinase activity regulates the intranuclear translocation of ATR and RPA following ionizing radiation.
ATR kinase activity regulates the intranuclear translocation of ATR and RPA following ionizing radiation. · Barr SM, Leung CG, Chang EE, Cimprich KA · Current Biology 13(12):1047-51 2003What this evidence cannot show
TITLE-LEVEL claim only. This record was surfaced by the '"replication protein A"[tiab] AND leukemia[tiab]' control list and its title was read by efetch; the abstract was NOT read in full and the full text was not opened, so nothing beyond the title assertion is claimed here. Author list, journal, volume and pages are from PubMed's record and were not independently re-verified beyond the esummary fetch. This IS one of UniProt's five ECO:0000269 attributions for RPA2's nuclear annotation and the sole attribution for its PML-body call - a case where the database's pointer and the relevant localisation study coincide, which rule 91 warns is not the norm. No haematopoietic cell and no fusion.
Evidence that argues against this (1)
- Reading only a title is the weakest form of citation this beat uses and it is flagged rather than hidden. What the title does NOT say is which RPA subunit was followed, what 'intranuclear translocation' means operationally (foci formation? chromatin fractionation? compartment change?), or whether the ATR dependence is on RPA's movement or on its retention. Any beat wanting to use this for a localisation claim must open it. It is cited here to establish that a damage-induced, ATR-dependent relocalisation of RPA exists and is documented, nothing more. UniProt Consortium 2026
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
DNA damage induces an increase in nuclear RAD51 and its accumulation into multiple nuclear foci colocalising with RAD51AP1 and RPA2, with RAD51 and FIGNL1 redistributed into discrete damage-induced foci after ionising radiation or camptothecin.
UniProtKB Q06609 (RAD51) subcellular-location comments as recorded in this repository's gene-rad51 entry, fetched live 2026-09-21 by that beat; six location values with nine ECO:0000269 attributions to Nucleus and eight to Chromosome, the majority of them in the damage-induced literature · UniProt Consortium / this repository 2026What this evidence cannot show
This item is INHERITED from gene-rad51's own fetch on 2026-09-21 and was NOT re-fetched by this beat, so it is a corpus read rather than a database read and could be stale by three days. It is included because the RPA2-RAD51 focus colocalisation is a localisation fact about RPA2 that lives in another beat's entry, and a localisation entry for RPA that omitted it would be incomplete. gene-rad51 is the entry of record. A colocalisation at immunofluorescence resolution is not a contact.
Evidence that argues against this (1)
- RPA and RAD51 colocalising in damage foci is the expected picture and is also the step where they are ANTAGONISTIC - RAD51 displaces RPA to form the recombinogenic filament, so colocalisation at focus resolution reflects a handover in progress rather than co-residence. Reading it as a stable association would invert the mechanism. At the KMT2A breakpoint hotspot specifically, INO80 is implicated in the exchange. Sun J et al. 2010
- E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
No RPA localisation of any kind - compartment, subnuclear, focus or genomic occupancy - has been measured in a KMT2A-rearranged cell; the one locus-level RPA measurement in the KMT2A literature is a ChIP at the 11q23 breakpoint cluster region in SV40-transformed fibroblasts and a CML-derived B-cell line, none of which carries a fusion.
PubMed esearch sweep of 15 RPA-versus-MLL/KMT2A queries via scripts/pubmed_safe.py with every querytranslation read back, plus PMC full text of PMID 21048951 for the cell lines used; both 2026-09-24 by Q192. Full working in ana-q192-the-complex-form-query-the-rpa2-decline-left-unrun-finds-rpa-at-the-kmt2a-breakpoint-hotspot. · NCBI PubMed E-utilities and this repository 2026What this evidence cannot show
Bounded by titles and abstracts, and that bound bites hard for localisation specifically: an RPA2 focus image in a supplementary figure of a KMT2A paper is exactly the kind of result this query cannot see, and RPA2 foci are a standard assay. The name list is recorded in the analysis entry. The claim is about RPA as a SUBJECT; RPA2 foci used as an assay readout in a KMT2A-rearranged cell would be a different thing and this corpus already proposes exactly that experiment without having run it.
Evidence that argues against this (1)
- This corpus's own exp-hr-competence-under-menin-inhibition-in-kmt2a-rearranged-blasts specifies RPA2 foci as its readout in KMT2A-rearranged blasts - so the measurement is designed, specified and unrun rather than unimagined, and describing the state of knowledge as a void would overstate it. The gap is that nobody has executed a designed experiment, not that nobody has conceived one. This repository 2026
subnuclear localisation record for the complex and its three subunits
RPA2 carries the most coded localisation evidence of the three subunits, plus the prereplication-foci observation
RPA3 has NO HPA subcellular-location call at all - recorded as a measured gap, not a nuclear assumption
menin did not mobilise into RPA2 prereplication foci at G1/S - the negative that constrains any replication-fork model of that contact