Selected horse accession: A0A9L0S5Z5. Source: ProtNLM2 API snapshot 2026-09-08, frozen in horse40-predictions.csv.
Component of the ribosome-associated complex (RAC), a complex involved in folding or maintaining nascent polypeptides in a folding-competent state. In the RAC complex, binds to the nascent polypeptide chain, while DNAJC2 stimulates its ATPase activity
PLI. Primary work identifies the human RAC pair as Mpp11 and Hsp70L1, while Apg2/HSPA4 experiments establish nucleotide exchange on Hsc70 in a chaperone/disaggregase system. Generic Hsp70-family similarity does not place HSPA4 in RAC. The selected horse protein is a near-identical HSPA4 homolog with its ATPase region retained and an internal C-terminal-region deletion; it is not thereby the RAC Hsp70 paralog. PMID:36604744(https://pubmed.ncbi.nlm.nih.gov/36604744/).
UNC for this selected sequence. Apg2/HSPA4 supports Hsc70-dependent aggregate reactivation as a nucleotide-exchange factor. This validates broad chaperone biology, not the particular ribosome-bound mechanism or direct nascent-chain-binding role asserted by the paragraph. At that broad level the claim is not novel for the horse gene: the full-length horse HSPA4 product F6YZJ4 already carries protein folding (GO:0006457) in GOA. A0A9L0S5Z5 is an exon-skipping isoform that lacks human residues 679–718 in the conserved alpha-helical C-terminal region, while retaining the nucleotide-binding domain (gene-model comparison). HSPA4 takes part in refolding through its exchange activity on Hsc70, and the distinctive C-terminal extension of Apg2 interacts with the Hsc70 nucleotide-binding domain and tunes this activity. No study has tested whether a form lacking this segment keeps exchange activity, so the contribution of this exact sequence is unresolved. PMID:30521813(https://pubmed.ncbi.nlm.nih.gov/30521813/); DOI:10.1016/j.jmb.2018.11.026; PMID:36167183(https://pubmed.ncbi.nlm.nih.gov/36167183/).
Human–horse sequence comparison records coverage, gaps and architecture caveats. Human experimental results support transfer where the relevant features are conserved; ARBA assertions and generated gene-review prose are not used as biological evidence. The human reference sequence and current horse sequence are not known to be the exact prediction-time inputs. Claim assessments are kept separate because a partly correct paragraph should not receive an undifferentiated verdict.