The research report should be a detailed narrative explaining the function, biological processes, and localization of the gene product. Citations should be given for all claims.

You should prioritize authoritative reviews and primary scientific literature when conducting research. You can supplement
this with annotations you find in gene/protein databases, but these can be outdated or inaccurate.

We are specifically interested in the primary function of the gene - for enzymes, what reaction is catalyzed, and what is the substrate specificity? For transporters, what is the substrate? For structural proteins or adapters, what is the broader structural role? For signaling molecules, what is the role in the pathway.

We are interested in where in or outside the cell the gene product carries out its function.

We are also interested in the signaling or biochemical pathways in which the gene functions. We are less interested in broad pleiotropic effects, except where these elucidate the precise role.

Include evidence where possible. We are interested in both experimental evidence as well as inference from structure, evolution, or bioinformatic analysis. Precise studies should be prioritized over high-throughput, where available.

Research Report: Mouse Serpinh1 / SERPINH1 (UniProt P19324) — Functional Annotation

0) Target verification (identity disambiguation)

The UniProt accession P19324 corresponds to mouse (Mus musculus) Serpin H1 (HSP47; colligin; CBP1), a serpin-family protein that is unusual among serpins in that it functions as an endoplasmic-reticulum (ER) luminal molecular chaperone rather than a serine-protease inhibitor. It is an ER-resident collagen-binding protein with a C-terminal ER retrieval signal and a well-defined collagen-binding specificity. (ito2017biologyofhsp47 pages 6-9, ito2017biologyofhsp47 pages 1-6)

1) Key concepts & definitions (current understanding)

1.1 What SERPINH1/HSP47 is (conceptual definition)

SERPINH1 encodes HSP47, an ER-resident, collagen-specific molecular chaperone that adopts a serpin fold but lacks serine protease inhibitory activity. Its canonical role is to bind newly synthesized triple-helical procollagen/collagen in the ER to promote correct folding and prevent misfolding/aggregation, enabling secretion and downstream extracellular matrix (ECM) assembly. (ito2017biologyofhsp47 pages 1-6, sakamoto2023hsp47atherapeutic pages 1-2)

1.2 Substrate specificity (what it binds)

A central feature of HSP47 is its sequence/structure selectivity for collagen triple helices. Mechanistic and structural evidence indicates HSP47 recognizes collagen triple-helical motifs centered on Gly-Xaa-Arg (also described as Yaa-Gly-Xaa-Arg-Gly in the triple helix) and shows additional preferences for nearby residues (e.g., the “Yaa-3” position). (sakamoto2023hsp47atherapeutic pages 1-2, ito2017biologyofhsp47 pages 9-12)

1.3 “Chaperone cycle” and trafficking terminology

HSP47’s chaperone function is coupled to a trafficking cycle in the early secretory pathway: it binds collagen in the neutral-pH ER and dissociates in more acidic early Golgi compartments (ERGIC/cis-Golgi), after which it is retrieved to the ER by a KDEL-receptor–dependent pathway driven by its C-terminal retrieval signal. (ito2017biologyofhsp47 pages 6-9, ito2017biologyofhsp47 pages 1-6)

2) Molecular function and mechanism (evidence-based)

2.1 Primary molecular function in the ER

HSP47 binds triple-helical procollagen in the ER and stabilizes the triple helix, thereby preventing unfolding and preventing aggregation within the ER lumen; this promotes efficient collagen secretion and correct downstream processing (including proteolytic processing), fibril formation, and ECM deposition. (ito2017biologyofhsp47 pages 15-19, ito2017biologyofhsp47 pages 1-6)

2.2 Binding mode: motif recognition and key residues

HSP47 binding requires the triple-helical conformation and collagen-like repeat length (e.g., binding appears when model repeats exceed a threshold), and depends on an Arg at the appropriate position in Gly-X-Y repeats (Gly-X-Arg). Structural work identifies key contacting residues such as Asp385 (contacting Arg in the motif) and Arg222 (involved in recognition of nearby positions). (ito2017biologyofhsp47 pages 9-12, wang2023naturalcompoundfraxinellone pages 1-2)

2.3 ER retention, release, and recycling

Mouse HSP47 has an N-terminal signal peptide and a C-terminal ER retrieval/retention motif (RDEL); deletion causes secretion. HSP47 binds procollagen in the ER and then releases it in ERGIC/cis-Golgi in a pH-dependent manner, and is returned to the ER via KDEL receptor recycling. (ito2017biologyofhsp47 pages 6-9)

3) Subcellular localization and where the protein acts

The dominant functional localization of SERPINH1/HSP47 is the ER lumen. It transiently accompanies procollagen into early Golgi compartments before dissociation, followed by retrieval to the ER, consistent with a chaperone that acts during ER folding and early secretory trafficking of procollagen. (ito2017biologyofhsp47 pages 6-9, ito2017biologyofhsp47 pages 1-6)

4) Biological processes and essentiality (mouse phenotypes)

4.1 Essential role in development

Genetic disruption of Hsp47/Serpinh1 in mice causes embryonic lethality (reported as no survival beyond ~11.5 dpc) with profound defects in basement membrane/collagen fibril formation and ER collagen accumulation/stress, indicating SERPINH1 is indispensable for collagen maturation during development. (ito2017biologyofhsp47 pages 9-12, ito2017biologyofhsp47 pages 1-6)

4.2 Skeletal development and cartilage

Conditional, tissue-specific loss (e.g., chondrocyte-specific) causes severe chondrodysplasia and major defects in endochondral ossification, consistent with HSP47 being required for correct folding/maturation of cartilage collagens (e.g., type II) and normal skeletal morphogenesis. (ito2017biologyofhsp47 pages 29-35)

5) Pathways and signaling: beyond “passive” collagen folding

5.1 Unfolded Protein Response (UPR) / ER stress integration: IRE1–XBP1

A 2024 mechanistic study in fibroblasts indicates HSP47 can influence collagen homeostasis through signaling: HSP47 overexpression increased type I collagen secretion/expression and was linked to activation of the IRE1α–XBP1 pathway (increased XBP1 splicing). IRE1 RNase inhibitors reduced HSP47-driven collagen induction, supporting pathway dependence. (ham2024hsp47increasesthe pages 1-2, ham2024hsp47increasesthe pages 2-4)

5.2 Cross-talk to Wnt/β-catenin

In the same 2024 system, HSP47 (or XBP1s) promoted β-catenin nuclear translocation and activated a TCF reporter, suggesting a mechanistic connection between ER proteostasis/IRE1 signaling and Wnt/β-catenin–dependent transcriptional regulation of collagen genes in fibroblasts. (ham2024hsp47increasesthe pages 1-2, ham2024hsp47increasesthe pages 2-4)

5.3 Integrin–FAK–PPARγ axis in adipose tissue biology

A 2023 Nature Communications study proposed HSP47 as a determinant of adiposity: HSP47 promotes collagen folding/secretion and collagen–integrin interactions, activating FAK signaling and preserving PPARγ protein from proteasomal degradation; adipose-specific knockout or pharmacological inhibition reduced adiposity and decreased pFAK/PPARγ in adipose tissue. (shin2023hsp47levelsdetermine pages 1-2, shin2023hsp47levelsdetermine pages 8-9)

6) Recent developments (prioritizing 2023–2024)

6.1 Expansion of client repertoire: neuroreceptor biogenesis (2024)

A 2024 eLife study substantially broadened SERPINH1 functional annotation by showing HSP47 promotes the biogenesis of multi-subunit Cys-loop neuroreceptors, including GABAA receptors. In HEK293T cells, HSP47 overexpression increased Endo H–resistant receptor glycoforms and increased functional surface expression (reported ~1.6-fold increase in peak GABAA currents). Mechanistically, HSP47 preferentially bound folded receptor conformations (acting after BiP), promoted assembly and anterograde trafficking, and reduced aggregation/ubiquitination without triggering a global UPR. Effects extended to other Cys-loop receptors (e.g., α4β2 nAChR, 5-HT3R), with reported assembly-efficiency improvement for α4β2 nAChR (FRET: 27.7% → 40.9%). (wang2024hsp47promotesbiogenesis pages 5-6, wang2024hsp47promotesbiogenesis pages 15-16)

6.2 Therapeutic targeting in fibrosis: siRNA/LNP and small molecules (2023–2024)

Authoritative 2023–2024 reviews emphasize HSP47 as a therapeutic node in fibrosis because it is required for efficient collagen maturation and secretion. In pulmonary-fibrosis models, Serpinh1 siRNA delivery approaches (e.g., vitamin A–coupled liposomes; LNP formulations such as ND-L02-s0201) reduced fibrosis-associated endpoints (e.g., collagen deposition, fibrosis scores), though the review notes context dependence across model systems. (sakamoto2023hsp47atherapeutic pages 8-9)

Small-molecule HSP47–collagen interaction inhibitors are also described, including compounds in the Col003 family; the pulmonary-fibrosis review reports ~85% and ~81% inhibition of HSP47–collagen binding for two Col003-related compounds at 1.9 µM (in vitro binding assay). (sakamoto2023hsp47atherapeutic pages 8-9)

6.3 Targeted RNA delivery as a real-world implementation direction (2023)

A 2023 Nature Communications study demonstrated ligand-tethered LNP delivery to activated fibroblasts in a liver fibrosis model, achieving ~65% HSP47 (Serpinh1) silencing and reducing collagen deposition and fibrosis. This is a concrete implementation of SERPINH1 functional knowledge into a therapeutic strategy (targeting collagen biosynthesis capacity rather than downstream ECM crosslinking alone). (han2023ligandtetheredlipidnanoparticles pages 9-10, han2023ligandtetheredlipidnanoparticles media 2fdf79e3)

7) Current applications and real-world implementations

7.1 Anti-fibrotic therapeutics (preclinical → early clinical)

HSP47 is actively pursued as an anti-fibrotic target using nucleic-acid therapeutics (siRNA/LNPs, targeted nanoparticles) and small-molecule inhibitors of the HSP47–collagen interaction. Reviews from 2023–2024 describe multiple delivery strategies and emphasize balancing efficacy with risks of disrupting essential collagen homeostasis and inducing ER stress. (sakamoto2023hsp47atherapeutic pages 8-9, khan2024hsp47inhuman pages 10-10)

7.2 Metabolic disease biology (adiposity)

SERPINH1/HSP47 has emerged as an adipose-tissue regulator connecting ECM collagen handling to integrin-FAK signaling and PPARγ stability, implying potential intervention points for obesity/metabolic phenotypes (with the caveat that systemic inhibition may have broad connective-tissue consequences). (shin2023hsp47levelsdetermine pages 1-2, shin2023hsp47levelsdetermine pages 8-9)

7.3 Neurological disease relevance (proteostasis of receptors)

The 2024 eLife findings imply that modulating HSP47 could influence surface expression and function of inhibitory neurotransmission receptors (GABAA and other Cys-loop receptors), framing SERPINH1 as part of the ER proteostasis network relevant to receptor-misfolding disorders (e.g., epilepsy-associated variants tested in the study). (wang2024hsp47promotesbiogenesis pages 1-2, wang2024hsp47promotesbiogenesis pages 5-6)

8) Relevant statistics and quantitative data (recent studies)

• ~65% silencing of HSP47 (protein level) in a mouse CCl4 liver-fibrosis model using ligand-tethered LNPs delivering pooled siRNA, with concomitant reduction in collagen deposition (Picosirius red) and fibrosis. (han2023ligandtetheredlipidnanoparticles pages 9-10, han2023ligandtetheredlipidnanoparticles media 2fdf79e3, han2023ligandtetheredlipidnanoparticles media c8e2d796)
• In vitro HSP47–collagen binding inhibition by small molecules: ~85% and ~81% inhibition at 1.9 µM for two Col003-related compounds (as summarized in a 2023 pulmonary-fibrosis review). (sakamoto2023hsp47atherapeutic pages 8-9)
• Neuroreceptor functional effects: ~1.6-fold increase in peak GABAA currents with HSP47 overexpression (HEK293T system) and improved receptor assembly metrics (α4β2 nAChR FRET 27.7% → 40.9%). (wang2024hsp47promotesbiogenesis pages 5-6, wang2024hsp47promotesbiogenesis pages 15-16)
• Adipose signaling readouts: adipose-specific HSP47 knockout significantly reduced HSP47, pFAK, and PPARγ in epididymal fat (p-values reported), consistent with the proposed collagen–integrin–FAK pathway and PPARγ proteostasis mechanism. (shin2023hsp47levelsdetermine pages 8-9)

9) Expert interpretation (authoritative analyses)

Reviews converge on a mechanistic interpretation that HSP47 is a “bottleneck” factor for collagen maturation (folding/quality control/export competence) and thus a rational target to reduce net collagen deposition in fibrotic disease, while emphasizing that systemic inhibition could have liabilities due to HSP47’s essential roles in collagen homeostasis and ER proteostasis. (sakamoto2023hsp47atherapeutic pages 8-9, khan2024hsp47inhuman pages 10-10)

10) Evidence-backed functional annotation summary (for database use)

Table: This table summarizes the core function, mechanism, localization, pathways, and recent research developments for mouse Serpinh1/SERPINH1 (HSP47). It is useful as a compact evidence-backed overview for functional annotation and prioritizes mechanistic and translational findings supported by the cited context IDs.

11) Source list (URLs and publication dates)

The report’s key sources (publication month/year and URL) include:

• Ito & Nagata, Feb 2017, Seminars in Cell & Developmental Biology: https://doi.org/10.1016/j.semcdb.2016.11.005 (ito2017biologyofhsp47 pages 1-6)
• Sakamoto et al., Aug 2023, Biomedicines: https://doi.org/10.3390/biomedicines11092387 (sakamoto2023hsp47atherapeutic pages 8-9)
• Shin et al., Nov 2023, Nature Communications: https://doi.org/10.1038/s41467-023-43080-x (shin2023hsp47levelsdetermine pages 1-2)
• Han et al., Jan 2023, Nature Communications: https://doi.org/10.1038/s41467-022-35637-z (han2023ligandtetheredlipidnanoparticles pages 9-10)
• Ham et al., Mar 2024, Cells: https://doi.org/10.3390/cells13060527 (ham2024hsp47increasesthe pages 1-2)
• Khan & Däinghaus, Aug 2024, Clinical and Translational Medicine: https://doi.org/10.1002/ctm2.1755 (khan2024hsp47inhuman pages 10-10)
• Wang et al., Oct 2024, eLife: https://doi.org/10.7554/eLife.84798 (wang2024hsp47promotesbiogenesis pages 5-6)

Scope note

SERPINH1/HSP47 is frequently discussed in human disease, but the mechanistic conclusions above are directly applicable to the mouse Serpinh1 protein because they describe core biochemical properties (collagen triple-helix binding, ER retrieval cycle) and/or include mouse in vivo models (e.g., liver fibrosis siRNA LNP; developmental essentiality). (ito2017biologyofhsp47 pages 1-6, han2023ligandtetheredlipidnanoparticles pages 9-10)

References

1. (ito2017biologyofhsp47 pages 6-9): Shinya Ito and Kazuhiro Nagata. Biology of hsp47 (serpin h1), a collagen-specific molecular chaperone. Seminars in cell & developmental biology, 62:142-151, Feb 2017. URL: https://doi.org/10.1016/j.semcdb.2016.11.005, doi:10.1016/j.semcdb.2016.11.005. This article has 286 citations and is from a peer-reviewed journal.

2. (ito2017biologyofhsp47 pages 1-6): Shinya Ito and Kazuhiro Nagata. Biology of hsp47 (serpin h1), a collagen-specific molecular chaperone. Seminars in cell & developmental biology, 62:142-151, Feb 2017. URL: https://doi.org/10.1016/j.semcdb.2016.11.005, doi:10.1016/j.semcdb.2016.11.005. This article has 286 citations and is from a peer-reviewed journal.

3. (sakamoto2023hsp47atherapeutic pages 1-2): Noriho Sakamoto, Daisuke Okuno, Takatomo Tokito, Hirokazu Yura, Takashi Kido, Hiroshi Ishimoto, Yoshimasa Tanaka, and Hiroshi Mukae. Hsp47: a therapeutic target in pulmonary fibrosis. Biomedicines, 11:2387, Aug 2023. URL: https://doi.org/10.3390/biomedicines11092387, doi:10.3390/biomedicines11092387. This article has 28 citations.

4. (ito2017biologyofhsp47 pages 9-12): Shinya Ito and Kazuhiro Nagata. Biology of hsp47 (serpin h1), a collagen-specific molecular chaperone. Seminars in cell & developmental biology, 62:142-151, Feb 2017. URL: https://doi.org/10.1016/j.semcdb.2016.11.005, doi:10.1016/j.semcdb.2016.11.005. This article has 286 citations and is from a peer-reviewed journal.

5. (ito2017biologyofhsp47 pages 15-19): Shinya Ito and Kazuhiro Nagata. Biology of hsp47 (serpin h1), a collagen-specific molecular chaperone. Seminars in cell & developmental biology, 62:142-151, Feb 2017. URL: https://doi.org/10.1016/j.semcdb.2016.11.005, doi:10.1016/j.semcdb.2016.11.005. This article has 286 citations and is from a peer-reviewed journal.

6. (wang2023naturalcompoundfraxinellone pages 1-2): Jie Wang, Mei Bai, Cui Zhang, Ning An, Li Wan, Xiao-ning Wang, Rong-hui Du, Yan Shen, Zhi-yao Yuan, Xu-dong Wu, Xue-feng Wu, and Qiang Xu. Natural compound fraxinellone ameliorates intestinal fibrosis in mice via direct intervention of hsp47-collagen interaction in the epithelium. Acta Pharmacologica Sinica, 44:2469-2478, Aug 2023. URL: https://doi.org/10.1038/s41401-023-01143-1, doi:10.1038/s41401-023-01143-1. This article has 9 citations and is from a peer-reviewed journal.

7. (ito2017biologyofhsp47 pages 29-35): Shinya Ito and Kazuhiro Nagata. Biology of hsp47 (serpin h1), a collagen-specific molecular chaperone. Seminars in cell & developmental biology, 62:142-151, Feb 2017. URL: https://doi.org/10.1016/j.semcdb.2016.11.005, doi:10.1016/j.semcdb.2016.11.005. This article has 286 citations and is from a peer-reviewed journal.

8. (ham2024hsp47increasesthe pages 1-2): So Young Ham, Min Ju Pyo, Moonkyung Kang, Yeon-Soo Kim, Dong Hun Lee, Jin Ho Chung, and Seung-Taek Lee. Hsp47 increases the expression of type i collagen in fibroblasts through ire1α activation, xbp1 splicing, and nuclear translocation of β-catenin. Cells, 13:527, Mar 2024. URL: https://doi.org/10.3390/cells13060527, doi:10.3390/cells13060527. This article has 16 citations.

9. (ham2024hsp47increasesthe pages 2-4): So Young Ham, Min Ju Pyo, Moonkyung Kang, Yeon-Soo Kim, Dong Hun Lee, Jin Ho Chung, and Seung-Taek Lee. Hsp47 increases the expression of type i collagen in fibroblasts through ire1α activation, xbp1 splicing, and nuclear translocation of β-catenin. Cells, 13:527, Mar 2024. URL: https://doi.org/10.3390/cells13060527, doi:10.3390/cells13060527. This article has 16 citations.

10. (shin2023hsp47levelsdetermine pages 1-2): Jihoon Shin, Shinichiro Toyoda, Yosuke Okuno, Reiko Hayashi, Shigeki Nishitani, Toshiharu Onodera, Haruyo Sakamoto, Shinya Ito, Sachiko Kobayashi, Hirofumi Nagao, Shunbun Kita, Michio Otsuki, Atsunori Fukuhara, Kazuhiro Nagata, and Iichiro Shimomura. Hsp47 levels determine the degree of body adiposity. Nature Communications, Nov 2023. URL: https://doi.org/10.1038/s41467-023-43080-x, doi:10.1038/s41467-023-43080-x. This article has 13 citations and is from a highest quality peer-reviewed journal.

11. (shin2023hsp47levelsdetermine pages 8-9): Jihoon Shin, Shinichiro Toyoda, Yosuke Okuno, Reiko Hayashi, Shigeki Nishitani, Toshiharu Onodera, Haruyo Sakamoto, Shinya Ito, Sachiko Kobayashi, Hirofumi Nagao, Shunbun Kita, Michio Otsuki, Atsunori Fukuhara, Kazuhiro Nagata, and Iichiro Shimomura. Hsp47 levels determine the degree of body adiposity. Nature Communications, Nov 2023. URL: https://doi.org/10.1038/s41467-023-43080-x, doi:10.1038/s41467-023-43080-x. This article has 13 citations and is from a highest quality peer-reviewed journal.

12. (wang2024hsp47promotesbiogenesis pages 5-6): Ya-Juan Wang, Xiao-Jing Di, Dong-Yun Han, Raad Nashmi, Brandon J. Henderson, Fraser J. Moss, and Ting-Wei Mu. Hsp47 promotes biogenesis of multi-subunit neuroreceptors in the endoplasmic reticulum. eLife, Oct 2024. URL: https://doi.org/10.7554/elife.84798, doi:10.7554/elife.84798. This article has 6 citations and is from a domain leading peer-reviewed journal.

13. (wang2024hsp47promotesbiogenesis pages 15-16): Ya-Juan Wang, Xiao-Jing Di, Dong-Yun Han, Raad Nashmi, Brandon J. Henderson, Fraser J. Moss, and Ting-Wei Mu. Hsp47 promotes biogenesis of multi-subunit neuroreceptors in the endoplasmic reticulum. eLife, Oct 2024. URL: https://doi.org/10.7554/elife.84798, doi:10.7554/elife.84798. This article has 6 citations and is from a domain leading peer-reviewed journal.

14. (sakamoto2023hsp47atherapeutic pages 8-9): Noriho Sakamoto, Daisuke Okuno, Takatomo Tokito, Hirokazu Yura, Takashi Kido, Hiroshi Ishimoto, Yoshimasa Tanaka, and Hiroshi Mukae. Hsp47: a therapeutic target in pulmonary fibrosis. Biomedicines, 11:2387, Aug 2023. URL: https://doi.org/10.3390/biomedicines11092387, doi:10.3390/biomedicines11092387. This article has 28 citations.

15. (han2023ligandtetheredlipidnanoparticles pages 9-10): Xuexiang Han, Ningqiang Gong, Lulu Xue, Margaret M. Billingsley, Rakan El-Mayta, Sarah J. Shepherd, Mohamad-Gabriel Alameh, Drew Weissman, and Michael J. Mitchell. Ligand-tethered lipid nanoparticles for targeted rna delivery to treat liver fibrosis. Nature Communications, Jan 2023. URL: https://doi.org/10.1038/s41467-022-35637-z, doi:10.1038/s41467-022-35637-z. This article has 202 citations and is from a highest quality peer-reviewed journal.

16. (han2023ligandtetheredlipidnanoparticles media 2fdf79e3): Xuexiang Han, Ningqiang Gong, Lulu Xue, Margaret M. Billingsley, Rakan El-Mayta, Sarah J. Shepherd, Mohamad-Gabriel Alameh, Drew Weissman, and Michael J. Mitchell. Ligand-tethered lipid nanoparticles for targeted rna delivery to treat liver fibrosis. Nature Communications, Jan 2023. URL: https://doi.org/10.1038/s41467-022-35637-z, doi:10.1038/s41467-022-35637-z. This article has 202 citations and is from a highest quality peer-reviewed journal.

17. (khan2024hsp47inhuman pages 10-10): Essak. S. Khan and Tobias Däinghaus. Hsp47 in human diseases: navigating pathophysiology, diagnosis and therapy. Clinical and Translational Medicine, Aug 2024. URL: https://doi.org/10.1002/ctm2.1755, doi:10.1002/ctm2.1755. This article has 16 citations and is from a peer-reviewed journal.

18. (wang2024hsp47promotesbiogenesis pages 1-2): Ya-Juan Wang, Xiao-Jing Di, Dong-Yun Han, Raad Nashmi, Brandon J. Henderson, Fraser J. Moss, and Ting-Wei Mu. Hsp47 promotes biogenesis of multi-subunit neuroreceptors in the endoplasmic reticulum. eLife, Oct 2024. URL: https://doi.org/10.7554/elife.84798, doi:10.7554/elife.84798. This article has 6 citations and is from a domain leading peer-reviewed journal.

19. (han2023ligandtetheredlipidnanoparticles media c8e2d796): Xuexiang Han, Ningqiang Gong, Lulu Xue, Margaret M. Billingsley, Rakan El-Mayta, Sarah J. Shepherd, Mohamad-Gabriel Alameh, Drew Weissman, and Michael J. Mitchell. Ligand-tethered lipid nanoparticles for targeted rna delivery to treat liver fibrosis. Nature Communications, Jan 2023. URL: https://doi.org/10.1038/s41467-022-35637-z, doi:10.1038/s41467-022-35637-z. This article has 202 citations and is from a highest quality peer-reviewed journal.

20. (khan2024hsp47inhuman pages 1-3): Essak. S. Khan and Tobias Däinghaus. Hsp47 in human diseases: navigating pathophysiology, diagnosis and therapy. Clinical and Translational Medicine, Aug 2024. URL: https://doi.org/10.1002/ctm2.1755, doi:10.1002/ctm2.1755. This article has 16 citations and is from a peer-reviewed journal.

21. (shin2023hsp47levelsdetermine pages 7-8): Jihoon Shin, Shinichiro Toyoda, Yosuke Okuno, Reiko Hayashi, Shigeki Nishitani, Toshiharu Onodera, Haruyo Sakamoto, Shinya Ito, Sachiko Kobayashi, Hirofumi Nagao, Shunbun Kita, Michio Otsuki, Atsunori Fukuhara, Kazuhiro Nagata, and Iichiro Shimomura. Hsp47 levels determine the degree of body adiposity. Nature Communications, Nov 2023. URL: https://doi.org/10.1038/s41467-023-43080-x, doi:10.1038/s41467-023-43080-x. This article has 13 citations and is from a highest quality peer-reviewed journal.

22. (wang2024hsp47promotesbiogenesis pages 18-19): Ya-Juan Wang, Xiao-Jing Di, Dong-Yun Han, Raad Nashmi, Brandon J. Henderson, Fraser J. Moss, and Ting-Wei Mu. Hsp47 promotes biogenesis of multi-subunit neuroreceptors in the endoplasmic reticulum. eLife, Oct 2024. URL: https://doi.org/10.7554/elife.84798, doi:10.7554/elife.84798. This article has 6 citations and is from a domain leading peer-reviewed journal.

23. (wang2024hsp47promotesbiogenesis pages 6-7): Ya-Juan Wang, Xiao-Jing Di, Dong-Yun Han, Raad Nashmi, Brandon J. Henderson, Fraser J. Moss, and Ting-Wei Mu. Hsp47 promotes biogenesis of multi-subunit neuroreceptors in the endoplasmic reticulum. eLife, Oct 2024. URL: https://doi.org/10.7554/elife.84798, doi:10.7554/elife.84798. This article has 6 citations and is from a domain leading peer-reviewed journal.

24. (wang2024hsp47promotesbiogenesis pages 14-15): Ya-Juan Wang, Xiao-Jing Di, Dong-Yun Han, Raad Nashmi, Brandon J. Henderson, Fraser J. Moss, and Ting-Wei Mu. Hsp47 promotes biogenesis of multi-subunit neuroreceptors in the endoplasmic reticulum. eLife, Oct 2024. URL: https://doi.org/10.7554/elife.84798, doi:10.7554/elife.84798. This article has 6 citations and is from a domain leading peer-reviewed journal.

25. (ito2017biologyofhsp47 pages 19-22): Shinya Ito and Kazuhiro Nagata. Biology of hsp47 (serpin h1), a collagen-specific molecular chaperone. Seminars in cell & developmental biology, 62:142-151, Feb 2017. URL: https://doi.org/10.1016/j.semcdb.2016.11.005, doi:10.1016/j.semcdb.2016.11.005. This article has 286 citations and is from a peer-reviewed journal.

26. (khan2024hsp47inhuman pages 8-10): Essak. S. Khan and Tobias Däinghaus. Hsp47 in human diseases: navigating pathophysiology, diagnosis and therapy. Clinical and Translational Medicine, Aug 2024. URL: https://doi.org/10.1002/ctm2.1755, doi:10.1002/ctm2.1755. This article has 16 citations and is from a peer-reviewed journal.

Citations

1. han2023ligandtetheredlipidnanoparticles pages 9-10
2. wang2023naturalcompoundfraxinellone pages 1-2
3. https://doi.org/10.1016/j.semcdb.2016.11.005;
4. https://doi.org/10.3390/biomedicines11092387;
5. https://doi.org/10.1038/s41467-022-35637-z
6. https://doi.org/10.3390/cells13060527
7. https://doi.org/10.1038/s41467-023-43080-x
8. https://doi.org/10.7554/eLife.84798
9. https://doi.org/10.1002/ctm2.1755;
10. https://doi.org/10.3390/biomedicines11092387
11. https://doi.org/10.1016/j.semcdb.2016.11.005
12. https://doi.org/10.1002/ctm2.1755
13. https://doi.org/10.1016/j.semcdb.2016.11.005,
14. https://doi.org/10.3390/biomedicines11092387,
15. https://doi.org/10.1038/s41401-023-01143-1,
16. https://doi.org/10.3390/cells13060527,
17. https://doi.org/10.1038/s41467-023-43080-x,
18. https://doi.org/10.7554/elife.84798,
19. https://doi.org/10.1038/s41467-022-35637-z,
20. https://doi.org/10.1002/ctm2.1755,

Deep Research Report: Serpinh1 and Serine-Type Endopeptidase Inhibitor Activity (GO:0004867)

Executive Judgment

Verdict: REFUTED / Over-annotated

The hypothesis that Serpinh1 (HSP47, UniProt P19324) has serine-type endopeptidase inhibitor activity (GO:0004867) is refuted. Despite belonging to the serpin superfamily and possessing the canonical serpin fold, HSP47 is a structurally and functionally non-inhibitory serpin whose established primary function is as a collagen-specific molecular chaperone resident in the endoplasmic reticulum. The GO:0004867 IBA annotation derives from phylogenetic annotation transfer (PANTHER, GO_REF:0000033) at a tree resolution too coarse to distinguish inhibitory from non-inhibitory serpin family members, and should be removed. This conclusion is supported by three independent and convergent lines of evidence: (1) structural incompatibility of the RCL hinge region with the serpin inhibitory mechanism, (2) explicit statements in authoritative literature denying serine protease inhibitory activity, and (3) direct provenance analysis revealing the annotation transfer error.

Summary

Serpinh1 encodes heat shock protein 47 (HSP47), a member of the serpin (serine protease inhibitor) superfamily. The seed hypothesis — that Serpinh1 has serine-type endopeptidase inhibitor activity — was evaluated through structural analysis, sequence comparison, literature review of 33 papers, and interrogation of the annotation provenance.

Computational analysis of the reactive center loop (RCL) hinge region — the structural feature that determines whether a serpin can function as a protease inhibitor — reveals that HSP47 has 0 of 6 small residues at the critical P14–P9 positions (Asn-Pro-Phe-Asp-Gln-Asp), compared to 5 of 6 in canonical inhibitory serpins such as alpha-1-antitrypsin (Thr-Glu-Ala-Ala-Gly-Ala). The presence of Pro at P13 and Phe at P12 is structurally incompatible with the conformational change (RCL insertion into β-sheet A) required for the serpin suicide-substrate inhibitory mechanism. This non-inhibitory hinge sequence is 100% conserved across mouse, human, rat, and chicken HSP47 orthologs, indicating strong purifying selection against protease inhibitory function.

Authoritative primary literature from multiple independent research groups explicitly and unequivocally states that HSP47 has no serine protease inhibitory activity (PMID: 27838364, PMID: 21683254, PMID: 12685865). No published study has ever demonstrated direct serine protease inhibition by HSP47 in any organism or assay system. Meanwhile, the annotation was traced to a PANTHER phylogenetic tree node (PTN000156127) whose "with/from" evidence field lists 12 genuinely inhibitory serpins, but the tree does not resolve non-inhibitory family members such as HSP47 from their inhibitory relatives.

Key Findings

Finding 1: HSP47 Lacks the Structural Requirements for Serine Protease Inhibition

The serpin inhibitory mechanism requires that the reactive center loop (RCL) insert into β-sheet A after cleavage by the target protease, trapping the protease in a covalent, kinetically stable complex. This conformational change is critically dependent on the hinge region (positions P14–P9 of the RCL), which must contain small, flexible residues (Ala, Gly, Ser, Thr) to permit smooth insertion. The importance of this requirement has been experimentally validated: mutagenesis studies have shown that replacing small hinge residues with bulky ones converts inhibitory serpins to substrates, while chimeric constructs with non-inhibitory hinge sequences show dramatically reduced inhibition efficiency (PMID: 9236002, PMID: 22312466).

Analysis of the HSP47 hinge region reveals a complete absence of the small residues required for this mechanism:

The Pro at P13 is particularly informative: proline's rigid pyrrolidine ring is fundamentally incompatible with the extended β-strand conformation required for RCL insertion. This single residue is sufficient to abolish inhibitory function, and its conservation across all HSP47 orthologs examined (mouse, human, rat, chicken) indicates that non-inhibitory status is an ancestral and deeply conserved feature, not a species-specific loss. The chicken ortholog shows a minor variation (Asn-Pro-Tyr-Asp-Ala-Asp) but retains the critical Pro at P13 and still scores only 1/6 small residues.

{{figure:hinge_region_comparison.png|caption=Comparison of RCL hinge region residues between canonical inhibitory serpins and HSP47. HSP47 has 0/6 small residues at the critical P14-P9 positions, with Pro at P13 and Phe at P12 being structurally incompatible with β-sheet A insertion required for the serpin inhibitory mechanism.}}

Finding 2: HSP47's Primary Function Is Collagen-Specific Molecular Chaperoning

Extensive experimental evidence from multiple laboratories and organisms establishes that HSP47 functions as an ER-resident, collagen-specific molecular chaperone. This function is supported by several types of gold-standard evidence:

Crystal structure evidence (PMID: 22847422): The structure of HSP47 in complex with a collagen model peptide, described as revealing that "Hsp47/SERPINH1 is a procollagen-specific molecular chaperone that, unlike other chaperones, specifically recognizes the folded conformation of its client," shows HSP47 specifically recognizes the folded triple-helical conformation of collagen via Gly-Xaa-Arg repeats. This is unique among chaperones — most chaperones recognize unfolded states. The collagen-binding interface is on a surface distinct from the RCL region, confirming that HSP47's functional binding site is not the serpin protease-interaction surface.

Knockout studies (PMID: 15522896): HSP47-knockout mice are embryonic lethal at E9.5–E11.5 due to catastrophic failure of collagen maturation. As reported, "In Hsp47(-/-) mouse embryos at 9.5 days postcoitus (dpc), immunostaining indicated the absence of type IV collagen, but not of laminin and nidogen-1, in the basement membrane." This demonstrates that HSP47 is essential and specific for collagen biosynthesis — non-collagen ECM components are unaffected.

Human and veterinary disease genetics (PMID: 25510505, PMID: 26004778): Mutations in SERPINH1 cause osteogenesis imperfecta (OI) type X in both humans and dachshunds — a collagen-processing disease. Biochemical analysis of the dachshund model (L326P mutation) showed that "type I collagen synthesized by mutant cells had decreased electrophoretic mobility" and "Procollagen was retained intracellularly with concomitant dilation of ER cisternae and activation of the ER stress response markers GRP78 and phospho-eIF2α" (PMID: 26004778). A separate human OI case revealed that HSP47 and FKBP65 cooperate in procollagen maturation, and that mutant HSP47 disrupts this cooperative pathway (PMID: 25510505). Population screening in dachshunds found a carrier frequency of 12.9% across European breeding populations (PMID: 23315765).

ER retention signal: HSP47 contains a C-terminal RDEL ER-retention signal (positions 414–417), consistent with its function as an ER-resident chaperone. Serine protease inhibitors typically function in the extracellular space or at the cell surface.

UniProt annotation: UniProt keywords for P19324 list "Chaperone" but NOT "Serine protease inhibitor," reflecting the curated expert consensus.

The authoritative review by the leading HSP47 research group synthesizes this evidence: "Hsp47, a collagen-specific molecular chaperone that localizes in the endoplasmic reticulum (ER), is indispensable for molecular maturation of collagen" (PMID: 27838364).

Finding 3: The GO:0004867 Annotation Is a Phylogenetic Transfer Error

The IBA (Inferred from Biological Aspect of Ancestor) annotation for GO:0004867 on Serpinh1 derives from the PANTHER phylogenetic tree via GO_REF:0000033. Investigation of the annotation provenance reveals:

• Source tree node: PTN000156127 in the PANTHER serpin family tree
• "With/from" evidence: Lists 12 genuinely inhibitory serpins including PAI-1 (SERPINE1), C1-inhibitor (SERPING1), alpha-2-antiplasmin (SERPINF2), SERPINB1, SERPINB6, SERPINB4, SERPINB8, SERPINB9, and neuroserpin (SERPINI1)
• Problem: The tree node encompasses both inhibitory and non-inhibitory serpin family members at insufficient resolution

The serpin superfamily contains approximately 30% non-inhibitory members that have repurposed the serpin fold for entirely different functions. Well-characterized non-inhibitory serpins include HSP47/SERPINH1 (collagen chaperone), ovalbumin/SERPINB14 (storage protein), maspin/SERPINB5 (tumor suppressor via non-inhibitory mechanism), and PEDF/SERPINF1 (neurotrophic factor). The PANTHER IBA pipeline does not adequately distinguish these functional subclasses from the inhibitory majority.

An additional IEA annotation exists from InterPro domain IPR000215 (Serpin family), which similarly does not distinguish inhibitory from non-inhibitory members. The InterPro2GO mapping IPR000215 → GO:0004867 is inherently too broad for this superfamily. Critically, no IDA, IMP, or IPI evidence supports serine protease inhibitor activity for HSP47 in any species.

{{figure:hsp47_go_curation_summary.png|caption=Summary of evidence refuting GO:0004867 annotation for HSP47/Serpinh1. The serpin fold is shared across the superfamily, but HSP47 has evolved as a non-inhibitory member functioning as a collagen-specific chaperone. The annotation derives from phylogenetic transfer at insufficient resolution.}}

Mechanistic Model and Interpretation

The Serpin Fold: Shared Structure, Divergent Function

The serpin superfamily represents a striking example of how a conserved protein fold can be repurposed for diverse functions during evolution. The canonical serpin inhibitory mechanism is among the most dramatic conformational changes known in protein biology:

INHIBITORY SERPIN MECHANISM (e.g., alpha-1-antitrypsin):

  Native (stressed) state           Protease binds RCL        Acyl intermediate forms
  ┌──────────────────┐              ┌──────────────┐          ┌──────────────────┐
  │  β-sheet A       │              │  β-sheet A   │          │  β-sheet A       │
  │  ┃┃┃┃┃           │   Protease   │  ┃┃┃┃┃       │  P1-P1'  │  ┃┃┃┃┃           │
  │  ┃┃┃┃┃  ←gap     │ ──────────→  │  ┃┃┃┃┃  P    │ ──────→  │  ┃┃┃┃┃  P        │
  │  ┃┃┃┃┃           │              │  ┃┃┃┃┃ /     │  cleaved │  ┃┃┃┃┃ ╱         │
  │         RCL ~~~~ │              │       RCL~~  │          │  ┃RCL inserted   │
  └──────────────────┘              └──────────────┘          └──────────────────┘

                                                                      │
                                                                      ▼

  Trapped complex (relaxed state)    KEY REQUIREMENT:
  ┌──────────────────┐               Hinge region (P14-P9) must have
  │  β-sheet A       │               small residues (A, G, S, T) to
  │  ┃┃┃RCL┃┃┃       │               permit RCL insertion as new
  │  ┃┃┃┃┃┃┃┃┃       │               β-strand into sheet A
  │  ┃┃┃┃┃┃┃┃┃       │
  │             P ←trapped           HSP47 FAILS this requirement:
  │           (distorted,            0/6 small residues, Pro at P13
  │            inactive)             ──→ CANNOT insert RCL
  └──────────────────┘               ──→ CANNOT trap proteases

HSP47: A Repurposed Serpin Fold for Chaperone Function

HSP47 retains the serpin fold but has repurposed it for a fundamentally different function:

HSP47 CHAPERONE MECHANISM:

  ER lumen (neutral pH)                          cis-Golgi (lower pH)
  ┌─────────────────────────────┐                ┌────────────────────────┐
  │                             │                │                        │
  │  HSP47 ──binds──→ Procollagen│  ──transit──→  │  HSP47 releases        │
  │  (serpin fold)   (Gly-X-Arg) │                │  procollagen           │
  │                   triple helix│                │  (pH-dependent)        │
  │                             │                │                        │
  │  Prevents:                  │                │  HSP47 recycled to ER  │
  │  • Premature aggregation    │                │  via RDEL signal       │
  │  • Local unfolding          │                │                        │
  │  • Misfolding               │                │  Procollagen continues  │
  │                             │                │  to secretory pathway  │
  └─────────────────────────────┘                └────────────────────────┘

The collagen-binding interface, revealed by crystal structure (PMID: 22847422), is on a distinct surface from the RCL, confirming that HSP47 uses a different face of the serpin fold for its function. The RCL region of HSP47, while still present, has accumulated mutations (especially Pro at P13) that would be catastrophic for an inhibitory serpin but are neutral for a chaperone.

Downstream Phenotypes Are Collagen-Related, Not Protease-Related

All known loss-of-function phenotypes for HSP47 are explained by impaired collagen maturation:
- Embryonic lethality in HSP47-knockout mice → basement membrane collapse from collagen IV deficiency
- Osteogenesis imperfecta in humans and dachshunds → defective bone collagen processing
- ER stress in mutant cells → procollagen accumulation and CHOP-mediated apoptosis
- Cancer-associated effects → altered ECM remodeling affecting tumor microenvironment

None of these phenotypes require or implicate serine protease inhibitory activity.

Evidence Matrix

GO Curation Implications

Recommended Action: REMOVE GO:0004867

The current GO:0004867 (serine-type endopeptidase inhibitor activity) annotation for mouse Serpinh1 should be removed. This is a curation lead requiring curator verification.

Rationale:
- Evidence type IBA (Inferred from Biological Ancestor via PANTHER phylogenetic tree) is inappropriate here because the ancestral node groups both inhibitory and non-inhibitory serpins under a single functional annotation
- An additional IEA annotation from InterPro (IPR000215, "Serpin family") similarly does not distinguish inhibitory from non-inhibitory members
- No IDA, IMP, IPI, or any experimentally supported evidence exists for serine protease inhibitor activity
- Multiple authoritative publications explicitly deny this activity

GO Annotation Decision Table

Possible New or Enhanced Annotations

• Consider whether a more specific chaperone MF term is available (e.g., "collagen-specific chaperone activity" if such a term exists in GO). GO:0051082 (unfolded protein binding) may be slightly misleading since HSP47 uniquely recognizes the folded triple-helical conformation of collagen.
• A NOT qualifier on GO:0004867, annotated with evidence from PMID:27838364 or PMID:21683254, would serve as a permanent guard against re-annotation by automated pipelines.

Broader IBA Quality Flag

This case exemplifies a known limitation of phylogenetic annotation transfer for superfamilies with functional divergence. The serpin superfamily has at least four well-characterized non-inhibitory members (HSP47/SERPINH1, ovalbumin/SERPINB14, maspin/SERPINB5, PEDF/SERPINF1). Curators should verify that IBA annotations for GO:0004867 have not been similarly propagated to these other non-inhibitory serpins via the same or related PANTHER tree nodes.

Conflicts and Alternatives

Primary Conflict: Family Nomenclature vs. Function

The seed hypothesis arises from a systematic confusion between superfamily membership and molecular function. The serpin superfamily (~1500 members across eukaryotes) derives its name from "serine proteinase inhibitor," but approximately 30% of members are non-inhibitory. HSP47 is among the most well-characterized non-inhibitory serpins. The gene name "Serpinh1" (Serpin Family H Member 1) reflects phylogenetic classification, not functional annotation — analogous to ovalbumin (SERPINB14), which is a storage protein with no protease inhibitory activity despite its serpin fold.

Cross-Class Inhibition Caveat

One publication (PMID: 12685865) reported that HSP47 (referred to as CBP2) can inhibit autolytic digestion of cysteine proteinases (cathepsins) in vitro. However, this paper explicitly states: "Although CBP2/Hsp47 is regarded as a molecular chaperone belonging to the serpin superfamily, this protein does not appear to inhibit serine proteinases." Key considerations:

• This is not serine protease inhibition (so not GO:0004867 regardless of validity)
• The mechanism is described as non-canonical (substrate pathway dominates, not stable complex formation)
• The finding has not been independently replicated in the 20+ years since publication
• The biological relevance is unclear given HSP47's ER localization
• Even if validated, it would correspond to a different GO term (GO:0004869, cysteine-type endopeptidase inhibitor activity)

Cancer Literature Naming Confusion

Several recent cancer papers (PMID: 39946769, PMID: 40334628, PMID: 40819105) refer to SERPINH1 as "serine protease inhibitor H1" or "a member of the serine protease inhibitor (serpin) superfamily" in their introductions. These descriptions reflect family nomenclature, not experimentally validated protease inhibitor function. None of these papers present any assay demonstrating direct serine protease inhibition by HSP47. The cancer-related effects described (proliferation, migration, invasion, EMT) are downstream consequences likely mediated through HSP47's collagen chaperoning activity and its effects on extracellular matrix remodeling.

No Organism-Specific Differences

The non-inhibitory hinge region is 100% conserved across mouse (P19324), human (P50454), rat, and chicken HSP47 orthologs. There is no evidence that any vertebrate HSP47 ortholog has serine protease inhibitory activity. This deep conservation rules out organism-specific functional divergence.

Knowledge Gaps

Discriminating Tests

Experimental Tests

1. Definitive negative inhibition assay: Express and purify recombinant mouse HSP47 (P19324). Test against a panel of serine proteases (trypsin, chymotrypsin, elastase, thrombin, plasmin, furin) using continuous chromogenic substrate assays. Measure apparent Ki values or demonstrate no inhibition at stoichiometric excess. Prediction: no inhibition. This would provide IDA-quality evidence for the absence of GO:0004867 function.

2. RCL insertion assay: Test whether HSP47's RCL can insert into β-sheet A by (a) limited proteolysis followed by SDS-PAGE mobility shift analysis, or (b) hydrogen-deuterium exchange mass spectrometry comparing native and RCL-cleaved HSP47. Canonical inhibitory serpins show a dramatic thermostability increase upon RCL cleavage and insertion; non-inhibitory serpins do not.

3. Hinge region swap experiment: Engineer HSP47 with a canonical inhibitory hinge region (replace P14–P9 Asn-Pro-Phe-Asp-Gln-Asp with α1-antitrypsin's Thr-Glu-Ala-Ala-Gly-Ala). Test whether this restores protease inhibitory activity. This would directly test whether the hinge region is the sole determinant of HSP47's non-inhibitory status.

Computational Tests

1. Systematic IBA audit: Query all IBA annotations for GO:0004867 across the serpin superfamily. Flag any non-inhibitory serpins (SERPINH1, SERPINB14/ovalbumin, SERPINB5/maspin, SERPINF1/PEDF) that carry this annotation. Report to GO Consortium for batch correction.

2. Molecular dynamics simulation: Simulate RCL insertion for HSP47 vs. α1-antitrypsin to quantify the energetic barrier imposed by Pro at P13 and Phe at P12.

Curation Leads

All leads below require curator verification.

Lead 1: Remove GO:0004867 (Highest Priority)

• Action: Remove the IBA annotation of GO:0004867 for Serpinh1 (P19324)
• Rationale: No experimental evidence supports this annotation; multiple authoritative sources explicitly deny it; computational analysis shows structural incompatibility
• Candidate references to verify:
• PMID: 27838364 — Exact quote: "Hsp47, which is encoded by the SERPINH1 gene, belongs to the serpin family and has the serpin fold; however, it has no serine protease inhibitory activity."
• PMID: 21683254 — Exact quote: "It belongs to the serpin family and contains a serpin loop, although it does not have serine protease inhibitory activity."
• PMID: 12685865 — Exact quote: "Although CBP2/Hsp47 is regarded as a molecular chaperone belonging to the serpin superfamily, this protein does not appear to inhibit serine proteinases."

Lead 2: Flag IEA Annotation from InterPro IPR000215

• Action: The IEA annotation of GO:0004867 from InterPro entry IPR000215 (Serpin family) should also be reviewed
• Rationale: IPR000215 matches all serpins regardless of inhibitory status; the InterPro2GO mapping IPR000215 → GO:0004867 is too broad for this superfamily

Lead 3: Consider NOT Qualifier

• Action: If the GO annotation framework supports a "NOT" qualifier, annotating NOT GO:0004867 with evidence from PMID:27838364 would explicitly block future automated re-annotation
• Rationale: Given the persistent automated annotation pipelines (IBA, IEA), a NOT annotation would serve as a permanent curation guard against re-introduction of this incorrect annotation

Lead 4: Verify Cross-Class Inhibition Relevance

• Question for curator: Should the cysteine proteinase cross-class inhibition from PMID:12685865 be annotated as GO:0004869 (cysteine-type endopeptidase inhibitor activity)?
• Concern: Single unreplicated study from 2002, non-canonical mechanism, uncertain biological relevance in the ER context. Annotation would require curator judgment on evidence quality and would need replication.

Lead 5: Ensure Chaperone Annotations Are Complete

• Action: Verify that all positive chaperone annotations are retained:
• GO:0005518 (collagen binding) — currently IEA
• GO:0051082 (unfolded protein binding) — currently IDA from MGI
• GO:0044183 (protein folding chaperone) — currently IDA from MGI
• GO:0005788 (endoplasmic reticulum lumen) — currently IEA
• Consider: Whether a more specific MF term for collagen-specific chaperone activity should be proposed

Lead 6: Report PANTHER Tree Issue

• Action: Report to PANTHER/GO that tree node PTN000156127 propagates inhibitory serpin annotations to non-inhibitory family members
• Affected proteins: At minimum HSP47/SERPINH1; potentially also ovalbumin, maspin, PEDF
• Suggested resolution: Tree refinement to separate inhibitory and non-inhibitory serpin clades for GO term propagation

Methodology Notes

This investigation combined:
- Sequence analysis: RCL hinge region residue composition at P14–P9, compared across HSP47 orthologs and canonical inhibitory serpins
- Literature review: 33 papers spanning HSP47 biology, serpin mechanism, osteogenesis imperfecta genetics, and cancer biology
- Annotation provenance: QuickGO query to trace the IBA annotation back to specific PANTHER tree nodes and "with/from" evidence
- UniProt/InterPro analysis: Feature annotation, keyword verification, and domain classification review

All computational analyses are preserved as provenance in the iteration logs.

Artifacts

• OpenScientist final report
• OpenScientist final report
• OpenScientist hinge region comparison
  
• OpenScientist hsp47 go curation summary
  
• OpenScientist plot 1
  
• OpenScientist plot 2