Skip to content

After reviewing all the documentation and relevant information, I'll now provide a detailed review of the GO-CAM model (gomodel:66b5638000001719) titled "AKT1 inhibits EIF2AK3/PERK (Human)":

Review of GOModel: 66b5638000001719 - "AKT1 inhibits EIF2AK3/PERK (Human)"

Overview

This GO-CAM model (gomodel:66b5638000001719) represents the regulatory relationship between AKT1 (protein kinase B) and EIF2AK3 (also known as PERK), which is a key component of the unfolded protein response (UPR) pathway. The model is focused on human proteins (NCBITaxon:9606) and depicts how AKT1 negatively regulates the PERK-mediated UPR.

Model Components

The model contains 5 molecular activities:

  1. AKT1 (UniProtKB:P31749) - protein serine/threonine kinase activity (GO:0004674)
  2. Part of negative regulation of PERK-mediated UPR (GO:1903898)

  3. Directly negatively regulates PERK/EIF2AK3 (RO:0002630)

  4. PERK/EIF2AK3 (UniProtKB:Q9NZJ5) - eukaryotic translation initiation factor 2alpha kinase activity (GO:0004694)

  5. Located in endoplasmic reticulum membrane (GO:0005789)
  6. Part of PERK-mediated UPR (GO:0036499)
  7. Directly negatively regulates EIF2S1/eIF-2-alpha translational activity (RO:0002630)

  8. Directly positively regulates ATF4 (RO:0002629)

  9. EIF2S1/eIF-2-alpha (UniProtKB:P05198) - translation initiation factor activity (GO:0003743)

  10. Located in cytosol (GO:0005829)

  11. Part of translational initiation (GO:0006413)

  12. ATF4 (UniProtKB:P18848) - DNA-binding transcription factor activity (GO:0003700)

  13. Located in nucleus (GO:0005634)

  14. Part of PERK-mediated UPR (GO:0036499)

  15. EIF2S1/eIF-2-alpha (UniProtKB:P05198) - translation initiation factor activity (GO:0003743)

  16. Located in cytosol (GO:0005829)
  17. Part of translational initiation (GO:0006413)
  18. Causally related to ATF4 transcription factor activity (RO:0002407)

Strengths of the Model

  1. The model accurately captures the core regulatory relationship where AKT1 inhibits PERK/EIF2AK3 kinase activity.

  2. It correctly depicts the PERK pathway where PERK activation leads to EIF2S1 phosphorylation, which inhibits general translation while promoting ATF4 translation.

  3. The model uses appropriate GO terms for activities, cellular components, and biological processes.

  4. The causal relationships are correctly represented using appropriate relationship ontology (RO) terms.

  5. The evidence provided for each association is robust, with multiple primary literature references.

Areas for Improvement

  1. Phosphorylation Sites Representation: The model doesn't explicitly indicate that AKT1 phosphorylates PERK at Thr-802, which is the mechanism by which it inhibits PERK activity (as described in the UniProt entry for PERK). Adding this molecular detail would enhance the model.

  2. Representation of EIF2S1 Phosphorylation: While the model shows PERK negatively regulating EIF2S1, it would be more precise to indicate that PERK phosphorylates EIF2S1 at Ser-52, which then leads to translational inhibition.

  3. Upstream Activators: The model doesn't include what activates AKT1 in the context of ER stress. Adding this context would provide a more complete picture.

  4. Missing Pathway Components: The model doesn't include some important components of the PERK pathway such as:

  5. GADD34/PPP1R15A which provides negative feedback

  6. CHOP/DDIT3 which is induced by ATF4 and involved in ER stress-induced apoptosis

  7. ATF4 Transcriptional Targets: The model stops at ATF4 activation but doesn't show any of its downstream targets such as SCAF1/COX7A2L (as described in the paper by Balsa et al., 2019), which would provide more context about the biological outcome.

  8. Integrated Stress Response Context: The model doesn't fully contextualize that PERK is part of the broader integrated stress response (ISR), which can be activated by multiple stressors beyond ER stress.

  9. Potential Duplication: There are two identical activities for EIF2S1 (entries #3 and #5 in the components list), which seems redundant unless they represent different instances/populations of the same protein.

Biological Accuracy

The model is biologically accurate in depicting the core regulatory relationship between AKT1 and PERK. The evidence provided comes from multiple peer-reviewed publications, strengthening the model's validity. The relationships shown align with current understanding of the PERK branch of the UPR pathway.

However, as noted above, including more molecular details (such as specific phosphorylation sites) and extending the model to include more downstream components would improve its comprehensiveness.

GO-CAM Modeling Guidelines Compliance

The model generally follows GO-CAM modeling guidelines:

  1. It uses appropriate molecular function GO terms for each protein.
  2. It accurately represents cellular locations.
  3. It uses appropriate causal relationship predicates (RO terms).
  4. It includes proper evidence for each assertion.

However, regarding how complexes are annotated, since the model doesn't explicitly represent any protein complexes, it doesn't fully leverage the guidance provided in the "How to annotate complexes in GO-CAM" document. For a more comprehensive representation, it might be worth considering if any of the proteins act as part of complexes (e.g., eIF2 is typically a complex of alpha, beta, and gamma subunits).

Recommendations

To enhance this GO-CAM model, I recommend:

  1. Add Phosphorylation Details: Specify that AKT1 phosphorylates PERK at Thr-802 and that PERK phosphorylates EIF2S1 at Ser-52.

  2. Resolve the Duplication: Remove or distinguish between the two identical EIF2S1 activities.

  3. Extend the Pathway: Consider adding:

  4. Upstream regulators of AKT1 in the context of ER stress
  5. Downstream targets of ATF4 (especially SCAF1/COX7A2L)

  6. The negative feedback loop involving GADD34/PPP1R15A

  7. Consider Complex Representation: Evaluate if representing eIF2 as a complex would be more appropriate than just showing the alpha subunit.

  8. Add Physiological Context: Include annotations that connect this regulatory pathway to its physiological outcomes, such as how AKT1-mediated inhibition of PERK affects cell survival during ER stress.

Conclusion

The GO-CAM model gomodel:66b5638000001719 "AKT1 inhibits EIF2AK3/PERK (Human)" is a valuable representation of a key regulatory relationship in the PERK branch of the UPR pathway. While it accurately captures the core elements of this regulation, adding more molecular details and extending the pathway representation would enhance its comprehensiveness and utility for the scientific community.