Refining the Structure of the Amot Coiled Coil Homology Domain (ACCH) to Understand Function in Epithelial Cell Cancers

Abstract

1 in 8 women in the United States will develop breast cancer in her lifetime. Epithelial cell cancers result from the overexpression of Angiomotins (Amot). Amots are a family of adaptor proteins that control the location of proteins responsible for cell proliferation, migration, and maintenance of normal cell structure. This trafficking function is related to the ability of Amot's coiled coil homology (ACCH) domain to associate with and remodel cellular membranes. We are interested in deciphering the structure of the ACCH domain to better understand this regulatory function. Our lab’s previous research proposed a theoretical structure of the ACCH. This project focuses on refining this structure, using modeling tools like RoseTTAFold, trRosetta, and GoogleCoLab, against globular envelope dimensions determined experimentally with small-angle X-ray scattering (SAXS). Approximately 80 models were generated and compared to current SAXS data. We also investigated mutations of the ACCH to better understand how single-amino acid changes affect structure and therefore functionality. Specific focus was on Arg-153 mutations and how structural differences cause previously documented reductions in association with cellular membranes. Refined structures are key to understanding how all documented Amot mutations lead to cancer initiation and metastasis to find effective cancer therapeutics.

Objectives

Use the Alphafold130 and Peck80 models of the ACCH domain as a tool to model how the domain interacts with cellular membranes
Explore which models are more refined through computational programs and ranking of each model
Compare structures to SAXS data to find the most refined structure

Background

The Kimble-Hill Lab is researching Triple-negative Breast Cancer. It is triple-negative due to its lack of estrogen receptors, progesterone receptors, and its low levels of human epidermal growth factor receptor 2 (HER2), which all regulate breast cell growth and division. Women who are Black are 2-3 times more likely to develop Triple-negative Breast Cancer than any other race, and spreads faster than any other type of breast cancer. We are interested in Angiomotin(Amot) proteins, which are a family of adaptor proteins that contribute to cell proliferation and migration in epithelial cells. We aim to understand the structure, function, and link of Amots to breast cancer.

From previous research, the Kimble-Hill Lab identified specific amino acids of Angiomotin (Amot) proteins that relate to membrane association. These proteins correlates with abnormal cell growth and proliferation and is overexpressed in breast cancers and other cancers Amots have a lipid-binding domain named the coiled coil homology (ACCH) domain. We hypothesize that disrupting the ACCH domain's ability to fuse membranes will lead to loss of normal cellular polarization and adhesion, which increases the rates of proliferation. Residues of four lysines and three arginines were identified to assist in lipid binding and shows affinity for phosphatidylinositols (PIs). Previous work in the lab determined theoretical structures of the ACCH domain to further understand the structure-function relationship of this domain. These models were used as a basis for this project to better refine the structure of this protein.

In this research project, the goal was to generate several models of this domain to be able to find the most stable version. To do this, we explored many computational programs to test different constraints and discover the most effective program that will generate a more refined model. These models are then used to compare how the structure affects the functionality of this protein in epithelial cells as it relates to membrane association.

Articles

Methodology

Computational programs were utilized throughout this project to form the models, rank the structures, and compare to SAXS data. PyMol was used to visualize and make the mutations of the domain. Mutations were made on the previously determined Alphafold130 and Peck80 models. We chose to mutate specific amino acids in relation to their ability to interact with the cellular membrane. There were 28 different mutations that were made from this program. Central focus was mutating Lysine(K) and Arginine(A) amino acids at various amino acids in the sequence.

Once the mutations were finalized, we then ran the protein data bank (pdb) files through multiple sequence alignment tools to check our mutations were correctly placed. We then ran models through multiple structure prediction programs including RobettaBakerLab, GoogleCoLab Alphafold2, GoogleCoLab RoseTTAFold, trRosetta, and Crysol to generate more models. Our models were then ran through ROSIE, a scoring function, to score each model and rank them based on scoring value.

These models could then be compared against previous SAXS data to determine the most refined structure. From SAXS, we generated Guinier analysis plots and confirmed that the suggested globular envelope dimensions data fit against the theoretical models.

Results

During the course of the internship, we have been able to generate ~80 models of the ACCH domain. This large number of results gives us confidence that we will find a more refined structure of this model to be compared against current SAXS data.

Top ranked wild type models:

1. trRosetta Alphafold Wt model 3 with a score of -471.897 REU

2. trRosetta Alphafold Wt model 5 with a score of -541.975 REU

The models were scored and measured using Rosetta Energy Units (REU) and scores of -100 to -300 REU are typical. The lower the total score, the more stable the structure is considered for the protein. Here we show our two top hits for the theoretical models of the wild type ACCH domain.

PyMol Mutation Models

Specific focus was on Arg-153 mutations and how structural differences cause previously documented reductions in association with cellular membranes. Here we focus on the Arg-153 mutations models included in the ACCH domain that have been refined. The three Arg-153 mutations studied are alanine(A), histidine(H), and glutamic acid(E). The images of these top ranked models showcase the differences between the mutations were made on the ACCH domain.

Top ranked models:

1. trRosetta Alphafold R153H model 4 with a score of -454.655 REU

2. trRosetta Alphafold R153E with a score of -480.132 REU.

3. trRosetta Alphafold R153A model 2 with a score of -501.725 REU

ACCH domain wild type is pink R153A mutation is purple

R153H mutation is orange R153E mutation is green

ACCH Domain wild type Only

ACCH domain wild type (pink) and R153A (purple)

ACCH domain (pink), R153A (purple), and R153H(orange)

These are two predicted histidine ring positions

ACCH domain (pink), R153A (purple), R153H (orange), and R153E(green)

Glutamic acid (E) is following the same directional pattern as the histidine

360° view of Alphafold mutations using PyMol

ACCH_Movie.mov

From the models, we are able to see that there are few structural differences between the mutations of Amot. We now hypothesize that membrane association occurs on the inside face of the ACCH domain "V structure". More computational modeling is needed to understand how the structure and amino acid residue location drives the membrane reorganization activity.

Future

In further experimentation, we plan to do more testing to understand the structure-function relationship of Amot80 by utilizing small angle x-ray scattering (SAXS) and computational modeling. We plan to use cryo-electro microscopy (cryo-EM) analysis to understand the role of mutations in a full length Amot80 structure.

Another area of interest is determining the role of lipid trafficking dysregulation on cellular proliferation rate, migration, invasiveness, and cellular adhesion. Further testing in this area will allow us to assess the loss of Amot130 association with membranes from the mutations.

Finding a more refined structure and understanding the function on a deeper level will lead to finding more effective cancer therapeutics.

References

Aase, K., Ernkvist, M., Ebarasi, L., Jakobsson, L., Majumdar, A., Yi, C., . . . Holmgren, L. (2007). Angiomotin regulates endothelial cell migration during embryonic angiogenesis. Genes & Development, 21(16), 2055-2068. doi:10.1101/gad.432007

Hall, L. C., Donovan, E., Araya, M., Idowa, E., Jiminez-Segovia, I., Folck, A., . . . Kimble-Hill, A. C. (2019). Identification of Specific Lysines and Arginines That Mediate Angiomotin Membrane Association. ACS Omega, 4(4), 6726-6736. doi:10.1021/acsomega.9b00165

Heller, B., Adu-Gyamfi, E., Smith-Kinnaman, W., Babbey, C., Vora, M., Xue, Y., . . . Wells, C. D. (2010). Amot Recognizes a Juxtanuclear Endocytic Recycling Compartment via a Novel Lipid Binding Domain. Journal of Biological Chemistry, 285(16), 12308-12320. doi:10.1074/jbc.M109.096230

Kimble-Hill, A. C., Petrache, H. I., Seifert, S., & Firestone, M. A. (2018). Reorganization of Ternary Lipid Mixtures of Non-Phosphorylated Phosphatidylinositol Interacting with Angiomotin. The Journal of Physical Chemistry B, 122(35), 8404-8415. doi:10.1021/acs.jpcb.7b12641

Lv, M., Lv, M., Chen, L., Qin, T., Zhang, X., Liu, P., & Yang, J. (2015). Angiomotin promotes breast cancer cell proliferation and invasion. Oncol Rep, 33(4), 1938-1946. doi:10.3892/or.2015.3780

Moleirinho, S., Guerrant, W., & Kissil, J. L. (2014). The Angiomotins–from discovery to function. FEBS Letters, 588(16), 2693-2703.

Peck, C. J., Virtanen, P., Johnson, D., & Kimble-Hill, A. (2018). Using the Predicted Structure of the Amot Coiled Coil Homology Domain to Understand Lipid Binding. IU Journal of Undergraduate Research, 4(1), 27-46. doi:10.14434/iujur.v4i1.24528

Ranahan, W. P., Han, Z., Smith-Kinnaman, W., Nabinger, S. C., Heller, B. D., Herbert, B.-S., . . . Wells, C. D. (2011). The Adaptor Protein AMOT Promotes the Proliferation of Mammary Epithelial Cells via the Prolonged Activation of the Extracellular Signal-regulated Kinases. Cancer Research. doi:10.1158/0008-5472.can-10-1995

Sears, S., Evans, B., & Kimble-Hill, A. (2021). Towards Understanding the Angiomotin Membrane Fusion Activity. ChemRxiv.

Wells, C. D., Fawcett, J. P., Traweger, A., Yamanaka, Y., Goudreault, M., Elder, K., . . . Pawson, T. (2006). A Rich1/Amot Complex Regulates the Cdc42 GTPase and Apical-Polarity Proteins in Epithelial Cells. Cell, 125(3), 535-548. doi:10.1016/j.cell.2006.02.045

Page updated

Report abuse