Poster H8 - Drug-induced Golgi Fragmentation and Sex-Specific Gene Expression Differences in Melanoma

Thursday, October 24, 2024

2:45 PM – 3:45 PM EST

Location: Exhibit Hall E, F & G

Presenting Author(s)

Iylan N. Howson (she/her/hers)

Undergraduate Researcher
North Carolina Agricultural and Technical State University
Decatur, Georgia, United States

Co-Author(s)

Shay Ladd

PhD Candidate
University of Virginia, United States

Primary Investigator(s)

SD

Sepideh Dolatchahi, PhD (she/her/hers)

Assistant Professor
University of Virginia, United States

Introduction: The Golgi apparatus is the primary site of glycosylation, a precisely regulated cell and microenvironment-specific post-translational modification. The Golgi is commonly fragmented in cancer cells, which can affect protein glycosylation. Glycosylation plays a crucial role in regulating intercellular signaling through modulation of protein localization, function, and interactions. This in turn influences various aspects of tumor progression including cell adhesion, migration, and immune evasion. Hypoxia is a common microenvironmental stressor in solid tumors, including metastatic melanoma, that is associated with tumor cell adaptation and poor patient prognosis. We hypothesize that one way in which hypoxia promotes malignant transformation is by Golgi fragmentation and subsequently modulating glycosyltransferase enzyme localization. In this project, we will explore the relationship between hypoxia and Golgi fragmentation in melanoma by first looking at the effects of Brefeldin A (BFA)-induced Golgi fragmentation in a BRAF-mutant melanoma cell line.

Hypoxia-regulated pathways are known to play a role in melanoma patient outcomes, but the interaction of this factor with sex has not yet been identified. Recently, sex hormone signaling has begun to explain some of the clinical outcomes of melanoma as well (Schwartz et al., 2019). This suggests that sex could play a significant role in the biology and prognosis of melanoma. Historically, men have been found to have a higher mortality rate and risk of developing melanoma than females. We hypothesized that differential activation of cancer-related pathways and differential expression of Golgi genes between sexes could further clarify determinants of sex-based differences in prognosis.

Materials and

Methods: We quantified brefeldin A (BFA)-induced changes in Golgi morphology, using immunofluorescent microscopy (IF). We treated A375 melanoma cells with 0.2μg or 2μg of BFA for 5, 10, 15, 30, and 45 minutes. We fixed cells with paraformaldehyde immediately after treatment and stained them for GM130 (cis-Golgi), TGN-46 (trans-Golgi), and DAPI (nucleus). We used CellProfiler software to identify the cis-Golgi, trans-Golgi, and nucleus. We then removed background noise, enhanced object quality, and segmented cells based on nuclei. Overlap between cis- and trans-Golgi and nuclei was removed. Objects were assigned to cells; those touching the border of the image were discarded. Convex hull and disconnected fragment areas were calculated for the cis- and trans-Golgi on a per-cell basis. We calculated the compactness metric by dividing the disconnected area by the convex hull on a per-cell basis.

To identify sex-based differences, we utilized publicly available RNA-seq data from the Genomic Data Commons (GDC). We built our cohort to include 578 patients (351 male, 227 female) with metastatic or primary melanoma. We used PROGENy (Schubert et al., 2018) to examine differences in pathway activation between male and female patients. PROGENy infers pathway activity from gene expression across a variety of conditions by using expression of downstream markers of pathway activation rather than direct components of the pathway. We examined the sex-specific survival associations of significant PROGENy genes identified by leading-edge analysis with TCGA data accessed through Protein Atlas. Lastly, we performed similar differential expression and survival analyses on pathways responsible for maintaining Golgi structure.

Results, Conclusions, and Discussions: We established the feasibility of immunofluorescent staining with confocal microscopy to obtain images of Golgi structure in melanoma cells. Using these images, we developed and refined a CellProfiler pipeline to extract object measurements needed for the calculation of quantitative metrics of Golgi morphology including cis- and trans-Golgi area and compactness. This pipeline will be used in future experiments to evaluate Golgi disruption caused by hypoxia.

Contrary to our expected results, we found that longer BFA treatment and higher concentrations have little impact on per-cell cis- and trans-Golgi compactness. The closer together the fragments are, the easier it is to traffic proteins between the compartments, which may correlate with less disruption to glycosylation. However, the compactness as calculated is likely artificially inflated by inaccuracies in convex hull tracing and reduced expression of Golgi markers caused by prolonged BFA treatment. Potential solutions include improving the CellProfiler pipeline to reduce dependence on convex hull measurements and treating at low concentrations and short times with BFA to ensure the Golgi does not fully collapse into the endoplasmic reticulum.

Using PROGENy, we discovered sex-specific activation of key cancer-related pathways and significant survival associations of genes driving these differences. The MAPK, PI3K, estrogen, and androgen pathways were more highly activated in men; NFkB, TNFa, and JAK.STAT were more highly activated in females. The upregulation of estrogen pathway activity in men was unexpected, so leading-edge genes were inspected. ASB13 was positively associated with estrogen and increased overall survival in males but not females. ABCC5 had opposite trends: negatively associated with estrogen pathway activation and overall survival in males, but not females.

Our results establish an experimental framework for quantitative metrics of Golgi disruption that can be used to investigate the Golgi as a possible intermediate for cancer-related glycosylation abnormalities. We also identify candidate genes for sex-specific prognostic indicators that are related to differentially activated oncogenic pathways and maintenance of Golgi structure. Future work may investigate sex-based differences in Golgi morphology in male- and female-derived melanoma cell lines and further probe connections between oncogenic pathway activation, Golgi disruption, and abnormal glycosylation.

Acknowledgements (Optional): Research reported in this publication was supported by the Cancer Systems Biology Cross Consortium award (University of Virginia Center for Systems Analysis of Stress-adapted Cancer Organelles, NCI U54 CA274499), Multi-Scale Systems Bioengineering & Biomedical Data Sciences REU at UVA.