Carolinas Research Day 2021

Predicting the Effects of Oxidative Stress on Disease Phenotypes Elizabeth Lawson OMS-I, B.S., Nicholas Kinney, Ph.D. Edward Via College of Osteopathic Medicine-Carolinas Campus, Spartanburg, SC 29303 .

BIOM-2

Abstract

Results

Conclusions

Figures 3 & 4

Future Directions: Analyze RNAseq databases for the disease phenotypes known to be affected by oxidative stress: 1. Diabetes 2. Down Syndrome 3. Autism 4. COVID-19 5. Breast cancer • Normalize gene expression data heterogeneous experiments: RNAseq data for healthy and disease cohorts. • Explore Human Variome Project and identify relevant RNAseq datasets • Investigate the effects of fasting and input fluctuations in diabetes • Extend the approach to mathematical models for other metabolic pathways.

Glutathione is an important metabolite related to resisting cell damage by reactive oxygen species, as well as other harmful reactants. Maintaining an appropriate GSH/GSSG ratio is key to avoiding cell damage and death considering a wide variety of diseases. Response to oxidative stress, and therefore appropriate glutathione response, is altered in various chronic diseases. Here we use gene expression data (RNAseq) to repurpose a published model of glutathione metabolism and predict effects of oxidative stress in medulloblastoma, the most common pediatric brain malignancy. Future work will expand the investigation to well- established chronic diseases such as diabetes and autism. There is also potential to link the investigation to the ongoing COVID-19 pandemic, as it is hypothesized that healthy glutathione levels could offset inflammatory effects of the virus.

Preliminary results suggest differences in response to oxidative stress in medulloblastoma patients compared to healthy controls. In particular, reduced concentration of blood cysteine is seen in the medulloblastoma. Further investigations will pursue this – and similar results – as potential companion diagnostics and prognostics. Blood glutathione is robust to amino acid changes in simulated healthy and medulloblastoma cohorts; specifically, both cohorts demonstrated similar response to fasting conditions and input fluctuations. Future work will investigate the effects of fasting and input fluctuations in more relevant cohorts, i.e., diabetes.

Introduction or Methods

Figure 1 . Glutathione metabolism schematic

References

Figure 2 .

1. Reed, Michael C., et al. "A mathematical model of glutathione metabolism." Theoretical biology and medical modelling 5.1 (2008): 1-16. 2.Lappalainen, Tuuli, et al. "Transcriptome and genome sequencing uncovers functional variation in humans." Nature 501.7468 (2013): 506-511. 3.Heath, Allison P., et al. "EPID-14. GABRIELLA MILLER KIDS FIRST DATA RESOURCE CENTER: COLLABORATIVE PLATFORMS FOR ACCELERATING RESEARCH IN PEDIATRIC CANCERS & STRUCTURAL BIRTH DEFECTS." Neuro-Oncology 22.Supplement_3 (2020): iii321-iii321 4. Pizzorno, J. (2014). Glutathione! IMCJ, 13. Retrieved March 7, 2021, from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4684116/ 5. Silvagno, F., Vernone, A., & Pescarmona, G. P. (n.d.). Antioxidants. doi:https://doi.org/10.3390/antiox9070624

+ 1 2 ,1 + 1 ,2 + 2 For most of the reactions, the original Reed model assumed Michaelis-Menten form with one substrate or random order Michaelis-Menten form with two substrates:

def Vmax(diseaseFPKM):

return 4500*(diseaseFPKM/normalFPKM)

7

2 0 2 1 R e s e a r c h R e c o g n i t i o n D a y