153: Proof of Mechanism & Pharmacodynamic Fingerprinting; Relating 7,000 Protein Measurements to Disease & Pharmacology: An Example Using the Impact of Peginterferon on Influenza A Pathway

Statement of Purpose: It has previously been shown that a single treatment with peginterferon (pegIFNβ-1a) can alter the abundance of numerous proteins in a time-dependent manner. The current study sought to investigate the effects of pegIFNβ-1a at the Influenza A pathway level with the goal of 1) demonstrating how groups of functionally related proteins respond in concert and 2) potentially understanding whether the effects of the drug on a pathway are directly or indirectly mediated.

Description of Methods & Materials: Healthy female and male participants were randomized to receive a single injection of pegIFNβ-1a [125 μg] (n=12) or placebo (n=6). EDTA plasma samples were collected prior to injection (baseline) and 11 timepoints over 13 days post-injection. The SOMAscan® assay v4.1 (SomaLogic, CO) was used to produce over 7,000 proteins measurements. The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) online database of protein-protein interactions was accessed to identify proteins that interact with IFNβ-1a, and the Kyoto Encyclopedia of Genes and Genomes (KEGG) was accessed to identify proteins in the Influenza A pathway. For each protein, relative fluorescence units were transformed to Log2 fold change from baseline (log2 FC). Separate composite variables were generated by averaging log2 FC values of Influenza A pathway proteins with primary (first shell) or secondary (second shell) interactions with IFNβ-1a, and a third composite variable was generated for Influenza A pathway proteins without known association with IFNβ-1a. Composites were then compared over time between drug and placebo groups using linear mixed-effects models.

Data & Results: Through STRING analysis, 26 first shell and 4 second shell proteins were found to overlap with the Influenza A KEGG pathway. An additional 69 Influenza A pathway proteins with no known association with IFNβ-1a were identified on the SOMAscan® assay v4.1 menu. Acute pegIFNβ-1a administration resulted in statistically significant increases from baseline in first and second shell composites (p < 0.001 for both). Maximum responses of 0.41 and 0.22 log2 FCs were observed for the first and second shell composites, respectively, at the 8-hour mark. In contrast, the composite for proteins without known-associations with IFNβ-1a did not differ between drug and placebo groups at any time. These results suggest that the effects of pegIFNβ-1a on the Influenza A pathway are mediated by proteins with known association with the drug target.

Interpretation, Conclusion or Significance: The current study used publicly available bioinformatics resources to produce a novel demonstration of how subsets of proteins within a canonical pathway are altered by a single dose of a drug. Importantly, the effects of the pharmacological intervention on the pathway were limited to only proteins that have known associations with the drug target. These findings and methodologies may be useful for future examinations of on/off target drug effects and could potentially lead to the development of a more targeted and mechanistic approach to investigating the pharmacodynamics of clinically relevant pathways.

Disclosures: JW, JD and JG are employees of SomaLogic Inc.