Naturally‐Occurring Mutations to Muscle‐Type Creatine Kinase Differentially Impact Canonical and Pharmacological Activities in Vitro

Abstract

Tenofovir (TFV) is a nucleotide analog reverse transcriptase inhibitor (NRTI) and is administered as a prodrug in HIV pre-exposure prophylaxis (PrEP). Current data shows that PrEP has reduced HIV infections, but has not led to complete protection and some individuals on PrEP still become infected. Similar to other NRTIs, TFV must be phosphorylated inside cells to be pharmacologically active. We have found that in colon tissue, a key site of HIV infection, muscle-type creatine kinase (CKM) is responsible for the second phosphorylation event, producing the pharmacologically active TFV-diphosphate (TFV-DP) metabolite. We hypothesize that person to person biological variation in CKM activity is a cause of some PrEP failures. We have sequenced the CKM gene in individuals enrolled in PrEP trials and found several previously unreported variants. To determine the impact of these mutations on enzyme activity, we have recombinantly expressed and purified wild-type CKM and 15 of these mutants and performed in vitro assays of both the endogenous and drug-activating reactions. We carried out analogous ADP and TFV-monophosphate (TFV-MP) phosphorylation assays and measured ATP and TFV-DP formation, respectively, using ultra-high performance liquid chromatography tandem mass spectrometry. We also performed ATP dephosphorylation kinetics experiments using an enzyme-coupled colorimetric assay. We found that 11 of the 15 mutations significantly reduced TFV-DP formation compared to wild-type (WT). The greatest effect was observed with three active site mutations R130H, R132C, and N286I, plus W211R, all reducing TFV-DP to less than 5% of WT levels. Reductions to 14-25% of WT activity were also observed in T35I, R43Q, I92M, H97Y, F169L, and Y173C mutants. The T52N, F250S, N274Y, and L317M mutations did not impact TFV-DP formation. Only the three active site mutations and W211R displayed significant decreases in ATP formation, with all four mutations reducing activity to less than 30% of WT, including R130H and R132C having less than 5% of WT ATP formation. Additionally, in the ADP formation kinetic assay, the catalytic efficiency was decreased to less than 10% of WT by the three active site mutations along with W211R and H97Y. The TFV-DP formation assay results suggest that individuals carrying certain CKM mutations may not achieve protective levels of TFV-DP in colon tissue as readily as others. Further, the activity assays using endogenous substrates suggest a substrate-specific effect of these mutations on CKM function, as mutations that impacted activity toward TFV did not impact canonical functions with the endogenous substrate. This could indicate it is possible that mutations to CKM could be benign and not endogenous normal function and physiology but be detrimental to PrEP outcomes.