Mathematical Modeling of Protracted HCMV Replication using Genome Substrates and Protein Temporal Profiles

Abstract

RationaleClinical manifestations of human cytomegalovirus (HCMV) infection, a major cause of morbidity and mortality in the immunocompromised, are associated with the lytic replication cycle. Lytic replication occurs over 96 hrs and is marked by viral DNA (vDNA) synthesis and expression of viral proteins, the latter of which can be generalized into distinct temporal classifications. Given the complexity of the lytic replication cycle, mathematical modeling can be used to probe this elaborate system. Recently, mathematical modeling has been used to study the early lytic replication cycle and the transition between lytic and latent cycles. To date, there appears to be few modeling studies focusing on the late phases of the lytic replication cycle.MethodvDNA replication kinetics were probed using qPCR assays with absolute quantification over varying multiplicities of infection (MOIs) in fibroblasts. An empiric model of vDNA replication was generated and parameterized using these data and then compared to qPCR data obtained post hoc. The vDNA model was used to drive a model of late viral infection that incorporated protein temporal class expression and culminated in virion production. This model was parameterized using data from immunoblots of representative proteins from the late temporal profile and viral titering data at three different MOIs. Predictions from this model were then compared to post hoc high‐temporal‐resolution, fluorescence‐based concentration measurements of one of the representative proteins over a 96‐hr replication cycle.ResultsA MOI‐dependent model of vDNA and late viral replication was developed from the experimental data. The model of late viral replication is mechanistic. It predicts feedback inhibition of late viral protein production and an increase in late viral protein degradation following an increase in vDNA. Both the vDNA and late replication models yielded 3 primary predictions: (1) presence of saturation kinetics at high MOIs, (2) inefficient replication at low MOIs, and (3) a range of MOIs in which vDNA replication and virus production are maximized. Saturation kinetics at high MOIs likely represents physical limitations of cellular machinery. Inefficient replication at low MOIs likely indicates a minimum MOI required to sustain infection. Identification of a maximal range of MOIs likely represents a set of conditions most favorable to HCMV lytic replication where extraneous viral material is minimized, while both the number of cells and the amount of viral material delivered to each cell is maximized.ConclusionA mathematical model of HCMV vDNA synthesis and late protein expression was generated and predicts a range of MOIs where maximal replication occurs, which can be extended to predict protein dynamics during infection.