Gene expression profiling of monkeypox virus-infected cells reveals novel interfaces for host-virus interactions

Abstract

Monkeypox virus (MPV) is a zoonotic Orthopoxvirus and a potential biothreat agent that causes human disease with varying morbidity and mortality. Members of the Orthopoxvirus genus have been shown to suppress antiviral cell defenses, exploit host cell machinery, and delay infection-induced cell death. However, a comprehensive study of all host genes and virus-targeted host networks during infection is lacking. To better understand viral strategies adopted in manipulating routine host biology on global scale, we investigated the effect of MPV infection on Macaca mulatta kidney epithelial cells (MK2) using GeneChip rhesus macaque genome microarrays. Functional analysis of genes differentially expressed at 3 and 7 hours post infection showed distinctive regulation of canonical pathways and networks. While the majority of modulated histone-encoding genes exhibited sharp copy number increases, many of its transcription regulators were substantially suppressed; suggesting involvement of unknown viral factors in host histone expression. In agreement with known viral dependence on actin in motility, egress, and infection of adjacent cells, our results showed extensive regulation of genes usually involved in controlling actin expression dynamics. Similarly, a substantial ratio of genes contributing to cell cycle checkpoints exhibited concerted regulation that favors cell cycle progression in G1, S, G2 phases, but arrest cells in G2 phase and inhibits entry into mitosis. Moreover, the data showed that large number of infection-regulated genes is involved in molecular mechanisms characteristic of cancer canonical pathways. Interestingly, ten ion channels and transporters showed progressive suppression during the course of infection. Although the outcome of this unusual channel expression on cell osmotic homeostasis remains unknown, instability of cell osmotic balance and membrane potential has been implicated in intracellular pathogens egress. Our results highlight the role of histones, actin, cell cycle regulators, and ion channels in MPV infection, and propose these host functions as attractive research focal points in identifying novel drug intervention sites.

Figure 1

Validation of microarray results by quantitative real-time PCR. Copy number of eight common cytokines and a house keeping gene (B2M) were calculated based on RT-PCR Ct values. Fold change of genes expression is plotted on Y-axis after normalization to mock-treated samples. Results plotted to compare calculated fold change in expression of each gene at 3 hpi (A) or at 7 hpi (B) using Microarray (grey bars) and RT-PCR (black bars).

Figure 2

(A) Gene expression overview of Macaca mulatta kidney epithelial cells infected with MPV. (A) Using one-way ANOVA, a total of 2702 elements exhibited reproducible change (P-value ≤ 0.05). This represented 5.7% of the total 47000 interrogated transcripts on the GenChip. About 89% of the differentially expressed transcripts were suppressed and less than 11% were up-regulated. (B) Expression data for significantly influenced transcripts were hierarchically clustered. Columns represent triplicate (A, B, C) of time points 3 and 7 hpi. Color intensity reflects fold change relative to control (mock-transfected) cells. Red and green indicate up- and downregulation respectively.

Figure 3

Up- and down-regulated transcripts follow different distribution patterns during infection progression. Each bar represents average fold-change ratios (FCRs) of gene expression at 7 hpi/3 hpi for either 100 consecutive downregulated genes (A), or 10 upregulated genes (B), after sorting all genes according to their FCR values from smallest to largest. Genes with average FCR values < 1 are in grey and those with values > 1 are in black. Results show most downregulated genes maintained an increasing suppression trend in the course of infection as more than 92% of the genes gave fold-change ratio > 1 (A). The 295 upregulated genes exhibited more balanced distribution between 7 and 3 hpi with about 51% of the genes having higher fold change at early time point.

Figure 4

Comparative functional analysis of differentially expressed genes at 3 and 7 hpi. (A) Represents functions exhibiting -log (P-values) differences ≥ 1.301 between 3 and 7 hpi time points. Few identified functions were unique to one time point only under adopted statistical constraints. (B) Equally present functions in both time points showing a -log (P-value) differences ≤ 1.301. All functions were identified with P-values ≤ 0.05 or -log (P-value) = 1.301 thresholds (dotted black lines).

Figure 5

Major influenced canonical pathways of MPV-infected MK2 cells. Relevant pathways are shown at 3 (A) and 7 (B) hpi. The primary Y-axis shows the -log (P-value) of the probability for genes in data set to associate with identified pathway by chance. A threshold P-value of 0.05 or a -log (P) = 1.3 is presented in dotted line. The ratio of the number of genes from the data set that map to given pathway divided by the total number of genes that map to the canonical pathway is shown in solid line. All enlisted pathways met ratios ≥ 0.1 in at least one of the two time points (secondary Y-axis).

Figure 6

Heat-map of differentially expressed genes in MPV infected MK2 cells that mapped to molecular mechanisms of cancer pathway. Columns represent expression at 3 and 7 hpi in a triplicate (A, B, C). Each row represents one gene that met analysis cutoff of average fold change ≥1.8 in at least one of the two time points and P-value ≤ 0.05 in both. A gradient of green and red colors represent low and high relative fold change of gene expression to mock- infected cells. Chance for random association of listed genes at 3 or 7 hpi in this pathway is 2.82e-6 or 1.25e-5, respectively.

Figure 7

Cluster Analysis of expression profile for genes of ephrin receptor signaling pathway influenced by MPV infection. Columns of the heat-map represent expression of genes transcripts at 3 and 7 hpi in a triplicate (A, B, C). Each row represents one gene that met the analysis parameters of ≥1.8 average fold change cutoff in at least one of the two time points and P-value ≤ 0.05 in both. Gradient green and red color represents low and high relative expression fold change to mock infected cells, respectively. Chance for random association of listed genes at 3 or 7 hpi with this pathway is 5.95e-5 or 1.37e-4, respectively.

Figure 8

Cluster analysis of expression profile for genes of ataxia telangiectasia mutated (ATM) pathway. Columns of the heat-map represent expression of genes transcripts at 3 and 7 hpi time points in a triplicate (A, B, C). Each row represents one gene that met the analysis criteria of ≥1.8 average fold change cutoff in at least one of the two time points and P-value ≤ 0.05 in both. Gradient green and red color represents low and high relative expression fold change to mock infected cells, respectively. Chance for random association of listed genes at 3 or 7 hpi with this pathway is 1.87e-3 or 4.53e-3, respectively.

Figure 9

Cluster analysis of differentially- expressed genes involved in cell cycle regulation. Heat-map of gene expression levels at 3 and 7 hpi in triplicate (A, B, C). Folds of change in gene expression represented as a gradient of green and red color for low- and high-expression intensity, respectively. Each row represents one gene that met analysis cutoff of average fold change ≥1.8 in at least one of the two time points and P-value ≤ 0.05 in both. Genes fell in three cell cycle pathways with some genes functioning in more than one pathway (A) cell cycle: G1/S checkpoint regulation pathway (B) cell cycle: G2/M checkpoint regulation pathway (C) cell cycle regulation by BTG family proteins pathway.