Adenovirus E4ORF1-induced MYC activation promotes host cell anabolic glucose metabolism and virus replication

Abstract

Virus infections trigger metabolic changes in host cells that support the bioenergetic and biosynthetic demands of viral replication. Although recent studies have characterized virus-induced changes in host cell metabolism (Munger et al., 2008; Terry et al., 2012), the molecular mechanisms by which viruses reprogram cellular metabolism have remained elusive. Here, we show that the gene product of adenovirus E4ORF1 is necessary for adenovirus-induced upregulation of host cell glucose metabolism and sufficient to promote enhanced glycolysis in cultured epithelial cells by activation of MYC. E4ORF1 localizes to the nucleus, binds to MYC, and enhances MYC binding to glycolytic target genes, resulting in elevated expression of specific glycolytic enzymes. E4ORF1 activation of MYC promotes increased nucleotide biosynthesis from glucose intermediates and enables optimal adenovirus replication in primary lung epithelial cells. Our findings show how a viral protein exploits host cell machinery to reprogram cellular metabolism and promote optimal progeny virion generation.

Figure 1. The E4 region is necessary for adenovirus-induced enhancement of glucose metabolism in host cells

(A) MCF10A cells were either mock infected, or infected with AD WT or AD ΔE4 virus for the indicated times. Yellowing of the pH indicator dye phenol red depicts media acidification from enhanced lactic acid production by the AD WT-infected cells. The average cell numbers at 72 hours for mock, WT, and ΔE4 infections are 5.8×10⁶, 3.2×10⁶, and 3.8×10⁶, respectively. In (B)–(D) MCF10A cells were infected for the indicated times, and metabolic measurements were taken. Glucose consumption rates (B), lactate production rates (C), and oxygen consumption rates (D) from cells infected with AD WT or AD ΔE4 were measured in triplicate samples. (E) Metabolic measurements from MCF10A cells constitutively expressing a vector control or the full E4 region. Error bars denote standard errors of the mean (n=3). * denotes p<0.05; ** denotes p<0.01.

Figure 2. The adenoviral gene product E4ORF1 is sufficient to promote increased glucose metabolism in cultured epithelial cells

(A) Immunoblotting of lysates from MCF10A cells stably expressing flag-tagged E4 ORF1, ORF2, ORF3, ORF4, or ORF6. The top panel was probed with FLAG antibody, and the bottom panel was probed with tubulin antibody to control for loading. (B) Glucose consumption and lactate production rates from MCF10A cells stably expressing the flag-tagged ORFs indicated in (A). (C) Immunoblotting of lysates from MCF10A cells stably expressing HA-tagged E4ORF6 along with an inducible empty vector, flag-tagged ORF1 WT, or flag-tagged ORF1ΔPDZ. Briefly, the cells were treated with 1ug ml⁻¹ doxycycline for 24 hours and subsequently starved for 24 hours in serum-free media with 1ug ml⁻¹ doxycycline. To test the activation of PI3K, lysates were probed for phospho-Ser473-AKT; total AKT was used as a loading control, and FLAG was used to indicate induced ORF1 expression. (D) Cells described in (C) were induced to express flag-tagged ORF1 WT or flag-tagged ORF1ΔPDZ, and glucose consumption and lactate production rates were measured in full serum conditions. Error bars denote standard errors of the mean (n=3). * denotes p<0.05; ** denotes p<0.01.

Figure 3. E4ORF1 enhances MYC transcriptional activation of glycolytic genes

(A) Immunoblotting (IB) of nuclear and cytoplasmic fractions from HA-tagged E4ORF6-expressing MCF10A cells induced to express vector control, flag-tagged E4ORF1 (ORF1WT), or the flag-tagged E4ORF1 D68A point mutant (ORF1D68A). Lysates were clonally selected to have equal molar amounts of E4ORF1 proteins. (B) Immunoprecipitations (IP) from nuclear fractions of HA-tagged E4ORF6-expressing MCF10A cells expressing FLAG-GFP or induced to express flag-tagged ORF1WT or flag-tagged ORF1D68A. (C) Heatmap depicts the level of MYC binding to the indicated genes by ChIP-qPCR in HA-tagged E4ORF6-expressing MCF10A cells induced to express vector control, flag-tagged ORF1WT, or flag-tagged ORF1D68A using antibodies towards MYC, FLAG, or IgG as a negative control. RLP32 is used as a negative control locus for MYC. The ChIP values represent the amount of immunoprecipitated DNA in each sample relative to the total amount of input chromatin (equivalent to 1) and color-coded with a gradient from blue for the minimum through red for the maximum percent of input. Each square is an independent experiment with 3 replicates. Specific values for % input can be found in Figure S3E. (D) Relative glucose consumption and lactate production rates from HA-tagged E4ORF6-expressing MCF10A cells induced to express equal levels of ORF1WT or ORF1D68A compared to respective uninduced samples. (E) Relative mRNA levels of glycolytic genes from HA-tagged E4ORF6-expressing MCF10A cells induced to express empty vector, E4ORF1 WT or E4ORF1 D68A. (F) Heatmap representing MYC binding to indicated gene targets by ChIP-qPCR in NHBE cells infected with AD WT or AD E4ORF1D68A for 4 hours at equal titers. Specific values for % input can be found in Figure S3F. (G) Relative mRNA levels of glycolytic genes from NHBE cells infected with AD WT or AD D68A for 4 hours at equal titers. For (D), (E) and (G), error bars denote standard errors of the mean (n=3). * denotes p<0.05; ** denotes p<0.01.

Figure 4. E4ORF1 activation of MYC promotes anabolic glucose metabolism and optimal adenovirus replication

(A) AD WT-infected NHBE cells stably expressing a MYC-directed shRNA exhibit less yellowing of the media pH indicator dye phenol red than AD WT-infected NHBE cells stably expressing a scrambled shRNA control. (B) Glucose consumption and lactate production rates from NHBE cells expressing scrambled shRNA (dark grey) or MYC shRNA (light grey) 24 hours post infection with AD WT versus mock infection. (C) Progeny virus was harvested from AD WT-infected NHBE cells expressing scrambled shRNA or MYC shRNA, and adenovirus titers were determined using serial plaque assays on HeLa monolayers. (D) AD D68A-infected NHBE cells exhibit less yellowing of the media pH indicator dye phenol red than AD WT-infected cells. (E) Glucose consumption and lactate production rates from NHBE cells mock infected or infected with AD WT or AD D68A. (F) Progeny virus from AD WT- and AD D68A-infected NHBE cells was harvested and adenovirus titers were determined as in (C). (G) NHBE cells mock infected or infected with AD WT or AD D68A were cultured in medium with [1,2-¹³C]-glucose for 24 hours. Extracted metabolites were analyzed by LC-MS/MS, and the percentage of ¹³C-labeled nucleotides was compared. (H) Relative mRNA levels of pentose phosphate pathway genes from NHBE cells infected with AD WT or AD D68A for 12 hours at equal titer. For (B), (C), (E), (F), (G), and (H), error bars denote standard errors of the mean (n=3). * denotes p<0.05; ** denotes p<0.01 (I) Schematic representation of the shift in glucose metabolism upon infection with adenovirus. Upon infection with adenovirus, E4ORF1 binds to MYC and activates MYC-dependent transcription of metabolic genes to increase glycolytic flux and potential shunting of glucose metabolites into the non-oxidative arm of the PPP (to facilitate ribose biosynthesis for viral RNA production and DNA replication).