Cytomegalovirus-induced inactivation of TSC2 disrupts the coupling of fatty acid biosynthesis to glucose availability resulting in a vulnerability to glucose starvation.

IMPORTANCE: Viruses modulate host cell metabolism to support the mass production of viral progeny. For human cytomegalovirus, we find that the viral UL38 protein is critical for driving these pro-viral metabolic changes. However, our results indicate that these changes come at a cost, as UL38 induces an anabolic rigidity that leads to a metabolic vulnerability. We find that UL38 decouples the link between glucose availability and fatty acid biosynthetic activity. Normal cells respond to glucose limitation by down-regulating fatty acid biosynthesis. Expression of UL38 results in the inability to modulate fatty acid biosynthesis in response to glucose limitation, which results in cell death. We find this vulnerability in the context of viral infection, but this linkage between fatty acid biosynthesis, glucose availability, and cell death could have broader implications in other contexts or pathologies that rely on glycolytic remodeling, for example, oncogenesis.

Cytomegalovirus-induced inactivation of TSC2 disrupts the coupling of fatty acid biosynthesis to glucose availability resulting in a vulnerability to glucose limitation.

Human cytomegalovirus (HCMV) modulates cellular metabolism to support productive infection, and the HCMV UL38 protein drives many aspects of this HCMV-induced metabolic program. However, it remains to be determined whether virally-induced metabolic alterations might induce novel therapeutic vulnerabilities in virally infected cells. Here, we explore how HCMV infection and the UL38 protein modulate cellular metabolism and how these changes alter the response to nutrient limitation. We find that expression of UL38, either in the context of HCMV infection or in isolation, sensitizes cells to glucose limitation resulting in cell death. This sensitivity is mediated through UL38's inactivation of the TSC complex subunit 2 (TSC2) protein, a central metabolic regulator that possesses tumor-suppressive properties. Further, expression of UL38 or the inactivation of TSC2 results in anabolic rigidity in that the resulting increased levels of fatty acid biosynthesis are insensitive to glucose limitation. This failure to regulate fatty acid biosynthesis in response to glucose availability sensitizes cells to glucose limitation, resulting in cell death unless fatty acid biosynthesis is inhibited. These experiments identify a regulatory circuit between glycolysis and fatty acid biosynthesis that is critical for cell survival upon glucose limitation and highlight a metabolic vulnerability associated with viral infection and the inactivation of normal metabolic regulatory controls.

Heat Capacity of DPPC/Cholesterol Mixtures: Comparison of Single Bilayers with Multibilayers and Simulations.

The excess heat capacity (Δ C p) of mixtures of dipalmitoylphosphatidylcholine (DPPC) and cholesterol (Chol) is examined in detail in large unilamellar vesicles (LUVs), both experimentally, using differential scanning calorimetry (DSC), and theoretically, using a three-state Ising model. The model postulates that DPPC can access three conformational states: gel, liquid-disordered (Ld), and liquid-ordered (Lo). The Lo state, however, is only available if coupled with interaction with an adjacent Chol. Δ C p was calculated using Monte Carlo simulations on a lattice and compared to experiment. The DSC results in LUVs are compared with literature data on multilamellar vesicles (MLVs). The enthalpy change of the complete phase transition from gel to Ld is identical in LUVs and MLVs, and the melting temperatures ( Tm) are similar. However, the DSC curves in LUVs are significantly broader, and the maxima of Δ C p are accordingly smaller. The parameters in the Ising model were chosen to match the DSC curves in LUVs and the nearest-neighbor recognition (NNR) data. The model reproduces the NNR data very well. It also reproduces the phase transition in DPPC, the freezing point depression induced by Chol, and the broad component of Δ C p in DPPC/Chol LUVs. However, there is a sharp component, between 5 and 15 mol % Chol, that the model does not reproduce. The broad component of Δ C p becomes dominant as Chol concentration increases, indicating that it involves melting of the Lo phase. Because the simulations reproduce this component, the conclusions regarding the nature of the phase transition at high Chol concentrations and the structure of the Lo phase are important: there is no true phase separation in DPPC/Chol LUVs. There are large domains of gel and Lo phase coexisting below Tm of DPPC, but above Tm the three states of DPPC are mixed with Chol, although clusters persist.