Employing distinct genetic and pharmacological approaches
Employing distinct genetic and pharmacological approaches, Divakaruni et al. now demonstrate that macrophages still acquire their IL-4-elicited phenotype in the absence of CPT1a (considered the main isoform in leukocytes) or CPT2, and highlight that the increased LC-FAO in M(IL-4) cells is less crucial than previously considered. They also observe that a high concentration of etomoxir (200 μM) has distinct off-target effects, including suppression of OXPHOS through inhibition of electron transport chain complex I and by blocking rvx nucleotide translocase (ANT). In contrast to previous studies that applied the ATP synthase inhibitor oligomycin to block OXPHOS (Tan et al., 2015, Van den Bossche et al., 2016, Vats et al., 2006), the current paper surprisingly observed that OXPHOS is dispensable for M(IL-4) activation. Instead, depletion of the intracellular CoA pool in the presence of high etomoxir concentrations was identified as the reason for suppressed M(IL-4) activation, revealing CoA homeostasis as a new regulator of macrophage responses. This is an area ripe for further investigations, and the main open question arising from those findings is why CoA is critical for M(IL-4) cells. The authors speculate that sequestration of CoA in etomoxiryl-CoA (the active form of etomoxir) may limit the availability of acetyl-CoA, which can alter the epigenetic landscape of the cell. Interestingly, succinyl-CoA and malonyl-CoA appear to be affected and, given their crucial role in regulating macrophage responses, future research should explore the mechanisms by which CoA may regulate macrophage fates. Another key question to be addressed is whether the effects of etomoxir on CoA homeostasis are specific for M(IL-4) cells, or whether other macrophage responses or T cells are similarly affected. Could disrupted CoA levels in inflammatory macrophages be another explanation for why they cannot be reprogrammed into M(IL-4) macrophages, as reported earlier (Van den Bossche et al., 2016)? As noted by the authors, a limitation of this study is that alternative activation of CPT1a-deficient macrophages was only assessed in vitro. An important question is how macrophage-specific CPT1a-deficient mice behave in vivo in situations like asthma, helminth infections, wound healing, and tumors, in which alternatively activated macrophages are known to play crucial roles, and where increased LC-FAO may provide a bioenergetic advantage. In T cells, the use of etomoxir (200 μM) and distinct acute genetic approaches (Cpt1a short hairpin RNA [shRNA] and overexpressing constructs) has revealed that increased LC-FAO in IL-15-induced CD8+ Tmem cells supports their enhanced spare respiratory capacity (SRC), and that CPT1a promotes Tmem development in vivo (van der Windt et al., 2012). Next, O’Sullivan et al. (2014) showed that Tmem cells do not take up many LC-FAs but use glucose to generate de novo LC-FAs and subsequently oxidize those. Pan et al. (2017) confirmed the relatively low LC-FA uptake by Tmem cells, and applied Cpt1a small interfering RNAs (siRNAs) to elegantly demonstrate that in contrast to conventional Tmem, tissue-resident Tmem do require exogenous lipid uptake to support LC-FAO for their survival in vivo. Raud et al. now explore the importance of LC-FAO in T cells using a Cre-induced genetic deletion of Cpt1a and investigate potential off-target effects of etomoxir in T cells. Recently, O’Connor et al. (2018) described that etomoxir concentrations >5 μM have CPT1-independent effects on proliferating human Teff cells, and that 50 μM etomoxir induces mitochondrial swelling and increases reactive oxygen species (ROS) production. Raud et al. observed that etomoxir at concentrations >100 μM inhibited proliferation and differentiation, reduced OXPHOS, and increased mitochondrial membrane potential in Treg and Tmem cells. While the authors suggest that etomoxir could inhibit ANT in T cells as it does in macrophages, further research is needed to validate this mechanism. Raud et al. found that CPT1a is not required for Treg homeostasis, differentiation, and suppressive functions. They also found no impairment in Tmem development in vivo, and correspondingly no reduction in SRC in CPT1a-deficient Tmem. However, it should be noted that the IL-15-induced Tmem cells that were assessed here did not show the well-documented increased SRC compared to IL-2-induced Teff cells (O’Sullivan et al., 2014, Pan et al., 2017, van der Windt et al., 2012), which complicates interpretation of these data. Importantly, their human data showed no significant deficiencies in Tmem populations in patients with inherited deficiencies in long- and medium-chain FAO, which nicely fits with the mouse CPT1a-deficient T cell data, indicating similar pathways can be engaged to get around the loss of LC-FAO in T cells in mouse and human.