Background: Major depressive disorder (MDD) is a debilitating mental disease, but its underlying molecular mechanisms remain obscure. Gut microbiome can modulate brain function and behaviors through the microbiota-gut–brain (MGB) axis in depression. Our previously established non-human primate model of naturally occurring depression-like behaviors, which is characterized by MGB axis disturbances, can be used to interrogate how a disturbed gut ecosystem may modulate the MDD onset. To better clarify the molecular interrelationships and downstream functional consequences in the MGB axis on MDD pathology, here, gut metagenomics were used to characterize how gut virus and bacterial species, and associated metabolites, change in depressive monkey model.Results: We identified a panel of 33 gut virus and 14 bacterial species that could discriminate the depression-like (DL) from control M. fascicularis. In addition, using lipidomic analyses of central and peripheral samples obtained from these animals, we found that the DL macaque were characterized by alterations in the relative abundance, carbon-chain length, and unsaturation degree of 1,2-diacylglyceride (DG) in the prefrontal cortex (PFC), in a brain region-specific manner. In addition, lipid-reaction analysis identified more active and inactive lipid pathways in PFC than in amygdala or hippocampus, with DG being a key nodal player in these lipid pathways. Significantly, co-occurrence network analysis showed that altered gut viral and bacterial species, and their interaction may be relevant to the onset of negative emotions behaviors by modulating the DG levels in PFC in the depressive macaque. Conclusions: Our findings suggest that altered gut virus and bacteria as well as DG levels and structure in the PFC are hallmarks of the DL macaque, thus providing a new framework for understanding the gut microbiome's role in depression.