Macrophage metabotypes in the hypoxic inflammatory environment assessed using metabolomic profiling

Background: Macrophages are abundant in the synovium of RA and are the major source of many inflammatory cytokines, including TNFα. Disease severity correlates strongly with both synovial macrophage abundance and activation status but inversely with oxygen levels. Given the importance of oxygen to cell metabolism, differentiation and function, we hypothesized that the unique metabolic and oxygen environment of the joint may drive the differentiation of aggressive macrophages.

Methods: Human monocytes were differentiated into macrophage subsets (M1 and M2c) under GM-CSF and M-CSF polarizing conditions and under a variety of oxygen conditions, and stimulated with lipopolysaccharide (LPS). Cell metabolites were extracted for metabolomic analysis using NMR spectroscopy and secreted cytokines assessed. NMR spectra were analysed using partial least squares discriminant analysis (PLS-DA) to build models describing differences in metabolites, and partial least squares regression (PLS-R) to correlate these to cytokine production. Metabolites differentiating between experimental groups and pathways were identified and analysed using bespoke metabolic analysis tools.

Results: Macrophage subtypes exhibited unique metabolic patterns that, while shifting during differentiation, remained distinct throughout stimulation and subsequent activation with LPS. Differentiation of monocytes to both M1 and M2c macrophages increased production of myo-inositol, a marker of calcium signaling, and the proteinogenic amino acid serine. These correlated strongly with LPS-induced IL-6 production. M2c macrophages showed increased arginine production which correlated with IL-10 production, a key M2c anti-inflammatory cytokine. M1 macrophages exhibited higher glycolytic flux and a reduction in citric acid cycle metabolites.

Differentiation under hypoxia gave an altered metabolic phenotype, with a reduced citric acid cycle, and increased glycolytic flux. However, both M1 and M2c subsets remained metabolically distinct. Strikingly, differentiation of both M1 and M2c macrophages under reperfusion conditions resulted in an activated metabolic profile, indistinguishable from that induced by LPS, but interestingly up-regulation of IL-6 was not seen.

Conclusions: Classically-activated M1 macrophages and regulatory M2c macrophages each possess a unique metabotype that, while altered by environmental stressors, remains unique to the cell. This stability may depend on tight regulation at the gene and protein level, and we suggest that this is closely coupled to the macrophage differentiation programme. The increased arginine in M2c, for example, suggests a down-regulation of the arginase pathway in this resolution-associated subset. Therefore, while a capacity to adapt to the environment clearly remains, differentiation constrains this. We propose that this may impair subset survival or function in abnormal environments such as the rheumatoid synovium.

Disclosures: The authors have declared no conflicts of interest.


Related posts