Enhanced Free Fatty Acid Uptake Via CD36 Promotes a Metabolic Switch to B-Oxidation within Hematopoietic Stem Cells in Response to Acute Infection

Acute infection is known to promote rapid differentiation of hematopoietic stem cells (HSC) and expansion of leukocytes. Metabolic changes in the HSC underpin the mammalian response to pathogenic stimuli however, knowledge of how this occurs is not fully understood. Here we investigate the immunometabolic changes which facilitate HSC expansion in response to acute infection.

We created a novel in vivo transplant model to understand if the immune response to infection involves the acquisition of free fatty acids (FFA) by HSCs in the bone marrow (BM). We transduced CD45.1 lineage negative, c-kit positive cells with firefly luciferase and transplanted into CD45.2 animals. Post engraftment animals were treated with lipopolysaccharide (LPS). Live animal imaging confirmed activation of luciferase in the BM, demonstrating in vivo, uptake of long chain FFA by haematopoietic cells in response to LPS. To confirm the specific haematopoietic stem and/or progenitor cells (HSPC) with increased lipids during infection, mice were treated with S.typhimurium for 72h then sacrificed Analysis of LSK, MPP and HSC populations (LN-, CD117+, Sca1+, CD48-, CD150+ and CD34+) showed increased intracellular BODIPY 493/503 neutral lipid dye and uptake of Bodipy FL-C12 (FFA linked to bodipy) at 72 h compared to control non-infected animals. Together, these experiments show that HSC, multipotent progenitor (MPP) and LSK (LN-, CD117+, Sca1+) cells all acquire FFA in response to bacterial infection.

Seahorse mitochondrial stress test confirmed increased oxygen consumption levels in HSC from LPS (16 hours) and S.typhimurium (72 hours) treated C57BL/6. LSK metabolic activity for different substrates was assessed LSKs from LPS treated animals had an increased dependency on β-oxidation when compared to control cells. Moreover, the β-oxidation inhibitor etomoxir blocks LSK basal metabolism and reduced HSC expansion as measured by Ki-67 staining in LPS and S.typhimurium treated animals. To understand the importance of β-oxidation in HSC expansion in response to infection we first measured CPT1A expression in HSC. RNA analysis confirmed CPT1A was increased in sorted HSCs in response to LPS. Next we transplanted WT or CPT1A KD cells into WT animals, and treated with LPS. HSC expansion as measured by Ki-67 staining was inhibited in the CPT1A KD cells in response to LPS, thus confirming the importance of β-oxidation based metabolism in HSC expansion.

To determine how FFA are transported into the HSC we analysed the expression of known genes to be involved in lipid uptake. HSC sorted from LPS and S.typhimurium treated animals had increased CD36 mRNA expression compared to control. C57BL/6 mice were pre-treated with the CD36 inhibitor sulfosuccinimidyl oleate (SSO) before injection with LPS. HSCs had had less lipids, lower maximal respiration and reduced cycling when compared to animals treated with LPS alone. In a similar way CD36 KO animals were treated with LPS for 16 h, no increased FFA uptake or increased lipid content or HSC cycling was observed. Seahorse mitochondrial stress test confirmed CD36^-/- LSK have increased basal ECAR but no change in basal OCR in response to S.typhimurium infection. Moreover, CD36^-/- LSK have reduced dependency on β-oxidation.

To determine if the uptake of FFA response is specific to HSC we transplanted CD36^+/+ (CD45.1) LK into CD36^-/- (CD45.2) animals. We found that CD36 expression was elevated in CD36^+/+ HSC transplanted into CD36^-/- treated with LPS, normal uptake of FFA, an increase in lipid content and cycling of HSCs. In addition, transplanted WT CD36^+/+ (CD45.1) LSK into CD36^-/- (CD45.2) animals reversed the metabolic phenotype observed in CD36^-/- in response to LPS. We transplanted WT^(CD36+/+) or CD36KO (WT^(CD36-/-)) LSK into WT animals infected with S. typhimurium for 4 days. WT^(CD36-/-) transplanted animals showed enhanced mortality, increased weight loss and increased liver injury. We also observed less HSC cycling and a reduced lipid content in HSC from WT^(CD36-/-) compared to WT^(CD36+/+).

These findings expand the mechanistic understanding of the interplay between HSC and the bone marrow microenvironment, and in doing so explains how immunometabolic reprogramming of HSCs during infection supports the metabolic demands of HSCs in response to a pathogenic challenge without which results in increased susceptibility to infection.