Purpose: Foam cell macrophage (FCM) formation is an early event in atherosclerosis that could contribute to fibrous cap development, although it has also been strongly implicated in cap rupture and thrombus formation. To understand this apparent paradox we compared the transcriptome of FCM and non-foamy macrophages (NFM) produced in vivo using Illumina bead chips.
Methods: Sponges were implanted into fat-fed ApoE null or chow-fed C57Bl6 mice to produce FCM or NFM, respectively. Cells were purified based on their buoyant density and/or differential adherence. Gene expression was examined using Ingenuity Pathway Analysis, GO annotation and clustering (DAVID Bioinformatics Resources). Data was confirmed using q-PCR and IHC.
Results: The functions enriched/upregulated in FCM included connective tissue development, function and disorders as well as immune response, cell signalling. The primary canonical pathway upregulated in the FCM was liver X receptor/retinoid X receptor (LXR/RXR) activation, consistent with a recent study using peritoneal FCM from LDLr knockout mice (Spann et al Cell 2012;151:138). We confirmed that FCM had more mRNA for LXR, RXR and target genes using qRT-PCR, and used IHC to identify LXRα containing FCM in the sponges. GO identified hepatic fibrosis pathways as highly upregulated in the FCM and Ingenuity suggested a role for transforming growth factor β1 (TGFβ1). For example, there was an increase in mRNA for many extracellular matrix proteins (including collagens 1, 4, 6, 8, biglycan and decorin), connective tissue growth factor (CTGF), BMP-1 and Fos/FosB/Jun/JunB. FCM also had more mRNA for matrix proteases (including cathepsins C and E and matrix metalloproteinases 2 and 23), without a corresponding difference in their inhibitors. Using IHC we established that many cells within the sponges and brachiocephalic arteries from fat-fed ApoE mice contained CTGF. As TGFβ1-induced CTGF expression is known to be regulated via SMADs, we examined these tissues for the presence of phosphoSMAD2 (pSMAD2) by IHC and found that it was present in the cytoplasm and nucleus of sponge FCM and plaque cells.
Conclusions: Our data confirm the paradox introduced by Spann and colleagues showing that FCMs are anti-inflammatory via LXR activation. We have extended this by showing that FCM can contribute to fibrosis and matrix deposition, possibly due to exposure to TGFβ1. Clearly additional factors, including mediators of innate and acquired immunity, must be responsible for converting FCM from an anti-inflammatory/pro-fibrotic phenotype into cells capable of mediating plaque rupture.