The hypoxia-sensor carbonic anhydrase IX affects macrophage metabolism, but is not a suitable biomarker for human cardiovascular disease

Abstract

Hypoxia is prevalent in atherosclerotic plaques, promoting plaque aggravation and subsequent cardiovascular disease (CVD). Transmembrane protein carbonic anhydrase IX (CAIX) is hypoxia-induced and can be shed into the circulation as soluble CAIX (sCAIX). As plaque macrophages are hypoxic, we hypothesized a role for CAIX in macrophage function, and as biomarker of hypoxic plaque burden and CVD. As tumor patients with probable CVD are treated with CAIX inhibitors, this study will shed light on their safety profile. CAIX co-localized with macrophages (CD68) and hypoxia (pimonidazole), and correlated with lipid core size and pro-inflammatory iNOS+ macrophages in unstable human carotid artery plaques. Although elevated pH and reduced lactate levels in culture medium of CAIX knock-out (CAIXko) macrophages confirmed its role as pH-regulator, only spare respiratory capacity of CAIXko macrophages was reduced. Proliferation, apoptosis, lipid uptake and expression of pro- and anti-inflammatory genes were not altered. Plasma sCAIX levels and plaque-resident CAIX were below the detection threshold in 50 and 90% of asymptomatic and symptomatic cases, respectively, while detectable levels did not associate with primary or secondary events, or intraplaque hemorrhage. Initial findings show that CAIX deficiency interferes with macrophage metabolism. Despite a correlation with inflammatory macrophages, plaque-resident and sCAIX expression levels are too low to serve as biomarkers of future CVD.

Figure 1

CAIX was present in human atherosclerotic plaque and co-localized with CD68⁺ cells. (A) Microphotograph of hematoxylin and eosin (H&E) stained human unstable carotid artery plaque. Red square represents region that was magnified in all stainings. (B) Double staining of CAIX (blue) and CD68 (red) in unstable human atherosclerotic plaque, magnification of red square is depicted in (C). Arrows indicate double positive cells. (D) Unstable human atherosclerotic plaque of patient injected with pimonidazole to detect hypoxia (brown) or (E) iNOS (red), and CAIX (blue), arrows indicate double-positive areas accompanied by an additional zoomed in area, (F) Heatmap of Pearson correlations of CAIX mRNA in unstable human plaque segments with plaque traits, determined on adjacent histology slides. All plaque sections originate from the MaasHPS cohort. Arg1; arginase 1. αSMA; alpha smooth muscle cell actin. D2-40; podoplanin. iNOS; inducible nitric oxide synthase. IPH; intra-plaque hemorrhage. MVD, microvessel density NA; not applicable. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 2

CAIXko reduced spare respiratory capacity in BMDMs. (A) pH and (B) lactate measurement of BMDM culture medium (n = 4) after 72 h of incubation in either normoxia or hypoxia (1% O₂). (C) Representative graph of Seahorse mitochondrial stress assay of BMDMs (n = 5/group, 1 experiments). Analysis was performed on data corrected for protein content in each well and with baseline measurement 4. (D) BMDM spare respiratory capacity displayed relatively to WT. (E) BMDM proliferation measured by EdU incorporation (n = 4) after 24 h incubation in normoxia or hypoxia (1% O₂). (F) Mean BMDM proliferation measured by real-time impedance (n = 5) over 72 h. Slope represents the increment of impedance over time. (G) Real time quantitative PCR of pro-inflammatory genes inducible nitric oxide synthase (iNOS), (H) tumor necrosis factor (TNF), (I) Tumor necrosis factor-induced protein 3 (A20) (J) interleukin 6 (IL6) and anti-inflammatory genes (K) mannose receptor (CD206) and (L) interleukin 10 (IL10) in non-stimulated (medium) and LPS + IFNγ stimulated BMDMs. All values are relative to WT unstimulated BMDMs of respective target gene. *Indicates a significant difference of stimulated WT compared to unstimulated WT BMDM. #Indicates significant difference of stimulated CAIXko compared to unstimulated CAIXko BMDMs. No difference was observed between WT and CIX-ko BMDM (white and black bars respectively). All results show mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3

CAIXko did not alter BMDM apoptosis or lipid uptake. Apoptosis was induced by incubating cells with 50 µM 7-ketocholesterol for 24 h in all experiments. (A) Representative photographs of apoptotic cells (annexin positive) in WT and CAIXko BMDMs in normoxia and hypoxia (1% O₂). (B) BMDM (n = 3) apoptosis in normoxia or (C) hypoxia (1% O₂, 24 h). (D) Representative photographs of BMDMs after lipid uptake (Topfluor positive) in WT and CAIXko BMDMs in normoxia and hypoxia (1% O₂). (E) BMDM lipid uptake in normoxia (n = 3) or (F) hypoxia. All results show mean ± SEM.

Figure 4

sCAIX and plaque resident CAIX was not changed in patients with events or plaque type. (A) sCAIX values in human carotid plaques with or without IPH from the MaasHPS cohort. (B) Representative cropped western blot and (C) quantification of CAIX protein in human carotid plaque lysates with (n = 27) and without a second event (n = 35) derived from Athero-Express biobank study. CAIX protein intensity is corrected for protein levels by ponceau S, and square root (sqrt) normalized. (D) Representative cropped western blot and (E) quantification of CAIX protein in atheromatous (26) and fibrous (n = 33) human carotid plaque lysates derived from Athero-Express biobank study. Arrow heads indicate two bands known to correspond to CAIX, 2nd, second. Full length western blots can be found in Supplemental Figures S3 and S4, (F) CAIX mRNA as derived from microarray analysis in human carotid plaques with or without IPH derived from the MaasHPS biobank.