Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Jul 1.
Published in final edited form as: Arterioscler Thromb Vasc Biol. 2017 May 4;37(7):e82–e91. doi: 10.1161/ATVBAHA.117.309522

The DBA/2J haplotype on distal chromosome 2 reduces Mertk expression, restricts efferocytosis and increases susceptibility to atherosclerosis

Yukako Kayashima 1, Natalia Makhanova 1, Nobuyo Maeda 1
PMCID: PMC5497749  NIHMSID: NIHMS871052  PMID: 28473436

Abstract

Objective

Arch atherosclerosis 4 (Aath4) is a quantitative trait locus for atherosclerotic plaque formation in the inner curve of the aortic arch previously identified in an F2 cross of Apoe−/− mice on DBA/2J and 129S6 backgrounds. Mertk, coding for a ligand-activated transmembrane tyrosine kinase, is a candidate gene within the same chromosomal region. Our objective was to determine whether strain differences in Mertk influence plaque formation.

Approach and Results

To dissect the strain effects of Mertk on atherosclerosis, we first established a congenic mouse line (Aath4aDBA/DBA) in which a 5′ region of Aath4 of DBA/2J, including Mertk, was backcrossed onto a 129S6-Apoe−/− background. The resulting Aath4aDBA/DBA male mice developed significantly larger plaques compared with control mice (Aath4a129/129), proving that the DBA/2J allele of Aath4a is pro-atherogenic. Thioglycollate-elicited peritoneal macrophages from Aath4aDBA/DBA mice express less than 50% of Mertk mRNA and cell-surface MERTK protein compared to those from the control mice. Moreover, both large and small peritoneal Aath4aDBA/DBA macrophages showed reduced phagocytosis of apoptotic cells. When Mertk cDNAs from 129S6 and DBA/2J mice were overexpressed in HEK293T cells, phagocytosis of apoptotic cells was equally enhanced in direct proportion to Mertk levels, indicating that phagocytosis is modulated by the amount of MERTK, but that it is not affected by MERTK amino acid differences between 129S6 and DBA/2J.

Conclusions

Reduced transcription of Mertk, rather than differences in MERTK protein structure, determines the reduced efficiency of apoptotic cell clearance in the Aath4aDBA/DBA mice, which in turn contributes to their increased susceptibility to atherosclerosis.

Keywords: mice, aorta, atherosclerosis, macrophage, phagocytosis

Subject codes: Atherosclerosis, Genetically Altered and Transgenic Models

Introduction

Atherosclerosis is a complex multifactorial disease, and individual susceptibility to plaque development is influenced by many genetic factors. We previously showed that plaque size in mice is dependent on the strain and vascular location in the early stage of atherosclerosis.1–4 At the aortic root, apolipoprotein E-deficient (Apoe−/−) mice on a DBA2/J background (DBA/2J-Apoe−/−) develop larger plaques than those on a C57BL/6J background (C57BL/6J-Apoe−/−), while Apoe−/− mice on a 129S6 background (129S6-Apoe−/−) are resistant to plaque development compared to DBA/2J-Apoe−/− or C57BL/6J-Apoe−/− mice. In the aortic arch area, on the other hand, 129S6-Apoe−/− and DBA/2J-Apoe−/− mice show larger lesion size than C57BL/6J-Apoe−/− mice. These observations clearly indicate that the location specificity of plaque development is genetically controlled.

Our QTL analysis using an intercross between DBA/2J-Apoe−/− and 129S6-Apoe−/− revealed Aath4, an atherosclerosis QTL for the aortic arch area on the distal part of chromosome 2 (Peak: 137 Mb, Confidence Interval: 123−148 Mb).4 The DBA/2J allele of Aath4 confers susceptibility to atherosclerosis, while the 129S6 allele confers resistance. Aath4 was not detected in a C57BL/6J-Apoe−/− × 129S6-Apoe−/− cross,2 indicating that Aath4 sequences unique to DBA/2J are responsible for the different phenotypes. Many candidate genes are present in the chromosomal region in which Aath4 resides, including several phagocytosis-related genes.

One of the candidates is c-mer proto-oncogene tyrosine kinase (Mertk) located at 128.5 Mb, which encodes a member of the TAM (Tyro3, Axl and Mer) receptor tyrosine kinase family. MERTK is primarily expressed in monocytes as well as in epithelial and reproductive tissues,5 and is involved in phagocytosis of apoptotic cells.6, 7 It has two Ig-like domains and two fibronectin type III domains in the extracellular region, and binds to apoptotic cells via bridging molecules such as GAS6 and protein S.5, 8, 9 The binding promotes phosphorylation of the tyrosine kinase domain located within the intracellular region of MERTK; this phosphorylation leads to activation of downstream signaling and induces structural changes in cytoskeletons, enabling the cell to engulf its target cells.10 Macrophages from mice with an inactivated Mertk kinase domain (Mertk−/−) are deficient in the clearance of apoptotic cells (efferocytosis).6 The Mertk−/− mice are viable, but show retinal degeneration, due to the failure of the retinal pigment epithelium to engulf the outer segments of photoreceptors.7, 11 MERTK also plays a pivotal role in atherosclerotic plaque development via its effects on efferocytosis: Mertk deficiency in Apoe−/− mice promotes accumulation of apoptotic cells and expansion of necrotic cores within plaques12; and low-density lipoprotein receptor-deficient (Ldlr−/−) mice with Mertk−/− bone marrow show accumulation of apoptotic cells and accelerated atherosclerosis.13

In the present study, we have generated and studied a mouse line, Aath4aDBA/DBA, in which a 5′ portion of the Aath4 from DBA/2J has been transferred onto a 129S6-Apoe−/− background. We show that these mice have elevated plaque susceptibility and reduced efferocytosis by macrophages. We also demonstrate that decreased transcription of Mertk, not DBA-unique amino acid alterations in MERTK, determines the limited efferocytosis that occurs in Aath4aDBA/DBA mice.

Material and Methods

Materials and Methods are available in the online-only Data Supplement.

Results

Generation of Aath4aDBA/DBA mice

Aath4aDBA/DBA congenic mice were constructed by backcrossing DBA/2J-Apoe−/− mice to the 129S6-Apoe−/− strain for more than seven generations. The backcrossed genomic segments in chromosome 2 (117 Mb to 137 Mb) are shown in Figure 1A. The DBA allele of SNP rs27446327 was used as a marker for Chr 2: 128 Mb. Heterozygous mice were mated to generate homozygotes. SNPs rs30884601, rs36935260, rs48427643, and rs30243498 were also typed at early generations to ensure that DBA alleles at other atherosclerosis QTLs (Chr 1: 155 Mb, Chr 1: 165 Mb, Chr 2: 143 Mb, and Chr 10: 86 Mb) were eliminated.

Figure 1. Aath4aDBA/DBA males develop larger atherosclerotic plaques.

Figure 1

Open in a new tab

A. Haplotype maps of Chr 2 taken from the Mouse Phylogeny Viewer (http://msub.csbio.unc.edu/#viewer). 129S6 sequence is colored in green (top); Genomic regions where DBA/2J share the same sequences as 129S6 are shown in green, and DBA/2J-specific sequences are highlighted with purple (middle); C57BL/6J unique regions are colored in peach (bottom). The backcrossed region in the congenic strain Aath4aDBA/DBA (red bar), position of Aath4 (black bar), and the location of some of the candidate genes (arrows) are indicated. B. Comparison of plaque size at the aortic arch (left) and root (right) between the control Aath4a129/129 (filled bars) and Aath4aDBA/DBA (open bars) mice at 5 months old. Plaque size is indicated in square root (sqrt) of area in μm2. Data are shown as the mean ± SD. Numbers of mice are indicated in the bars.

Aath4aDBA/DBA male mice develop larger plaques with increased calcium deposits

As summarized in Table 1, Aath4a129/129 control mice and Aath4aDBA/DBA mice showed no significant differences in body weight. In males, plasma cholesterol levels were almost the same between controls and Aath4aDBA/DBA, whereas triglycerides were significantly higher in Aath4aDBA/DBA (P = 0.002). In female Aath4aDBA/DBA mice, plasma total cholesterol was lower (P = 0.010) and HDL-cholesterol was higher (P = 0.016) compared to Aath4a129/129 controls, but triglycerides were not significantly different.

Table 1.

Body weights and plasma lipids in the control and Aath4aDBA/DBA mice.

Aath4a129/129 (n) Aath4aDBA/DBA (n)
Body Weight, g Male 28.59 ± 0.37 (33) 28.08 ± 0.42 (23)
Female 21.30 ± 0.38 (27) 21.37 ± 0.37 (20)
T-Chol, mg/dL Male 680.9 ± 23.0 (37) 738.2 ± 29.2 (23)
Female 572.3 ± 27.5 (26) 482.2 ± 31.3* (20)
HDL-C, mg/dL Male 79.4 ± 5.0 (33) 77.1 ± 9.5 (9)
Female 45.1 ± 3.7 (26) 60.2 ± 4.2* (20)
TG, mg/dL Male 79 ± 9 (33) 125 ± 8† (23)
Female 48 ± 4 (25) 58 ± 5 (19)
Open in a new tab

Data are shown as the mean ± SE.

*

P < 0.05,

†

P < 0.01 vs. Aath4a129/129.

At 5 months of age, Aath4aDBA/DBA male mice developed approximately 50% larger plaques at the inner curve of the aortic arch compared with control mice (P < 0.05) (Figure 1B, left). Plaque size at the aortic root was also significantly larger in Aath4aDBA/DBA males (P < 0.01) (Figure 1B, right). In contrast, females did not show significant differences in the size of either aortic arch or aortic root plaques. Male Aath4aDBA/DBA mice also developed significantly larger brachiocephalic artery lesions compared with the control (P < 0.05), but did not differ in the left common carotid and left subclavian arteries. In females, no significant difference was observed in any of these branches between Aath4aDBA/DBA mice and controls (Figure I in the online-only Data Supplement).

The plaques developing in the aortic roots of Aath4aDBA/DBA and control mice ranged from a monolayer of foam cells to complex lesions with inflammation and necrotic cores (Figure 2A and Figure II in the online-only Data Supplement). When plaques of similar size and locations were examined histologically, the early raised lesions of Aath4aDBA/DBA contained more scattered basophilic calcium deposits (Figure 2A and 2B) than the Aath4a129/129 control plaques. Multiple calcium deposits were detectable, often associated with the acellular necrotic cores, in the raised plaques in 28 of 36 Aath4aDBA/DBA mice compared to 6 of 28 Aath4a129/129 control mice. In more mature plaques, calcium deposits were detected equally in both Aath4aDBA/DBA mice and Aath4a129/129 controls, and they were located deeper in the plaques, near the internal lamina (Figure IIE and IIF in the online-only Data Supplement, arrowheads). Enhanced cell death and/or reduced clearance of dead cells could be the source of these calcium deposits. Although very few, TUNEL-positive nuclei were detected within the plaques of both Aath4a129/129 and Aath4aDBA/DBA mice (Figure III in the online-only Data Supplement).

Figure 2. Aath4aDBA/DBA mice develop plaques with increased calcium deposits, larger necrotic core and less collagen content.

Figure 2

Open in a new tab

A. Representative early but raised plaques at the aortic root of control Aath4a129/129 (129/129) and Aath4aDBA/DBA (DBA/DBA) mice at 5 months of age. The plaques with similar sizes stained with SudanIVB and hematoxylin are shown. Aath4aDBA/DBA mouse (right) contains extensive calcium deposits (arrowheads). B. Calcium deposits detected by von Kossa stain. Numerous deposits were observed in the early plaques of Aath4aDBA/DBA mice (right), whereas rarely seen in those of control Aath4a129/129 mice (left). Bar = 100 μm. C. Necrotic core size in the plaques of Aath4a129/129 (129/129) and Aath4aDBA/DBA (DBA/DBA) mice at 5 months old. Advanced lesions that are similar in total size were selected and necrotic area was measured in the selected lesions (n = 7). D. Fibrous cap thickness of advanced plaques. In the seven advanced lesions selected in C, fibrous cap thickness was measured at three sites per lesion and averaged. E. Collagen content in the advanced lesions detected by trichrome staining. The Aath4aDBA/DBA (DBA/DBA) lesion contains less collagen (blue). Bar = 100 μm.

Since average lesions were significantly larger in size and more advanced in Aath4aDBA/DBA mice than in Aath4a129/129, we selected seven raised lesions with similar size from Aath4a129/129 and Aath4aDBA/DBA mice to compare the components of the early to intermediate stage plaques. Necrotic area was approximately five times larger in Aath4aDBA/DBA than in control mice (Figure 2C). Fibrous cap thickness was variable and the difference was not statistically significant between the two groups (Figure 2D). However, collagen content in these lesions was less in the Aath4aDBA/DBA mice, suggesting that resolution of inflammation associated with the plaque development is likely delayed in Aath4aDBA/DBA (Figure 2E). General inflammation in these mice was low, judged by the low concentration of TGFβ1 and undetectable IL-10 in plasma. No significant differences in mRNA levels of IL-1β and TGFβ1 in the aorta were observed between control Aath4a129/129 and Aath4aDBA/DBA mice or in plasma TGFβ1 levels (Figure IVA and IVB in the online-only Data Supplement), probably because these mice were fed with normal chow.

Together these results indicate that the DBA/2J allele of the 5′ region of Aath4 enhances atherosclerosis in both the aortic arch and root. Increased cell death in the early stages of plaque development may be associated with this enhancement.

Reduced Mertk expression in Aath4aDBA/DBA macrophages

The Aath4a QTL includes Mertk, which is important for efferocytosis, the phagocytotic removal of apoptotic cells.4 MERTK is localized on cell surface where it is proteolytically cleaved by ADAM metallopeptidase domain 17 (ADAM17) to produce soluble MER (sMER) (see below in Figure 6A). sMER attenuates MERTK-triggered intracellular signaling by blocking bridging molecules.14 In thioglycollate-elicited peritoneal macrophages isolated from Aath4aDBA/DBA, Mertk mRNA levels were less than 50% of that in Aath4a129/129 control macrophages (Figure 3A). In parallel with the mRNA expression, the amount of MERTK protein in cultured macrophages was also reduced in Aath4aDBA/DBA (Figure 3B), as well as cell-surface expression of MERTK (Figure 3C).

Figure 6. Amino acid substitutions in DBA-MERTK do not affect phagocytosis of apoptotic cells.

Figure 6

Open in a new tab

A. Schematic structure of mouse MERTK indicating the nine amino acids which differ between 129S6 and DBA/2J proteins. Substitutions are shown as 129S6-amino acid position-DBA/2J. Ig-like, immunoglobulin-like domains; FN-III, fibronectin type-III domains; TM, transmembrane domain; TK, tyrosine kinase domain. B. Expression of MERTK in HEK293T cells and sMER in the conditioned medium detected by western blotting. UT, untransfected cells; EV, empty vector control; 129, 129S6-Mertk transfected cells; DBA, DBA/2J-Mertk transfected cells. C. Phagocytosis of apoptotic Jurkat T cells by Mertk-transfected HEK293 cells with or without 25 nM of protein S (n = 3). D. Phagocytosis of apoptotic Jurkat T cells by peritoneal macrophages isolated from Mertk+/+ and Mertk+/− mice. (n = 4). Data are shown as the mean ± SE. E. Correlation between the MERTK levels and phagocytosis in HEK293T cells transfected with varying amounts of Mertk cDNA.

Figure 3. Expression of MERTK is decreased in Aath4aDBA/DBA.

Figure 3

Open in a new tab

A. Mertk mRNA expression in thioglycollate-elicited peritoneal macrophages from control Aath4a129/129 (129/129) and Aath4aDBA/DBA (DBA/DBA) mice measured by quantitative RT-PCR (n = 4). B. MERTK protein expression in cultured peritoneal macrophages and soluble form of MERTK (sMER) in their medium detected by western blotting (left). Signal intensity ratios of sMER in the conditioned medium and cell-associated MERTK (right). C. Cell-surface expression of MERTK in peritoneal macrophages analyzed by flow cytometry. An overlay of representative histograms showing MERTK positive cells is displayed (left). Median fluorescence intensity (MFI) of MERTK in CD11b positive macrophages (right). D. sMER in the plasma detected by western blotting. Equal amounts of plasma (2.5 μl per lane) were loaded in each lane. Data are shown as the mean ± SE. Numbers in the bar diagrams indicate the number of animals analyzed.

sMER in the conditioned medium of Aath4aDBA/DBA macrophage and plasma sMER in Aath4aDBA/DBA mice were also reduced compared to those in controls, consistent with the reduction of protein amounts in macrophages (Figure 3B and 3D). MERTK protein was abundantly present within the atherosclerotic plaques of the Aath4a129/129 control mice, bordering the macrophage marker CD68 positive area, while the amount was significantly decreased in Aath4aDBA/DBA plaques (Figure V in the online-only Data Supplement).

Reduced phagocytosis in Aath4aDBA/DBA macrophages

MERTK is important for the normal execution of efferocytosis during atherosclerotic plaque development.12, 13 We therefore examined phagocytosis of apoptotic Jurkat T cells by peritoneal macrophages of Aath4aDBA/DBA mice and found that the uptake of apoptotic cells was reduced to approximately 40% of Aath4a129/129 controls (Figures 4A through 4C and Figure VI in the online-only Data Supplement).

Figure 4. Efferocytosis is reduced in Aath4aDBA/DBA macrophages.

Figure 4

Open in a new tab

A. Phagocytosis of apoptotic Jurkat T cells by peritoneal macrophages. Apoptotic Jurkat T cells labeled with pHrodo Red were added to macrophages isolated from control Aath4a129/129 (129/129) and Aath4aDBA/DBA (DBA/DBA) mice, and incubated for 1 hour at 37°C. Apoptosis of Jurkat T cells was confirmed as shown in Figure IIA in the online-only Data Supplement. B. Representative flow cytometry panels showing phagocytosis of apoptotic Jurkat T cells by peritoneal macrophages from control Aath4a129/129 (129/129) and Aath4aDBA/DBA (DBA/DBA) mice. Detailed gating strategy is shown in Figure IIB in the online-only Data Supplement. Apoptotic Jurkat T cells and macrophages were labeled with pHrodo Red and anti-CD11b-FITC, respectively. Gates indicate pHrodo Red positive macrophages phagocytosing apoptotic cells. C. An overlay of representative histograms showing pHrodo Red positive phagocytosing cells in the CD11b positive macrophages (left). Phagocytosis assessed by the median fluorescent intensity (MFI) of pHrodo Red (right, n = 4). Data are shown as the mean ± SE.

Ghosn et al. have reported that there are two types of peritoneal macrophages: large peritoneal macrophages (LPMs) and small peritoneal macrophages (SPMs).15, 16 LPMs predominate in the steady state peritoneal cavity but are decreased by inflammatory stimuli such as LPS and thioglycollate, which induce infiltration of SPMs into the peritoneal cavity from myeloid progenitor cells in the bone marrow. We found that 3 days after stimulation by thioglycollate, the percentage of LPM was significantly smaller in Aath4aDBA/DBA than in control mice (0.8% vs. 3.0% in Aath4aDBA/DBA and control mice, respectively; P < 0.001) (Figures 5A and 5B). Both SPMs and LPMs express MERTK, although in Aath4aDBA/DBA macrophages, MERTK was markedly reduced in both SPMs and LPMs (Figure 5C). Consistent with a previous report, phagocytosis was higher in LPMs than in SPMs,17 but importantly, both LPMs and SPMs showed similarly reduced phagocytosis in Aath4aDBA/DBA macrophages (Figure 5D).

Figure 5. SPMs and LPMs in Aath4aDBA/DBA mice.

Figure 5

Open in a new tab

A. Representative flow cytometry panels showing SPM and LPM populations in control Aath4a129/129 (left) and Aath4aDBA/DBA (right) mice. SPMs are defined as a F4/80lowCD11blow and LPMs as a F4/80highCD11bhigh population. B. Percentage of SPMs and LPMs in control Aath4a129/129 (129/129) and Aath4aDBA/DBA (DBA/DBA) macrophages. Data are representative of five independent experiments each using 4–5 mice of each genotype. C. Cell-surface expression of MERTK on SPMs and LPMs measured by flow cytometry (n = 3). D. Phagocytosis in SPMs and LPMs assessed by MFI of pHrodo Red (n = 3). Data are shown as the mean ± SE.

MERTK amino acid differences do not affect efferocytosis

There are nine amino acid substitutions between 129S6 and DBA/2J MERTK, three of which (W25G, T80E and S479R) are predicted to potentially alter protein function (Figure 6A).4 The S479R substitution located adjacent the ADAM17 cleavage site does not appear to have a significant effect on the shedding of MERTK, since the proportion of sMER:MER was not altered between control and Aath4aDBA/DBA macrophages (Figure 3B). Likewise, when plasmids carrying Mertk cDNA of 129S6 or DBA/2J were transfected into HEK293T cells, MERTK protein expression was similar, and comparable amounts of sMER were detected in the respective culture medium (Figure 6B). These results suggest that W25G, which is located close to the cleavage site of the secretion signal peptide, does not affect the protein maturation process.

We next examined the ability of the HEK293T cells expressing MERTK to phagocytose apoptotic Jurkat T cells. Overexpression of MERTK increased phagocytosis of apoptotic Jurkat T cells compared with the empty vector-transfected control cells, but no significant difference was observed between the 129S6 and DBA/2J (empty vectors: 0.8 ± 0.5%, 129S6-MERTK: 28.4 ± 1.5%, DBA/2J-MERTK: 29.5 ± 1.5%) (Figure 6C). Addition of protein S, which bridges apoptotic cells and MERTK, stimulated phagocytosis in all of them, but a genetic effect was not observed (empty vectors: 4.1 ± 1.6%, 129S6-MERTK: 45.3 ± 1.5%, DBA/2J-MERTK: 48.7 ± 2.7%) (Figure 6C). Thus, the T80E substitution in the first immunoglobulin-like domain, which contains a ligand binding site, does not affect the binding to bridging protein S. These results indicate that the DBA/2J-specific amino acid substitutions in MERTK do not affect phagocytotic function.

Differences in Mertk expression levels modulate efferocytosis

Previous works have shown that mice lacking Mertk and Apoe (Mertk−/−Apoe−/−) have increased accumulation of apoptotic cells and expanded necrotic cores in their plaques when compared to Mertk+/+Apoe−/− mice,12 and that peritoneal macrophages from Mertk deficient mice (Mertk−/−) phagocytose apoptotic cells at a much reduced level (less than 20% compared to Mertk+/+ macrophages).6 This raises the question whether differences in the level of expression of MERTK affect efferocytosis in a graded manner. To test this, we examined phagocytosis of apoptotic Jurkat T cells by peritoneal macrophages isolated from wild-type (Mertk+/+) and Mertk heterozygous (Mertk+/−) mice. We found that Mertk heterozygous macrophages, which are supposed to express 50% of MERTK, show approximately 70% phagocytosis compared to wild-type macrophages (P = 0.002) (Figure 6D), indicating that the amount of MERTK modulates efficiency of phagocytosis.

In agreement, when the amount of Mertk-plasmid transfected to HEK293 cells was gradually decreased, cell surface 129-MERTK and DBA-MERTK and phagocytosis were equally reduced (Figure 6E). We conclude that the amount of MERTK determines the efficiency of phagocytosis in a dose-dependent, not an all or none manner, and that decreased Mertk expression in Aath4aDBA/DBA macrophages is the cause of the reduced phagocytosis.

Transcription activity of the proximal promoter region of Mertk

There are a large number of nucleotide polymorphisms throughout the Mertk gene which could affect steady state levels of mRNA by affecting its stability. Accordingly, to clarify the cause of reduced Mertk mRNA in Aath4aDBA/DBA macrophages, we asked whether the degradation of Mertk mRNA is enhanced in Aath4aDBA/DBA macrophages. We found that the estimated half-life of Mertk mRNA was 5.6 hours in the control Aath4a129/129 macrophages and 5.7 hours in Aath4aDBA/DBA macrophages, suggesting that the lower amount of Mertk mRNA in Aath4aDBA/DBA macrophages is likely caused by reduced Mertk transcription, rather than by faster degradation of Mertk mRNA (Figure VII in the online-only Data Supplement).

We next searched for DBA/2J-specific differences in the promoter region of the Mertk gene that could affect the efficiency of transcription. Within the 1.1 kb immediately upstream of the Mertk transcription start site, there are ten SNPs and one deletion that are unique to DBA compared with B6 and 129 sequences (Figure VIIIA in the online-only Data Supplement). The Mertk gene has a CpG island promoter and does not contain a typical TATA box or CAAT box.

To test whether the sequence differences in the proximal promoter region of Mertk affect the transcription level, we compared promoter activity of 129S6 and DBA/2J by a luciferase reporter assay. Mertk promoters of both 129S6 and DBA/2J enhanced luciferase expression in HEK293T cells approximately 5 times compared to empty vector, but the transcription activities of the Mertk promoter from 129S6 and DBA2/J were not significantly different (Figure VIIIB in the online-only Data Supplement). Furthermore, addition of factors that are thought to regulate Mertk expression18 including GM-CSF, M-CSF and IL-4 did not significantly change the promoter activities (Figure VIIIC in the online-only Data Supplement). Our experiments suggest that the genomic variants affecting the expression of Mertk must reside outside of this proximal 1.1 kb promoter region.

Indeed, when we searched for eQTLs associated with the expression levels of Mertk from Hybrid Mouse Diversity Panel (HMDP) by UCL A (https://systems.genetics.ucla.edu/data/hmdp),19 more than 15 SNPs within the Aath4a interval were significantly associated with Mertk expression levels (Table I in the online-only Data Supplement). The SNPs that are shared by 129 and B6 but unique in DBA, and located between 63 kb to 2.6 Mb upstream of the Mertk gene, appear to regulate its expression in the liver and adipose tissue, while SNPs located 2.3 Mb to 2.5 Mb downstream were broadly associated with the expression levels in macrophages (Table I in the online-only Data Supplement). Two SNPs at −325 kb and −202 kb upstream also showed strong association with the Mertk expression in macrophages, but 129 and DBA share the same variants at these positions. Together, it is likely that distant genetic variants affect the expression of Mertk.

Discussion

Each QTL identified in the crosses of inbred mice generally spans a large genomic distance, sometimes almost an entire chromosome. In complex phenotypes such as atherosclerosis, where a large number of genes are involved, transferring a target region onto an inbred background and creating congenic line is a powerful step towards identifying causative genes. Here we have analyzed the effect of the atherosclerosis QTL Aath4 by establishing a congenic line (Aath4aDBA/DBA), where the 5′ region of Aath4 DBA was backcrossed onto a 129S6-Apoe−/− background. As expected, the resulting Aath4aDBA/DBA males had significantly larger plaques, and macrophages isolated from these mice exhibited reduced efferocytosis as a consequence of allele-specific decrease in MERTK expression. Together, our results provide strong evidence that the increased susceptibility to atherosclerosis determined by the DBA allele of Aath4 is at least in part due to decreased MERTK expression.

MERTK is known to play a significant role in efferocytosis and the resolution of inflammation during atherosclerosis.12, 13 In this report, we have demonstrated that nine non-synonymous SNPs in Mertk, which are uniquely different in DBA/2J compared to C57BL/6J and 129S6, are not responsible for the reduced efferocytosis observed in peritoneal macrophages from Aath4aDBA/DBA. Notably, our experiments demonstrated that the level of MERTK expression controls phagocytosis in Mertk+/− and Aath4aDBA/DBA macrophages in a dose-dependent, not an all or none manner. This is consistent with many published reports that disease-related SNPs identified by GWAS are usually located in introns or inter-gene regions rather than in coding sequences. Unlike diseases in which mutations in a single gene cause a drastic phenotype, the pathogenesis of atherosclerosis is more complex and involves numerous factors. SNPs or other alterations in regulatory regions typically lead to small changes in gene expression, which cumulatively influence susceptibility to disease. Our experiments show that Mertk expression is likely reduced at the transcription level, since the stability of Mertk mRNA was unchanged in the Aath4aDBA/DBA macrophages. Furthermore, our reporter assay tests show that the genetic differences that cause different Mertk expression in DBA/2J and 129S6 are likely to be outside the proximal 1.1 kb promoter region, and search of eQTL database suggests that expression of Mertk is affected by distant genetic variants. In humans, a SNP rs869016 in intron 1 of MERTK is associated with decreased risk of carotid atherosclerosis, although it is unknown whether the SNP modulates Mertk mRNA expression.20 Further investigation of the causative variants is clearly required.

Efferocytosis, the phagocytosis of apoptotic cells by macrophages, is critical in preventing progression of atherosclerosis.21 Consistently, it has been shown that Mertk−/−Apoe−/− mice fed a Western-type diet have increased numbers of apoptotic cells and expanded necrosis in advanced lesions.12 In early plaques, removal of apoptotic cells is normally very efficient; although, we have observed signs of basophilic (dystrophic) calcium deposits even in the very early raised plaques of Aath4aDBA/DBA mice. Our attempt to detect apoptotic cells in vivo in plaques by TUNEL assay was not productive to make a comparison, because TUNEL-positive nuclei were so few per section in both Aath4a129/129 and Aath4aDBA/DBA. This is because mice were fed with normal chow in order to eliminate additional effects caused by high-fat diet, and most of the lesions we observed were at their early stages. However, the calcium deposits are likely the remnants of dead/dying cells via apoptotic and necrotic processes, giving strong evidence that the normal process of removing dead cells is restricted in the Aath4aDBA/DBA mice. Delayed apoptotic cell clearance is expected to cause acceleration of plaque development and necrotic core formation. The calcium deposits in the early plaques, however, appear to be short-lived, since larger calcified areas in advanced Aath4aDBA/DBA plaques are not very different from the control plaques and are seen mostly deeper in the intima near the internal elastic lamina. Although the thickness of fibrous caps was not significantly different, the Aath4aDBA/DBA plaques contained less interstitial matrix proteins than similar sized plaques in the control mice. In advanced plaques, efferocytosis by macrophages becomes less effective than in early plaques, but an alteration in the balance between synthesis and degradation of matrix protein would be expected to contribute to vulnerable plaque morphology. Although we have not encountered any premature deaths associated with the Aath4aDBA/DBA mice, a more detailed examination of later stage plaques would be worthwhile.

Despite the well-known heterogeneity of mouse peritoneal macrophages,15, 16 we found that MERTK is expressed in both LPMs and SPMs, and that it is similarly reduced in both populations of Aath4aDBA/DBA macrophages. We also observed that the LPMs, which have higher phagocytic activity than SPMs,17 were consistently fewer in Aath4aDBA/DBA than in controls. Fewer LPMs and more SPMs could be another factor contributing to the reduced phagocytosis in Aath4aDBA/DBA peritoneal macrophages. SPMs are induced from circulating monocytes by inflammatory stimuli, while LPMs are thought to be tissue-resident macrophages, maintained locally by differentiation from precursor cells. Macrophages in atherosclerotic plaques are similarly heterogeneous,22, 23 and the role of circulating monocyte-derived macrophages in the development of plaques has been well-documented. Evidence is also beginning to accumulate that tissue-resident macrophages originating from the adventitial precursor cells contribute to atherogenesis. Factors that control the relative balance of these populations are not known, but a shift in the population of the Aath4aDBA/DBA macrophages in their plaques is conceivable. Whether MERTK expression directly affects this shift or not requires further study.

Although we identified Aath4 as a QTL specific to the arch lesion, the Aath4aDBA/DBA mice have significantly larger plaques in both the aortic arch and root. This is in agreement with our prediction that the reduced efferocytosis of Aath4aDBA/DBA macrophages should equally affect both aortic arch and root. Our previous studies of F2 population between 129S6-Apoe−/− and C57BL/6-Apoe−/− indicated that 129S6 genome carries the sequence that affects the aortic arch geometry which are also associated with increased aortic arch atherosclerosis.2 Therefore, the variant in Mertk expression is a risk factor that is additional to the 129S6 sequences that determine the aortic arch development, since the Aath4aDBA/DBA strain is based upon the129S6 genome. Moreover, the contribution of Mertk expression to the 50% increase in root lesion is, although significant, relatively small, considering that the parental DBA/2J show 10 times larger plaques in the root than 129S6 mice.

The genomic region of DBA/2J carried in the Aath4aDBA/DBA mice is large, and the effect of DNA variants in other locations in this region must be considered. For example, Siglec1 at 131 Mb, which encodes sialoadhesin (CD169), is involved in the retention of hematopoietic stem cells in the bone marrow niche.24 It is also expressed in a subset of tissue-resident macrophages, preventing excessive inflammation upon injury.25 Similarly, CD169 may also play a role in phagocytosis.26, 27 Altered expression of Bcl2l11 at 128 Mb, encoding for a proapoptotic protein Bim, could affect early cell death in the Aath4aDBA/DBA plaques. Additionally, since the backcrossed region in the Aath4aDBA/DBA line is the proximal half of Aath4, the distal half may include genes that are pro-atherogenic only in the arch, or are athero-suppressive specifically in the root. One of the candidates in distal Aath4 is Cd93 at 148 Mb, a transmembrane glycoprotein which is expressed in endothelial and myeloid cells. CD93-deficient mice show a defect in the clearance of apoptotic cells in vivo,28 and a SNP in Cd93 gene is associated with an increased risk of coronary heart disease.29 Analysis of a congenic line carrying the distal half of Aath4 of DBA/2J is currently under way.

We also note that the plaque size difference was observed only in Aath4aDBA/DBA males, whereas the male/female differences in plaque size were not seen in the parental strains: DBA/2J-Apoe−/− mice show larger root lesions compared with 129S6-Apoe−/− in both males and females;3 arch lesion size was comparable between the two strains in both males and females.4 The QTLs on Chr 2 were detected in both males and females in the F2 population from DBA/2J-Apoe−/− and 129S6-Apoe−/− mice.3, 4 Since no QTL was detected on Chr 2 for plasma lipids in the intercross between the two strains, the plasma lipid differences between males and females are not likely the strong determinant of the plaque size differences.3 The major determinants for the plaque sizes in females must therefore lie outside of the Aath4a interval. The gender dimorphism in the Aath4aDBA/DBA mice requires further investigation.

In summary, our experiments have shown that the DBA allele of the 5′ portion of Aath4 causes inefficient efferocytosis via lower expression of MERTK, and that difference contributes to the enhanced plaque development observed in Aath4aDBA/DBA mice. Since the effects of reduced MERTK and efferocytosis are seen in the both aortic arch and root areas, further studies will be required to fully understand the factors that cause location specificity in atherosclerosis.

Supplementary Material

Graphic Abstract
Materials and Methods
Supplemental Figures

Highlights.

  • Aath4aDBA/DBA, a congenic line of an atherosclerosis QTL Aath4 on distal Chr 2, was generated by transferring DBA/2J alleles of Aath4 to the129S6-Apoe−/− strain.

  • Aath4aDBA/DBA males develop larger plaques and peritoneal macrophages isolated from Aath4aDBA/DBA showed reduced phagocytosis of apoptotic cells.

  • Lower Mertk transcription, rather than DBA/2J-specific amino acid substitutions, causes restricted efferocytosis in Aath4aDBA/DBA.

Acknowledgments

We thank Dr. Glenn Matsushima and Akhil Patel for Mertk+/− mice; Dr. Hyung-Suk Kim for qPCR; Longquan Xu, Svetlana Zhilicheva, Sylvia Hiller and Jennifer Wilder for technical assistance; and the UNC CGIBD Histology Core, Microscopy Services Laboratory and Flow Cytometry Core Facilities for technical support. We also thank Drs. Marlon Lawrence, Glenn Matsushima, Jonathon Homeister and Oliver Smithies for critical reading of the manuscript; and Dr. Robert Reddick at the University of Texas Health Science Center at San Antonio for discussions on evaluating atherosclerotic plaques.

Sources of funding

This research was supported by a National Institutes of Health (NIH) grant HL042630. The UNC Flow Cytometry Core Facility is supported in part by P30 CA016086 Cancer Center Core Support Grant to the UNC Lineberger Comprehensive Cancer Center.

Abbreviations

QTL (s)

quantitative trait locus (loci)

Aath4

arch atherosclerosis 4

MERTK

c-mer proto-oncogene tyrosine kinase coded by the Mertk gene

SPMs

small peritoneal macrophages

LPMs

large peritoneal macrophages

Footnotes

Disclosures

None.

References

  • 1.Maeda N, Johnson L, Kim S, Hagaman J, Friedman M, Reddick R. Anatomical differences and atherosclerosis in apolipoprotein E-deficient mice with 129/SvEv and C57BL/6 genetic backgrounds. Atherosclerosis. 2007;195:75–82. doi: 10.1016/j.atherosclerosis.2006.12.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Tomita H, Zhilicheva S, Kim S, Maeda N. Aortic arch curvature and atherosclerosis have overlapping quantitative trait loci in a cross between 129S6/SvEvTac and C57BL/6J apolipoprotein E-null mice. Circ Res. 2010;106:1052–1060. doi: 10.1161/CIRCRESAHA.109.207175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Kayashima Y, Tomita H, Zhilicheva S, Kim S, Kim HS, Bennett BJ, Maeda N. Quantitative trait loci affecting atherosclerosis at the aortic root identified in an intercross between DBA2J and 129S6 apolipoprotein E-null mice. PLoS One. 2014;9:e88274. doi: 10.1371/journal.pone.0088274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Kayashima Y, Makhanova NA, Matsuki K, Tomita H, Bennett BJ, Maeda N. Identification of aortic arch-specific quantitative trait loci for atherosclerosis by an intercross of DBA/2J and 129S6 apolipoprotein E-deficient mice. PLoS One. 2015;10:e0117478. doi: 10.1371/journal.pone.0117478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Graham DK, Dawson TL, Mullaney DL, Snodgrass HR, Earp HS. Cloning and mRNA expression analysis of a novel human protooncogene, c-mer. Cell Growth Differ. 1994;5:647–657. [PubMed] [Google Scholar]
  • 6.Scott RS, McMahon EJ, Pop SM, Reap EA, Caricchio R, Cohen PL, Earp HS, Matsushima GK. Phagocytosis and clearance of apoptotic cells is mediated by MER. Nature. 2001;411:207–211. doi: 10.1038/35075603. [DOI] [PubMed] [Google Scholar]
  • 7.Seitz HM, Camenisch TD, Lemke G, Earp HS, Matsushima GK. Macrophages and dendritic cells use different Axl/Mertk/Tyro3 receptors in clearance of apoptotic cells. J Immunol. 2007;178:5635–5642. doi: 10.4049/jimmunol.178.9.5635. [DOI] [PubMed] [Google Scholar]
  • 8.Paul SR, Merberg D, Finnerty H, Morris GE, Morris JC, Jones SS, Kriz R, Turner KJ, Wood CR. Molecular cloning of the cDNA encoding a receptor tyrosine kinase-related molecule with a catalytic region homologous to c-met. Int J Cell Cloning. 1992;10:309–314. doi: 10.1002/stem.5530100509. [DOI] [PubMed] [Google Scholar]
  • 9.Graham DK, Bowman GW, Dawson TL, Stanford WL, Earp HS, Snodgrass HR. Cloning and developmental expression analysis of the murine c-mer tyrosine kinase. Oncogene. 1995;10:2349–2359. [PubMed] [Google Scholar]
  • 10.Uehara H, Shacter E. Auto-oxidation and oligomerization of protein S on the apoptotic cell surface is required for Mer tyrosine kinase-mediated phagocytosis of apoptotic cells. J Immunol. 2008;180:2522–2530. doi: 10.4049/jimmunol.180.4.2522. [DOI] [PubMed] [Google Scholar]
  • 11.Duncan JL, LaVail MM, Yasumura D, Matthes MT, Yang H, Trautmann N, Chappelow AV, Feng W, Earp HS, Matsushima GK, Vollrath D. An RCS-like retinal dystrophy phenotype in mer knockout mice. Invest Ophthalmol Vis Sci. 2003;44:826–838. doi: 10.1167/iovs.02-0438. [DOI] [PubMed] [Google Scholar]
  • 12.Thorp E, Cui D, Schrijvers DM, Kuriakose G, Tabas I. Mertk receptor mutation reduces efferocytosis efficiency and promotes apoptotic cell accumulation and plaque necrosis in atherosclerotic lesions of apoe−/− mice. Arterioscler Thromb Vasc Biol. 2008;28:1421–1428. doi: 10.1161/ATVBAHA.108.167197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Ait-Oufella H, Pouresmail V, Simon T, Blanc-Brude O, Kinugawa K, Merval R, Offenstadt G, Leseche G, Cohen PL, Tedgui A, Mallat Z. Defective mer receptor tyrosine kinase signaling in bone marrow cells promotes apoptotic cell accumulation and accelerates atherosclerosis. Arterioscler Thromb Vasc Biol. 2008;28:1429–1431. doi: 10.1161/ATVBAHA.108.169078. [DOI] [PubMed] [Google Scholar]
  • 14.Thorp E, Vaisar T, Subramanian M, Mautner L, Blobel C, Tabas I. Shedding of the Mer tyrosine kinase receptor is mediated by ADAM17 protein through a pathway involving reactive oxygen species, protein kinase Cdelta, and p38 mitogen-activated protein kinase (MAPK) J Biol Chem. 2011;286:33335–33344. doi: 10.1074/jbc.M111.263020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Ghosn EE, Cassado AA, Govoni GR, Fukuhara T, Yang Y, Monack DM, Bortoluci KR, Almeida SR, Herzenberg LA, Herzenberg LA. Two physically, functionally, and developmentally distinct peritoneal macrophage subsets. Proc Natl Acad Sci U S A. 2010;107:2568–2573. doi: 10.1073/pnas.0915000107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Cassado Ados A, D’Imperio Lima MR, Bortoluci KR. Revisiting mouse peritoneal macrophages: heterogeneity, development, and function. Front Immunol. 2015;6:225. doi: 10.3389/fimmu.2015.00225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Cain DW, O’Koren EG, Kan MJ, Womble M, Sempowski GD, Hopper K, Gunn MD, Kelsoe G. Identification of a tissue-specific, C/EBPbeta-dependent pathway of differentiation for murine peritoneal macrophages. J Immunol. 2013;191:4665–4675. doi: 10.4049/jimmunol.1300581. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Zizzo G, Hilliard BA, Monestier M, Cohen PL. Efficient clearance of early apoptotic cells by human macrophages requires M2c polarization and MerTK induction. J Immunol. 2012;189:3508–3520. doi: 10.4049/jimmunol.1200662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Bennett BJ, et al. A high-resolution association mapping panel for the dissection of complex traits in mice. Genome Res. 2010;20:281–290. doi: 10.1101/gr.099234.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Hurtado B, Abasolo N, Munoz X, Garcia N, Benavente Y, Rubio F, Garcia de Frutos P, Krupinski J, Sala N. Association study between polymorphims in GAS6-TAM genes and carotid atherosclerosis. Thromb Haemost. 2010;104:592–598. doi: 10.1160/TH09-11-0787. [DOI] [PubMed] [Google Scholar]
  • 21.Tabas I. Consequences and therapeutic implications of macrophage apoptosis in atherosclerosis: the importance of lesion stage and phagocytic efficiency. Arterioscler Thromb Vasc Biol. 2005;25:2255–2264. doi: 10.1161/01.ATV.0000184783.04864.9f. [DOI] [PubMed] [Google Scholar]
  • 22.Moore KJ, Sheedy FJ, Fisher EA. Macrophages in atherosclerosis: a dynamic balance. Nat Rev Immunol. 2013;13:709–721. doi: 10.1038/nri3520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Peled M, Fisher EA. Dynamic Aspects of Macrophage Polarization during Atherosclerosis Progression and Regression. Front Immunol. 2014;5:579. doi: 10.3389/fimmu.2014.00579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Chow A, Lucas D, Hidalgo A, Mendez-Ferrer S, Hashimoto D, Scheiermann C, Battista M, Leboeuf M, Prophete C, van Rooijen N, Tanaka M, Merad M, Frenette PS. Bone marrow CD169+ macrophages promote the retention of hematopoietic stem and progenitor cells in the mesenchymal stem cell niche. J Exp Med. 2011;208:261–271. doi: 10.1084/jem.20101688. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.O’Neill AS, van den Berg TK, Mullen GE. Sialoadhesin - a macrophage-restricted marker of immunoregulation and inflammation. Immunology. 2013;138:198–207. doi: 10.1111/imm.12042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Guo M, Hartlova A, Dill BD, Prescott AR, Gierlinski M, Trost M. High-resolution quantitative proteome analysis reveals substantial differences between phagosomes of RAW 264.7 and bone marrow derived macrophages. Proteomics. 2015;15:3169–3174. doi: 10.1002/pmic.201400431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Xiong YS, Yu J, Li C, Zhu L, Wu LJ, Zhong RQ. The role of Siglec-1 and SR-BI interaction in the phagocytosis of oxidized low density lipoprotein by macrophages. PLoS One. 2013;8:e58831. doi: 10.1371/journal.pone.0058831. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Norsworthy PJ, Fossati-Jimack L, Cortes-Hernandez J, Taylor PR, Bygrave AE, Thompson RD, Nourshargh S, Walport MJ, Botto M. Murine CD93 (C1qRp) contributes to the removal of apoptotic cells in vivo but is not required for C1q-mediated enhancement of phagocytosis. J Immunol. 2004;172:3406–3414. doi: 10.4049/jimmunol.172.6.3406. [DOI] [PubMed] [Google Scholar]
  • 29.van der Net JB, Oosterveer DM, Versmissen J, Defesche JC, Yazdanpanah M, Aouizerat BE, Steyerberg EW, Malloy MJ, Pullinger CR, Kastelein JJ, Kane JP, Sijbrands EJ. Replication study of 10 genetic polymorphisms associated with coronary heart disease in a specific high-risk population with familial hypercholesterolemia. Eur Heart J. 2008;29:2195–2201. doi: 10.1093/eurheartj/ehn303. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Graphic Abstract
NIHMS871052-supplement-Graphic_Abstract.pdf (37.2MB, pdf)
Materials and Methods
NIHMS871052-supplement-Materials_and_Methods.pdf (279.6KB, pdf)
Supplemental Figures
NIHMS871052-supplement-Supplemental_Figures.pdf (1.5MB, pdf)

RESOURCES