Animal experiments were approved by the Laboratory Animal Welfare and Ethics Review Committee of University of South China (approval no. SYSK2013-0010). The present study was performed according to the requirements in the Guide for the Care and Use of Laboratory Animals (20). Specific pathogen-free (SPF) C57BL/6J wild-type mice were purchased from Shanghai Model Organisms Center, Inc. (SMOC) and raised under SPF conditions. A total of 45 mice (30 females and 15 males; age, 6–8 weeks; weight, 17–20 g) were housed in standard plastic cages at 20–26°C, 40–70% humidity with a 12 h light/dark cycle and had free access to water and food. The mice were given 2 weeks to acclimate for breeding before any treatment.

The goal of the present study was to subclone mouse CXCL16 cDNA into the pGL3-basic mammalian expression vector. PCR amplification was used to multiply Murine CXCL16 cDNA and simultaneously as part of the primer a sequence (5′ to 3′) of flag tag (GATTACAAGGATGACGACGATAAG) was attached to the N terminus. Later scavenger receptor (SR)-enhancer, SR-promoter and CXCL16-flag were all respectively inserted into pGL3-basic between restriction sites KpnI and NheI, XhoI and HindIII, HindIII and XbaI. A genomic DNA isolation kit (Bio Basic Inc.) was used to extract and purify DNA (the yield of 5–10 µg) from miniprep according to the manufacturer's protocols. Verification was performed through Sanger sequencing (21) using an ABI 3730XL automatic sequencer (Applied Biosystems; Thermo Fisher Scientific, Inc.). A triple cut site ClaI was set to cut the vector in to three parts that were respectively 1.8, 2.7 and 0.1 kbp. The 1.8 kbp fragments including CXCL16 cDNA were collected and purified from agarose gel electrophoresed in TBE running buffer using the gel recovery kit (Bio Basic Inc.), according to the manufacturer's instructions, for further use.

A total of 30 C57BL/6J × CBA F1 hybrid female mice (age, 2–4 weeks; weight, 7–10 g) supplied by SMOC were housed in standard plastic cages at 20–26°C and 40–70% humidity with a 12 h light/dark cycle. These mice were used as their oocytes exhibit high survival and maturation rates after microinjection (22). At the age of ~6–8 weeks old, 12 h before dark cycle, they were injected with 10 IU PMSG and ~48 h afterwards with 10 IU human chorionic gonadotropin i.p. In the same afternoon they were mated 1:1 with fertile male C57BL/6J mice. For all mouse experiments, the mice were anesthetized with 50–60 mg/kg pentobarbital sodium (Sigma-Aldrich; Merck KGaA) and sacrificed by cervical dislocation. The next day, if a vaginal plug was present, they were set as donors and fertilized eggs were excised from oviducts into M2 culture medium (Sigma-Aldrich; Merck KGaA). After washing, screening and removal of cumulus cells, the zygotes were transferred into M16 medium (Sigma-Aldrich; Merck KGaA) prior to microinjection. Linear DNA constructs (1.8 kbp) prepared as aforementioned were injected into the pronuclei of zygotes using micromanipulator (Narishige Group). After ~20–30 min rest in M16, zygotes (20–30 each) were transferred into the oviducts of pseudopregnant C57BL/6J mice prepared beforehand. Integration of the transgene was determined by a double check of transgene-specific PCRs with genomic DNA isolated from tail biopsies of the progenies. To distinguish genomic CXCL16 from transgenic CXCL16, special upstream and downstream primers were located in SR-promoter and CXCL16-flag respectively. The primer sequences (purchased from Sangon Biotech Co., Ltd.) were given in Table I.

PCR positive offspring were mated with wild-type C57BL/6J mice according to female/male ratio (1:1 or 1:2). To generate homozygous transgenic strain, sib-mating was applied most of the time and backcrossing when necessary. Genomic DNA was extracted from the descendants to sequence for the verification of transgene presence. PCR was used as a screen method. When all progenies (n>5) of two consecutive generations were all homozygous, the mouse was considered homozygous. After 10 generations of selective breeding, a CXCL16⁺/⁺ transgenic line was successfully established. In total, 16 CXCL16⁺/⁺ C57BL/6J mice from F11 combined with 16 wild-type C57BL/6J mice were chosen as subjects.

CXCL16⁺/⁺ and wild-type mice were randomly selected and divided into two treatment groups. One group was fed with high-fat diet (HFD, 67.5% standard chow, 15% lard, 10% yolk powder, 3% cholesterol, 4% whole milk powder and 0.5% Sodium cholate; Amresco, LLC) and the other standard chow, termed low-fat diet (LFD) for 18 weeks (23–25). Body weight, plasma total cholesterol (TC), triglycerides (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL) were measured at week 18. Mice in each group were intraperitoneally injected with sodium pentobarbital (50 mg/kg) and were sacrificed by cervical dislocation followed by decapitation.

ELISA kits for Mouse CXCL16 (cat. no. 10030301; Cusabio Technology LLC), Mouse matrix metalloproteinase-9 (MMP-9; cat. no. EK1462), monocyte chemoattractant protein-1 (MCP-1; cat. no. EK1351) and intercellular adhesion molecule-1 (ICAM-1; cat. no. EK1726; Wuhan Boster Biological Technology, Ltd.) were used according to the manufacturer's protocols. Absorbance was measured at 450 nm using an 800 TS Absorbance Reader (BioTek Instruments, Inc.). ELISA calc software (BioTNT; V0.1) was used to draw standard curve and calculate concentration.

Biopsies were obtained from the aortic root and artery sinus and were fixed overnight in 4% formaldehyde solution at room temperature. Aortic root HE staining was performed to examine the degree of fatty acid infiltration among the groups. The sections were stained in hematoxylin solution for 10 min and eosin solution for 1 min at room temperature. Immunohistochemical staining of the lesions was performed at room temperature for 1 h using rabbit polyclonal MMP-9 antibody (1:1,000; cat. no. ab38898), rabbit polyclonal MCP-1 antibody (1:1,000; cat. no. ab73680) (both Abcam) and rabbit polyclonal ICAM-1 antibody (1:200; cat. no. PB9081; Wuhan Boster Biological Technology, Ltd.).

Each experiment was repeated in triplicate. Data analysis was performed with SPSS v17.0 (SPSS, Inc.) and presented as the mean ± standard deviation. Results were determined by unpaired Student's t test and one-way ANOVA (post hoc test: Tukey's multiple comparison test) or two-way ANOVA (post hoc test: Bonferroni). P<0.05 was considered to indicate a statistically significant difference. All diagrams were generated by GraphPad Prism7 (GraphPad Software, Inc.).

Following several rounds of insertion, a plasmid with the backbone of pGL3-basic, SR-enhancer, SR-promoter, CXCL16 CDS and flag was successfully constructed (Fig. 1) and confirmed by sequencing.

A total of 426 microinjected eggs were obtained and 20 eggs as one unit were transferred to oviducts of 20 pseudo mothers. Among which 12 became pregnant and 171 young were born. Following PCR screening, 26 were proved to be carriers of the target gene. After 4 weeks, another double-blind PCR was performed on 26 transgenic and four wild-type mice, and confirmed the stability of transgene expression. Founder mice were mated with wild-type C57BL/6J mice. The PCR-positive offspring were then mated with positive siblings in order to obtained homozygous transgenic mice. When all progenies (n>5) of two consecutive generations were all homozygous, the parent mice were considered homozygous. F8-1–2 (female) and F9-2–3 (male) were thus proved homozygous and, when mated with wild type mice, their offspring were all positive. Backcross F8-1–2 with F9-2-3, 5 descendants were obtained and termed C10-1–1 to C10-1-5. Sib-mate C10-1 and 19 descendants were obtained and eight were randomly chosen as CXCL16⁺/⁺ transgenic subject mice. In addition, eight from the 11th generation of wild-type mice were randomly selected as subjects.

CXCL16 mRNA expression on artery AS lesion and plasma protein expression was measured at week 18. In the two HFD groups or the two LFD groups, CXCL16⁺/⁺ transgenic mice had a higher CXCL16 mRNA expression (Fig. 2A). CXCL16 mRNA expression of CXCL16⁺/⁺ transgenic mice or wild-type mice fed with HFD was significantly higher than those fed with LFD (Fig. 2A). CXCL16 plasma protein expression was result similar to CXCL16 mRNA expression on artery AS lesion (Fig. 2B).

Body weight and plasma lipid profile were measured at the end of the week 18. No significant difference was observed between the two LFD groups, but the average body weight of CXCL16⁺/⁺ transgenic mice fed with HFD was significantly highly (Fig. 3A) compared with that of wild-type mice fed with the same regimen. Similar results were observed in plasma TC level, LDL-cholesterol (LDL-C) level, non HDL-C level and TC/HDL-cholesterol (HDL-C) ratio (Fig. 3B-E). Another significant change was identified between CXCL16⁺/⁺ transgenic mice fed with HFD and LFD. When fed with HFD, CXCL16⁺/⁺ transgenic mice tend to maintain much higher plasma lipoprotein levels such as LDL-C and HDL-C compared with the LFD groups (Fig. 3C and F), even when no significant differences were observed between the expression levels of the wild-type mice groups. This suggested that CXCL16 exerted an effect on fatty acid metabolism, but it was only significant when a higher intake of fatty acid was present.

Visceral fat accumulation was recorded in the form of images (data not shown). Fibrosis status was determined according to the Ishak staging system (26). Table II shows the percentage of mice with different degree of liver disease; after 18 weeks of HFD, 100% of mice with CXCL16⁺/⁺ developed liver diseases, while only 50% of wild-type mice in HFD did. The degree of fibrosis status was also relatively less. None of mice in the LFD groups developed fatty liver disease.

For aortic root HE staining, CXCL16 double positive mice in the HFD groups demonstrated the most marked contrast compared with healthy aorta (Fig. 4).

Plasma MCP-1 and ICAM-1 expression levels were determined by ELISA at the end of the week 18. Fig. 5A demonstrates the overall expression level of MCP-1 in the four groups. MCP-1 expression pattern among the LFD groups showed no observable discrepancy (Fig. 5A) while among the HFD groups, a significant difference was observed; with a higher expression of CXCL16, MCP-1 expression levels were also higher (Fig. 5A).

ICAM-1 expression levels showed an incremental gradient from LDF wild-type, LDF CXCL16⁺/⁺, HDF wild-type to HDF CXCL16⁺/⁺. This had no statistical significance except for the difference between CXCL16^+/+ transgenic mice fed with HFD and LFD (Fig. 5B).

Immunohistochemistry staining also indicated a higher local expression of MCP-1 and ICAM-1 in transgenic mice at the aortic root and artery sinus (Fig. 6). Positive MCP-1 signals were seen within the AS lesions, while accumulation of ICAM-1 was mainly located on the aortic intima surface (Fig. 6E-H). An overall higher expression of MCP-1 and ICAM-1 was observed at the artery sinus compared with the aortic root (Fig. 6).

Plasma MMP-9 level was tested at the end of the week 18. It showed high specificity among the four groups. HDF CXCL16⁺/⁺ mice had the highest plasma level of all, and it was statistically significant when compared with HDF wild-type or LDF CXCL16⁺/⁺ mice (Fig. 7). Additionally, LDF CXCL16⁺/⁺ mice also expressed a higher level of plasma MMP-9 than LDF wild-type mice (Fig. 7).

Local expression of MMP-9 in artery sinus was also tested among the HFD groups (Fig. 8A and B). Atherosclerotic plaques of CXCL16⁺/⁺ mice possessed a higher percentage of cells that were MMP-9 positive (Fig. 8C).