Functional characterization of human IL-8 in vascular stenosis using a novel humanized transgenic mouse model

Abstract

IL-8 (aka interleukin 8, CXCL8) is a prototypic cytokine that is highly expressed in the diseased vessel wall and its plasma concentration is strongly associated with cardiovascular events. However, whether IL-8 plays a causative role in cardiovascular diseases remains largely unknown. In this study we used a human IL-8 transgenic (Tg) mouse strain with a bacterial artificial chromosome (BAC) integrated into its genome. This BAC encompasses 166 kb of sequence encompassing the human IL-8 gene locus as well as upstream and downstream DNA sequences containing regulatory elements. This BAC ensured a pathophysiologically regulated, rather than forced constitutive, expression of human IL-8 in the mouse. Tg mice were subjected to complete carotid ligation injury. IL-8 was highly expressed in the ligation-injured carotid artery from 3 days until 2 weeks after injury. As a result, exacerbated neointimal hyperplasia and increased Mac2 and PCNA positive cells were observed in Tg mice. To further confirm its role in promoting neointimal formation, IL-8 was neutralized by anti-IL8 treatment at the ligation site. Consequently, the size of neointima was significantly reduced. Our results provided new insights into the regulation and function of IL-8 in response to vascular insult and during neointima formation.

1. Introduction

Interleukins are a group of chemokines critical to the initiation and progression of vascular wall remodeling [1]. Upon injury, vascular mural cells such as vascular smooth muscle cells (VSMCs) secret interleukins that recruit immune cells to the site of insults for tissue repair [2]. Infiltrated immune cells in turn release more chemokines and cytokines to trigger the inflammation, proliferation, and trans-differentiation of VSMCs.

Interleukin 8 (IL-8) or CXCL8 is the prototypical member of the CXC subfamily of chemokines. Serving as a marker of inflammation, IL-8 is a reliable prognostic predictor of cardiovascular events such as myocardial infarction, worsening heart failure and mortality [3–5]. IL-8 functions through the binding to receptors CXCR1 or CXCR2. Because these two receptors are highly expressed in neutrophils and monocytes, IL-8 is well known as a recruiter of neutrophils and monocytes, which are the major responders for acute inflammation. While virtually absent in healthy vessel walls, IL-8 was substantially boosted in atherosclerotic plaques, where it was produced by macrophages, VSMCs and endothelial cells (ECs) [6–8]. The expression and secretion of IL-8 were triggered by proinflammatory stimuli or oxidative stress, which promotes all types of vascular disorders. However, it remains largely unexplored whether IL-8 plays a causative role in cardiovascular disease (CVD), due to the lack of a mouse ortholog that otherwise necessitates a prospective study.

Transgenic mice that carry a human gene is a powerful tool for gene function study. IL-8 transgenic mice exhibited a significant neutrophil accumulation in the tissues where IL-8 was overexpressed [9–11]. However, the constitutive overexpression of IL-8 lacks physiological relevance and thus is not the best for a pathogenesis study. Physiologically, the transcription of IL-8 is repressed at basal conditions. The repression is dependent on a long range of regions that reside outside of the immediate IL-8 (CXCL8) gene locus [12]. Bacterial artificial chromosome (BAC) mice provide a broad genomic landscape containing most, if not all, regulatory regions for proper spatiotemporal expression of human genes [13]. Asfaha etc. generated a transgenic BAC mouse line carrying 166 kb of human DNA surrounding the IL-8 gene. This BAC transgenic mouse exhibited a physiological regulation of IL-8 [14]. Once induced by the inflammatory stimuli, IL-8 promoted colitis- and gastritis-induced tumorigenesis in the mouse body.

Inflammation initiates and exacerbates vascular pathogenesis, including the neointima formation [15,16]. We recently reported a novel long non-coding RNA named INKILN that promotes VSMC inflammation and aggravates neointimal hyperplasia [17]. INKILN is the neighbor gene of IL-8. These two share similar features such as human-specificity and proinflammatory activity. We hypothesized that IL-8 is not only a passive inflammatory marker but an active promoter in neointima formation. We measured the neointima size in the ligated carotid artery of IL-8 transgenic (Tg) mice and compared it to the Control (Ctrl) mice. Furthermore, we applied antibodies against IL-8 at the ligated artery and evaluated the effect of neutralization of IL-8 on the neointima formation.

2. Methods

2.1. Animals

Animal use and experimental protocols have been approved by Augusta University Animal Care and Use Committee (IACUC, protocol 2019–0999). All the mice used for present work are in the C57BL/6 J background. BAC IL-8 transgenic mice [14] were bred with ApoE− /− mice for gaining of a hyperlipidemic background. In this work, Tg stands for the IL8+/−, ApoE− /− strain and Ctrl stands for the ApoE− /− littermate control. Both male mice and female mice were studied. Mice were fed with a standard Chow diet and regular water ad libitum. Vascular ligation surgeries were performed on the left carotid arteries when animals were at 8–10 weeks of age, as previously described [17]. In brief, a mouse under isoflurane anesthesia was placed at the surgery platform in supine position. After hair removal, a midline incision was made at neck skin and then muscles were bluntly dissected to expose the left carotid artery. A complete ligation at the bifurcation was made using a 4–0 silk suture. Skin incision was then closed using a 6–0 vicryl absorbable suture. For animals that subjected to antibody treatment, F127 gel mixed with antibody was applied peri-vascularly at the ligated bifurcation and a proximal region of carotid artery. Once the gel was solidified (normally occurred within 30s), the skin incision was closed with suture.

2.2. Preparation of Pluronic F127 hydrogel and in vivo application with antibodies

F127 powder (2.5 g) was added to 5 mL PBS solution in a 50 mL falcon tube. The tube was placed on a rotator shaker and rotated overnight at 4 °C to obtain a well-mixed F127 gel. On ice, 25 mg of polycarbophil (PCB) and 50 mg of Trypsin were added into the freshly prepared F127 gel. The tube was placed on a rotator shaker and rotated at 4 °C for another night.

The lyophilized recombinant human IL-8 antibody and its isotype control: monoclonal mouse IgG₁ were purchased from R&D system (Cat. No. MAB208 and MAB002). The antibody was reconstituted at 0.5μg/uL in sterile PBS. On ice, 2ul of antibody solution (=1μg) was mixed with 200uL freshly prepared F127-based hydrogel. The antibody-containing F127 gel was used for in vivo application on the same day of preparation. For each animal around 20uL of F127 gel (which contains around 100 ng of antibody) was applied on the surface of the ligated left carotid artery, using a wide bore pipette tip.

2.3. Tissue preparation and carotid remodeling evaluation

Two weeks after surgery, animals were anesthetized with isoflurane. Then the chest cavity was open, and a cut was made at the right atrium of heart to release the blood. A needle was inserted into the left ventricle for a transcardial perfusion. Animals were perfused by gravity-fed saline for blood clearing. A quick dissection was executed on aortic arch, the ligated left and the unligated right carotid arteries. The quick dissection may allow some perivascular fat and small amount of surrounding tissue attached on the vessel until fixed. Tissue was fixed by immersing in 4 % PFA with gentle shaking at 4 °C for 24–36 h. The fixed tissue was transferred into cold PBS and were ready for the following steps: either further dissected for a clearance of perivascular fat or immediately underwent tissue processing followed by paraffin embedding. Level 1 starts at where the suture knot was trimmed off. Consecutive cutting was executed with a thickness of 5um for each section. Total 40 sections were collected continuously on 10 glass slides (4 sections/slide). The first 200um (5um X 40 sections) segment of artery was level 1 and the following 200um was level 2, and so on. The 5th slide of each level was subjected to H&E staining and morphological analysis.

2.4. Immunofluorescent staining and microscopy

Before staining, Formalin-Fixed Paraffin-Embedded (FFPE) sections first underwent a xylene-based deparaffinization and then an antigen-retrieval (citrate-based solution H-3300, Vector Labs). After a brief wash with PBS and air-dry, a hydrophobic circle around tissue was drawn using the Pap pen (Cole-Parmer). Blocking reagent (Agilent, X090930–2) was added dropwise onto the tissue and incubated for 1 h at room temperature. Primary antibodies (IL-8, Invitrogen PA5–47289; Mac-2, ThermoFisher 14–5301–82) were diluted with the antibody diluent (Sigma-Aldrich, 938B-03) at 1:13 and 1:200, respectively. Diluted primary antibody was incubated with the tissue section overnight at 4 °C. Tissue section was then washed by adding PBST dropwise and incubated for 5 min. The wash was repeated for 2 more times. Secondary fluorescent antibodies (Alexa Fluor) were diluted in the antibody diluent reagent at 1:500 and incubated with tissue for 1 h at room temperature. Then section was washed 3 times and a mounting media with nuclear counter stain (ThermoFisher, P36981) was applied onto the section. Fluorescence Images were taken with a Zeiss LSM900 confocal microscope. Images were exported in a TIFF format. Analysis such as area measurement was performed in Image J.

2.5. RNA extraction and RT-qPCR

The frozen mouse carotid artery was ground to a fine powder with the pellet pestles (Fisherbrand^™ Pellet Pestles^™) in the presence of liquid nitrogen. The powder was equilibrated on ice and then proceeded to RNA extraction using the QIAGEN RNeasy Mini Kit. Tissue RNA was extracted following the QIAGEN protocol. Yield and purity were determined on the NanoDrop Spectrophotometer (ThermoFisher Scientific). The extracted RNA was then used for the cDNA synthesis using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Real Time PCR was performed using Bio-Rad SYBR Green Supermix and CFX386 Real-Time PCR Detection System. Sequences of primers used for target genes are listed in Table 1.

Table 1.

primers for RT-PCR.

Gene symbol	sequence
CXCL8 (IL-8)	Forward Reverse	AGCCTTCCTGATTTCTGCAG GTCCACTCTCAATCACTCTCAG
Il6	Forward Reverse	ACAAAGAAATGATGGATGCTACC GTATCTCTCTGAAGGACTCTGG
Ccl2	Forward Reverse	TCTCTCTTCCTCCACCAC CTCTCCAGCCTACTCATTG
Cxcl5	Forward Reverse	TCAGTCATAGCCGCAACG GGGTCAGAGTCCTCAGAAATC
Cxcl1	Forward Reverse	GTCATAGCCACACTCAAGAATG GAACAAGCAGAACTGAACTACC
Il1b	Forward Reverse	CTACAGGCTCCGAGATGAAC TTCTTCTTTGGGTATTGCTTGG
Hprt1	Forward Reverse	TGGCCCTCTGTGTGCTCAA TGATCATTACAGTAGCTCTTCAGTCTGA

2.6. Western blot analysis

Mouse carotid artery was ground to a fine powder as described in 2.5. The powered sample was equilibrated on ice and then suspended in 80uL of protein lysis buffer (CST 9803S) supplemented with Protease Inhibitor Cocktail (Research Products International, P50600–1). Bio-Rad TGX protein gel was used for electrophoresis and Trans-Blot Turbo PVDF transfer pack was used for protein transfer. Antibodies were used as follows: PCNA (Invitrogen or ThermoFisher Scientific, PA5–32541), ACTA2 (Sigma, A2547), and TUBA (Sigma, T5168).

2.7. Human saphenous vein sample analysis

The human saphenous vein (HSV) RNA samples were kindly shared by the laboratory of Dr. Neal Weintraub. RNA samples were extracted from segments of fresh HSV or subjected to 7 days of ex vivo culture following the Weintraub lab protocol [18]. RNA samples were processed to cDNA synthesis and Real-Time PCR as described in 2.5.

2.8. Bioinformatic analyses of public microarray and RNA-seq datasets

Microarray datasets GSE57691, GSE66240 and GSE43292 were accessed in GEO database and the IL-8 expression value in individual samples were extracted using GEO2R platform.

Raw reads of RNA-seq datasets GSE163244 and GSE69637 were downloaded from GEO database. Low quality reads were removed using Trimmomatic 0.38 to get the pass-filtered reads. The latter were then aligned to the human reference genome hg38 using TopHat 2.1.0. Raw counts of IL-8 were estimated using featureCounts 1.6.2. Counts per Million (CPM) or Fragments Per Kilobase of exon per Million fragments mapped (FPKM) values were calculated and plotted for the graphs represented in Fig. 1D and E, respectively. Differential expression and FDR adjusted p-value were analyzed using DESeq2 R package. FDR < 0.05 was considered significant.

Fig. 1.

IL-8 expression in diseased human arteries, pathologically stimulated human VSMCs, and ex vivo cultured human veins. (A-C) IL-8 expression assessed by microarrays in (A) specimens from 10 control organ donors, 20 small AAAs and 29 large AAAs (GSE57691), (B) 7 unruptured cerebral aneurysm and 10 healthy arteries. (C) 32 carotid endarterectomy specimens comparing self-healthy part with the atheroma part (GSE43292). (D, E) IL-8 level in bulk RNA seq data from HCASMCs treated with cholesterol (GSE163244) (D) or PDGF (GSE69637) (E). (F) QPCR results of IL-8 mRNA level in cultured HSV segments compared to the uncultured segments. Statistical analysis: A, ordinary one-way ANOVA; C, paired t-test; B and F, Mann-Whitney test (unpaired and non-parametric). D and E, adjusted P values.

2.9. Statistical analyses

Data are presented as mean value scatter plot. The error bar represents SD. Statistical analyses were performed by Mann Whitney test (nonparametric, unpaired) for comparisons of skewed data (Fig. 1B, C, F, 2D and 4E); unpaired t-test (parametric) for Fig. 2B, 3B, D, 4F and Fig. S1; One-way ANOVA for Fig. 1A and D. P < 0.05 were considered significant and indicated with asterisk.

Fig. 2.

IL-8 aggravated neointimal hyperplasia after carotid ligation. (A) schematic for the generation of BAC IL-8 Tg mice. (B) IL-8 expression in carotid arteries 3 days after ligation. (C) Representative macroscopic view of the ligated carotid arteries, levels of sections and representative images of H&E-stained cross sections from each level. (D) Areas of neointima and media were measured in ImageJ (Ctrl = 8, Tg = 9, mixed gender). An unpaired, nonparametric Mann-Whitney test was used for individual levels. (E) IF staining of IL-8 on cross sections, showing representative 1 pair out of 4 pairs.

Fig. 4.

Neutralization of IL-8 by perivascular antibody administration at the ligated carotid artery. (A) Schematic shows application of pluronic F127 gel. (B) Macroscopic view of representative carotid arteries 2 weeks after ligation. (C) Levels of sections at the proximal region of ligation. (D) Representative images of cross-sections from level 1 with H&E staining. (E) neointima areas at cross sections from each level were measured in Image J (IgG = 5, Anti-IL-8 = 6, mixed gender). An unpaired, nonparametric Mann-Whitney test was used for individual levels. (F) Total volume of neointima calculated from Image J measurements. (G) IF staining of IL-8 on cross sections, showing representative 1 pair out of 3 pairs.

Fig. 3.

IL-8 promotes tissue inflammation and cell proliferation in the ligated carotid artery. (A-B), Upregulated Mac2 staining in the neointima of Tg mice. L: Lumen, M: Media, N: Neointima. (C–D) Significantly increased PCNA levels and decreased ACTA2 levels in the ligated carotid artery of Tg mice.

3. Results

3.1. IL-8 positively associates with vascular diseases and pathological stimuli in VSMCs

To gain more insights into the association between IL-8 and vascular diseases, we analyzed three sets of microarray data from the Gene Expression Omnibus (GEO) database, which were collected by researchers worldwide [19–21]. IL-8 mRNA level was significantly higher in the specimen of diseased vessels than in the healthy control vessels. The elevated expression of IL-8 was measured in vessels of different origins or diseases including abdominal aortic aneurysm (AAA) (Fig. 1A), cerebral aneurysm (Fig. 1B), and carotid atheroma (Fig. 1C). In agreement with these microarray results, our recent QPCR data showed the upregulated IL-8 levels in both AAA and atherosclerosis specimens [17].

VSMCs are the major source of cells and extracellular matrix in both healthy and diseased blood vessel walls. In the pathological milieu VSMC undergoes a phenotypic modulation and results in a massive change in its transcriptome and proteome. We reported that VSMCs produce excessive IL-8 when exposed to pathological stimuli such as IL1β and TNFα [17]. In the present study, we analyzed two sets of publicly available bulk RNA sequencing data from independent groups (GSE163244 and GSE69637). IL-8 level was significantly increased when VSMCs were exposed to cholesterol or PDGF (Fig. 1D, E). IL-8 upregulation was also found in the ex vivo cultured human saphenous vein (HSV), as shown in Fig. 1F. HSV underwent a wall thickening after culture, majorly attributed to VSMC proliferation [22].

3.2. Carotid neointimal hyperplasia is exacerbated by IL-8

Given the results shown in Fig. 1 that IL-8 is positively associated with vascular wall pathology and highly upregulated in the phenotypically changed VSMCs, we hypothesized that IL-8 promotes VSMC phenotypic modulation and neointimal formation. To prove this hypothesis, we performed carotid ligation in the BAC IL-8 transgenic (IL-8 Tg) mouse which was a generous gift from Dr. Timothy Wang [14]. BAC RP11–997 L11 encompassing the entire IL-8 gene locus along with upstream and downstream regulatory elements, was integrated into the mouse genome (Fig. 2A). IL-8 expression in the ligated(L) left carotid artery was substantially elevated as early as 3 days (Fig. 2B) and sustained until 2 weeks (Fig. 2E). In the unligated (UL) right carotid of the same mouse, the counterpart of the ligated one, a trace amount of IL-8 was also detected, likely attributed to systemic inflammation (Fig. 2B, E). At 2 weeks, the vessel wall of the ligated left carotid artery was significantly thickened as the neointima formed. At the proximal region to the ligation site (<400um), which is level 1 and level 2 shown in Fig. 2C, the cross-section area of neointima in the artery of Tg is comparable to that of Ctrl. However, at a distal region(>400um) to the ligation site, shown as level 3 in Fig. 2C the neointima area diminished in Ctrl but sustained in Tg. In Fig. 2D, the increase neointima area was observed at all 3 levels but the significance was only observed at level 3. The averaged value for the thickness of media layers was slightly higher in Tg than Ctrl. The lumen areas showed no difference at level 1, but a significant decrease at level 2 and level 3, comparing Tg to Ctrl. In summary, at level 3 the more distal region to the ligation site, there was an increased neointimal area and a decreased lumen area in IL-8 Tg mice.

3.3. IL-8 promotes tissue inflammation and cell proliferation in the ligated carotid artery

To test our hypothesis that IL-8 promotes tissue inflammation and cell proliferation in the injured arterial wall, we measured the levels of Mac2 and PCNA. In the injured vessel wall Mac2-expressing cells are the infiltrated macrophages or monocytes, indicating a tissue inflammation. At two weeks after ligation, the ratio of Mac2 staining to DAPI staining in the total areas of neointima and media was significantly higher in Tg mice compared to that in Ctrl mice (Fig. 3A,B).

PCNA was barely detectable in the unligated carotid artery, however, was substantially upregulated in the ligated arteries at two weeks post-injury. In Tg mice, the PCNA level was further increased to two folds of that in Ctrl mice (Fig. 3C, D). ACTA2 is a contractile marker selectively expressed in VSMCs and myofibroblasts. Repeatably, ACTA2 protein level is lower in the ligated carotid compared to the unligated counterpart. The decrease of ACTA2 after ligation was more dramatic in IL-8 Tg mice than that in Ctrl mice (Fig. 3C, D). The decreased ACTA2 was attributed to two factors: first, decreased VSMC contractility resulting from phenotypic modulation; second, decreased VSMC portion in the vessel wall due to the increased non-SMC components.

3.4. Neutralization of IL-8 inhibits neointima formation

To further confirm the causative role of IL-8 in neointima formation and to investigate the therapeutic potential of anti-IL-8, we applied antibodies at the ligation site to neutralize IL-8. Mouse anti-human IL-8 or the isotype IgG control was mixed with Pluronic gel and applied topically on the ligated artery (Fig. 4A). Two weeks after ligation and antibody treatment, carotid arteries were harvested for evaluation. From the macroscopic perspective, in the IgG-treated control arteries the outward remodeling was remarkable (Fig. 4B, left). In contrast, after anti-IL-8 treatment the outward remodeling was dramatically diminished (Fig. 4B, right). At level 1 the proximal region (<200um) (Fig. 1C), neointimal area was significantly less in the anti-IL-8 group compared to IgG group (Fig. 4D, E). At level 2 and the more distal region of the anti-IL-8 group neointima formation was almost abolished (Fig. 4B, E). In contrast, arteries from the IgG control group exhibited wall remodeling from bifurcation to the root. Thus, we calculated the total neointimal volume of the entire carotid artery. Shown in Fig. 4F, the total neointima volume in anti-IL-8 artery was reduced by 60 % compared to IgG control. The neutralization of IL-8 was validated as the density of IL-8 signal was drastically decreased after anti-IL-8 treatment (Fig. 4G). These results demonstrated that the neutralization of IL-8 at the ligated carotid artery inhibited neointima formation.

4. Discussion

The findings presented here, either interrogated from public data or the data generated on our own, showed the upregulation of IL-8 in the diseased blood vessel walls and in the phenotypically modulated VSMCs. More importantly, we provided the first in vivo evidence that the insult-triggered IL-8 expression exacerbated neointimal hyperplasia. The strategy of BAC transgene necessitates a pathophysiological regulation of human IL-8 expression in the mouse, i.e. a suppressed transcription under the basal or physical condition and an activated expression under the pathological condition.

The complete ligature placed at carotid bifurcation introduces both mechanical stress and flow stagnation to the common carotid artery. These two factors are the main drivers of pathogenesis, which however, attenuated with distance. The decreased severity of stenosis and the reduced development of neointima was observed when one moves away from the ligature. At the most proximal region, i.e. level 1, the pathological responses reached near a maximal level and IL-8 failed to further worsen it. In contrast, in the more distal regions at level 2 or 3, where the pathology was moderate, IL-8 substantially exacerbated it.

In the Tg mouse carotid artery at 3 days post-injury, when IL-8 was largely elevated there was no increase of mouse Cxcl1, or any other major cytokines compared to the Ctrl (Sup F1). This suggests a nonredundant role of IL-8 in promoting neointimal hyperplasia. In literature, mouse GRO-a (CXCL1) was sometimes considered to be the murine ortholog of IL-8. This is because IL-8 shares common features with GRO-a in terms of structure and function. However, accumulated evidence showed that these two are functionally distinct. IL-8 but not GRO-a activated phospholipase D [23], while only GRO-a but not IL-8 induced a calcium influx in neutrophil [24]. A recent clinical report confirmed the positive association between IL-8 and carotid intima-media thickness (c-IMT) but not between GRO-a and c-IMT [25]. This report emphasizes the divergent roles between IL-8 and GRO-a in vascular diseases, even though they are highly alike. Accordingly, cautions should be used when a conclusion was drawn about the in vivo function of IL-8 based on the interpretation of the roles of GRO-a or CXCR2, the common receptor for IL-8 and GRO-a.

The uniqueness of IL-8 as a cytokine is known to be more resistant, than others, to proteolysis, temperature, and acidic environment at the site of inflammation [26,27]. This suggests its potential role in prolonged chronic inflammation which is the culprit in cardiovascular diseases. We here showed that in the ligation-injured carotid artery wall, IL-8 promoted the infiltration of Mac2+ cells and the arterial cell proliferation, consistent with previous reports that IL-8 was a chemoattractant for monocytes/macrophages [28] and a mitogen for VSMCs [29].

There are some limitations in this study. First, we did not characterize which type(s) of cells was the predominant source of IL-8 at the neointimal lesion. Second, we did not characterize which was the major type of immune cells recruited by IL-8 in response to injury. Third, we did not characterize which receptor of IL-8 played the dominant role in the IL-8-promoted neointimal hyperplasia.

Myeloid-derived suppressor cells (MDSCs) were identified to be the major type of immune cells recruited by IL-8 to tumor [14]. MDSCs play critical role in anti-tumor immunity and thus promotes tumor growth. In both pre-clinical and clinical studies it was shown that the neutralization of IL-8 mitigated MDSCs recruitment [35] and increased immune surveillance at the tumor microenvironment [36]. Whether or not this is the case in vascular diseases such as atherosclerosis or aneurysm need to be precisely evaluated in BAC Tg mice. MDSCs play complex roles in vascular disease. They can be either detrimental or beneficial to vascular health, depending on the nature of disease. Of note, the systemic administration of anti-IL-8 in the above-mentioned studies has shown to meet the safety needs, with mild to moderate adverse effects.

A general concern about BAC transgene is the cross-species differences between mouse and human. Although the transcriptional machinery for human chromosome 21 in mouse Tc1 hepatocytes resembles that in human cells [30], the in vivo transcription of hIL-10 in the hIL-10 BAC mice exhibited a cell type-dependent manner that is different from human [31].

Alternative strategies have been developed to study the tissue-specific function of human IL-8 in mouse body. One transgenic mouse was designed to selectively express IL-8 in its intervertebral disc (IVD), based on their observation that in human patients, lL-8 expression in IVD was associated with back pain [32]. To achieve the goal of a tissue-specific IL-8 expression, the authors constructed a plasmid that contains the IL-8 cDNA driven by a cre-dependent promoter. The plasmid was then integrated into the mouse genome. Mice that express human IL-8 in their IVD showed similar behavior with human patients, i.e., they were less active and tended to eat more food [32,33]. Another transgenic mouse expresses hIL-8 specifically in the bronchial epithelial cells, under the control of cell specific promoter [34]. This lung-targeted hIL-8 expression resulted in airway neutrophilia and a protective effect against bacterial infection in lung. These reports suggest that IL-8 plays active roles in pathogenesis than being merely a marker of disease.

In summary, our results suggested the regulation and expression of human IL-8 in BAC Tg mice resembled that in human. We provided new insights into the in vivo role of human IL-8 in vascular remodeling and stenosis. Further understanding of the role of IL-8 in cardiovascular diseases will set a stage for novel therapeutic approaches.

Data availability

Data will be made available on request.