Background
Colorectal cancer is the third most common cancer and the second leading cause of cancer death worldwide [
1]. Environmental, lifestyle and genetic factors are related to the occurrence of colorectal cancer [
2,
3]. Recently, however, it was determined that imbalance of the gut microbiota and immunity was the cause of 90% of colorectal cancer [
1,
4]. A large amount of epidemiological and experimental evidence has shown that chronic inflammation is crucial to the induction and progression of colorectal cancer [
4,
5], but the underlying mechanism is still unclear. Chronic inflammation often leads to the continuous functional destruction of mucosal and gut microbiota protective barriers, which directly induce the mutation of intestinal epithelial cells and trigger the development and progression of colorectal cancer [
6‐
8]. Therefore, in addition to chemotherapy or targeted drug therapy, immunotherapy and gut microbiota therapy are presently being considered for colorectal cancer patients [
6]. However, there is still a lack of research on the regulatory mechanism and targeting strategy of immunity and the gut microbiota in the development of colorectal cancer, which greatly limits research on treatment strategies for colorectal cancer, especially colitis-associated colorectal cancer.
Neutrophils are the one of infiltrating immune cells in the tumor microenvironment that maintain tumor progression, especially in colitis-associated tumorigenesis [
2,
9]. Previous studies have shown that CD11b
+Gr1
+ tumor-associated neutrophils are the main infiltrating immune cells of colitis-associated colorectal cancer [
1,
9‐
11]. However, their precise immune regulation mechanism in colitis-associated tumorigenesis and their relationship with the gut microbiota have not been elucidated.
Aryl hydrocarbon receptor nuclear translocator (ARNT) is a transcription factor and belongs to the basic helix loop helix (bHLH)-per Arnt sim (PAS) family [
12]. It is generally believed that ARNT plays a key role in two different fields: the aryl hydrocarbon receptor (AHR) and hypoxia inducible factor (HIF) pathways [
13‐
15]. HIF1α and AHR have been shown to play important roles in tumorigenesis and proliferation, but the role of ARNT is still unclear. It is generally believed that its expression is constant, and it may not play a critical role [
12]. However, recent studies suggest that ARNT may also play an important role, especially in tumor growth [
13‐
15].
In this study, we found that ARNT deficiency in neutrophils, drives their recruitment, neutrophil extracellular trap (NET) formation, inflammatory cytokine secretion and suppressive activities in a gut microbiota-dependent manner in colitis-associated tumorigenesis.
Materials and methods
Mice
C57BL/6
Arntfl/fl and
Lyz-Cre mice were obtained from the Jackson Laboratory and extensively backcrossed to the C57BL/6 background. Wild-type (WT) controls from ARNT knockout mice included Cre
+ mice (
Lyz-Cre) to account for the effects of
Cre, as described previously [
16‐
18]. Sex-matched littermates at 8 to 10 weeks of age were used in the experiments described in this study. All mice were bred and maintained under specific pathogen-free conditions. All experiments used cohoused littermates unless otherwise indicated so that consistency of common microbiota and genetic background/alterations would be ensured. All animal experimental protocols were approved by the Animal Ethics Committee of Beijing Institute of Microbiology and Epidemiology and Beijing Normal University (IACUC-DWZX-2017-003 and CLS-EAW-2017-002; Beijing, China). Animals were maintained in pathogen-free conditions.
Tumor model and histological analysis
The colitis-associated colorectal cancer mouse model was established in our lab and other labs as described previously [
19‐
21]. Mice of each genotype were given a single intraperitoneal administration of azoxymethane (AOM; Sigma–Aldrich, St. Louis, MO, USA; 10 mg/kg body weight). Five days later, these mice were randomly divided into two groups fed 2% dextran sulfate sodium salt (DSS; MP Biomedicals, Santa Ana, CA, USA) in drinking water for 3 cycles or plain water as a control. The experiment was terminated 80 days after the first injection of AOM. Mice were weighed and observed daily. Compared with baseline, the percentage of mouse weight loss was used to evaluate the severity of colitis, and the progress of the disease was observed in combination with daily observation of clinical symptoms such as rectal bleeding, diarrhea or prolapse. If there was a significant weight loss of more than 20% or obvious clinical symptoms, the experiment was terminated, and a humane end point was applied to the mice. As the AOM/DSS model is quite reliable in producing tumors, it is usually unnecessary to observe tumor growth with endoscopy, as described previously [
21]. To establish subcutaneous tumors, 5 × 10
6 MC38 colon cancer cells were injected into C57BL/6 mice, half male and half female, randomization group. These cells formed a tumor of 1–2 cm diameter within 2–3 weeks of injection and double blinding detecting of mouse tumor size, as described previously [
22].
For neutrophils adoptive transfer experiments, CD11b+Gr1+ cells were isolated from femurs of AOM/DSS-treated WT mice by a Cell Isolation Kit (Miltenyi Biotec, Cologne, BG, Germany) after lysis of red blood cells (RBCs) according to the manufacturer’s instructions. The first flow through from the column contains unbound immune cells, which were used as controls (non-neutrophils). CD11b+Gr1+ cells were eluted twice, and the purity of CD11b+Gr1+ cells was more than 95% by flow cytometry. Isolated cells were subjected to neutrophil adoptive transfer experiments. A total of 5 × 106 purified neutrophils was intravenously injected into one recipient mouse twice a week from the beginning of DSS treatment to the end of the experiments. To deplete CD11b+Gr1+ cells in vivo, 0.5 mg of depleting anti-Gr1 mAb (RB6-8C5; Biolegend, San Diego, CA, USA) was i.p. injected into the recipient mice on day − 1 before tumor induction, after which they were injected twice a week. To delete the mouse gut microbiota, broad-spectrum antibiotics were added to water, including 10 g/l ampicillin, 10 g/l neomycin sulfate, 10 g/l metronidazole, and 5 g/l vancomycin. The mixed antibiotic water was given by gavage in 0.5 ml once every two days.
At the end of the experiments, some colons from each group were used to count tumors and determine inflammation scores, and the rest were used to examine the profiles of immune cells. For histologic analysis, 4 μm thick sections from all groups were stained with hematoxylin and eosin (HE) to examine colonic inflammation and tumor morphology. Histological scoring of inflammation was determined as described previously [
18,
19]. Briefly, inflammatory scores were graded by the amount of inflammation, the depth of inflammation, and the amount of crypt damage or regeneration. The unstained sections were subjected to immunostaining.
Immunohistochemistry (IHC)
The colon or colorectal tumor tissues from mouse models were collected and fixed in 10% formalin overnight and embedded in paraffin. Formalin-fixed paraffin-embedded tissue was cut into 4 μm sections. The sections were stained using standard protocols for xylene and an alcohol gradient for deparaffinization. After antigen retrieval and unmarking procedures, the primary rat anti-mouse Gr1 antibody (1 ng/ml in 50 µl volume; RB6-8C5; Biolegend, San Diego, CA, USA) was incubated and stained. Sections were incubated with HRP goat anti-rat IgG antibody (2.0 µg/ml; Poly4054; Biolegend, San Diego, CA, USA) for 40 min and then developed with an Ultravision DAB Plus Substrate Detection System (TA-125-QHDX, Thermo Fischer Scientific, Waltham, MA, USA) for 2–5 min at room temperature, followed by hematoxylin staining, dehydration and coverslipping with Permount. Immunohistochemistry (IHC) slides were scanned with a Pannoramic Digital Slide Scanner (SDHISTECH, Budapest, Hungary), and images were cropped from virtual slides in Pannoramic Viewer.
Neutrophil isolation and cell culture
At eighty days after the first AOM injection, the mouse colon was taken. Tumor tissue was carefully isolated, and the rest of the nontumor colon tissue was cut into pieces and washed with RPMI 1640. Then, the colon fragments were resuspended in a 2 ml solution of 1 mg/ml collagenase XI (Sigma–Aldrich, St. Louis, MO, USA) containing 20 U/ml DNase I (Sigma–Aldrich, St. Louis, MO, USA) and incubated at 37 °C for 30 min. PBS with 1% FBS and 5 mM ethylene diamine tetraacetic acid (EDTA) was used to neutralize digestion. Cells were washed twice (452 g, 5 min), resuspended in RPMI 1640, and filtered to remove clumps. Murine bilateral femurs and tibias were taken, and 10 ml syringes were used to flush the medullary cavity with RPMI 1640. After passing through the 200-mesh net, the RBCs were lysed and washed twice (452 g, 5 min), and the cells were resuspended in RPMI 1640. After peripheral blood was obtained from the retroorbital vein, the RBCs were lysed, washed twice and resuspended in RPMI 1640. The spleens were collected and ground. The RBCs were lysed and washed twice (452 g, 5 min), and the cells were resuspended in RPMI 1640. Mesenteric lymph nodes (MLNs) were taken, ground and washed twice (452 g, 5 min), and the cells were resuspended in RPMI 1640. Spleen and tumor tissues were dissected and digested with 0.7 mg/ml collagenase XI (Sigma–Aldrich, St. Louis, MO, USA) and 30 mg/ml type IV bovine pancreatic DNase (Sigma–Aldrich, St. Louis, MO, USA) for 45 min at 37 °C. Finally, single-cell suspensions were prepared from blood, spleen, BM, MLN, colon or tumor samples for further assay. CD11b
+Gr1
+ neutrophils were isolated from single-cell suspensions of the spleen, BMs, colon or tumor tissues by cell sorting on a FACSAria (BD Biosciences, Franklin Lake, NJ, USA), as previously described [
10,
23].
BM-derived CD11b
+Gr1
+ cells were generated as described previously [
24‐
26]. In brief, BM cells were cultured in complete DMEM supplemented with 2 mM l-glutamine, 10 mM HEPES, 20 mM 2-ME, 150 U/ml streptomycin, 200 U/ml penicillin, and 10% FBS and stimulated with combinations of recombinant murine GM-CSF (10 ng/ml, Peprotech, Rocky Hill, NJ, USA). The cultures were maintained at 37 °C in a 5% CO
2-humidified atmosphere for 4 days. Normal human peripheral blood neutrophils (915,410, Beijing Nuowei Biology, Beijing, China) were treated by GNF351 (50–500 nM, MCE) and stimulated for 12 h with LPS for further analysis.
Immunosuppressive activities assay
CD11b
+Gr1
+ cells were sorted from the spleen, BM, colon or tumor tissues by flow cytometry. The suppressive activities of neutrophils were assessed by determining their abilities to inhibit T-cell proliferation as described previously [
16,
27]. C57BL/6 CD4
+ T cells (1 × 10
5 cells/well) were cocultured with 15 µg/ml mitomycin C (Sigma–Aldrich, St. Louis, MO, USA)-pretreated BALB/c splenocytes (1 × 10
5 cells/well), and neutrophils (1 × 10
5 cells/well) were sorted at different ratios in a flat-bottom 96-well plate at 37 °C in 5% CO
2. Cell proliferation was determined 72 h after incubation with [
3 H] thymidine for the last 16 h of culture.
Flow cytometry
Single-cell suspensions were prepared from the blood, spleen, BMs, colon or tumor tissues. For flow cytometric analysis of cell surface markers, cells were stained with mAbs in PBS containing 2% (w/v) BSA and 0.1% NaN
3 for 30 min at 4 °C, as described previously [
28]. The following mAbs were obtained from eBioscience (San Diego, CA, USA): FITC rat anti-mouse CD11b (M1/70), PE rat anti-mouse CD11b (M1/70), APC rat anti-mouse CD11b (M1/70), PE rat anti-mouse Gr1 (RB6-8C5), APC rat anti-mouse Gr1 (RB6-8C5), FITC rat anti-mouse Gr1 (RB6-8C5), FITC rat IgG 2b (eB149/10H5), PE rat IgG 2b (eB149/10H5), and APC rat IgG 2b (eB149/10H5). The following mAbs were obtained from Biolegend (San Diego, California, USA): PE/Cyanine7 rat anti-mouse CD45 (30-F11). The following antibodies were obtained from R&D system (Minnesota, USA): PE rat anti-mouse CXCR2 (242,216), APC rat anti-mouse CXCR2 (242,216), PE rat anti-mouse CD115 (460,615), PE mouse anti-human CXCR2 (48,311), PE rat IgG2a (54,447), APC rat IgG2a (54,447) and PE mouse IgG2a (20,102). The following antibodies and reagents were obtained from Abcam (Cambridge, UK): APC rat anti-mouse Ly6G (RB6-8C5), FITC rat anti-mouse Ly6C (HK1.4), FITC rat IgG2c (RTK4174), and 7-AAD staining solution.
FACS-based intracellular staining of cytokines was performed as previously described [
28]. Cells were stimulated with LPS (100 ng/ml, Sigma–Aldrich, St. Louis, MO, USA) and GolgiPlug (BD Pharmingen, Lake Franklin, NJ, USA) for 5 h. BD Cytofix/Cytoperm and BD Perm/Wash buffer sets were used according to the manufacturer’s instructions (BD Pharmingen, Lake Franklin, NJ, USA). APC rat anti-mouse IL-10 (JES5-16E3) and its APC rat IgG2b isotype control (RTK4530); APC rat anti-mouse tumor necrosis factor α (TNFα; MP6-XT22) and its APC rat IgG1 isotype control (RTK2071) were obtained from Biolegend (San Diego, CA, USA). The same amount of isotype control staining used as the negative control.
Flow cytometry data were acquired on an ACEA NovoCyte (ACEA Biosciences, Inc., San Diego, CA, USA), and data were analyzed with NovoExpress or Flow Jo (TreeStar, San Carlos, CA, USA). The viability dye was used to exclude dead cells and isotype matched immunoglobulins staining was performed to set up the regions and quadrants of the negative control, and then the percentages of antibody staining of various targets were determined.
Quantitative RT–PCR
RNA was extracted with a RNeasy kit (QIAGEN, Dusseldorf, Germany), and cDNA was synthesized using SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA, USA). An ABI 7900 real-time PCR system was used for quantitative PCR, with primer and probe sets (Supplementary Table
1) obtained from Applied Biosystems (Carlsbad, CA, USA). The results were analyzed using SDS 2.1 software (Applied Biosystems, Foster City, CA, USA). The expression of each target gene is presented as the fold change relative to that of control samples, as described previously [
22,
28].
AHR knockdown with RNA interference
A gene-knockdown lentiviral construct was generated by subcloning gene-specific short hairpin RNA (shRNA) sequences into lentiviral shRNA expression plasmids (pLL3.7). The AHR shRNA sequence was 5′-AAG UCG GUC UCU AUG CCG CTT-3′, and the control shRNA sequence was 5′-GCG CGC UUU GUA GGA UUC GTT-3′. Lentiviruses were harvested from the culture supernatant of 293T cells transfected with shRNA vector. BM cells were cultured in complete DMEM supplemented in the presence of recombinant murine GM-CSF for 4 days to induce CD11b+Gr1+ neutrophils. Neutrophils were infected with recombinant lentivirus, and GFP-expressing cells were isolated using fluorescence sorting 48 h later. AHR expression was confirmed using qPCR. The sorted CD11b+Gr1+ cells with control or shRNA vectors were used for further assays.
Retroviral transduction of ARNT
Retroviral transduction ARNT were cloned into the MSCV retroviral vector (Clontech Laboratories, Mountain View, CA), as previously described [
25]. Phoenic-Eco packaging cells (ABP-RVC-10,001; Allele Biotechnology, San Diego, CA) were transfected with Lipofectamine (Invitrogen), and recombinant retrovirus was collected 48 h after transfection. After 2 days of differentiation, BM cells were cultured in complete DMEM supplemented in the presence of recombinant murine GM-CSF for 4 days to induce CD11b
+Gr1
+ neutrophils. Neutrophils were transduced with retroviral supernatant by spin inoculation and GFP-expressing cells were isolated by flow cytometry and performed the further analysis.
Western blot
BM cells were cultured in complete DMEM in the presence of recombinant murine GM-CSF for 4 days to induce CD11b+Gr1+ neutrophils. The sorted CD11b+Gr1+ cells were washed twice with cold PBS and lysed in RIPA buffer (50 mM Tris-HCL, pH 7.4, 1% NP-40, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM EDAT, pH 7.4) for 10 min on a rocker at 4 °C. The protein concentration was determined via bicinchoninic acid assay (BCA; Beyotime, Shanghai, China). The protein samples were separated by 10% SDS–PAGE and then transferred onto 0.22 μm polyvinylidene fluoride membranes (Merck Millipore, Bedford, MA, USA). The membranes were blocked with 5% nonfat dried milk for 1 h at room temperature and incubated with rabbit anti-mouse primary antibodies (1:200) overnight on a shaker at 4 °C. Subsequently, HRP-coupled secondary goat anti rabbit antibody (1:10000; Beyotime, Shanghai, China) was added for 1 h at room temperature. After sufficient washing, protein samples were detected with an eECL Western Blot Kit (Cat: CW0049M, CWBIO, Taizhou, Jiangsu, China) using AllDoc-x software with a Tanon 5200 Imager (Tanon, Shanghai, China). The following primary Abs were used: anti-ARNT (D28F3) was obtained from Cell Signaling Technology (Danvers, MA, USA), anti-β-actin (AC-15) was obtained from Sigma–Aldrich (St. Louis, MO, USA), anti-AHR (A3) and anti-CYP1A1 (B4) were obtained from Santa Cruz Biotechnology (Santa Rosa, CA, USA), anti-cyclooxygenase (COX) 2 was obtained from Abcam (Cambridge, UK), and anti-CYP1B1 (K008135P) and anti-AHRR (K007675P) were purchased from Beijing Solarbio Science & Technology Company (Beijing, China).
ELISA
Colorectal cancer induction for 80 days after the injection of AOM and the addition of DSS to the drinking water as described in the tumor model section. The concentrations of serum CXCL1 and CXCL2 at day 80 were quantified by sandwich ELISA. Before execution, fresh mouse feces were collected and partly resuspended with PBS. Fecal supernatant was used to stimulate RAW264.7 cells for 3 h, and the concentrations of TNFα and IL-1β in the culture supernatant were detected by sandwich ELISA. Mouse CXCL1 (MKC00B) and CXCL2 (MM200) ELISA kits were obtained from R&D Systems (Minneapolis, MN, USA), and mouse IL-1β (#abs520001), mouse TNFα (#abs552812), human TNFα (#abs510006) and human IL-10 (#abs510005) ELISA kits were obtained from Absin Biotechnology Co., Ltd (Shanghai, China), as described previously [
22].
Fecal collection and 16 S rRNA gene sequencing
Colorectal cancer induction for 80 days after the injection of AOM and the addition of DSS to the drinking water. At day 80, fresh stool pellets were obtained and immediately frozen at -80 °C. Fecal DNA was extracted from the feces using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions.
The V4 region of the 16 S rRNA gene was selected as the sequencing region to compare the diversity and structure of the bacterial species in each of the samples. The primers for the V4 region were 515 F (GTGCCAGCMGCCGCGGTAA) and 806R (GGACTACHVGGGTWTCTAAT). Sequencing was performed by Illumina MiSeq at QualityHealth Bioinformatics Technology Co., Ltd. (Beijing, China).
Sequence analysis
Paired-end reads 250 bp in length in each direction were generated, and the overlapping reads were stitched together. The raw paired-end reads were assembled using Pandaseq. A self-written script was run to discard low-quality reads with an average quality score < 20, containing > 3 nitrogenous bases, and with lengths of < 220 or > 500 nucleotides. The high-quality sequences were then treated to filter singletons and remove chimeras.
The 16 S rDNA sequence data were analyzed using UPARSE software (UPARSE v7.0.1001,
http://drive5.com/uparse/), which divided the sequences into OTUs using a similarity threshold of 97%. A similarity score of < 97% was considered indicative of a different species, and a score of < 93–95% was considered indicative of a different genus. The SSUrRNA database was used to annotate the species, and MUSCLE (Version 3.8.31,
http://www.drive5.com/muscle/) was used to BLAST the sequences. Finally, all data were normalized for further analysis.
Diversity within samples (α-diversity) and between samples (β-diversity) was estimated on the basis of the OTUs. α-Diversity refers to the diversity of a particular region or ecosystem and is an expression of the species diversity in a single sample. β-Diversity indices were used to estimate the distance between samples based on the evolutionary relationship of the sample sequence and its abundance. β-Diversity indices are expressed in terms of the differences between the sample groups by means of principal coordinate analysis (PCoA). Metastat and LDA effect size (LefSe) analyses were used to detect significant differences among groups in the microbial community biomarkers.
Fresh stool pellets were collected from colorectal cancer model mice at the day 80, and fecal supernatant was obtained after feces were resuspended in PBS. All the fecal supernatant was filtered. Real-time cellular analysis (RTCA) was used to assess the effect of fecal metabolites on intestinal epithelial cells. Briefly, HT-29 cells (6 × 104/well) were plated on E-plates, then put the E-plates into XCELLigene (ACEA Bioscience, USA), which was set the automatic detection interval every 5 min. Five hours later, fecal filtrate (100 µl/well) was added to E-plates. The E-plates were detected for 20 h. The cell response curves and IC50 values of different time periods can be obtained by continuous real-time dynamic detection.
Neutrophils were sorted from the single-cell suspensions of the BM or tumor tissue. A total of 2 × 105 cells was plated in 200 µl in a 96-well flat bottom plate and incubated for the indicated time, and 5% CO2 and Sytox Green (25 nM, Thermo Fisher, Waltham, MA, USA) were added and incubated for 5 min. Cells were fixed with 16% paraformaldehyde (PFA) and kept at 4 °C until confocal microscopy was performed to quantify NET formation. Z-Stacks (10–30 μm 40x magnification) were taken using an LSM800 equipped with a 488 diode and a Plan-Apochromat 1,3 N/An Oil DIC III objective. For NET area quantification, FIJI software and the particle analysis plugin were used. Only structures depicting NET morphology and positive for Sytox green were selected for area quantification, and intact granulocyte nuclei were excluded from the analysis. Triplicate wells of each condition were included.
Statistical analysis
All data are presented as the mean ± SD. Student’s unpaired t test was used to compare two sets of parametric data. When comparing three or more datasets, one-way analysis of variance with Dunnett’s post-hoc test was applied for parametric data, and a Kruskal–Wallis test was applied for nonparametric data. A comparison of survival curves was performed using the log-rank (Mantel–Cox) test. A P value of less than 0.05 was considered statistically significant.
Discussion
In the present study, we established that an underappreciated function of the transcription factor ARNT in colorectal cancer is to limit the recruitment and function of neutrophils coupled with the gut microbiota to enable colitis-associated colorectal cancer progression (Fig. S
12). ARNT deficiency enhances the migration and functional activities of neutrophils under physiological conditions and pathological tumor microenvironment. Although ARNT plays an important role in regulating the function and recruitment of tumor-infiltrating neutrophils, it is obvious that these effects depend on alterations in gut microbiota homeostasis in the tumorigenic microenvironment. Therefore, these data collectively display the essential relationship between the recruitment and function of neutrophils and homeostasis of the gut microbiota with implications for the combination of immune and gut microbiota regulation as a therapeutic approach for colorectal cancer.
Colorectal cancer includes hereditary, sporadic and colitis-associated colorectal cancers [
1,
3,
29]. A large amount of epidemiological and experimental evidence strongly supports the view that chronic inflammation leads to the occurrence, development and metastasis of colorectal cancer [
5,
26,
30]. Ulcerative colitis is the most common form of inflammatory bowel disease (IBD) and is associated with an increased risk of colorectal cancer. Because chronic inflammation is related to immunosuppression, a reasonable explanation for the development of chronic inflammation into cancer is a tumor microenvironment induced by immunosuppression and immune tolerance to tumor cells[
5,
10]. Neutrophils are one of the key regulatory factors in immune activities [
5,
10,
30]. Although the role and mechanism of neutrophils in tumors have been widely studied, the precise effects of ARNT on neutrophils remain unclear. Herein, our data showed that
Arnt−/− mice significantly promoted local accumulation of neutrophils in tumors, NET formation and secretion of anti-inflammatory cytokines by tumor-infiltrating neutrophils, all of which contribute to the formation of an immune-tolerant tumor microenvironment and promote the occurrence and development of colorectal cancer. Importantly, CD11b
+Gr1
+ neutrophils depletion significantly reversed the tumor-promoting effects of
Arnt−/− mice, suggesting that the absence of ARNT in neutrophils plays a key role in tumorigenesis. CXCR2 is critical for inducing the recruitment and NET formation of neutrophils in cancer [
9,
31]. We further investigated the expression of CXCR2 and found that the absence of ARNT significantly promoted the expression of CXCR2 in neutrophils under tumor or physiological conditions. However, blocking or inhibiting CXCR2 expression significantly downregulated local neutrophil accumulation, NET formation and proinflammatory cytokine production. These results collectively suggest that CXCR2 is required for the regulatory effect of ARNT on neutrophil recruitment and suppressive activities. Additionally, these data also suggest that ARNT is important for all neutrophil populations with regulatory activities under physiological or tumor conditions. However, the present phenotypic analysis cannot distinguish different neutrophils’ populations, and neutrophil origin cells should also be included in CD11b
+Gr1
+ neutrophils, such as myeloid-derived suppressor cells (MDSCs) from neutrophils origin.
ARNT is a transcription factor and plays a key role in signaling processes involving AHR and HIF [
12,
32]. AHR is a cytosolic bHLH PAS transcription factor that consists of domains similar to ARNT [
33]. Appropriate compounds are bound in the PAS domain of AHR, which leads to conformational changes and reveals a complete domain [
34]. AHR is then transferred to the nucleus, where it is immortalized with ARNT. The transcriptionally active AHR/ARNT complex binds to exogenous response elements in the regulatory region of the target gene and initiates transcription [
35]. AHR has been shown to play an important role in neutrophils induction and mobilization [
36,
37], but the precise role of ARNT in neutrophils remains unclear. It is generally believed that its expression is constant and may not play an important role. However, recent studies suggest that ARNT may also play an important role, especially in tumor growth [
38]. One study overexpressed ARNT in Hepa-1 cells in vitro, and according to the kinetic data, ARNT probably plays an important role in the early stage of tumor growth. In addition, the authors suggested that ARNT as a drug target is superior to HIF1α expression in some tumors of the control group [
32,
38,
39]. A recent study confirmed the role of ARNT as a potential target in the treatment of cancer, which linked the expression of the bHLH PAS transcription factor with cisplatin resistance [
40]. Cisplatin resistance is mediated by the ARNT/SP1-dependent transcription of the multidrug resistance 1 gene, which encodes an ATP-binding protein cassette transporter that promotes the action of this anticancer drug [
40]. Herein, our data demonstrated that ARNT negatively regulated the recruitment and function of neutrophils under physiological conditions or in the tumor microenvironment. ARNT deficiency displayed similar effects to AHR deficiency in neutrophils which is based most likely on a disrupted canonical AHR signaling pathway by the lack of its dimerization partner ARNT.
Recently, a large number of studies have shown that changes in the gut microbiota are key regulatory factors in the occurrence and development of colorectal cancer [
4]. Using 16 S rRNA sequencing of the gut microbiome, we noted that MDSC
Arnt−/− mice exhibited alterations in the gut microbiome and metabolome in colorectal cancer. It was found that the feces of ARNT-deficient mice could also replicate the tumor-promoting effect of ARNT-deficient neutrophils. Cohousing WT and
Arnt−/− mice to eliminate ARNT-deficient mice with different flora or antibiotics to eliminate the flora significantly weakened the protumor effect of ARNT-deficient neutrophils. These results suggest that the functional changes in neutrophils induced by ARNT deficiency could promote the occurrence and development of tumors by changing the gut microbiota. These results also provide experimental evidence for the future combination of immune and microbiological treatment studies of colitis-associated colorectal cancer.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.