Gut bacteria are required for the benefits of black raspberries in Apc Min /+ mice

2.1 Animals and BRBs

All animal study protocols were approved by the Medical College of Wisconsin Animal Care and Use Committee under the animal protocol AUA00002430 “ Identification of specific berry types and berry components that exhibit anti-inflammatory and anti-cancer activities in different animal cancer models.” Breeding pairs of Apc^Min/+ mice were purchased from the Jackson Laboratory (Bar Harbor, ME).

A synthetic diet from the American Institute of Nutrition (AIN-76A; Dyets Inc., Bethlehem, PA) was used as the control diet. BRB powder was purchased from Berri Products LLC (Corvallis, OR) and stored at 4° C. The sugar and starch content of the BRB diet was adjusted to create an isocaloric diet [37–40, 46].

2.2 Animal experiments

Four- to five-week-old Apc^Min/+ mice were randomly assigned to three study groups. The mice in groups 1 and 2 (G1 and G2) were fed regular drinking water and the control AIN-76A diet for 4 weeks. Then the mice in G1 continued on the control diet, while the mice in G2 changed to 5% BRB diet. The mice in group 3 (G3) were first given the control diet and antibiotics in the drinking water (1 g/L ampicillin, 1 g/L neomycin, 1 g/L metronidazole, and 0.5 g/L vancomycin) for 4 weeks. For the next 4 weeks, they were fed 5% BRBs along with the antibiotic treatment (Fig. 1A). At the end of the study, all the mice were euthanized by CO₂ asphyxiation, and the number of colonic and intestinal polyps was determined. Whole tissues of the colon and small intestine of all the mice were collected, fixed in formalin, and embedded in paraffin (FFPE). Hematoxylin and eosin (H& E)-stained tissue sections were evaluated histopathologically by our pathologists.

OPEN IN VIEWER

Fig.1

(A) Animal experimental protocol of the current study. Apc^Min/+ mice were randomly assigned to three study groups. The mice in G1 were fed regular drinking water and the control diet for 8 weeks. The mice in G2 were fed regular drinking water and the control diet for 4 weeks, and then change to 5% BRBs for additional 4 weeks. The mice in G3 were first given the control diet and antibiotics (1 g/L ampicillin, 1 g/L neomycin, 1 g/L metronidazole, and 0.5 g/L vancomycin) in the drinking water for 4 weeks. For the next 4 weeks, they were fed 5% BRBs along with the antibiotic treatment. (B) Antibiotics in the drinking water substantially decreased gut bacterial populations in Apc^Min/+ mice. wk: week.

RNA was isolated from FFPE samples of colon and small intestine according to the manufacturer’ s instructions (Recover All Total Nucleic Acid Isolation Kit for formalin-fixed and paraffin-embedded tissues, Ambion, Grand Island, NY). Quantitative PCR was performed to measure the relative expression levels of TLR4 (Mm.PT.58.41780308.g), NF-κ B1 (Mm.PT.58.30400172), and COX2 (Mm.PT.58.17730756). Respective primers were purchased from Integrated Device Technology (San Jose, CA). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal reference gene. Δ Ct-method was performed to analyze the relative expression levels of the respective genes.

2.4 Measurements of fecal microbial populations by quantitative real-time PCR

Fecal specimens were collected every week for 4 weeks from the antibiotic-treated Apc^Min/+ mice. Fecal DNA was isolated according to the manufacturer’ s instructions (PowerSoil^® DNA Isolation Kit, MO Bio laboratories, Carlsbad, CA). Universal primers were designed to measure the overall populations of bacteria (Forward: ACTCCTACGGGAGGCAGCAGT; Reverse: ATTACCGCGGCTGCTGGC) as previously described [47–51].

2.5 Statistical analysis

One-way ANOVA and post-hoc analysis were performed using SigmaPlot (Systat Software, San Jose, CA) to analyze polyp number and relative gene expression. A p value less than 0.05 was considered statistically significant.

3.1 Gut bacteria are required for BRBs’ anti-tumor effects in Apc^Min/+ mice

The Adenomatous polyposis coli (Apc) gene, a tumor suppressor, functions to induce β -catenin degradation and suppress the Wnt signaling pathway, and mutations in the Apc gene contribute strongly to CRC [52, 53]. Moreover, multiple intestinal neoplasia mice (Apc^Min/+), which carry a truncating mutation at codon 850 of the Apc gene, develop multiple colonic and intestinal polyps, making them a widely used model of human CRC [53]. We fed Apc^Min/+ mice with either a control diet (G1) or 5% BRBs from week 4 to week 8 (G2). Another group of Apc^Min/+ mice were first given the control diet and antibiotics in the drinking water for 4 weeks. Starting at week 4, the mice received the 5% BRB diet along with antibiotics for 4 weeks (G3) (Fig. 1A). After 1 week of the antibiotic treatment, there was a substantial decrease in the overall fecal bacterial population (Fig. 1B). At the end of study, we examined the number of colonic and intestinal polyps. The results confirmed those of our previous studies [38, 39]: that 5% BRBs in the diet significantly suppress colonic (Fig. 2A) and intestinal (Fig. 2B) tumor development in Apc^Min/+ mice. However, antibiotics completely abolished the anti-tumor effects of the BRB diet in small intestine (Fig. 2), suggesting that gut microbiota is required for BRBs’ beneficial effects in Apc^Min/+ mice.

OPEN IN VIEWER

Fig.2

Gut bacteria are required for the benefits of BRBs in Apc^Min/+ mice. BRBs significantly decreased the number of polyps in colon (A) and small intestine (B) of Apc^Min/+ mice, but antibiotics abolished those BRB-mediated chemoprotective effects. Ctrl: control diet; Abx: antibiotics. n = 5 per group; * p < 0.05; ** p < 0.01; *** p < 0.001.

Although large epidemiologic studies have suggested a positive association between frequent use of antibiotics and risk of CRC in humans [18–21], this conclusion is controversial in animal models as it is dependent on the model. In one study, antibiotics promoted intestinal polyp development in Apc^Min/+ mice [54], while in another study that used the CAC model, antibiotics suppressed tumor growth induced by AOM/DSS [55]. The underlying mechanism of the latter study suggests that antibiotics might inhibit bacteria-induced inflammation, thereby suppressing tumor development. Other studies have reported similar results: severe colonic inflammation exacerbated polyposis and antibiotics rescued that symptom [56, 57]. Interestingly, one study observed that antibiotics depleted the gut microbiota but induced mild inflammation in the gastrointestinal track of wild-type mice [58], suggesting that colonic inflammation and dysbiosis cannot be cured simply by depleting gut bacteria.

The dynamic interactions between a host and its gut microbiota and between diet and the microbiota play critical roles in the prevention and treatment of CRC [59–61]. A healthy dietary pattern has been shown to associate with a lower risk of CRC [33]. Microbiome studies have identified several “ good” bacterial species, such as Akkermansia muciniphila, and some “ bad” bacterial species, including Escherichia coli and Fusobacterium spp [7]. Therefore, maintaining a balanced and healthy gut microbiota by adopting a healthy lifestyle would help lower the risk of developing CRC [6, 62].

TLRs are important regulators of intestinal epithelial homoeostasis. They belong to the interleukin (IL)-1 superfamily of transmembrane receptors that recognize pathogen-associated molecular patterns (PAMPs). In particular, lipopolysaccharide (LPS) fragments from degraded outer membrane of gram-negative bacteria interact with TLR4, leading to the recruitment of downstream adaptor molecules, such as myeloid differentiation factor 88 (MyD88), IL-1 receptor-associated kinase (IRAK), and tumor necrosis factor receptor-associated factor 6 (TRAF6) [63]. The complex then phosphorylates the inhibitor of NF-κ B kinases (IKKs), activating NF-κ B. NF-κ B signaling plays an essential role in inflammation and cell survival. One of its downstream targets is COX2 [63]. Therefore, we examined the expression of TLR4/NF-κ B/COX2 pathway in BRB-treated Apc^Min/+ mice.

We found that 5% BRBs in the diet significantly decreased mRNA levels of TLR4 (Fig. 3A), NF-κ B (Fig. 4A), and COX2 (Fig. 5A) in colon of the Apc^Min/+ mice. However, the TLR4/NF-κ B/COX2 pathway was not significantly changed by BRBs (Figs. 3B, 4B, and 5B) in small intestine, suggesting that other mechanisms contributed to the BRBs’ anti-tumor effects in small intestine. Intriguingly, significantly enhanced mRNA levels of TLR4 (Fig. 3B) and NF-κ B (Fig. 4B) were observed in small intestine of the BRB-treated Apc^Min/+ mice fed antibiotics. Our results agree with those of Grasa L et al., who reported that antibiotics induced mild inflammation and TLR4 expression in the gastrointestinal tract of wild-type mice [58]. It is likely that the antibiotic treatment eliminated majority of the gut bacteria, whereas some other (or unknown) bacteria were selected and able to grow and induce a mild inflammatory response. Further investigations need to examine the mechanisms underlying the increased expression of TLR4 and NF-κ B promoted by antibiotics in the small intestine, as well as whether this enhancement contributes to antibiotics’ ability to counter the beneficial effects of BRBs.

OPEN IN VIEWER

Fig.3

mRNA expression levels of TLR4 in colon (A) and small intestine (B) of Apc^Min/+ mice. Ctrl: control diet; Abx: antibiotics. ** p < 0.01; *** p < 0.001; **** p < 0.0001.

OPEN IN VIEWER

Fig.4

mRNA expression levels of NF-κ B1 in colon (A) and small intestine (B) of Apc^Min/+ mice. Ctrl: control diet; Abx: antibiotics. ** p < 0.01; **** p < 0.0001.

OPEN IN VIEWER

Fig.5

mRNA expression levels of COX2 in colon (A) and small intestine (B) of Apc^Min/+ mice. Ctrl: control diet; Abx: antibiotics. ** p < 0.01.

The TLR4/NF-κ B/COX2 pathway has been examined in CRC patients and animal models, but the results have been surprisingly inconsistent. For example, studies showed that human intestinal epithelial cells normally express low levels of TLR4 [63], while specimens from UC [41] or CRC patients [42, 43], as well as from the CRC mouse model [41], showed increased expression of TLR4. In addition, high expression of TLR4 and MyD88 has been shown to associate with liver metastasis and poor prognosis in patients with CRC [44]. In contrast, another study demonstrated that normal colon epithelium and CRC cell lines were positive for TLR4, and specimens from metastatic CRC patients showed either loss of expression or strong downregulation of TLR4 in comparison with normal tissue and non-metastatic tumors [64]. Interestingly, one group found that TLR4 expression by tumor cells associated with a lower rate of tumor recurrence, whereas TLR4 expression by fibroblasts associated with a high rate of tumor recurrence [65]. Therefore, the TLR4/NF-κ B/COX2 pathway might play different roles in different compartments of the tumor microenvironment.