Chemopreventive effect of a milk whey by-product derived from Buffalo (Bubalus bubalis) in protecting from colorectal carcinogenesis

Background Several studies show that natural foods are a source of compounds with anticancer properties that affect the gut microbiota and its metabolites. In the present study, we investigate the effect of a delactosed buffalo milk whey by-product (DMW) on colorectal carcinogenesis. Methods The effect of DMW on colorectal carcinoma (CRC) was investigated in the established mouse model of azoxymethane (AOM)-induced colon carcinoma, which closely resembles the human clinical condition of CRC. The effect of DMW on CRC immortalized cell lines was also evaluated to further identify the antineoplastic mechanism of action. Results Pretreatment of AOM-treated mice with DMW significantly (P < 0.05) reduced the percentage of mice bearing both aberrant crypt foci with more than four crypts (which are early precancerous lesions that progress to CRC) and tumors. In addition, DMW completely counteracted the effect of AOM on protein expression of caspase-9, cleaved caspase-3 and poly ADP-ribose polymerase in colonic tissue. Administration of DMW alone (i.e. without AOM) resulted in changes in the composition of the gut microbiota, leading to enrichment or depletion of genera associated with health and disease, respectively. DMW was also able to restore AOM-induced changes in specific genera of the gut microbiota. Specifically, DMW reduced the genera Atopobiaceae, Ruminococcus 1 and Lachnospiraceae XPB1014 and increased the genera Parabacteroides and Candidatus Saccharimonas, which were increased and reduced, respectively, by AOM. Blood levels of butyric acid and cancer diagnostic markers (5-methylcytidine and glycerophosphocholine), which were increased by AOM treatment, were reduced by DMW. Furthermore, DMW exerted cytotoxic effects on two human CRC cell lines (HCT116 and HT29) and these effects were associated with the induction of apoptotic signaling. Conclusions Our results suggest that DMW exerts chemopreventive effects and restores the gut microbiota in AOM-induced CRC, and induces cytotoxic effect on CRC cells. DMW could be an important dietary supplement to support a healthy gut microbiota and reduce the prevalence of CRC in humans. Video Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s12964-023-01271-5.


Methods
The effect of DMW on colorectal carcinoma (CRC) was investigated in the established mouse model of azoxymethane (AOM)-induced colon carcinoma, which closely resembles the human clinical condition of CRC.The effect of DMW on CRC immortalized cell lines was also evaluated to further identify the antineoplastic mechanism of action.

Results
Pretreatment of AOM-treated mice with DMW significantly (P < 0.05) reduced the percentage of mice bearing both aberrant crypt foci with more than four crypts (which are early precancerous lesions that progress to CRC) and tumors.In addition, DMW completely counteracted the effect of AOM on protein expression of caspase-9, cleaved caspase-3 and poly ADP-ribose polymerase in colonic tissue.Administration of DMW alone (i.e.without AOM) resulted in changes in the composition of the gut microbiota, leading to enrichment or depletion of genera associated with health and disease, respectively.DMW was also able to restore AOM-induced changes in specific genera of the gut microbiota.Specifically, DMW reduced the genera Atopobiaceae, Ruminococcus 1 and Lachnospiraceae XPB1014 and increased the genera Parabacteroides and Candidatus Saccharimonas, which were increased and reduced, respectively, by AOM.Blood levels of butyric acid and cancer diagnostic markers (5-methylcytidine and glycerophosphocholine), which were increased by AOM treatment, were reduced by DMW.Furthermore, DMW exerted cytotoxic effects on two human CRC cell lines (HCT116 and HT29) and these effects were associated with the induction of apoptotic signaling.

Delactosed and defatted buffalo milk whey by-product
Milk from Mediterranean Italian dairy buffaloes (Bubalus bubalis) was obtained from 30 commercial buffalo farms in the Caserta region of southern Italy.The milk was subjected to enzymatic hydrolysis with the enzyme lactase (neutral pH and at 25 °C) to obtain delactosed milk.Then the whey obtained by adding curd to the milk was used to make ricotta.Finally, the whey residue obtained from the ricotta was subjected to reverse osmosis to obtain a concentrated solution of 60% delactosed milk whey byproduct (DMW).
For the in vitro experiments, the DMW was defatted to obtain DMW without fat (DMW/WF).Briefly, the DMW was centrifuged three times at 23,000 g (4°C for 30 min) and then centrifuged at 23,000 g for another 1 h [26].The solution was aliquoted and frozen until use.Before use, the DMW/WF was added to the complete cell culture medium and the solution was finally passed through a 0.22 µm Millipore filter (Merck Millipore, Milan, Italy).

Milk whey by-product metabolomic profiling
The metabolomic profile of DMW has been reported previously [19].DMW/WF metabolomic analysis was performed under the same experimental conditions as reported by Cacciola et.al [19].Quantification of δ-valerobetaine (δ-VB) content was performed by HPLC-ESI MS/MS [27].
All experimental procedures were assessed and approved by the Institutional Animal Ethics Committee for the use of experimental animals and complied with the guidelines for the safe use and care of experimental animals according to the Italian D.L. no.116 of 27 January 1992 and associated guidelines in the European Communities Council (86/609/ECC and 2010/63/UE).

Preneoplastic lesions and tumors
Aberrant crypt foci (ACF) and tumors were evaluated in mice colons as previously described [29].Briefly, the colons were quickly removed after euthanasia, washed, opened longitudinally, laid flat and fixed with 10% buffered formaldehyde.After staining with 0.2% methylene blue in saline, the colons were examined for the detection of ACF and tumors using a light microscope at 20 × magnification (Leica Microsystems, Milan, Italy) by two independent investigators in a blinded fashion.

Histological evaluation
After sacrifice, colon segments (1.5 cm) were fixed in 10% formaldehyde for 24 h.The samples were then embedded in paraffin and cut into 4 µm sections.For histological evaluation, the samples were stained with hematoxylin and eosin (H-E) (Carlo Erba Reagents SAS, Val de Reuil, France).Histological evaluation and counting of mitotic figures (MF) were performed with a microscope (E-400; Nikon Eclipse, Tokyo, Japan) by two independent observers blinded to the treatments.MF were counted from 10 randomly selected high-power fields (HPF; 40 × objective 10 × ocular) for each colon sample, as previously described [30].

Microbiota identification by 16S rRNA sequencing
Fecal samples (300 mg) from mice were collected at time 0 (before the start of treatment, week 0) and after 13 weeks of treatment (week 13) and total DNA was extracted using the QIAamp DNA Stool Mini Kit (QIA-GEN) according to the manufacturer's instructions.Partial 16S rRNA gene sequences were amplified using the primer pair Probio_Uni and Probio_Rev, targeting the V3 region of the 16S rRNA gene [31].Amplicon checks were performed as previously described [31].Sequencing of the 16S rRNA gene was performed using a MiSeq (Illumina) in the GenProbio srl DNA sequencing facility (www.genpr obio.com) following the protocol described previously [30].After sequencing and demultiplexing, the reads obtained from each sample were filtered to remove low-quality and polyclonal sequences.All quality-approved, trimmed and filtered data were exported as.fastq files.The.fastq files were processed using a script based on the QIIME software suite [32].Paired-end reads pairs were assembled to reconstruct the complete Probio_Uni / Probio_Rev amplicons.After quality control, sequences with a length between 140 and 400 bp and mean sequence quality score > 20 were retained.Sequences with homopolymers > 7 bp and mismatched primers were omitted.
To calculate downstream diversity measures (alpha and beta diversity indices, Unifrac analysis), 16S rRNA Amplicon Sequence Variants (ASVs) were defined at ≥ 100% sequence homology using DADA2 [33] and ASVs that did not encompass at least 2 sequences from the same sample were removed.All reads were classified to the lowest possible taxonomic rank using QIIME2 [32,34] and the SILVA database v. 132 as the reference dataset [35].The biodiversity of the samples (alpha-diversity) was calculated using the Chao1 and Shannon indexes.The similarities between the samples (beta-diversity) were calculated using the weighted uniFrac [36].The range of similarities is calculated between the values 0 and 1. PCoA representations of beta-diversity were performed with QIIME2 [33,35].

Measurement of blood short chain fatty acid (SCFA)
At the end of the experiments, blood was taken from each animals by cardiac puncture and serum was obtained by centrifugation (4 °C, 10 min, 800 g).The serum was analyzed by gas-chromatography mass spectrometry (GC-7890A, Agilent Technologies; MS-5977A MSD, Agilent Technologies).In brief, serum samples were acidified with 20 μl phosphoric acid (H 3 PO 4 ) 85% (w/v), vortexed for 5 min and incubated on ice for 5 min.The acidified samples were extracted by adding diethyl ether (DE) for acetic and propionic acids (1:1, v/v), and ethyl acetate (EA) for butyric acid (1:1, v/v), vortexed for 5 min and then centrifuged at 12,000 × g for 20 min at RT. Anhydrous sodium sulphate was added to the organic extract (containing the SCFAs) to remove the residual water and the samples were transferred to a new glass-tube for GC-MS analysis.A standard curve (0.1-3 µg/ml) was generated for each SCFA at the beginning of the run.A blank solvent (DE or ethyl acetate) was injected between each sample to ensure that no memory effects occurred.The GC was programmed to achieve the following run parameters: initial temperature of 90 °C, hold of 2 min, ramp of 2 °C/min up to a temperature of 100 °C, hold of 10 min, ramp of 5 °C/min up to a final temperature of 110 °C for a total run time of 21 min.

Untargeted metabolomic analysis of blood
The water residues of the serum after extraction with ethyl acetate and petroleum ether (see section Measurement of SCFA concentration in serum) were used for LC-MS Q-TOF analysis.The untargeted metabolomic study was performed using an Agilent HP 1260 Infinity Series liquid chromatography coupled to a Q-TOF mass spectrometer and equipped with a DAD system (Agilent Technologies, Santa Clara, CA, USA).The MS system was equipped with a Dual Electrospray Ionization (ESI) source which operated in positive mode.Chromatographic separation was performed using an Adamas C18-X-Bond column (50 mm X 4.6 mm, 3.5 µm, Sepachrom Srl) held at a constant temperature of 25 °C.Analyses were performed at a flow rate of 0.5 ml min-1, using a linear gradient system composed of 0.1% (v/v) formic acid in water (eluent A), and 0.1% (v/v) formic acid in acetonitrile (eluent B).The gradient program was set as follows: starting from 5 to 70% eluent B in 4 min, isocratic at 70% of eluent B from 4 to 5 min; from 70 to 80% of eluent B from 5 to 8 min and from 80 to 100% eluent B from 8 to 10 min; lowering to starting condition (5% eluent B) from 10 to 15 min to achive the equilibration after 1 min.The injection volume was 10 µL.UV spectra were recorded from DAD every 0.4 s from 190 to 750 nm, with a resolution of 2 nm.The capillary was maintained at 2000 V, fragmentor voltage at 180 V, cone 1 (skimmer 1) at 45 V, Oct RFV at 750 V.The gas flow rate was set to 11 L min-1, at 350 °C, and the nebulizer was set at 45 psig.Mass spectra were recorded within the m/z range 100-1700 as centroid spectra, with three scans per second.To perform real time mass-lock correction, a solution consisting of purine (C 5 H 4 N 4 , m/z 121.050873, 10 µmol L-1), and hexakis (1H,1H,3H-tetrafluoropentoxy)-phosphazene (C 18 H 18 O 6 N 3 P 3 F 24 , m/z 922.009798, 2 µmol L-1) was constantly infused with an Isocratic pump (1260 Infinity Series, Agilent Technologies) at a flow rate of 0.06 ml min-1.All MS and HPLC parameters were determined using Agilent MassHunter Data Acquisition Software, rev.B.05.01.

Cell viability
The cell viability was determined using the MTT assay [28].Briefly, cells were seeded in 96-well plates (at a density of 1 × 10 4 cells per well for HCT116 and HCEC and 1.5 × 10 4 for HT29) and allowed to adhere for 24 h.The cells were then exposed to increasing concentrations of DMW/WF from 2 to 20% v/v for 24 h.Cells were then incubated with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (250 μg/ml, Merck Millipore) for 1 h at 37 °C, and the formation of formazan was measured at 570 nm using a microplate reader (BioTek ™ Cytation ™ 3, Winoosky, USA).Results are expressed as percentage of cell viability (n = 3 experiments with 3-6 replicates for each treatment).Morphological changes of cells, treated or not with DMW/WF (0.3-20% v/v) for 24 h, were analyzed by scanning and phase contrast photography with a Zoe ™ Fluorescent Cell Imager (Biorad, Germany).

Cell apoptosis
Apoptosis was assessed by cytofluorimetric analysis using the Annexin V and propidium iodide (PI) apoptosis assay Kit (Dojindo Molecular Technologies Inc., Munich, Germany) as previously described [37].Briefly, HCT116 and HCEC cells were seeded in six-well plates (4 × 10 5 cells/well and 4.5 × 10 5 cells/well for HCT116 and HCEC, respectively) and cultured in the absence or presence of DMW/WF (1-10% v/v) for 24 h.Subsequently, cells were harvested, washed twice with PBS, resuspended in Annexin V binding buffer and incubated with 5 μL Annexin V and 5 μL PI for 15 min at room temperature (RT).Finally, for HCT116 cells, apoptotic cells were analyzed using the BD FACSCanto II system and analysis performed using DIVA software (BD Biosciences, Milan, Italy).For HCEC cells, apoptotic cells were analyzed using the BriCyte E6 (Mindray, P.R. China).At least 10,000 events were recorded for each condition.

Cell and tissue lysate preparation
HCT116 cells were seeded in 6 cm 2 Petri dishes at a density of 1 × 10 6 cells and allowed to adhere for 24 h.Cells were then incubated with DMW/WF (6% or 10% v/v) for 0, 3, 6, 12, and 24 h.At each endpoint, cells were washed twice with ice-cold 1 × PBS, collected by scraping and lysed in RIPA buffer (Cell Signaling Technology, USA).After 30 s of sonication, lysates were centrifuged at 14,000 g (4 °C for 15 min), the supernatant was collected, and protein concentration was determined using the Bio-Rad DC protein assay (Bio-Rad, Munich, Germany).Total protein lysates from colon tissues were extracted as previously described [38].In brief, colon tissues were washed with chilled 1 × PBS to remove all blood contaminants.The colon tissues were then dissected into small pieces, lysed in RIPA buffer using an Ultra Turrax ™ device (Ika-Werke GmbH, Staufen, Germany) and sonicated for 30 s. Finally, the lysates were centrifuged at 14,000 g (4 °C for 15 min), the supernatant was collected and the protein concentration was determined using the Bio-Rad DC protein assay (Bio-Rad).

Statistical analyses
Statistical analyses were performed using GraphPad Prism 9.0 (GraphPad software, San Diego, CA, USA).Statistical significance was determined using either Student's t test for comparing a single treatment mean with a control mean or One-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test for analysis of multiple treatment means.Two-way ANOVA followed by the Bonferroni multiple comparisons post-test was used to compare different cumulative concentration effect curves.The DMW concentration that produced 50% inhibition of cell viability (EC 50 ) or maximum inhibitory effect (E max ) was calculated by non-linear regression analysis using the equation for a sigmoid concentrationresponse curve.
To determine the SCFA concentration in µg/ml using Graphpad Prism, the data were inserted in "XY" form, with the "X" field indicating the values of the straight concentration-response curve, while the "Y" field indicated the values of the area under the curve (AUC) with respect to the peaks obtained from the mass gas.The AUC values of the individual samples (obtained from the mass gas) were interpolated with the line X (concentration-response relationship) to determine the corresponding concentration.
For untargeted metabolomic analysis, raw data were analyzed using Mass Profile Professional 13.1.1software (Agilent Technologies) and divided into the following groups: AOM + DMW, AOM, DMW and compared with the control group.In addition, the groups AOM + DMW and AOM were compared, with AOM serving as the control group.For statistical analysis all groups were subjected to One-way ANOVA (p-value < 0.05) and to fold change ≥ 2.0.Principal components analysis (PCA) were performed to compare metabolic profiles of serum samples.Statistically relevant compounds were putatively identified using an in-house serum metabolites database, a freely available electronic database, Human metabolome database (HMDB).
Analysis of fecal microbiota composition was performed using SPSS software 25 (www.ibm.com/ softw are/ it/ analy tics/ spss/) and differences between experimental groups, including differences in abundance of bacterial genera, were determined by One-way or Twoway ANOVA analyses followed by Tukey's or LSD (least significant difference) multiple comparisons tests.For all analyses p < 0.05 was considered significant.Results are presented as means ± SEM of n experiments.

Milk whey by-product metabolome profiling
The metabolome profiling of milk whey by-product was obtained by HPLC-MS analysis.Mass spectra and chromatograms were used to putatively identify metabolites by comparison with public databases, data from the literature and, where possible, by comparison with authentic standards.The main constituents identified in DMW/WF are listed in Table 1.Eight metabolites were identified in DMW/WF samples, belonging to the classes amino acids, fatty acids, tricarboxylic acids, hydroxycoumarin derivatives, triterpenes, amino alcohols and O-glycoside.Compared to DMW, defatted DMW resulted in a loss of lipophilic compounds such as erucamide and butyryl-L-Carnitine [19].To titrate the DMW/WF, we determined the δ-valerobetaine content to be 284 µM.

In vivo induction of colon cancer Effect of AOM and DMW on colon cancer
After 3 months of treatment, AOM triggered the appearance of aberrant crypt foci (ACF) and tumors (Fig. 1  B).Pretreatment of AOM mice with DMW decreased (p < 0.05) the percentage of mice bearing i) ACF containing more than four crypts, ii) tumors and iii) more than one tumor (Table 2).DMW also reduced (P < 0.05) the number of tumors (Fig. 1B); there was only a trend to reduce ACF formation (30% inhibition, p = 0.12).No formation of ACF and tumors was observed in control and DMW treated mice (data not shown).

Effect of AOM and DMW on colon histology
No histological changes were observed in the colonic mucosa of the control and DMW-treated mice.In these mice, the crypts were well organized, aligned and parallel, and characterized by a single layer of columnar and goblet cells (Fig. 1C).The colon tissues of the AOM-treated mice showed pathological changes including focal atypia, dysplasia and hyperplasia, as well as adenomatous polyps characterized by branched tubules in the lamina propria (Fig. 1C).In addition, the tissues were disorganized and covered by epithelium with nuclear pseudostratification and signs of dysplasia (basal cells disorganization, hypercromasia, loss of polarity) (Fig. 1C).In mice treated with AOM plus DMW, the colons had areas of polypoid hyperplasia, characterized by a simple tubular architecture with elongated and straight crypts with nuclear hypercromasia and maintenance of polarity (Fig. 1C).DMW-treated mice had a similar number of mitotic figures (MF) as control mice (Fig. 1D).AOM treatment increased (p < 0.0001) the number of MF, and this effect was reduced (p < 0.0001) by DMW (Fig. 1D).

Effect of AOM and DMW on apoptotic pathway
To investigate the mechanism responsible for the chemopreventive effects of DMW, we examined the protein expression of caspase-9, cleaved caspase-3, and PARP in colonic tissues of mice treated with AOM.Western blot analysis of three randomly selected samples from three individual mice showed increased protein expression levels of cleaved caspase-3, caspase-9, and PARP after AOM administration (p < 0.0001).Pretreatment with DMW completely reversed the effect of AOM on these protein markers (Fig. 2A).The effect of DMW on PARP protein expression was confirmed by immunofluorescence microscopy analysis.Quantification of PARPpositive cells showed a strong signal in the AOM group (p < 0.0001) which was partially counteracted by DMW treatment (Fig. 2B).DMW alone did not change any of the markers analyzed (Fig. 2).

Effect of AOM and DMW on intestinal microbiota
To characterize intestinal microbial composition, 16S-based sequencing analysis was performed on total DNA extracted from mice fecal samples.PCoA analysis based on Bray-Curtis distance revealed that the fecal microbiota in 18 of 24 untreated mice (week 0) clustered together, with 6 mice outside the cluster (indicated by arrows in Supplementary Figure 1).The microbiota composition in the feces of the mice (week 13) for the DMW, AOM and AOM + DMW treatments clustered independently and showed only partial overlap, clearly indicating that the different treatments influenced the fecal microbiota composition (Supplementary Figure 1).The ASVs representation curves showed that the microbial diversity of the samples was completely covered (Supplementary Figure 2 A), while the alpha-diversity analysis revealed similar diversity in the control and treated mice as measured by the Chao1 and Shannon indexes (Supplementary Figure 2 B and C).At week 13, the fecal microbiota composition of untreated mice (control group) did not cluster with that of the same animals at time 0 (week 0 versus week 13, Supplementary Figure 1).This finding suggests a possible influence of age on microbial composition.The analysis of the total microbiota at the phylum, family and genus levels is shown in Supplementary Figure 3.
When restricting the analysis to the bacterial genera with relative abundances greater than 0.05%, the genus composition of the fecal microbiota is represented differently in DMW and AOM mice (Fig. 3).Specifically, DMW or AOM treatment (at week 13) alone altered (p < 0.05) the microbial composition of 26 (DMW) and 24 (AOM) genera compared with the control group (at week 13) (Fig. 3A  and B).Analysis of the genera in the fecal microbiota of the AOM plus DMW group compared to the AOM group also revealed that 20 genera differed (p < 0.05) between the two groups (Fig. 3C).Of these 20 genera, five genera were altered by AOM (compared to control) and were restored to control levels by DMW, indicating that DMW counteracted the AOM effect (Fig. 4).Notably, the relative abundance of the three genera Atopobiaceae, Ruminococcus 1 and Lachnospiraceae XPB1014, which was increased in AOM-treated mice, was maintained at the control level by DMW treatment (Fig. 4).An opposite trend was observed for members of the genera Parabacteroides and Candidatus Saccharimonas, whose relative abundance decreased in AOM-treated mice and was maintained at the control level by DMW treatment (Fig. 4).Another six genera were specifically affected by DMW treatment, and the DMW-induced changes persisted in the presence of AOM (Fig. 5).For five genera, DMW treatment decreased their abundance compared to the control group, and these low abundances were maintained in the AOM plus DMW group (Fig. 5).In the case of the Eubacterium brachy group, the DMW-induced reduction (compared to the control) was no longer significant in the AOM plus DMW group (vs control), although an effect of DMW was observed in suppressing the increasing trend of AOM (Fig. 5).The gram-negative bacterium cTPY-13 was increased in DMW treated mice (Fig. 5).For the remaining nine genera that differed between the AOM and AOM plus DMW groups, eight genera were not altered by administration of AOM or DMW alone (compared to control) (Fig. 6).For three out of nine genera, an increase in genus abundance was observed only in the AOM plus DMW group compared with the AOM group (Fig. 6).For the other six genera, differences in genus abundance were observed between the AOM plus DMW group and all other groups (Fig. 6).Of these six genera, four genera were increased by AOM plus DMW (Fig. 6) while the remaining two genera (Helicobacter and Roseburia) were reduced (Fig. 6).AOM alone caused an increase in the genus Akkermansia which was further enhanced by simultaneous treatment with DMW (Fig. 6).

Short chain fatty acids (SCFAs)
Analysis of the fecal microbiota had shown that AOM and DMW induced changes in the genera that produce SCFAs.AOM treatment was associated with an increase in butyric and propionic acid, in the blood, with a significant effect for butyric acid (Supplementary Figure 4).DMW administration returned SCFAs to levels comparable to those of the control group (Supplementary Figure 4).No effect of acetic acid has been observed (Supplemenrary Figure 4).

Untargeted metabolomic analysis
After extraction for SCFAs quantification, the serum samples were subjected to LC-MS analysis.The resulting total ion chromatograms (TIC) and mass spectra were used for putative identification of compounds by comparing the spectra with data from the literature, freely available databases and /or, where possible, by comparison with authentic standards.The raw data obtained by LC-MS analysis were subjected to statistical analysis (both univariate and multivariate analyses).The statistically different metabolites are listed in Table 3 together with their differential accumulation obtained after comparing the samples treated with AOM and the sample treated with AOM plus DMW.The graphs in Supplementary Figure 5 represent the principal component analysis (PCA) and show a clear separation of the samples depending on the treatment applied.Specifically, in supplementary Figure 5A, the untreated mouse serum (control group) extracted with petroleum ether is separated from the treated groups along component 1. Separation between the treated samples is evident along component 2, with the DMW and AOM plus DMW groups closer together.

Effect of DMW/WF on HCT116 cell morphology
Untreated HCT116 cells (control) or cells exposed to 0.3% v/v DMW/WF spread regularly in the cell culture multiwell plates and grew to semiconfluence (Fig. 8).In contrast, HCT116 cells treated with 1-2% v/v DMW/WF exhibited a thin and flattened shape compared to control cells (Fig. 8).Treatment with 4% v/v or 6% v/v DMW/WF was associated with a decrease in cell size; cells shrank and lost contact with neighboring cells (Fig. 8).Cells treated with the highest DMW/WF concentrations (10% or 20% v/v) were very small and rounded (hallmarks of apoptotic cells), detached from the surface of the culture plate, and floated in the cell culture medium (Fig. 8).
To better understand the mechanism by which DMW/ WF-treated HCT116 cells undergo apoptosis, we performed a time course (0-24 h) and concentration-dependent experiment (6% v/v or 10% v/v DMW/WF concentrations).Exposure of HCT116 cells to 6% v/v DMW/WF resulted in a time-dependent increase (p < 0.05-0.0001) in caspase-3, caspase-9, and PARP cleaved fragments (Fig. 10).At the highest concentration (10% v/v), the increase in the caspases-and PARP-cleaved fragments was more pronounced but not time-dependent, suggesting that the time at which caspases and PARP cleavages occurred was related to both concentration and exposure time (Fig. 10 A and B).

Discussion
We had previously shown that non-protein metabolites in buffalo milk whey by-product have antiproliferative and pro-apoptotic effects on human cancer cell lines in vitro, including colorectal cancer cells [16].We had also shown in a preliminary study that buffalo milk whey upregulated apoptotic pathways in a HCT116 xenograft mouse model of colorectal cancer [19].We have now extended our previous work by examining the effects of delactosed buffalo milk whey by-product (DMW) in an in vivo model of CRC.We provided evidence that DMW treatment exerts chemopreventive effects by reducing tumor formation and this effect was associated to changes in the gut microbiota and blood markers.Moreover, we demonstrated that defatted DMW exerted a strong cytotoxic effect on human colorectal cancer cell lines by affecting both the extrinsic and intrinsic apoptotic pathways thus providing compelling evidence of a possible valuable role of DMW in colorectal carcinogenesis development.
The development of CRC occurs in a multistage process in which healthy epithelium slowly evolves into preneoplastic lesions, namely aberrant crypt foci (ACF), which over time develop into polyps and eventually malignant carcinomas.In this study, using the AOM mouse model of CRC, which induces the formation of CRC similar to the pathogenesis of sporadic human CRC, we showed that orally administered DMW significantly reduced the percentage of mice with ACF containing more than four crypts.These are the early precancerous lesions that more often progress to CRC.In addition, DMW also reduced the percentage of mice with tumours, mice with more than one tumour and tumour growth induced by AOM.AOM treatment alone was associated with increased mitotic figures in colon tissue, which were reduced in mice receiving pre-treatment with DMW.These results provided compelling evidence that buffalo DMW reduces the incidence and severity of colorectal cancer.
Mechanistically, we found increased expression of cleaved caspase-3, caspase-9 and PARP-1 in the colon tissue of AOM-treated mice, which was significantly reduced by DMW.It is generally believed that increased expression of caspase-3 correlates with the induction of apoptosis and therefore caspase-3 is considered an indicator of cancer treatment efficacy [41].However, there is increasing evidence for a non-apoptotic role of caspase-3 in cancer-induced cell migration, invasiveness, tumor angiogenesis and stress-induced growth of cancer cells [42][43][44][45].In addition, it has been reported that colorectal cancer patients with high expression of caspase-3 have a worse prognosis than patients with lower expression [46][47][48][49].In view of these findings, our results suggest a pro-tumor role of caspase-3 in the AOM-induced colon cancer model.Few data are available for the potential role of caspase-9 in malignancies, although increased expression of caspase-9 has been reported in lung cancer progressing to a metastatic phenotype [50] and in colorectal cancer patients with a poor prognosis [51].PARP-1 has been shown to play a role in the development of colorectal cancer, as PARP-1 expression is significantly higher in colorectal cancer than in healthy tissue and directly correlates with tumor size and histopathology [52,53].In a colitis-associated colorectal cancer model, AOM was associated with greater tumor number in wild-type mice than in PARP-1 −/− mice [54].Our results on PARP-1 expression (Western blot and immunofluorescence) are consistent with these data, confirming its tumor-promoting "burden".
The gut microbiota has been shown to be a critical factor in the development, progression and metastasis of CRC [55][56][57].Dysbiosis of the gut microbiota has been found in colorectal cancer patients as well as in mice with spontaneous and chemically induced colon tumorigenesis [58][59][60].Here, we found that DMW administration in healthy mice resulted in an increase in bacterial genera known for their beneficial effects.In particular, DMW treatment increased the relative abundance of members of the genera Eubacterium, Lactobacillus, Prevotellaceae and Stomatobaculum, which are well-established probiotics with health-promoting roles and strictly species capable of producing short-chain fatty acids (SCFA) such as butyrate and propionate, respectively [61][62][63].In addition, DMW supplementation also led to an increase in Blautia members, which were recently identified as a new genus with 20 species.Remarkably, bioinformatic analysis of 12 Blautia species has enabled the identification of potential probiotics capable of preventing infections, improving glucose and lipid homeostasis and reducing obesity-related disorders [64][65][66].Similarly, an increase in the gram-negative bacterium cTPY-13, which belongs to the Bacteroides class and was previously identified in the large intestine of mice, was observed in mice treated with DMW [67].Finally, we also found that DMW caused a decrease in members of the genera Turicibacter, Eubacterium, Clostridium and Lachnoclostridium, whose presence in the gut was previously associated with inflammation and cancer development [68].Remarkably, the DMW-induced changes in the gut microbiota were maintained after treatment with AOM.Taken together, all these data suggest that DMW supplementation has a positive effect on the gut microbial community, leading to an enrichment or depletion of genera associated with health or disease status.
According to the literature [69], administration of the carcinogenic agent AOM resulted in dysbiosis of the gut microbiota, which was partially restored by DMW in mice.Specifically, AOM administration led to an increase in three gram-positive genera, namely Atopobiaceae, Ruminococcus 1 and Lachnospiraceae XPB1014, and this effect was reversed by DMW.Remarkably, elevated levels of these three genera were found in CRC and/or in patients with squamous cell carcinoma of the oesophagus, but not in healthy individuals [63,70], suggesting a detrimental role of these bacteria in the development of CRC.An opposite trend was observed for members of the genera Parabacteroides and Candidatus Saccharimonas, whose relative abundance decreased in AOMtreated mice and was completely reversed by DMW treatment.The role of Parabacteroides, gram-negative bacteria commonly found in the gut of healthy animals, is controversial as their abundance has been associated with health and disease status.Some reports have found high abundance of Parabacteroides in CRC patients [70,71], while other papers have reported a negative correlation between Parabacteroides and CRC development [72,73].Members of the genus Candidatus Saccharimonas have been shown to be greatly reduced in inflammatory diseases [74] and to suppress the production of inflammatory markers in macrophages, indicating an immunosuppressive effect [75].
The taxa that are increased in the faeces of AOM plus DMW-treated mice are genera known for their healthpromoting effects in the gut.For example, Eubacterium spp.can produce SCFA and thus contribute to the modulation of intestinal inflammation [76].In particular, the species E. coprostanoligenes has recently received much attention for its protective effects against colitis and carcinogenesis [76].In addition, members of the Prevotellaceae taxa play an important role in the degradation of complex dietary polysaccharides, leading to the production of SCFA, the increase of which has been reported after the use of probiotics [62,63].In agreement with a recent article [77], our results showed that administration of AOM led to an increase in the genus Akkermansia.In the present study, this increase was further enhanced by AOM + DMW.Akkermansia is considered an important symbiont of the gut microbial population, and its abundance is associated with health-promoting effects.Interestingly, high Akkermansia levels were found in CRC patients compared to healthy individuals and increased after chemotherapy, suggesting a positive correlation with therapeutic effect [78,79].Two genera, Helicobacter and Roseburia, were specifically reduced by DMW in AOM-treated mice.Species of the genus Helicobacter, which are able to colonise the intestinal environment, are commonly associated with diarrhoea and the development of inflammation and CRC [80].Therefore, the DMW-induced Helicobacter reduction in AOM-treated mice supports our hypothesis that the antitumour effect of DMW is partly due to the shift in the composition of the gut microbiota towards a "healthier" community.Strikingly, however, our results also showed a decrease in the abundance of Roseburia (considered a health-promoting genus that can produce butyrate and other SCFAs [81]) in mice with tumours treated with DMW.Since a decrease in Roseburia bacteria has been reported in response to increased carbohydrate intake [82,83], we can speculate that the presence of carbohydrates in DMW may be responsible for the effect of DMW on Roseburia abundance.The above findings highlight that the gut microbiota is influenced by many factors and explain why it can be difficult to establish clear links between a particular treatment, the microbiota and colorectal cancer [84].
Considering the effect of DMW on bacteria producing SCFAs, in this study we also measured the content of SCFAs in mouse sera.SCFAs, which include acetate, butyrate and propionate, are the major microbial end products of the fermentation process in the colon.Of these, butyrate has been reported to have a beneficial effect against cancers, including colorectal cancer [85,86].However, it has also been reported that the effect of butyrate varies depending on the dosage.Low doses of butyric acid stimulated the proliferation of colon epithelial cells, while it was inhibited at high concentrations [87].In our study, we show increased butyric acid levels in the sera of AOM-treated mice, which were restored to baseline levels by DMW, probably due to the effect of DMW on the incidence of Roseburia.The effect of DMW on butyric acid levels seems to indicate a beneficial effect of butyric acid on the development of colon cancer.Further experiments are needed to clarify the role of butyric acid in the development of CRC triggered by AOM.
Further evidence for the preventive effect of DMW on the development of CRC came from the untargeted metabolomic analysis performed on mouse sera.AOM treatment resulted in an increase in two lipophospholipids known to be potential colorectal cancer biomarkers ( [88][89][90].DMW was able to lower the lipophospholipid levels increased by AOM. We also investigated the cytotoxic effect of DMW without fat (DMW/WF) on isolated CRC cell lines as well as on a healthy human colonic epithelial cell line (HCEC).We demonstrated that DMW/WF treatment affected the viability of both CRC cell lines (HCT116 and HT29) and, to a lesser extent, HCEC in a concentration-dependent manner, suggesting a more selective effect on CRC cells than on healthy colonic epithelial cells.Since δ-VB has been reported to exert a cytotoxic effect on HT-29 cells, it could be hypothesized that the cytotoxic effect of DMW is due to the presence of this compound.However, such a hypothesis can be ruled out as our results showed that 20% v/v DMW (containing 55.5 µM δ-VB) inhibited cell viability by ~ 99%, which is not in agreement with our previous results where δ-VB inhibited cell viability by ~ 30% [17].As further confirmation that the cytotoxic effect of DMW is not only due to δ-VB (but also to other compounds), a recent study showed that a purified fraction (3KDa) of whey by-product containing δ-VB (138 mM) exerted a less pronounced cytotoxic effect than our results (cell viability by ~ 30%) [16].
Morphological analysis of the cells showed that the DMW/WF-induced cytotoxic effect is probably due to an apoptotic cell death process, as the phenotypic changes (i.e.cell shrinkage, fragmentation into membrane-bound apoptotic bodies) are features of an apoptotic process.These data were further supported by the flow cytometry assay, which showed that treatment of HCT116 cells with DMW resulted in an increase in early and/or late apoptotic events without significant necrotic signs.
To determine the molecular pathway involved in DMW/ WF-induced apoptotic effects in HCT116 cells, we examined the expression of proteins related to apoptosis by Western blot analysis.Apoptosis is a finely tuned process initiated by a cellular intrinsic or/and cellular extrinsic pathway.
Cell death signals from the extrinsic pathway lead to activation of caspase-8, which in turn can directly cleave and activate effector caspases, such as caspase-3.Similarly, the intrinsic apoptosis pathway leads to activation of effector caspases mediated by caspase-9 [91].Since our results show increased cleavage of caspase-3, -8 and -9 after DMW/WF treatment, the intrinsic extrinsic pathways may likely be involved in the apoptotic effect of DMW/WF.It is well known that cleaved caspase-3 activates the cleavage of PARP, a well-known cellular hallmark of apoptosis.This fits with our results showing an increase in PARP cleavage.

Conclusions
In conclusion, the present study provides the first evidence that DMW, a by-product of the dairy processing industry, has chemopreventive effects in an AOM-induced sporadic mouse model of colorectal cancer by reducing the number of tumors, and that this effect is associated with a modulation of the composition of the gut microbiota and blood metabolome that can be considered beneficial to health.Furthermore, we have shown that defatted DMW has a direct cytotoxic effect on human colorectal adenocarcinoma cell lines by triggering a pro-apoptotic mechanism.Identification of the bioactive compounds, which may be a single compound or a mixture of these compounds, is necessary to better understand the mechanisms underlying the chemopreventive effects of DMW and to identify potential candidates for therapeutic strategies targeting the microbiota.Buffalo milk whey by-product has great translational potential as an important dietary supplement (prebiotic) to support a healthy gut microbiota and reduce the prevalence of CRC in humans.

Fig. 1
Fig. 1 Oral administration of delactosed milk whey by-product (DMW) from Mediterranean Italian dairy buffaloes prevents carcinogenesis development in the mouse model of azoxymethane (AOM)-induced colon cancer.A Experimental design related to the AOM-induced tumor model; B Number of aberrant crypt foci (ACF) and tumors induced in the mouse colon by AOM (inserts above the graphs show an example of ACF and tumor).AOM (40 mg/kg in total, intraperitoneally) was administered, at the single dose of 10 mg/kg, at the beginning of the first, second, third and fourth week.DMW was given (by oral gavage), at the 10 ml/kg dose, three times a week for the whole duration of the experiment starting 1 week before the first administration of AOM.Measurements were performed at the end of experiment (i.e., 13 weeks after the first injection of AOM).Data represent the mean ± SEM of 16 mice.*p < 0.05 vs AOM alone.C Histological features of colon from control and treated mice.Control and DMW mice colon showed normal morphology; AOM treated mice colon showed adenomatous polyp with signs of dysplasia; AOM plus DMW treated mice colon showed polypoid hyperplasia (H-E; Original magnification × 40; scale bar 50 µm).D The number of mitotic figures (%) was counted in colon sample (obtained from control, DMW, AOM and AOM plus DMW treated mice) considering 10 randomly selected high-power fields (HPF; 40X objective 10X ocular).Data represent the mean ± SEM of 4/5 mice.#p < 0.0001 vs control (Ctrl), ****p < 0.0001 vs AOM alone

Fig. 2 Fig. 3
Fig.2Oral administration of delactosed milk whey by-product (DMW) from Mediterranean Italian dairy buffaloes affects apoptosis-related proteins in the colon of azoxymethane (AOM)-treated mice.A Western blot analysis showing the protein expression levels of cleaved caspase-3, caspase-9 and poly (ADP) ribose polymerase (PARP) in colon of animals treated or not with AOM or/and DMW.AOM (40 mg/kg in total, intraperitoneally) was administered, at the single dose of 10 mg/kg, at the beginning of the first, second, third and fourth week.DMW was given (by oral gavage), at the 10 ml/kg dose, three times a week for the whole duration of the experiment starting 1 week before the first administration of AOM.Measurements were performed at the end of experiment (i.e., 13 weeks after the first injection of AOM).The right side of the panel shows the densitometry analysis of the above proteins.Signals were normalized on the housekeeping proteins β-actin.Data represent the mean ± SEM of three individual mice tissues.#p < 0.01 vs control (Ctrl, untreated mice), *p < 0.05, and **p < 0.01 vs AOM alone.B Representative immunofluorescence images performed on deparaffinized sections of colon from mice treated or not with AOM or/and DMW showing PARP fluorescence intensity expressed as arbitrary fluorescence units (AFU).Results are mean ± SEM of three independent experiments.#p < 0.0001 vs control (ctrl, untreated mice) and **p < 0.01 vs AOM alone

Fig. 4
Fig. 4 Taxa affected by the azoxymethane (AOM) treatment and restored by delactosed milk whey by-product (DMW) from Mediterranean Italian dairy buffaloes.Graphs reporting the relative abundance of the genera altered by AOM.Data represent the mean ± SEM of 5/7 mice; *p < 0.05, **p < 0.01 and ***p < 0.001

Fig. 7
Fig. 7 Delactosed milk whey by-product without fat (DMW/WF) from Mediterranean Italian dairy buffaloes reduces cell viability in human colorectal cancer cell lines HCT116 and HT29.HCT116 (A) and HT29 (B) and a healthy human colonic epithelial cell line HCEC (C) were cultured in the absence or presence of increasing concentrations of DMW/WF (2-20% v/v) for 24 h.The viability rate of DMW/WF-treated cells was measured using the MTT assay, as described in Material and Methods.The insert shows the difference between the curves representing the inhibitory effect of DMW/WF in HCT116, HT29 and Data are expressed as mean ± SEM of three independent experiments.****p < 0.0001 vs control (untreated cells)

Fig. 8
Fig. 8 Delactosed milk whey by-product without fat (DMW/WF) from Mediterranean Italian dairy buffaloes triggers phenotypic changes in the human colorectal cancer cell line HCT116.Representative bright field images indicating morphological changes in HCT116 cells after 24 h incubation with or without increasing concentrations of DMW/WF (0.3-20% v/v).Cell morphology was evaluated by fluorescence microscopy using a Zoe ™ Fluorescent Cell Imager (Biorad).Magnification, × 100

Fig. 10
Fig. 10 Delactosed milk whey by-product without fat (DMW/WF) from Mediterranean Italian dairy buffaloes activates apoptosis-related protein expression in the human colorectal cell line HCT116.Representative images of the protein expressions of cleaved caspase-3, caspase-9, caspase-8 and poly (ADP) ribose polymerase (PARP) measured by western blot analysis.HCT116 cells were treated or not with DMW/WF [6% v/v (A) and 10% v/v (B) at different time points (0-24 h)].The right side of the panel shows the results of immunoblot densitometric analysis of the protein expression levels in HCT116 cells (at different time points) after DMW/WF (6% v/v or 10% v/v concentrations) treatment.Cleaved caspase-3, caspase-9, caspase-8 and PARP signals were normalized on the housekeeping proteins β-actin or vinculin.Data are expressed as mean ± S.E.M of three independent experiments.*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 vs control (Ctrl, untreated cells)

Table 1
Compounds identified in delactosed milk whey by-product without fat (DMW/WF) from Mediterranean Italian dairy buffaloes (Bubalus bubalis) by HPLC-MS analysis

Table 2
Effect of delactosed milk whey by-product (DMW; 10 ml/kg) from Mediterranean Italian dairy buffaloes (Bubalus bubalis) on the percentage of mice bearing aberrant crypt foci (ACF), ACF with more than four crypts, tumors, or more than one tumor, in the mouse model of azoxymethane (AOM)-induced colon cancer

Table 3
Putatively identified metabolites that are differentially accumulated in sera from mice treated with azoxymethane (AOM) and delactosed milk whey by-product (DMW) vs AOM