Food Chemistry 344 (2021) 128669

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Green feed increases antioxidant and antineoplastic activity of buffalo milk:

A globally significant livestock
Angela Salzano a, Gianluca Neglia a, Nunzia D’Onofrio b, *, Maria Luisa Balestrieri b,
Antonio Limone c, Alessio Cotticelli a, Raffaele Marrone a, Aniello Anastasio a,
Michael J. D’Occhio d, Giuseppe Campanile a
a
Department of Veterinary Medicine and Animal Production, University of Naples Federico II, 80137 Naples, Italy
b
Department of Precision Medicine, University of Campania Luigi Vanvitelli, 80138 Naples, Italy
c
Istituto Zooprofilattico Sperimentale del Mezzogiorno, 80055 Portici, Italy
d
School of Life and Environmental Sciences, Faculty of Science, The University of Sydney, Sydney, New South Wales 2000 Australia

A R T I C L E I N F O A B S T R A C T

Keywords: The effect of green feed on health-promoting biomolecules in milk was examined in dairy buffaloes. Buffaloes
Buffalo received a total mixed ration (TMR) (Control, C; n = 40) or TMR + alfalfa green feed (30% of diet) (Treated, T; n
Green feed = 40). Biomolecules and functional activity were measured in milk obtained twice-monthly. Treated buffaloes
Milk
had higher milk L-carnitine, acetyl-L-carnitine, propionyl-L-carnitine and δ-valerobetaine (P < 0.01). They also
δ-Valerobetaine
Short-chain acylcarnitines
had higher antioxidant activity (P < 0.01). Compared with C buffaloes, milk of T buffaloes improved the viability
Antioxidant of endothelial cells exposed to high-glucose (P < 0.01), and reduced intracellular lipid peroxidation, reactive
Antineoplastic oxygen species (ROS), and cytokine release (P < 0.01). Milk of T buffaloes inhibited with greater potency the
viability of human HCT116 and Cal 27 cancer cells (P < 0.001). The findings show that including green feed in
the diet of dairy buffaloes enhances health-promoting biomolecules and the antioxidant and antineoplastic
properties of milk.

1. Introduction The meat and milk of buffaloes have distinct health-promoting

properties compared with other livestock. The milk of Mediterranean
Buffaloes (Bubalus bubalis) are highly adaptable which makes them a buffaloes has more conjugated linoleic acid (CLA) than bovine milk
globally significant livestock species across diverse climatic zones, in (Pegolo et al., 2017). CLA is beneficial in cardiovascular disease, cancer,
developed and developing countries (D’Occhio et al., 2020). In devel­ and energy metabolism (Lehnen, da Silva, Camacho, Marcadenti, &
oping countries, buffaloes are an important source of affordable meat Lehnen,2015). Buffalo milk also has more δ-valerobetaine than bovine
and milk, they provide draught power, and they have social and cultural milk (Servillo, D’Onofrio, & Giovane et al., 2018). δ-Valerobetaine in
importance (Wanapat, & Kang, 2013). In developed countries, buffaloes buffalo milk had antioxidant and anti-inflammatory actions on endo­
are used mainly as dairy animals for products such as mozzarella which thelial cell in vitro (D’Onofrio et al., 2019). A cytotoxic effect of
is targeted at premium food markets. The largest numbers of buffaloes δ-valerobetaine was reported in HT-29 and LoVo human colon adeno­
are in Asia which also has approximately 60% of the world’s population carcinoma colon cancer cells (D’Onofrio, Cacciola et al., 2020) and in
(Cruz, 2007). Hence, buffalo food products have the potential for a head and neck squamous cell carcinomas (D’Onofrio, Mele et al., 2020).
major influence in nutrition and global health. For example, buffalo milk The antineoplastic activity of buffalo milk occurs by apoptotic cell death
has double the energy content and lower cholesterol than bovine milk, it involving the activation of sirtuin 1 and 6 (D’Onofrio, Cacciola et al.,
is very high in calcium, and it also has high amounts of minerals 2020; D’Onofrio, Mele et al., 2020).
including magnesium, phosphorus, potassium and zinc. Buffalo milk also contains γ-butyrobetaine and L-carnitine precursors

* Corresponding author at: Department of Precision Medicine, University of Campania L. Vanvitelli, via L. De Crecchio 7, 80138 Naples, Italy.
E-mail addresses: angela.salzano@unina.it (A. Salzano), gianluca.neglia@unina.it (G. Neglia), nunzia.donofrio@unicampania.it (N. D’Onofrio), marialuisa.
balestrieri@unicampania.it (M.L. Balestrieri), antolim@izsmportici.it (A. Limone), raffaele.marrone@unina.it (R. Marrone), aniello.anastasio@unina.it
(A. Anastasio), michael.docchio@sydney.edu.au (M.J. D’Occhio), giuseppe.campanile@unina.it (G. Campanile).

https://doi.org/10.1016/j.foodchem.2020.128669
Received 2 June 2020; Received in revised form 11 November 2020; Accepted 14 November 2020
Available online 19 November 2020
0308-8146/© 2020 Published by Elsevier Ltd.
A. Salzano et al. Food Chemistry 344 (2021) 128669

which are broadly present in ruminant meat and milk (Servillo, day) recorded in the 10 days before commencement of the study. The
D’Onofrio, & Neglia et al., 2018). We recently provided the first evi­ diet of control (C) buffaloes was a standard TMR whilst treated (T)
dence that combined δ-valerobetaine and γ-butyrobetaine display a buffaloes received TMR + fresh feed which comprised green alfalfa
synergistic action in inhibiting proliferation and inducing apoptosis in fodder (approximately 30% of the diet).
oral squamous Cal 27 cells (D’Onofrio, Mele et al., 2020). Betaine and L- The forage was cut twice a day to avoid any fermentation. Alfalfa at
carnitine are both essential in human health (Flanagan, Simmons, the phenological stage of re-blossoming was cut, placed into the mixing
Vehige, Willcox, & Garrett, 2010). As for betaines, their anti- wagon, chopped and administered to animals without storage. The
inflammatory and antioxidant properties are well known (Zhao et al., forage to concentrates ratio of C buffaloes was 56:44 and that of T
2018). The above beneficial effect of milk biomolecules has led con­ buffaloes was 69:31. The two diets were isonitrogenous and isoenergetic
sumers to consider milk as a highly important source of specific nutri­ and differed only in the inclusion of green feed in T buffaloes (Supple­
ents with health-promoting properties (Bauman, Mather, Wall, & Lock, mentary Material, Table S1). Animals were fed twice daily in the
2006). morning and evening. Refusals were recorded and then removed. Indi­
Lactating Holstein and Jersey dairy cows that received a total mixed vidual feedstuff and refusals from 10 animals/group were sampled
ration (TMR) and grazed grasses (Agrostis tenuis, Dactylis glomerata, weekly and analysed according to AOAC (AOAC, 1980). Energy values
Lolium perenne) and legumes (Trifolium repens, Viccia cracca) had (milk forage units = 1,700 kcal) were calculated using equations pro­
elevated beneficial CLA, linolenic acid and vaccenic acid in milk (Mo­ vided by INRA (INRA, 2007). Average feed intake for each pen was
rales-Almaraz et al., 2010). Also, lactating Holstein cows on a diet of determined daily from unconsumed feed before the next feeding. The
forage (alfalfa hay, Chinese wildrye, corn silage) + TMR had a greater amount and composition of refusals was used to calculate DMI and diet
content of C16:0 and saturated fatty acids (FA) in milk compared with composition. Individual feed intake and differences (Δ) between nutri­
cows fed a TMR alone (Zhang, Ao, Khas-Erdene, & Dan, 2019). In tive intake and relative requirements were estimated as previously re­
addition, lactating Holstein cows that had high grazing (ryegrass) + low ported (Campanile et al., 1998):
concentrate (1–2 kg) showed greater levels of CLA and omega-3 FA in Dry matter (DM) intake = 91 g × MW + 0.27 kg × kg ECM
milk than cows with low grazing + greater concentrate (6–8 kg) (Marín, ΔCP = g CP intake – (80 g CP × 100 kg live weight + 2.7 g CP × g
Meléndez, Aranda, & Ríos, 2018). Furthermore, lactating Friesian cows milk protein yield)
that grazed exclusively perennial ryegrass or perennial ryegrass/white ΔMFU = MFU intake – [(1.4 + 0.6 × 100 kg live weight) × 1.1 +
clover had a greater than 2-fold increase in CLA C18:2 cis-9, trans-11 in 0.44 MFU × kg ECM]
milk compared with cows that received just a TMR (O’Callaghan et al., The body condition score (BCS) of each buffalo was recorded weekly
2016). using a 1 to 9 scale (Wagner, Lusby, Oltjen, Rakestraw, Wettemann, &
Buffaloes with a proportion of green feed (pasture) in the diet had a Walters, 1988).
beneficial profile of milk FA and other nutrients (Secchiari et al., 2003).
In beef cattle, grass feeding was associated with greater amounts of CLA 2.2. Milk sampling and composition
and ω-3 FA in meat (Daley, Abbott, Doyle, Nader, & Larson, 2010).
As noted above, the amounts of beneficial FA and other nutrients in The composition of milk (fat and protein) was determined for indi­
milk were increased in both bovine dairy cows and buffaloes that had vidual buffaloes daily in milk samples taken in the morning and after­
green feed included in their diet. In studies where Italian Mediterranean noon and proportionally mixed according to milk yield at each milking
buffaloes had greater amounts of CLA (Pegolo et al., 2017) and (60% morning and 40% afternoon milk). Samples were analysed in
δ-valerobetaine (Servillo, D’Onofrio, & Giovane et al., 2018) in milk triplicate using IR spectroscopy (Milkoscan 139, Foss Electric, Hillerød,
compared with bovine dairy cows, buffaloes were fed a TMR diet. Given DK) that was calibrated with a buffalo standard. Energy corrected milk
the health-promoting benefits of CLA, δ-valerobetaine and L-carnitine, (ECM = 740 kcal) was calculated using the formula for buffalo cows
which occur naturally in buffalo milk, it is important to establish if the (Campanile, De Filippo, Di Palo, Taccone, & Zicarelli, 1998) ([{fat
amounts of these significant functional nutrients can be further (g⋅kg–1) – 40 + protein (g⋅kg–1)– 31} × 0.01155] + 1) × milk yield.
increased in buffaloes. The inclusion of green feed in the diet of milking Pooled samples of the morning and afternoon milking were obtained
buffaloes could be cost-effective and would have consumer preference on days 0, 14, 28, 42 and 56, separately for C and T buffaloes (total of 10
for more natural livestock production. Also, the majority of buffaloes are pooled samples each for C and T buffaloes). The protein content from
in Asia where fresh plant material is the main source of feed. The aim in pooled milk samples for C and T buffaloes was spectrophotometrically
the present study was to determine if the inclusion of green feed in the evaluated. Pooled samples were split into aliquots and stored at − 80 ◦ C
diet of Italian Mediterranean buffaloes further increased the milk con­ until required for analyses.
tent of betaines such as δ-valerobetaine, L-carnitine (L-carnitine, acetyl-
L-carnitine, propionyl- L-carnitine), and FA. 2.3. Betaine and L-carnitine determination

2. Materials and methods The content of γ-butyrobetaine, glycine betaine, δ-valerobetaine, L-

carnitine, acetyl-L-carnitine and propionyl-L-carnitine was determined in
2.1. Animals and dietary treatment duplicate pooled soluble milk extracts as previously described (Servillo,
D’Onofrio, & Giovane et al., 2018). Briefly, samples of milk were
The study followed standard veterinary practices and had institu­ centrifuged at 3,000 × g for 15 min at 4 ◦ C to recover fat globules. The
tional approval from the Ethical Animal Care and Use Committee of the latter were filtered through a 5 μm Millipore filter to remove high mo­
University of Naples Federico II (Protocol No. 996072017). The coop­ lecular weight (MW) components and any precipitate that may have
erating commercial buffalo dairy in southern Italy provided informed been transferred when recovering fat globules. This was followed by
consent for the use of animals. The study was carried out over 56 days at filtration through an Amicon Ultra 0.5 mL centrifugal filter with a 3 kDa
a commercial buffalo dairy in southern Italy. Italian Mediterranean molecular weight cut-off to yield low MW components, including short-
dairy buffaloes (n = 80; 526 ± 16 kg live weight; 4.6 ± 0.3 years old) chain acylcarnitines (between 161.2 and 245.3 molecular weight),
that had undergone a 14-day adaptation were used over a period of 56 glycine betaine (153.6 molecular weight), γ-butyrobetaine (146.1 mo­
days. They were maintained in pens with a concrete floor and were lecular weight) and δ-valerobetaine (159.23 molecular weight). Anal­
milked twice daily in the morning and afternoon. Animals were ysis involved HPLC-ESI-MS/MS with an Agilent LC-MSD SL quadrupole
randomly assigned to two equal groups according to days in milk (80 ± ion trap and a 1100 series liquid chromatograph (Supelco Discovery C8
7 days), parity (2.5 ± 0.2), and average milk production (10.6 ± 0.5 kg/ column, 250 × 3.0 mm, particle size 5 μm) under isocratic conditions,

2
A. Salzano et al. Food Chemistry 344 (2021) 128669

0.1% formic acid in water, at flow rate of 100 μL/min. Quantification of cell viability. After treatments, a MTT solution in phosphate-buffered
each compound involved comparison of the peak area of its most intense saline (PBS) was added in culture medium to a final concentration of
MS2 fragment with the respective calibration curve built with solutions 0.5 mg/mL and then incubated for 4 h at 37 ◦ C, until intracellular purple
of standards products (L-carnitine, acetyl-L-carnitine, propionyl-L-carni­ formazan crystals were visible under the microscope. After removing
tine, glycine betaine, δ-valerobetaine and γ-butyrobetaine) from Sigma- MTT solution, 100 µL of solubilization solution (dimethylsulfoxide) was
Aldrich (Milan, Italy) (Servillo, D’Onofrio, & Neglia et al., 2018). added to each well and incubated at 37 ◦ C for 30 min on a shaking table
δ-Valerobetaine was prepared as previously described (Servillo, to dissolve formazan crystals. The absorbance at 570 nm was measured
D’Onofrio, & Neglia et al., 2018). Standard solutions were prepared by using a microplate reader (Model 680 Bio-Rad).
serial dilution of standard stock solutions (2 mg/L) with water con­
taining 0.1% formic acid. Linearity was assessed by correlation co­ 2.7. Reactive oxygen species assay
efficients (r2) greater than 0.99 for all compounds. Precision and
accuracy for all compounds in milk ranged from 95% to 105%. Intracellular reactive oxygen species (ROS) were assayed according
to the manufacturer’s instructions (Abcam, ab186028), as previously
2.4. Ferric reducing antioxidant power and total antioxidant assay described (D’Onofrio et al., 2019). Bovine endothelial cells (6 x103
cells/well) were seeded in 96-well plates and pre-treated for 12 h with
Milk ferric reducing antioxidant power (FRAP; Assay Kit extracts (10 μL) of day 56 pooled milk for C and T buffaloes, before
(MBS169262) and total antioxidant capacity (TAC; Assay Kit (#K274- starting high-glucose (30 mM) treatment for 48 h (4 replicates). After
100) were determined in pooled milk samples using the manufacturer’s treatment, endothelial cells were incubated with ROS orange working
instructions. Milk samples (4 replicates) were diluted 1:10 (v/v) with solution (100 µL/well), a sensor to quantify ROS in live cells, at 37 ◦ C
H2O and a 1 µL aliquot was used for each assay. Samples were incubated and 5% CO2 for 60 min. The fluorescence intensity was monitored at
at room temperature for 90 min protected from light, before measuring 540 nm excitation wavelength and 570 nm emission wavelength using a
the absorbance at 570 nm, or monitoring the increase in absorbance at Tecan Infinite 2000.
594 nm for 1 h at 37◦ C for TAC and FRAP assay, respectively. The
positive control comprised ascorbic acid (1 µg/mL). FRAP value was 2.8. Cytokine assay
expressed as Fe2+ iron equivalents (µM; range = 10–250 × µg/mL of
ascorbic acid) and TAC was expressed as Trolox equivalent capacity Assay kits were used to determine TNF-α (Cymax TNF-alpha ELISA,
(range = 10–250 × nmol µg/mL of ascorbic acid). YIF-LF-EK0193), IL-6 (Cymax IL-6 ELISA, YIF-LF-EK0260) and IL-1β
(Cymax IL-1beta ELISA, YIF-LF-EK0276 ELISA), according to the man­
2.5. Cell culture and treatment ufacturer’s instructions as previously described (D’Onofrio et al., 2019)
Briefly, bovine endothelial cells (4 × 105 cells/well) were seeded in a 6-
Bovine endothelial cells (ATCC, CCL 209), HCT116 human colon well plate and incubated with high-glucose (30 mM) for 48 h (4 repli­
cancer cells (ATCC, CCL-247), and Cal 27 human oral squamous carci­ cates). At the end of high-glucose treatment, endothelial cell lysates
noma cells (ATCC, CRL-2095), were grown in minimum essential me­ (100 μL) were added to pre-coated assay kit well plates and incubated
dium (MEM, Gibco, 11095-080), McCoy’s 5A Medium (ATCC, 30-2007) for 1 h at room temperature and protected from light. At the end of
and Dulbecco’s Modified Eagle Medium (DMEM, Gibco, 21969035), incubation, the excess of cytokines, along with other components of the
respectively. All media were supplemented with 20% heat-inactivated samples, were removed by washing and polyclonal antibodies against
fetal bovine serum, penicillin (100 U/mL), and streptomycin (0.1 mg/ TNF-α, IL-6 and IL-1β were added. This incubation was followed by the
mL). Conditions were 37 ◦ C in a humidified atmosphere, 95% air, 5% addition to each well of anti-rabbit IgG horseradish peroxidase (HRP) for
CO2. Cell lines were exposed to milk extract prepared by removal of fat 20 min at room temperature. Quantitative analysis was performed by
globules by centrifugation at 3,000 × g for 15 min at 4 ◦ C. The skimmed measuring absorbance at 450 nm by using a Tecan Infinite 2000.
samples were filtered through a 5 μm Millipore filter followed by
filtration through an Amicon Ultra 0.5 mL centrifugal filter with a 3 kDa 2.9. Lipid peroxidation assay
MW cut-off. Before being administered to cells, milk extracts were
filtered through 0.22 μm Millipore filters. Milk extracts (10 µL) from C Lipid peroxidation refers to the oxidative degradation of lipids.
and T buffaloes were added to endothelial cells before the exposure of Quantification of lipid peroxidation is essential to assess oxidative stress.
cell to normal-glucose (5.5 mM) (nGluc) or high-glucose (30 mM) Lipid peroxidation forms reactive aldehydes such as malondialdehyde
(hGluc) for 48 h. HCT116 and Cal 27 cells were exposed to milk of C and (MDA) which is used as a marker of oxidative damage and stress. The
T buffaloes (up to 30% v/v) added in complete culture medium up to 48 MDA from lipid peroxidation was determined using an MDA assay kit
h. Time-dependent experiments were performed up to 72 h by using (Abcam, ab118970), according to the manufacturer’s instructions as
30% (v/v) of milk extracts. Control cells were treated with corre­ previously described (D’Onofrio et al., 2019). Briefly, after high-glucose
sponding volumes (% v/v) of Hanks’ balanced salt solution (HBSS)-10 treatment (with or without extracts of day 56 pooled milk for C and T
mM Hepes. Before performing in vitro cell assays, medium without buffaloes) (4 replicates) endothelial cells (2 × 106) were harvested and
phenol red (RPMI 1640, Thermo Fisher, 11835030) was added. homogenized in 303 µL of MDA lysis solution. To generate
malondialdehyde-thiobarbituric acid (MDA-TBA) adduct, 600 µL of TBA
2.6. MTT assay reagent were added to each well containing 200 µL of sample (n = 4 in
triplicates), incubated at 95 ◦ C for 60 min, and then cooled to room
The MTT assay of cell metabolic activity (Sigma-Aldrich) was used as temperature in an ice bath for 10 min. The MDA-TBA adduct was
an indicator of cell viability, as previously described (D’Onofrio et al., quantified colorimetrically, measuring the absorbance on a microplate
2019). Briefly, endothelial cells (4 × 103 cells/well), HCT116 cells (2 × reader Model 680 Bio-Rad at 532 nm. Total MDA levels (nmol/µg of
103 cells/well), and Cal 27 cells (2 × 103 cells/well) were seeded into protein) were calculated based on the standard curve and normalized to
96-well plates and cultured in complete medium overnight at 37 ◦ C in a protein levels.
humidified atmosphere (95% air and 5% CO2) the day before starting
treatments with extracts of pooled milk for C and T buffaloes obtained at 2.10. Catalase and superoxide dismutase assay
day 56 (6 replicates). This colorimetric assay uses reduction of a yellow
tetrazolium salt (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium Catalase (CAT) (Catalase Activity Assay Kit, Abcam, ab83464) and
bromide, or MTT) to measure cellular metabolic activity as a proxy for superoxide dismutase (SOD) (Superoxide Dismutase Activity Assay Kit,

3
A. Salzano et al. Food Chemistry 344 (2021) 128669

Abcam, ab65354) activities were measured in lysates of control and pool of fecal samples for each animal was prepared by mixing the in­
high-glucose treated endothelial cells, either exposed or not exposed to dividual samples taken over 5 days. To avoid fermentation the daily
extracts of day 56 pooled milk from C and T buffaloes (3 replicates). samples were stored at − 20 ◦ C until preparation of the pooled sample.
After detaching and lysing the adherent endothelial cells, aliquots (150 Samples of TMR and feces were analyzed as per AOAC (AOAC, 1980).
μL) of cell lysate were transferred to wells of a 96-well microplate. For
SOD assay, each well contained tetrazolium salt (WST-1) and xanthine 2.14. Statistical analyses
oxidase and the reaction mixture was incubated at 37 ◦ C for 20 min. The
rate of WST-1 formazan formation (inversely proportional to SOD ac­ Statistical analyses were performed using SPSS (23.0) for Windows
tivity) was measured at 450 nm. The SOD activity (U/ µg of protein) was 10 (SPSS Inc., Chicago, IL). The buffalo was used as the experimental
calculated by the formula: % inhibition = (Abs control - Abs sample) / unit. Data on milk yield and quality, energy corrected milk (ECM), were
Abs control × 100. In the CAT activity assay, CAT present in the sample analyzed by ANOVA for repeated measures with treatment as the main
reacts with hydrogen peroxide (H2O2) to produce water and oxygen. The factor. Buffaloes were tested within treatments. Day of sampling was the
unconverted H2O2 reacts with probe to generate a product that can be repeated measure. One-way ANOVA was used to compare data for TAC
measured colorimetrically at 570 nm. The CAT activity was obtained by and FRAP of plasma and milk, and also FA and functional molecules of
averaging the absorbances of the samples and interpolating them with milk. Scheffe’s test was used for pair-wise comparisons. Results are
the formaldehyde, originated by the CAT, used as standard. Both assays mean ± standard error mean (sem). A statistically significant difference
used a microplate reader (Model 680 Bio-Rad) to reveal the colorimetric was accepted at P < 0.05. Multiple linear regression was performed with
reactions. the SPSS (23.0) stepwise procedure in order to evaluate the relationship
between ECM, BCS and the dietary characteristics.
2.11. Lipid assay
3. Results
Milk fat extraction was performed according to a modified Rose-
Gottlieb method. Briefly, ammonia 25% (0.4 mL), ethyl alcohol 95% 3.1. Dry matter intake and body condition score
(1 mL), and hexane (5 mL) were added in duplicate in analytical por­
tions of 1 g of sample. Samples were centrifuged at 900 × g for 20 min Total dry matter intake (DMI) was similar for C buffaloes (16.3 ± 0.3
and the upper layer was collected. Extraction was repeated using ethyl kg/day) and T buffaloes (16.2 ± 0.2 kg/day) (Supplementary Material,
alcohol 95% (1 mL) and hexane (5 mL). Morning and afternoon pooled Table S1). Body condition score (BCS) was also similar for C Buffaloes
milk samples were again centrifuged at 900 × g for 20 min and the upper (7.7 ± 0.2) and T buffaloes (7.6 ± 0.3).
layer was collected. A third extraction was undertaken using 5 mL of
hexane and again samples were centrifuged at 900 × g for 20 min and 3.2. Milk yield and quality
the upper layer was collected. The extracted lipid phase was then dis­
solved in hexane and purified using sodium chloride saturated solution Milk yield was approximately 10 kg/day throughout the study for
(3 mL). The hexane phase containing purified lipids was dried over both C and T buffaloes (Table 1). There were no differences for protein
anhydrous sodium sulphate and under nitrogen. The total lipid obtained percentage and energy corrected milk yield, while the fat percentage
was determined gravimetrically using the AOAC 1980 method. was higher (9.1 ± 0.1 vs 9.5 ± 0.1 respectively in T and C group; P <
0.05) in C buffaloes (Table 1). Milk total protein content did not differ
2.12. Fatty acid assay between C (41.4 ± 3.7 mg/mL) and T (49.0 ± 2.1 mg/mL) buffaloes.
The relationship between ECM and ΔCP, ΔMFU and BCS in C and T
The fatty acid methyl esters (FAME) of morning and afternoon buffaloes is expressed by the following equations:
pooled milk samples were prepared in duplicate with a base-catalysed Control: ECM (kg) = -20.044 + 0.19 (ΔCP);R2 = 0.859
transesterification according to the FIL-IDF standard procedure (Inter­ Treated: ECM (kg) = -2.979 – 1.984 (ΔMFU) + 0.11 (ΔCP); R2 =
national Dairy Federation Milk Fat, 1999). A Dani Instruments GC 1000- 0.707
Gas Chromatograph equipped with a programmed temperature vapor­ Energy production with milk (ECM) was influenced negatively by
izer (PVT) and flame ionization detector (FID) was used. The FAME was ΔCP in C buffaloes whereas it was influenced positively by ΔMFU and
separated on a capillary column (CP-select CB for Fame; 100 m × 0.32 negatively by ΔCP in T buffaloes. The BCS did not influence the energy
mm i.d., 0.25-μm film thickness). The identification and quantification production.
of peaks was performed using the Supelco 37 Component FAME MIX
(Supelco Bellefonte, PA) as external standards and GC retention time 3.3. Milk biomolecules
data available in the literature. Fatty acid concentration was calculated
through response factors to convert peak areas into weight percentages. Pooled milk of T buffaloes had higher (P < 0.01) concentrations of L-
The atherogenic index (log(triglycerides/high density lipoproteins); carnitine (42.0 ± 0.5 vs 31.5 ± 0.7 mg/L, respectively, for T and C
index of cardiovascular benefit) was calculated as follows: buffaloes), propionyl-L-carnitine (21.1 ± 0.4 vs 14.8 ± 1.3 mg/L) and
[C12:0 + (4 * C14:0) + C16:0)]/(ω3 + ω6 + mono-unsaturated FA) acetyl-L-carnitine (49.8 ± 0.8 vs 39.1 ± 0.5 mg/L) (Table 2). Treated

2.13. Plasma antioxidant capacity and metabolic status Table 1

Average milk production data and quality for control buffaloes that received a
Ten C buffaloes (63.6 ± 6.1 days in milk) and ten T buffaloes (64.5 ± total mixed ration (TMR) and treated buffaloes that received TMR + fresh feed
6.2 days in milk) underwent blood sampling every 14 days to assess (approximately 30% of the diet). Results are expressed as mean ± sem.a,b, P <
plasma antioxidant capacity and metabolic status. Serum and plasma, 0.05.
collected separately, were obtained by centrifugation at 1,800 × g for Control buffaloes Treated buffaloes
20 min and stored at − 80 ◦ C until analysis. Total antioxidant activity (group C) (group T)
was assessed as described above for the milk. Milk yield (kg/day) 9.8 ± 0.2 9.7 ± 0.2
The organic matter and cell wall digestibility were evaluated every Fat (%) 9.5 ± 0.1 a 9.1 ± 0.1b
month in the same ten C buffaloes and ten T buffaloes by using acid Protein content (PC) 4.5 ± 0.0 4.5 ± 0.0
insoluble ashes as internal marker (Van Keulen and Young, 1977) on Energy corrected milk 17.4 ± 0.4 16.8 ± 0.3
(ECM)
fecal samples (200 g) collected for 5 consecutive days for each animal. A

4
A. Salzano et al. Food Chemistry 344 (2021) 128669

Table 2 Saturated FA (SFA) (P < 0.01), mono-unsaturated FA (MUFA), and poly-

Functional biomolecules in milk from control buffaloes that received a total unsaturated FA (PUFA) (P < 0.05) were higher for T buffaloes (Table 3).
mixed ration (TMR) and treated buffaloes that received TMR + fresh feed Significant differences were found between C and T buffaloes for ω-3 (P
(approximately 30% of the diet). Results (mg/L) are expressed as mean ± sem.A, < 0.01), ω-6 (P < 0.05), ω-6/ω-3 (P < 0.05) and the ratio of PUFA/SFA
B
, P < 0.01.
Control buffaloes (group Treated buffaloes (group Table 3
C) T)
Fatty acid composition (% of total fatty acid) of bulk milk from control buffaloes
L-carnitine 31.5 ± 0.7 A
42.0 ± 0.5B that received a total mixed ration (TMR) and treated buffaloes that received
A
acetyl-L-carnitine 39.1 ± 0.5 49.8 ± 0.8B TMR + fresh feed (approximately 30% of the diet). Results are mean ± sem.a,b, P
A
propionyl-L- 14.8 ± 1.3 21.1 ± 0.4B < 0.05; A,B, P < 0.01. SFA, saturated fatty acids; MUFA, mono-unsaturated fatty
carnitine
acids; PUFA, and poly-unsaturated fatty acids.
γ-butyrobetaine 4.8 ± 0.2 4.1 ± 0.2
δ-valerobetaine 18.2 ± 0.6 A 22.2 ± 0.5B Control buffaloes Treated buffaloes
glycine betaine 7.0 ± 0.02 7.2 ± 0.3 (group C) (group T)

Fatty acids
SFA 76.0 ± 0.3 A 71.9 ± 1.1B
buffaloes also had higher milk δ-valerobetaine (22.2 ± 0.5 vs 18.2 ± 0.6 MUFA 20.9 ± 0.3 A 23.5 ± 0.7B
mg/L, respectively, for T and C buffaloes; P < 0.01). Concentrations of PUFA 3.1 ± 0.1 a 4.5 ± 0.4b
glycine betaine and γ-butyrobetaine did not differ between C and T PUFA/SFA 0.0 ± 0.0 A 0.1 ± 0.0B
buffaloes (Table 2). ω-3 0.3 ± 0.0 A 0.6 ± 0.0B
ω-6 2.8 ± 0.1 a 3.8 ± 0.3b
ω-6/ω-3 8.7 ± 1.1 a 6.0 ± 0.2b
3.4. Milk and plasma antioxidant capacity Atherogenic index 3.9 ± 0.1 A 3.1 ± 0.2B
Caproic acid C6:0 3.1 ± 0.0 a 2.7 ± 0.2b
Myristic acid C14:0 14.1 ± 0.0 A 12.4 ± 0.4B
Fresh feed influenced the antioxidant capacity of milk. The TAC and Palmitic acid C16:0 32.1 ± 0.1 a 31.2 ± 0.3b
FRAP assays showed that milk antioxidant activity was higher (P < 0.01) Margaric acid C17:0 0.7 ± 0.0 A 0.8 ± 0.0B
for T buffaloes, starting from day 14 until the end of the study (Fig. 1). Oleic acid C18:1 n9c 16.7 ± 0.3 a 18.8 ± 0.6b
Vaccenic acid C18:1 trans-11 0.2 ± 0.0 A 0.3 ± 0.0B
Green feed also influenced the antioxidant power of blood plasma. Both
Rumenic acid CLA C18:2 n6t 0.3 ± 0.0 0.3 ± 0.0
TAC and FRAP of plasma were higher (P < 0.01) for T buffaloes, from Linoleic acid CLA C18:2 n6c 2.3 ± 0.1 a 3.3 ± 0.3b
day 14 until the end of the study (Fig. 1). γ-Linolenic acid C18:3 n6 0.2 ± 0,0 0.2 ± 0.0
α-Linolenic acid C18:3 n3 0.3 ± 0.0 A 0.6 ± 0.0B
Arachidic acid C20:0 0.3 ± 0.0 a 0.2 ± 0.0b
3.5. Fatty acids Docosahexaenoic acid C22:6 0.0 ± 0.0 a 0.0 ± 0.0b
n3
The profile of FA showed different trends between C and T buffaloes.

Fig. 1. In vitro antioxidant activity of buffalo milk and plasma. Total antioxidant activity (TAC) and Ferric Reducing Antioxidant Power (FRAP) were used to
determine the antioxidant activity of (A, B) milk and (C, D) plasma of control buffaloes (group C) that received TMR and treated buffaloes (group T) that received
TMR + fresh feed (approximately 30% of the diet). Ascorbic acid (10–250 Trolox equivalent capacity) was used as control. *P < 0.01 vs Ctr; § P < 0.05 vs C at 0 d; ¶ P
< 0.01 vs C at 14 d; ǂ P < 0.01 vs T at d 0; ¶¶ P < 0.01 vs C at 28 d; ǂǂ P < 0.01 vs T at 14 d; ¤ P < 0.01 vs C at 42 d; Ɨ P < 0.01 vs T at 28 d; ɨ P < 0.01 vs C at 56 d.
Results are mean ± sem.

5
A. Salzano et al. Food Chemistry 344 (2021) 128669

(P < 0.01) (Table 3). In addition, the atherogenic index was lower in T 3.7. Endothelial antioxidant and anti-inflammatory action
buffaloes compared to C buffaloes (P < 0.01).
The cytoprotective effect of buffalo milk extract from C and T buf­
3.6. Endothelial cell cytoprotective effect during high-glucose treatment faloes was accompanied by a reduction in intracellular ROS. ROS were
greatest for high-glucose and less for high-glucose + C buffaloes (P <
The effects of buffalo milk extracts on cell viability was tested by 0.05) and high-glucose + T buffaloes (P < 0.01) (Fig. 2B). Cytokine
incubating endothelial cells for 48 h with milk extracts (10 μL) to a final (TNF-α, IL-1β, IL-6) release was reduced in high-glucose + C buffaloes
volume of 100 μL of incubation media (D’Onofrio et al., 2019). The (P < 0.05) and further reduced in high-glucose + T buffaloes (P < 0.01)
addition of high-glucose concentration (hGluc, 30 mM) was used to (Fig. 2C). Lipid peroxidation under high-glucose was decreased by co-
induce the hyperglycemic oxidative microenvironment which, in turn, expose to buffalo milk extracts (Fig. 2D). The increased activity of
determines ROS generation, cell lipid peroxidation and inflammatory catalase (CAT) and superoxide dismutase (SOD), the most important
cytokine release (D’Onofrio et al., 2019). Endothelial oxidative damage anti-oxidative enzymes that protect cells from free radicals, confirmed
induced by high-glucose was reduced by co-incubation with milk the cytoprotective effects of the milk extracts for C and T buffaloes (P <
extract. The greatest protective effect from oxidative damage was 0.05 vs high-glucose) (Fig. 2E, F).
observed with 10 μL of milk extract for T buffaloes (hGluc + T milk) at
48 h of co-incubation, and this was greater (P < 0.01) than for C buf­
3.8. Cytotoxic effects on cancer cell viability
faloes (hGluc + C milk) (Fig. 2A).
The cytotoxic effect of milk extracts for C and T buffaloes was
evaluated in HCT116 and Cal 27 cells. Results showed a dose-dependent
capacity of milk to inhibit cancer cell viability (Fig. 3). The highest

Fig. 2. Cytoprotective effect of milk extracts against endothelial damage induced by hyperglycemia. Endothelial cells were treated for 48 h with normal-glucose
(nGluc), high-glucose (hGluc), or group C and group T milk extract (10 µL), alone or in combination with hGluc. (A), cell viability (MTT assay); (B), intracellular
ROS; (C), cytokine levels; (D), lipid peroxidation (LPO); (E), superoxide dismutase (SOD) activity; (F), catalase (CAT) activity. Results are mean ± sem; ‡P < 0.001 vs.
nGluc, *P < 0.05 vs. hGluc; **P < 0.01 vs. hGluc.

6
A. Salzano et al. Food Chemistry 344 (2021) 128669

Fig. 3. Inhibition of cancer cell viability. Dose- and time- dependent cell viability of (A, B, C) Cal 27 human oral squamous carcinoma cells and (C, D, F) HCT116
human colon cancer cells exposed to milk extracts from group C and group T buffaloes at 24, 48 and 72 h. Control cells (Ctr) were grown in medium containing the
same volume (% v/v) of HBSS-10 mM Hepes. The MTT assay was used to determine cell viability. Morphology under inverted microscope (Axiovert-10, Zeiss
Microscope) of Cal 27 (C) and HCT116 cells (F) treated for 72 h with 10 μL milk extracts from group C and group T buffaloes. Magnification = 10X. Results are mean
± sem. *P < 0.05 vs Ctr, **P < 0.01 vs Ctr, †P < 0.001 vs Ctr; • P < 0.05 vs C.

reduction in cell viability was observed after incubation with 30% (v/v) 3.9. Diet digestibility
milk extracts for T buffaloes (P < 0.001). Time-dependent experiments
(up to 72 h) confirmed that the highest potency was achieved by milk for Diet digestibility was generally higher (P < 0.01) in C buffaloes
T buffaloes (30% v/v) added for 48 h to culture media of HCT116 cells compared to T buffaloes. However, hemicellulose (P < 0.05) and ash (P
(46% of inhibition of cell viability) (P < 0.001 vs C buffaloes) and Cal 27 < 0.01) digestibility were higher in T buffaloes (Supplementary Mate­
cells (48% of inhibition of cell viability) (P < 0.05 vs C buffaloes) rial, Table S2). There were no apparent differences between C and T
(Fig. 3). buffaloes in energy digestibility (GE) and, hence, MFU/ kg of dry matter
intake also did not differ.

7
A. Salzano et al. Food Chemistry 344 (2021) 128669

4. Discussion exposed to high-glucose. The milk extracts also improved the cytotoxic
effects on human colon cancer cells and tongue squamous carcinoma
Future livestock production will be characterized by ethical and cells (D’Onofrio, Cacciola et al., 2020; D’Onofrio, Mele et al., 2020). The
sustainable systems that produce food with functional properties that greater bioactivity in milk extracts of buffaloes that received green feed
help support human health and wellbeing. Studies have tended to report confirmed recent evidence that the antioxidant and anti-inflammatory
only the positive functional properties of animal and vegetable derived effects of milk extracts were improved when enriched with pure
food without considering potential negative (dysfunctional) effects of δ-valerobetaine (D’Onofrio et al., 2019).
some food components on human health. The manner in which livestock Vaccenic acid is produced by the rumen and had antiatherogenic and
are managed can modify the balance between functional and dysfunc­ hypolipidemic actions in animal models (Bassett et al., 2010). Vaccenic
tional components in animal food (Salzano et al., 2019). The diet of acid is the dietary precursor of CLA and potential health benefits in
livestock is an important strategy to influence the health properties of humans may be a direct effect of vaccenic acid, or through CLA (Song
animal food. For example, the provision of green feed can help improve et al., 2019). The greater amount of vaccenic acid in the milk of buf­
milk quality whilst contributing to livestock sustainability. In the pre­ faloes that received green feed could be presumed to be beneficial to
sent study, the inclusion of green feed in the TMR diet of dairy buffaloes humans. The FA composition of buffalo milk has not been as extensively
increased the concentration in milk of all the health-promoting bio­ studied as FA in bovine milk (Secchiari et al., 2003; Pegolo et al., 2017).
molecules examined. This included δ-valerobetaine, L-carnitine, short- The milk of buffaloes fed the TMR diet had a higher overall fat per­
chain acylcarnitines, vaccenic acid, and ω-3 long-chain PUFA (linoleic centage than the milk of buffaloes that received TMR + green feed. The
acid, linolenic acid). L-Carnitine, acetyl-L-carnitine, propionyl-L-carni­ greater fat percentage in the former buffaloes may have been due to
tine, and δ-valerobetaine are heat resistant as they are present in ricotta higher fibre digestibility of the TMR and more acetic acid production by
cheese that is produced from buffalo milk whey at 100 ◦ C (Salzano et. the rumen (Zebeli, Tafaj, Steingass, Metzler, & Drochner, 2006). The
al., 2019). The presence of betaines and carnitines in the milk of buf­ milk of buffaloes fed TMR + green feed had higher amounts of beneficial
faloes, together with proteins of high nutrient value, makes buffalo milk long-chain PUFA which confirmed previous studies in ruminants (Uzun
an important nutrient dense food. Betaines and carnitines are also pre­ et al., 2018). Greater long-chain PUFA in the milk of ruminants
sent in the ‘Healthy’ Nordic and Mediterranean diets which are plant- consuming green feed is due to relatively high amounts of α-linoleic
based but include some animal protein (Tuomainen et al., 2019). acid, linolenic acid and oleic acid in green feed (Uzun et al., 2018).
Proteins in buffalo milk include several peptides with recognized Buffaloes fed TMR + green feed had lesser amounts of detrimental short-
health benefits (antimicrobial, immunomodulating, antitumor, antidia­ chain fatty acids (SFA) in milk. The provision of green feed favours
betic, antihypertensive and antioxidant properties) (Basilicata et al., PUFA at the expense of SFA and medium chain FA in the milk of ru­
2018). δ-Valerobetaine is synthesized from dietary Nε-trimethyl lysine minants (Secchiari et al., 2003). The combined findings for δ-valer­
(TML) and was recently identified in ruminant milk and meat (Servillo, obetaine, vaccenic acid, and ω-3 long-chain-PUFA, lead to the
D’Onofrio, & Giovane et al., 2018; Servillo, D’Onofrio, & Neglia et al., conclusion that the milk of buffaloes fed TMR + green feed would have
2018). This betaine has both antioxidant and anti-inflammatory prop­ greater health-promoting properties than the milk of buffaloes fed TMR
erties and was reported to suppress pro-inflammatory cytokines such as alone (Chaudhry, 2008).
TNF-α, IL-1β and IL-6 through multiple pathways involving NAD- The inclusion of green feed in the diet of high-producing bovine dairy
dependent deacetylase sirtuin-1 (SIRT1) and SIRT6 modulation cows was associated with reduced dry matter intake (DMI) and failure to
(D’Onofrio et al., 2019). The role of δ-valerobetaine as an epi-nutrient reach milk production potential (Kolver, 2003). In the present study,
was strengthened by recent evidence of its antineoplastic effects in DMI was not reduced in buffaloes fed TMR + green feed. This finding
human colon and in head and neck squamous carcinoma cell lines was consistent with other studies in ruminants where the inclusion of
(D’Onofrio, Cacciola et al., 2020; D’Onofrio, Mele et al., 2020). The green feed to approximately 30% of the total diet did not reduce DMI
greater δ-valerobetaine content of milk and plasma in buffaloes that (Morales-Almaráz et al., 2010). Notwithstanding, reduced production in
received green feed in the present study provides one explanation for the ruminants fed green feed could be due to a combination of changes in
greater total antioxidant capacity compared with buffaloes fed the TMR nutrient ingestion, rumen fermentation, and gastrointestinal digestion
diet. Greater TAC of milk may, therefore, be a common feature in ru­ (Bargo, Muller, Delahoy, & Cassidy, 2002).
minants that consume green feed. TMR diets are a poor source of TML as Whilst the present study was undertaken in Italian Mediterranean
the concentration of protein-bound TML in protein sources (soybean dairy buffaloes, large numbers of different sub-species of buffaloes occur
protein, casein and wheat gluten) used in feed rations is negligible throughout Asia, and in particular South Asia. Asia is home to around
(Rebouche, Lehman, & Olson,1986). Leafy vegetables, and in particular 60% of the world’s human population and buffalo derived food,
alfalfa, contain relatively high amounts of TML which participates in L- particularly fresh milk, is an important part of the diet.
carnitine biosynthesis (Servillo, Giovane, Cautela, Castaldo, & Bales­ In conclusion, the present study provides further evidence that the
trieri, 2014). Clover and ryegrass produced favourable nutrient profiles consumption of buffalo milk can be beneficial to human health and
in cattle in terms of gross composition, macroelements, and trace ele­ wellbeing. The high content of energy and minerals, together with
ments (O’Callaghan et al., 2016; Gulati et al., 2018). protein, would be particularly beneficial to infants and young children.
L-Carnitine and short-chain acylcarnitines are present in the milk of Particularly relevant at older ages are the reported beneficial effects in
Italian Mediterranean dairy buffaloes and have a fundamental role in cardiovascular disease and cancer, energy metabolism and body
human health (Bene, Hadzsiev, & Melegh, 2018). These biomolecules composition, and allergy and asthma. The functional biomolecules
have antioxidant and anti-inflammatory effects on endothelial cells and examined in the present study are conserved during the processing of
platelets and, additionally, a neuroprotective potential in both the buffalo milk into mozzarella cheese, yogurt, and ricotta cheese (Salzano
central and peripheral nervous systems (Servillo, D’Onofrio, & Giovane et. al; 2019). Further research on strategies to optimize beneficial bio­
et al., 2018). The relatively high amounts of short-chain acylcarnitines molecules in buffalo milk has potential for major impact on global
in buffalo milk result from the production of volatile fatty acid (VFA) health. Most buffalo milk production in Asia is by smallholder farmers
and isoacids during anaerobic ruminal fermentation (Bergman, 1990). and an increase in consumption should additionally benefit rural
The benefits of functional food depend on bioavailability and the ad­ livelihoods.
ditive and/or additional interactions with other food bioactive compo­
nents. The in vitro findings in the present study showed that milk extracts CRediT authorship contribution statement
of buffaloes that received green feed enhanced the cytoprotective,
antioxidant, and anti-inflammatory responses in bovine endothelial cells Angela Salzano: Investigation, Methodology, Writing - original

8
A. Salzano et al. Food Chemistry 344 (2021) 128669

draft. Gianluca Neglia: Writing - review & editing. Nunzia D’Onofrio: SIRT1-Mediated Apoptosis in Human Oral Squamous Cell Carcinoma Cal 27.
Cancers, 12(9), 2468. https://doi.org/10.3390/cancers12092468.
Investigation, Methodology, Writing - original draft. Maria Luisa
Flanagan, J. L., Simmons, P. A., Vehige, J., Willcox, M. D., & Garrett, Q. (2010). Role of
Balestrieri: Conceptualization, Project administration, Supervision. carnitine in disease. Nutrition & Metabolism, 16, 7–30.
Antonio Limone: Writing - review & editing. Alessio Cotticelli: Formal Gulati, A., Galvin, N., Lewis, E., Hennessy, D., O’Donovan, M., McManus, J. J.,
analysis. Raffaele Marrone: Formal analysis. Aniello Anastasio: Fenelon, M. A., & Guinee, T. P. (2018). Outdoor grazing of dairy cows on pasture
versus indoor feeding on total mixed ration: Effects on gross composition and
Writing - review & editing. Michael J. D’Occhio: Supervision, Writing - mineral content of milk during lactation. Journal of Dairy Science, 101(3),
original draft. Giuseppe Campanile: Conceptualization, Project 2710–2723. https://doi.org/10.3168/jds.2017-13338.
administration, Supervision. INRA 2007. Alimentation des bovins, ovins et caprins – Besoins des animaux – Valeurs
des aliments – Tables INRA 2007. Editions Quae, Versailles, France. pp. 307.
International Dairy Federation (FIL-IDF) Milk Fat. (1999). Preparation of fatty acid
Declaration of Competing Interest methyl esters. Standard, 182:1999 International Dairy Federation, Brussels, Belgium.
Kolver, E. S. (2003). Nutritional limitations to increased production on pasture-based
systems. Proceedings of the Nutrition Society, 62(2), 291–300. https://doi.org/
The authors declare that they have no known competing financial 10.1079/PNS2002200.
interests or personal relationships that could have appeared to influence Lehnen, T. E., da Silva, M. R., Camacho, A., Marcadenti, A., & Lehnen, A. M. (2015).
A review on effects of conjugated linoleic fatty acid (CLA) upon body composition
the work reported in this paper. and energetic metabolism. Journal of the International Society of Sports Nutrition, 12,
36. https://dx.doi.org/10.1186/s12970-015-0097-4.
Acknowledgements Marín, M. P., Meléndez, P. G., Aranda, P., & Ríos, C. (2018). Conjugated linoleic acid
content and fatty acids profile of milk from grazing dairy cows in southern Chile fed
varying amounts of concentrate. Journal of Applied Animal Research, 46(1), 150–154.
This work was supported by VALERE 2019 Program, University of https://doi.org/10.1080/09712119.2016.1277729.
Campania L. Vanvitelli. The work is also supported by the co-financing Morales-Almaráz, E., Soldado, A., González, A., Martínez-Fernández, A., Domínguez-
Vara, I., de la Roza-Delgado, B., & Vicente, F. (2010). Improving the fatty acid profile
of PON I&C 2014-2020 - CAPSULE [grant number F/200016/01-03/ of dairy cow milk by combining grazing with feeding of total mixed ration. Journal of
X45]. Dairy Research, 77(2), 225–230. https://doi.org/10.1017/S002202991000004X.
O’Callaghan, T. F., Hennessy, D., McAuliffe, S., Kilcawley, K. N., O’Donovan, M.,
Dillon, P., Ross, R. P., & Stanton, C. (2016). Effect of pasture versus indoor feeding
Appendix A. Supplementary data systems on raw milk composition and quality over an entire lactation. Journal of
Dairy Science, 99(12), 9424–9440. https://doi.org/10.3168/jds.2016-10985.
Supplementary data to this article can be found online at https://doi. Pegolo, S., Stocco, G., Mele, M., Schiavon, S., Bittante, G., & Cecchinato, A. (2017).
Factors affecting variations in the detailed fatty acid profile of Mediterranean buffalo
org/10.1016/j.foodchem.2020.128669. milk determined by 2-dimensional gas chromatography. Journal of Dairy Science, 100
(4), 2564–2576. https://doi.org/10.3168/jds.2016-11696.
References Rebouche, C. J., Lehman, L. J., & Olson, L. (1986). ε-N-Trimethyllysine availability
regulates the rate of carnitine biosynthesis in the growing rat. The Journal of
Nutrition, 116, 751–759. https://doi.org/10.1093/jn/116.5.751.
AOAC. (1980). AOAC Official methods of analysis (13th). Washington, DC: Association of
Salzano, A., Licitra, F., D’Onofrio, N., Balestrieri, M. L., Limone, A., Campanile, G.,
Official Analytic Chemists.
D’Occhio, M. J., & Neglia, G. (2019). Short communication: Space allocation in
Bargo, F., Muller, L. D., Delahoy, J. E., & Cassidy, T. W. (2002). Performance of High
intensive Mediterranean buffalo production influences the profile of functional
Producing Dairy Cows with Three Different Feeding Systems Combining Pasture and
biomolecules in milk and dairy products. Journal of Dairy Science, 102(9),
Total Mixed Rations. Journal of Dairy Science, 85(11), 2948–2963. https://doi.org/
7717–7722. https://doi.org/10.3168/jds.2019-16656.
10.3168/jds.S0022-0302(02)74381-6.
Secchiari, P., Antongiovanni, M., Mele, M., Serra, A., Buccioni, A., Ferruzzi, G.,
Basilicata, M. G., Pepe, G., Sommella, E., Ostacolo, C., Manfra, M., Sosto, G., Pagano, G.,
Paoletti, F., & Petacchi, F. (2003). Effect of kind of dietary fat on the quality of milk
Novellino, E., & Campiglia, P. (2018). Peptidome profiles and bioactivity elucidation
fat from Italian Friesian cows. Livestock Production Science, 83(1), 43–52. https://doi.
of buffalo-milk dairy products after gastrointestinal digestion. Food Research
org/10.1016/S0301-6226(03)00043-5.
International, 105, 1003–1010. https://doi.org/10.1016/j.foodres.2017.12.038.
Servillo, L., Giovane, A., Cautela, D., Castaldo, D., Balestrieri, M. L., & Pastore, A. (2014).
Bassett, C.M., Edel, A.L., Patenaude, A.F., McCullough, R.S., Blackwood, D.P., Chouinard,
Where Does Nε-Trimethyllysine for the Carnitine Biosynthesis in Mammals Come
P.Y., Paquin, P., Lamarche. B., & Pierce, G.N. (2010). Dietary vaccenic acid has
from? PLoS ONE, 9(1). https://doi.org/10.1371/journal.pone.0084589.t001.
antiatherogenic effects in LDLr-/- mice. The Journal of Nutrition, 140, 18–24.
e84589.
https://doi.org/10.3945/jn.109.105163.
Servillo, L., D’Onofrio, N., Giovane, A., Casale, R., Cautela, D., Castaldo, D.,
Bauman, D. E., Mather, I. H., Wall, R. J., & Lock, A. L. (2006). Major Advances Associated
Iannaccone, F., Neglia, G., Campanile, G., & Balestrieri, M. L. (2018). Ruminant meat
with the Biosynthesis of Milk. Journal of Dairy Science, 89(4), 1235–1243. https://
and milk contain δ-valerobetaine, another precursor of trimethylamine N-oxide
doi.org/10.3168/jds.S0022-0302(06)72192-0.
(TMAO) like γ-butyrobetaine. Food Chemistry, 260, 193–199. https://doi.org/
Bene, J., Hadzsiev, K., & Melegh, B. (2018). Role of carnitine and its derivatives in the
10.1016/j.foodchem.2018.03.114.
development and management of type 2 diabetes. Nutrition & Diabetes, 8(1). https://
Servillo, L., D’Onofrio, N., Neglia, G., Casale, R., Cautela, D., Marrelli, M., Limone, A.,
doi.org/10.1038/s41387-018-0017-1.
Campanile, G., & Balestrieri, M. L. (2018). Carnitine Precursors and Short-Chain
Bergman, E. N. (1990). Energy contributions of volatile fatty acids from the
Acylcarnitines in Water Buffalo Milk. Journal of Agriculture and Food Chemistry, 66
gastrointestinal tract in various species. Physiological Reviews, 70(2), 567–590.
(30), 8142–8149. https://doi.org/10.1021/acs.jafc.8b02963.
https://doi.org/10.1152/physrev.1990.70.2.567.
Song, J., Wang, Y., Fan, X., Wu, H., Han, J., Yang, M., Lu, L.u., & Nie, G. (2019). Trans-
Campanile, G., De Filippo, C., Di Palo, R., Taccone, W., & Zicarelli, L. (1998). Influence of
vaccenic acid inhibits proliferation and induces apoptosis of human nasopharyngeal
dietary protein on urea levels in blood and milk of buffalo cows. Livestock Production
carcinoma cells via a mitochondrial-mediated apoptosis pathway. Lipids in Health
Science, 55(2), 135–143. https://doi.org/10.1016/S0301-6226(98)00123-7.
and Disease, 18(1). https://doi.org/10.1186/s12944-019-0993-8.
Chaudhry, A. S. (2008). Forage based animal production systems and sustainability, an
Tuomainen, M., Kärkkäinen, O., Leppänen, J., Auriola, S., Lehtonen, M., Savolainen, M.
invited keynote. Revista Brasileira de Zootecnia, 37(spe), 78–84. https://doi.org/
J., Hermansen, K., Risérus, U., Åkesson, B., Thorsdottir, I., Kolehmainen, M.,
10.1590/S1516-35982008001300010.
Uusitupa, M., Poutanen, K., Schwab, U., & Hanhineva, K. (2019). Quantitative
Cruz, L. C. (2007). Trends in buffalo production in Asia. Italian Journal of Animal Science,
assessment of betainized compounds and associations with dietary and metabolic
6(sup2), 9–24. https://doi.org/10.4081/ijas.2007.s2.9.
biomarkers in the randomized study of the healthy Nordic diet (SYSDIET). The
Daley, C. A., Abbott, A., Doyle, P. S., Nader, G. A., & Larson, S. (2010). A review of fatty
American Journal of Clinical Nutrition, 5, 1108–1118. https://doi.org/10.1093/
acid profiles and antioxidant content in grass-fed and grain-fed beef. Nutrition
ajcn/nqz179.
Journal, 9(1). https://doi.org/10.1186/1475-2891-9-10.
Uzun, P., Masucci, F., Serrapica, F., Napolitano, F., Braghieri, A., Romano, R., Manzo, N.,
D’Occhio, M. J., Ghuman, S. S., Neglia, G., della Valle, G., Baruselli, P. S., Zicarelli, L.,
Esposito, G., & Di Francia, A. (2018). The inclusion of fresh forage in the lactating
Visintin, J. A., Sarkar, M., & Campanile, G. (2020). Exogenous and endogenous
buffalo diet affects fatty acid and sensory profile of mozzarella cheese. Journal of
factors in seasonality of reproduction in buffalo: A review. Theriogenology, 150,
Dairy Science, 101(8), 6752–6761. https://doi.org/10.3168/jds.2018-14710.
186–192. https://doi.org/10.1016/j.theriogenology.2020.01.044.
Van Keulen, J., & Young, B.A. (1977) Evaluation of Acid-Insoluble Ash as a Natural
D’Onofrio, N., Balestrieri, A., Neglia, G., Monaco, A., Tatullo, M., Casale, R., Limone, A.,
Marker in Ruminant Digestibility Studies. Journal of Animal Science, 4(2), 282–287.
Balestrieri, M. L., & Campanile, G. (2019). Antioxidant and Anti-Inflammatory
https://doi.org/10.2527/jas1977.442282x.
Activities of Buffalo Milk δ-Valerobetaine. Journal of Agriculture and Food Chemistry,
Wagner, J. J., Lusby, K. S., Oltjen, J. W., Rakestraw, J., Wettemann, R. P., & Walters, L. E.
67(6), 1702–1710. https://doi.org/10.1021/acs.jafc.8b07166.s001.
(1988). Carcass composition in mature Hereford cows: Estimation and effect on daily
D’Onofrio, N., Cacciola, N. A., Martino, E., Borrelli, F., Fiorino, F., Lombardi, A.,
metabolizable energy requirement during winter. Journal of Animal Science, 66,
Neglia, G., Balestrieri, M. L., & Campanile, G. (2020). ROS-Mediated Apoptotic Cell
603–612. https://doi.org/10.2527/jas1988.663603x.
Death of Human Colon Cancer LoVo Cells by Milk δ-Valerobetaine. Scientific Reports,
Wanapat, M., & Kang, S. (2013). World Buffalo Production: Challenges in Meat and Milk
10(1). https://doi.org/10.1038/s41598-020-65865-6.
Production, and Mitigation of Methane Emission. Buffalo Bulletin, 32(1), 1–21. htt
D’Onofrio, N., Mele, L., Martino, E., Salzano, A., Restucci, B., Cautela, D., Tatullo, M.,
p://ibic.lib.ku.ac.th/e-Bulletin/32-sp1.
Balestrieri, M. L., & Campanile, G. (2020). Synergistic Effect of Dietary Betaines on

9
A. Salzano et al. Food Chemistry 344 (2021) 128669

Zebeli, Q., Tafaj, M., Steingass, H., Metzler, B., & Drochner, W. (2006). Effects of acid composition and gene expression of mammary lipogenic enzymes in mid-
Physically Effective Fiber on Digestive Processes and Milk Fat Content in Early lactating dairy cows. Animal Science Journal, 90(2), 214–221. https://doi.org/
Lactating Dairy Cows Fed Total Mixed Rations. Journal of Dairy Science, 89(2), 10.1111/asj.13154.
651–668. https://doi.org/10.3168/jds.S0022-0302(06)72129-4. Zhao, G., He, F., Wu, C., Li, P., Li, N., Deng, J., … Peng, Y. (2018). Betaine in
Zhang, H., Ao, C., Khas-Erdene, & Dan, N.i. (2019). Effects of isonitrogenous and Inflammation: Mechanistic Aspects and Applications. Frontiers in Immunology, 9,
isocaloric total mixed ration composed of forages with different quality on milk fatty 1070. https://doi.org/10.3389/fimmu.2018.01070.

10