Milk glucose-6-phosphate dehydrogenase activity and glucose-6-phosphate are associated with oxidative stress and serve as indicators of energy balance in dairy cows

Abstract

Early lactation in high-producing dairy cows is associated with negative energy balance (EB) and with massive lipolysis to support the energy demands for milk production. The large influx of free fatty acids and glycerol increases oxidative stress in the mammary gland. Milk concentrations of glucose and glucose-6-phosphate (G6P), activity of glucose-6-phosphate dehydrogenase (G6PDH), level of the oxidative stress marker malondialdehyde (MDA), total antioxidant capacity of milk, calculated EB and various parameters reflecting EB (plasma concentrations of nonesterified fatty acids and body condition score), and glucose metabolism (plasma insulin concentration) were measured in 12 high-yielding dairy cows in early lactation (3–57 days in lactation), once a week for 4 weeks. Weekly averages of milk glucose concentration increased from 81 to 184 μM, and of milk G6P decreased from 223 to 81 μM. The activity of milk G6PDH decreased from 902.8 to 256.4 mU ml⁻¹, so that the G6P/glucose ratio in milk decreased from 3.5 to 0.5. A significant correlation between milk G6PDH activity and milk G6P concentration, and an inverse relationship between milk MDA concentration and days in lactation suggest that G6P is shunted to the pentose phosphate pathway in the mammary gland in early lactation, as part of a homeostatic adaptation to counterbalance the excess oxidative stress during early lactation in dairy cows. Milk G6P concentration and G6P/glucose ratio may serve as objective, accurate and noninvasive indicators of EB in dairy cows and potentially in other mammals subjected to negative EB and oxidative stress.

---

Introduction

Modern high yielding dairy cows can produce 60 l of milk per day, and consume up to 5 times the maintenance-energy requirements.¹ The transition from late gestation to lactation presents a great metabolic challenge for the modern dairy cow because it is associated with a sudden shift of energy demand for milk production. The increased energy demand induces fat mobilization, as the cow is not able to consume sufficient nutrients to meet the energy requirements for milk production, resulting in negative energy balance (EB).^2,3 The transition period is also characterized by immune dysfunction,⁴ adverse effects on fertility⁵ and increased incidence of health and metabolic disorders.^6,7 These adverse effects of negative EB during the transition period have been generally related to the burden of oxidative stress, which is correlated to the pro-inflammatory effects of fatty acids released during the massive lipolysis in early lactation.^7,8 There are limited means of detecting the physiological state of postpartum dairy cows, among them calculated EB, blood metabolites (i.e. nonesterified fatty acids, NEFA), and body condition score (BCS; a subjective score of body fattening), all of which require complicated data collection (individual feed intake and body weight), or invasive blood collection, along with trained personnel. Therefore, there is a need for accurate, objective and noninvasive biomarkers that can indicate the physiological status of early postpartum dairy cows.

Glucose is a central metabolite in mammary epithelial cells. In dairy cows, the mammary gland extracts 80–85% of the whole body's glucose production.⁹ Mammary epithelial cells do not synthesize glucose because they lack the enzyme glucose-6-phosphatase.¹⁰ Hence, glucose concentration in mammary epithelial cells depends on the quantity of glucose extracted from the blood. Glucose concentration in milk reflects its concentration in the mammary epithelial cell cytoplasm.^9,11–13 Glucose-6-phosphate (G6P) is a central metabolite in the glycolytic axis because it serves as an intermediate compound during lactose synthesis, and participates in the first step of both glycolysis and the pentose phosphate pathway (PPP).⁹ Previous studies have shown that the concentration of metabolites produced by mammary epithelial cells in the milk of humans^14,15 and ruminants^{11–13,16–18} is closely associated with milk secretion. Those studies highlighted the potential of glucose and citrate concentrations in milk for predicting changes in milk yield. Because the large influx of free fatty acids and glycerol increases oxidative stress in the mammary gland in early lactation, and since the PPP plays an essential role in antioxidant defense by serving as a source of NADPH production,^19,20 in the present work, we hypothesized that oxidative stress in early lactation is associated with the shunting of glucose entering the mammary epithelial cells into the PPP. Therefore, the objectives of this study were: (i) to test the hypothesis that oxidative stress in early lactation is associated with changes in the activity of glucose-6-phosphate dehydrogenase (G6PDH), the first enzyme in the PPP, and with milk G6P concentrations, resulting in the shunting of glucose entering mammary epithelial cells into the PPP; (ii) to examine the relationships between milk concentrations of G6P and glucose, as well as G6PDH activity and oxidative stress markers in milk and dry matter intake (DMI), calculated EB, plasma NEFA and insulin concentrations in early lactation, to determine whether concentrations of G6P and glucose in milk can be used as noninvasive biomarkers of the physiological status of dairy cows during early lactation.

Results

During the study (1–60 days in lactation), average milk yield was 46.0 ± 9.1 kg (Fig. 1A) and average DMI was 23.5 ± 4.8 kg (Fig. 1B). Average milk fat percentage was 3.82 ± 0.61, average milk protein was 3.11 ± 0.32%, and average milk lactose was 5.02 ± 0.20%. The average body weight was 604.8 ± 74.0 kg (Fig. 1C), and average calculated EB was −2.17 ± 7.9 Mcal per day until 60 days in lactation (Fig. 1D). The average body condition score (BCS) in week 1 of lactation was 3.4 ± 0.2, and declined to 2.95 ± 0.2 in week 2 (P < 0.1), to 2.65 ± 0.2 in week 3 (P < 0.01) and to 2.8 ± 0.2 in week 4 of lactation (P < 0.02). These values are typical for modern high-yielding dairy cows.²¹

Fig. 1 Weekly averages of milk yield (A), dry matter intake (B), body weight (C) and calculated energy balance (D) during the study. Data are presented as weekly average ± SEM.

The average concentration of glucose in the milk increased from week 1 to week 3 in lactation (P < 0.01), and then remained constant (∼180 μM) until week 6 in lactation (Fig. 2A). Milk glucose was positively correlated to days in lactation (P < 0.0001; Table 1). The free glucose levels in milk in the present study were lower than those reported elsewhere for cows (331 μM;²² 220 μM;²³ 254 μM¹⁸). This can be explained by the early stage of lactation and by the high milk yield in the present study.

Fig. 2 Milk concentration of glucose and glucose-6-phosphate (G6P)/glucose ratio (A), and milk concentration of G6P and G6P-dehydrogenase (G6PDH) activity (B) in early lactation cows. Milk samples were collected once a week during the noon milking for a period of 4 weeks. Data are presented as weekly average ± SEM.

Table 1 Interrelationships among key measures reflecting response to oxidative stress, dynamic changes of those measures in early lactation, and their relationship to energy balance and the indices that reflect it [N (cows) = 12, n (individual measures) = 48]

Y	Range	X	Range	Type^a	r =	P <
a E, exponential interrelationship, y values transformed to their natural log form; L, linear interrelationship; MDA, malondialdehyde; ORAC, oxygen radical antioxidant capacity.
Milk G6P (μM)
Milk G6P	51.6–338.7	Days in lactation	3–57	E	−0.50	0.0003
Milk G6P	51.6–338.7	Milk ORAC (μM)	418.8–901.6	L	0.25	0.1
Milk G6P	51.6–338.7	Dry matter intake (kg per day)	4.8–35.9	L	−0.51	0.01
Milk G6P	51.6–338.7	Milk G6PDH activity (mU ml^−1)	38.8–4250.4	E	0.68	0.0001
Milk G6P	51.6–338.7	Plasma NEFA (μeq. l^−1)	190.2–1489.1	L	0.60	0.02
Milk G6P	51.6–338.7	EB (Mcal per day)	−30.3–20.7	L	−0.45	0.02
Milk G6P/glucose
Milk G6P/glucose	0.3–8.9	Dry matter intake (kg per day)	4.8–35.9	L	−0.65	0.0007
Milk G6P/glucose	0.3–8.9	Milk G6PDH activity (mU ml^−1)	38.8–4250.4	E	0.55	0.0001
Milk G6P/glucose	0.3–8.9	Days in lactation	3–57	E	−0.69	0.0001
Milk G6P/glucose	0.3–8.9	Plasma NEFA (μeq. l^−1)	190.2–1489.1	L	0.81	0.0003
Milk G6P/glucose	0.3–8.9	EB (Mcal per day)	−30.3–20.7	L	−0.52	0.0045
Plasma insulin (pg ml^−1)	162.7–2416.5	Milk G6P/glucose	0.3–8.9	L	−0.68	0.02
Milk G6PDH activity (mU ml^−1)
Milk G6PDH activity	38.8–4250.4	Dry matter intake (kg per day)	4.8–35.9	E	−0.57	0.0003
Milk G6PDH activity	38.8–4250.4	EB (Mcal per day)	−30.3–20.7	E	−0.50	0.001
Milk G6PDH activity	38.8–4250.4	Milk G6P/glucose	0.3–8.9	L	0.53	0.0003
Milk G6PDH activity	38.8–4250.4	Days in lactation	3–57	E	−0.37	0.01
Additional oxidative stress and metabolic status indicators
Milk MDA^a (nM)	101.3–948.5	Days in lactation	3–57	E	−0.36	0.01
Milk glucose (μM)	16.6–439.4	Days lactation	3–57	E	0.56	0.0001
Milk ORAC^a (μM)	418.8–901.6	Days in lactation	3–57	E	−0.29	0.06
Milk ORAC (μM)	418.8–901.6	EB (Mcal per day)	−30.3–20.7	E	−0.30	0.07
Plasma insulin (pg ml^−1)	162.7–2416.5	Milk glucose (μM)	16.6–439.4	L	0.49	0.06

---

In contrast, the average concentration of G6P in the milk (Fig. 2B) was highest (223.2 ± 29.2 μM) for week 1 in lactation, remained high in week 2 and then decreased in week 3 (P < 0.02), and continued to decrease until week 5 in lactation (81.0 ± 23.2 μM; P < 0.01). As a result, milk G6P concentration was negatively correlated to days in lactation (Table 1). Consistent with our results, a recent study using a similar enzymatic–fluorimetric method to determine milk G6P in over 3200 milk samples from cows on a commercial farm reported an average mean G6P of 81 μM (during weeks 1–70 in lactation²²). In addition, a similar decrease in G6P and increase in glucose concentrations in milk in early lactation (weeks 1–3) compared to later stages of lactation were also reported.²² However, that report lacked data on cows' DMI, blood parameters and EB. Nevertheless, the agreement of our findings with this previous report strengthens the validity of the former and adds important information on the relationship between these milk metabolites and indicators of metabolic status, as the present study is the first to examine the association between milk G6P and DMI, EB, and plasma NEFA and insulin concentrations.

The G6P/glucose ratio (Fig. 2A) was highest at week 1 in lactation (3.5 ± 0.5, P < 0.02) and reached its lowest levels at week 5 in lactation (0.5 ± 0.4, P < 0.04). The G6P/glucose ratio was exponentially negatively correlated with days in lactation (Table 1). To investigate the possible mechanism underlying this phenomenon, we examined the activity of milk G6PDH, which is the first enzyme in the PPP. Indeed, the activity of G6PDH was highest in weeks 1–2 of lactation, and then decreased, similar to milk G6P concentration, until week 5 in lactation (Fig. 2B). Moreover, G6PDH activity was linearly correlated with milk G6P (P < 0.0001) and with G6P/glucose ratio (Table 1). In addition, milk G6PDH activity was negatively correlated to days in lactation, DMI and EB (Table 1).

Milk MDA concentration was highest in early lactation and then exponentially decayed (Table 1). It was inversely correlated with days in lactation (Table 1), which is consistent with the pattern and levels of milk MDA in early lactation.²⁴ Milk antioxidative capacity (ORAC values, Table 1) tended to be negatively and exponentially correlated with days in lactation and EB (Table 1), and to be positively and linearly correlated to milk G6P (Table 1).

The concentrations of milk G6P were positively and linearly correlated to plasma NEFA concentrations (Table 1), and the G6P/glucose ratio was highly linearly correlated to plasma NEFA concentrations (Table 1). These correlations are in accordance with that between G6P and calculated EB, as NEFA concentrations reflect body reserve mobilization in early lactation. Moreover, the positive correlation between plasma NEFA and milk G6P is in accordance with that previously reported between milk beta-hydroxyl butyrate and G6P.²²

A negative correlation was found between milk G6P and DMI (Table 1). Similarly, the G6P/glucose ratio was closely correlated to DMI. In addition, a negative linear correlation between G6P in milk and EB was found, as well as between the G6P/glucose ratio and EB (Table 1). The G6P/glucose ratio was negatively correlated to plasma insulin concentration (Table 1). As the concentration of insulin in the blood increases with increasing DMI, the negative correlation between plasma insulin and milk G6P/glucose ratio is consistent with the negative relationships between G6P and G6P/glucose and DMI and EB.

Discussion

In general, milk production is associated with oxidative stress.²⁵ However, the results of this study support our core hypothesis that extra oxidative stress in the mammary gland of early lactating cows, which is a consequence of extensive fat mobilization, diverts glucose entering into mammary epithelial cells toward the PPP. Furthermore, we highlight the potential of using milk G6P concentration and G6PDH activity as noninvasive biomarkers of metabolic status and EB in early lactating cows.

Only a few studies have reported on G6PDH activity in milk. Among them, Mellenberger and Bauman²⁶ found that milk G6PDH activity increases from mid-pregnancy to lactation in rabbits. In humans, G6PDH is synthesized in the mammary gland in response to hormonal stimuli during pregnancy and lactation.²⁷ Grigor and Hartmann²⁸ examined G6PDH activity in the milk of sows, rats and rabbits, and concluded that milk G6PDH activity reflects that found in the mammary gland, which is consistent with our hypothesis. Based on our findings, we hypothesize that elevated G6PDH activity in early lactation reflects increased shunting of G6P to the PPP. Because G6P is derived from glucose, the only possible explanations for a G6P/glucose ratio > 1 in milk at the beginning of lactation (up to week 2, Fig. 2A) are: (i) inhibition of its passage through the downstream stages of glycolysis in the mammary epithelial cells, or (ii) recycling of fructose-6-phosphate formed in the PPP to G6P.²⁰ The increase of G6P in the mammary epithelial cell cytosol likely brings it closer to the optimal Km of G6PDH. According to this explanation, the enhanced diversion of G6P through the PPP and conversely, inhibition of its passage through the glycolytic pathway gradually fade, so that the G6P/glucose ratio in milk decreases on average from ∼3.5 immediately postpartum to ∼0.5 at established lactation (once a positive EB is obtained; Fig. 3). The low SCC in the cows used in the present experiment strongly suggests that the milk MDA and ORAC levels reflected cellular events and not post-secretion events due to bacterial infection. It is consistent with previous reports associating milk synthesis with fat and protein oxidation due to oxidative stress by free radicals.^24,29,30 The shunting of G6P to PPP and resulting formation of NADPH balanced, to some extent, the overloading of oxidative stress, as reflected in the positive relationship between G6P and ORAC levels. However, it could not prevent the increase in MDA formation because radical formation is most likely the cause of this response (Fig. 3) and because radical activity is very rapid (a few split seconds²⁵).

Fig. 3 Schematic model to explain the changes in milk G6P/glucose ratio in early lactation.

We suggest a biochemical–physiological model to explain the results of the present study (Fig. 3). The high G6P/glucose ratio in milk at the beginning of lactation may reflect the need to shunt G6P through the PPP to counterbalance the oxidative stress in the mammary epithelial cells (Fig. 3). Consistent with our model, it is known that PPP regulation during the oxidative stress response is a conserved paradigm from yeast to mammals of the cellular antioxidant defense mechanism via reduction of NADP⁺ to NADPH.²⁰ The fastest response (on the order of seconds) is made possible through oxidative inhibition of glycolytic enzymes supported by post-translational modifications (on the order of minutes) that increase G6PDH activity.²⁰ Thus the increase in G6PDH activity and G6P/glucose ratio in early lactation in association with increased oxidative stress are consistent with those basic homeostatic responses.

Early lactation is associated with negative energy balance and extensive lipolysis, which in turn induces oxidative stress in central and peripheral tissues. The mammary gland is the most active peripheral tissue and is subjected to oxidative stress leading to increased concentration of H₂O₂ in its epithelial cells.⁷ When it interacts with minerals such as Fe and Cu, H₂O₂ tends to react with peroxides and produce free radicals that may reach toxic levels in the cells. Glutathione is an important antioxidant that prevents damage to important cellular components caused by reactive oxygen species such as free radicals, peroxides, lipid peroxides and heavy metals.¹⁹ Conversion of oxidized glutathione to reduced glutathione by glutathione reductase is the central reaction for degradation of H₂O₂.²⁰ This reaction requires NADPH as a cofactor, increasing the cell's requirement for NADPH during oxidative stress.²⁰ Our and Bouwstra et al.²⁴ finding of a negative correlation between milk MDA and days in lactation supports this model. Two molecules of NADPH are produced per molecule of G6P shunted to the PPP. Under non-oxidative stress, most of the G6P is metabolized via the glycolytic pathway, which most likely reflects the lower Km of the immediate downstream enzymes of the glycolytic axis rather than the Km of G6P dehydrogenase, which is responsible for diverting G6P to the PPP. Thus, diversion of G6P to the PPP can be enhanced by inhibiting the downstream flow of G6P along the glycolytic axis, and its accumulation then allows its diversion into the PPP. The mechanism of inhibition of downstream glycolytic enzyme requires further study: it may be a direct effect of H₂O₂,²⁰ or mediated by an intracellular signal that is sensitive to H₂O₂ concentration. The gradual conversion from negative energy balance to positive energy balance progressively increases the flow of G6P through the glycolytic axis.

Experimental procedures

Animals

The experimental procedures of this study were approved by the Volcani Center Animal Care Committee, and included 12 multiparous Israeli Holstein cows at the Volcani Center experimental farm (Bet Dagan, Israel). For training and adjustment, cows were group-housed at 256 days of pregnancy in a shaded outdoor pen with adjacent outside yard, which was equipped with a real-time electronic individual feeding system. Each station was equipped with an individual identification system, which allowed recording individual food intake (SAE Afikim, Kibbutz Afikim, Israel). Cows were fed standard diets both prepartum (from 256 days of pregnancy until parturition) and postpartum (1–60 days in lactation) according to NRC (1989) recommendations. The pre- and postpartum diets contained 1.49 and 1.78 Mcal kg⁻¹ dry matter net energy for lactation, and 13.3 and 16.5% crude protein, respectively. Postpartum, cows were milked 3 times a day and body weight was measured electronically in the milking parlor (SAE Afikim). Milk yields and milk solids (protein, fat and lactose contents) and somatic cell count (SCC) were measured electronically from the Afilab system (SAE Afikim). The cows used in this study had SCC < 200 000 (an indication that the cows were free of bacterial infections in the udder) at the beginning of the study and throughout the experiment. Body condition score (BCS) was determined once a week on a scale of 1 to 5 by a trained technician.

EB calculation

EB was calculated according to NRC (2001)³¹ as follows:

NE_C = NE_L/kg of DM × DMI;

NE_M = BW^0.75 × 0.08 × 1.1;

NE_P = milk, kg × [(0.0929 × fat, %) + (0.0547 × protein, %) + (0.0395 × lactose, %)];

EB = NE_C − (NE_M + NE_P)

where NE_C = net energy consumed, NE_L = net energy for lactation, DM = dry matter, DMI = dry matter intake, BW^0.75 = body weight, NE_P = net energy output in milk, and NE_M = net energy for maintenance.

Blood sampling and analysis of plasma NEFA and insulin concentrations

Blood samples were collected thrice weekly at 0700 h, after the morning milking, from parturition until 30 days in lactation from the jugular vein into evacuated tubes containing lithium heparin (Becton Dickinson System, Cowley, UK). Blood samples were centrifuged at 3000g for 20 min and stored at −25 °C for subsequent analysis. Plasma NEFA concentration was determined by Wako NEFA C test kit (Wako Chemicals GmbH, Neuss, Germany). Plasma insulin concentration was determined by radioimmunoassay (MP Biomedicals, Solon, OH, USA), with intra- and interassay coefficients of variation of 7.2 and 5.1%, respectively.

Milk sampling and analysis

Milk samples were collected once a week during the noon milking (at 1300 h) for a period of 4 weeks. At each sampling time (ranging from 3–57 days in lactation), milk was collected from all cows participating in the study, for a total 48 milk samples collected from 12 cows, and analyzed. Milk samples were centrifuged at 3000 g for 20 min at 4 °C to remove the fat layer, and the skim milk was stored at −25 °C for subsequent analysis. Milk glucose and G6P contents were measured by fluorimetric assay applying enzymatic reactions as previously described.¹⁸ The activity of G6PDH in milk was analyzed by a modification of a classical enzymatic assay procedure³² in which the reduction of NADP⁺ to NADPH is coupled to form a fluorimetric chromophore.¹⁸ The concentration of NADPH at linear stages of the reaction (2 and 7 min after the start of the reaction in the present case) was calculated from standard containing variable levels (between 10 and 1000 μM) of NADPH, which was converted to NADP⁺ by diaphorase and coupled to conversion of resazurin to highly fluorimetric resorufin (r² = 0.988). The differences in NADPH concentration were divided by time and were converted to activity, where 1 U of G6PDH activity will oxidize 1.0 μmol of D-G6P to 6-phospho-D-gluconate per minute in the presence of NADP⁺ at pH 7.4, 25 °C. Milk malondialdehyde (MDA) concentration was measured according to the fluorimetric thiobarbituric acid reactive substances assay.³³ The oxygen radical antioxidant capacity (ORAC) in milk serum was analyzed by fluorimetric procedure.³⁴

Statistical analysis

The average DMI, milk yield, and milk solids were calculated from the daily data until 60 days in lactation. The average concentrations of glucose and G6P, the G6P/glucose ratio, averages of G6PDH activity and MDA concentrations in milk were calculated on a weekly basis. Weekly averages of BCS, milk glucose, G6P and the G6P/glucose ratio, MDA, ORAC and G6PDH activity were analyzed with PROC GLM of SAS (version 9.2, 2002). Least square means ± SE are presented. Linear or exponential correlation analysis, according to the best fit, was performed using the PROC REG procedure of SAS (version 9.2, 2002). For exponential correlations, the independent variables were transformed to their log-natural form (ln(x)). P < 0.05 was accepted as significant, and P < 0.1 was accepted as a tendency.

Conclusions

In conclusion, the significant correlation between milk G6PDH activity and G6P content, and the inverse relationship between MDA concentration and DIM, support our proposed biochemical model (Fig. 3). They suggest that G6P is shunted to the PPP in the mammary gland in early lactation as part of homeostatic adaptations to negative EB at the onset of lactation, to counterbalance oxidative stress. Our data highlight the potential of using milk G6P concentration and the G6P/glucose ratio as an objective, accurate and noninvasive technique to evaluate metabolic status and EB in early lactation. These parameters may have an advantage over BCS, calculated EB and plasma NEFA as indicators of the physiological state of postpartum cows. Further studies with a larger number of animals and more frequent milk sampling will most likely improve the predictive power in those interrelationships. Our results may also help identify situations associated with acute oxidative stress and negative EB in lactating women and other mammals.

Acknowledgements

This research was financially supported by the Israeli Dairy Board, Yehud, Israel (362-0496). The authors would like to thank the staff of the Volcani dairy farm for their assistance in handling the animals.

References

1. C. T. Kadzere, M. R. Murphy, N. Silanikove and E. Maltz, Livest. Prod. Sci., 2002, 77, 59–91.
2. A. W. Bell and D. E. Bauman, J. Mammary Gland Biol. Neoplasia, 1997, 2, 265–278.
3. J. K. Drackley, J. Dairy Sci., 1999, 82, 2259–2273.
4. B. A. Mallard, J. C. Dekkers, M. J. Ireland, K. E. Leslie, S. Sharif, C. Lacey Vankampen, L. Wagter and B. N. Wilkie, J. Dairy Sci., 1998, 81, 585–595.
5. W. R. Butler and R. D. Smith, J. Dairy Sci., 1989, 723, 767–783.
6. J. P. Goff and R. L. Horst, J. Dairy Sci., 1997, 80, 1260–1268.
7. B. J. Bradford, K. Yuan, J. K. Farney, L. K. Mamedova and A. J. Carpenter, J. Dairy Sci., 2015, 98, 6631–6650.
8. L. M. Sordillo, G. A. Contreras and S. L. Aitken, Anim. Health Res. Rev., 2009, 10, 53–63.
9. F. Zhao, J. Mammary Gland Biol. Neoplasia, 2014, 19, 3–17.
10. R. A. Scott, D. E. Bauman and J. H. Clark, J. Dairy Sci., 1976, 59, 50–56.
11. A. Faulkner, N. Chaitabutr, M. Peaker, D. T. Carrick and N. J. Kuhn, J. Dairy Res., 1981, 48, 51–56.
12. A. Faulkner and M. Peaker, J. Dairy Res., 1982, 49, 159–169.
13. M. Peaker and A. Faulkner, Proc. Nutr. Soc., 1983, 42, 419–425.
14. P. G. Arthur, J. C. Kent and P. E. Hartmann, J. Pediatr. Gastroenterol. Nutr., 1987, 19, 100–108.
15. M. C. Neville, W. W. Hay and P. Fennessy, Protoplasma, 1990, 159, 118–128.
16. N. Silanikove, A. Rauch-Cohen, F. Shapiro, S. Blum, A. Arieli and G. Leitner, J. Dairy Sci., 2011, 94, 4468–4475.
17. N. Silanikove, U. Merin, F. Shapiro and G. Leitner, J. Dairy Sci., 2013, 96, 6400–6411.
18. N. Silanikove, U. Merin, F. Shapiro and G. Leitner, J. Dairy Res., 2014, 81, 358–363.
19. A. Pompella, A. Visvikis, A. Paolicchi, V. Tata and A. F. Casini, Biochem. Pharmacol., 2003, 66, 1499–1503.
20. A. Stincone, A. Prigione, T. Cramer, M. M. C. Wamelink, K. Campbell, E. Cheung, V. Olin-Sandoval, N. M. Gruning, A. Kruger and M. T. Alam, Biol. Rev., 2015, 90, 927–963.
21. M. Zachut, A. Arieli, H. Lehrer, L. Livshitz, S. Yakoby and U. Moallem, J. Dairy Sci., 2010, 93, 5877–5889.
22. T. Larsen and K. M. Moyes, Animal, 2015, 9, 86–93.
23. R. J. Marschke and B. J. Kitchen, J. Dairy Res., 1984, 51, 233–237.
24. R. J. Bouwstra, R. M. A. Goselink, P. Dobbelaar, M. Nielen, J. R. Newbold and T. van Werven, J. Dairy Sci., 2008, 91, 977–987.
25. N. Silanikove, U. Merin and G. Leitner, RSC Adv., 2014, 4, 26476–26486.
26. R. W. Mellenberger and D. E. Bauman, Biochem. J., 1974, 138, 373–379.
27. K. M. Shahani, A. J. Kwan and B. A. Friend, Am. J. Clin. Nutr., 1980, 33, 1861–1868.
28. M. R. Grigor and P. E. Hartmann, J. Dairy Res., 1985, 52, 501–506.
29. N. Silanikove, F. Shapiro, M. Silanikove, U. Merin and G. Leitner, J. Agric. Chem. Food Sci., 2009, 57, 8018–8025.
30. N. Silanikove, U. Merin, F. Shapiro and G. Leitner, J. Dairy Sci., 2014, 97, 3449–3455.
31. NRC, Natl. Acad. Sci., Washington, DC, 7th rev. edn, 2001.
32. K. A. Gumaa, F. Novello and P. McLean, Biochem. J., 1969, 114, 253–263.
33. E. Feldman, Animal models of diabetic complications consortium (AMDCC protocols), 2004, version 1: 1–3, http://www.amdcc.org/shared/showFile.aspx?doctypeid=3%26docid=33.
34. A. Zulueta, M. J. Esteve and A. Frígola, Food Chem., 2009, 114, 310–316.

---