Comparison of different organic selenium supplementations on selenium status and serum biomarkers in dairy cows

1 INTRODUCTION

Selenium (Se) plays an important role in human nutrition, not only via direct influence on the maintenance of immune and hormonal systems, protection against heavy metals and xenobiotics, and prevention of chronic diseases, but also in the key balance of benefits and competition between host and gut microbial flora (Ferreira et al., 2021). For Chinese residents, a reduction in grain consumption was correlated with a decrease in hair Se measurements (Li et al., 2014), a factor that may have contributed to adverse health impacts in Se-deficient areas. Moreover, a previous study demonstrated that 39–61% of Chinese residents consumed less Se than the WHO/FAO recommended level (26–34 μg/d) (Dinh et al., 2018). Therefore, as suggested by Givens et al. (2004), increasing Se-enriched milk consumption could relieve the Se deficiency in China.

A meta-analysis of Se intake by dairy cows showed that oral Se supplementation resulted in an increase in milk Se level by 9.24–16.34 μg/l (95% confidence interval) (Ceballos et al., 2009). The source and dosage of Se and their interaction explained approximately 70% of the variance. Compared with inorganic Se, organic Se was more efficient in increasing milk Se content using either selenized yeast or Se-fertilized forage (Séboussi et al., 2016). As the most common organic Se supplementation for dairy cattle, selenized yeast contains approximately 60–70% selenomethionine (Se-Met) (Calamari et al., 2010). A novel selenoamino acid, selenohomolanthionine (Se-Hlan), has been identified in Japanese pungent radishes (Ogra et al., 2007). Se-Met is the Se analog of methionine, and as such, it is similar to methionine in chemical structure and metabolic pathways. Se-Hlan is the common name of 4,4′-selenobis [2-aminobutanoic acid], and it is utilized only in the trans-selenation pathway for selenoprotein synthesis (Tsuji et al., 2010). Se-Hlan has been used in broilers to evaluate growth performance, nutrient utilization, oxidative stress, and Se tissue concentrations, and comparisons between Se-Met and Se-Hlan have also been made (Celi et al., 2013). However, to date, there are no comparable data available for the two organic Se supplements in dairy animals.

We hypothesized that the addition of Se-Hlan in dairy cattle feed would be as effective as Se-Met as a substitute for inorganic Se. This study aimed to evaluate the effects of different organic Se supplements, Se-Hlan and Se-Met, on milk production and composition, serum biochemical parameters, antioxidant indicators, immune parameters, and the Se concentration in milk and blood of dairy cows. According to the specification of feed additives issued by the Ministry of Agriculture of China (Ministry of Agriculture of China, 2009), the recommended and maximum additive amounts for Se are 0.1–0.3 and 0.5 mg/kg in total mixed rations for ruminants, respectively. To evaluate the potential milk performance and health issues affected by Se intake, low (0.3 mg/kg) and high (0.5 mg/kg) levels of Se were set in each organic Se treatment.

2 MATERIALS AND METHODS

3 RESULTS

3.3 Total se concentration in milk and blood samples

The effect of Se supplementation on milk Se is shown in Figure 1. Extreme impacts (P < 0.001) of Se supplementation, week, and Se supplementation × week interaction on milk Se level were shown in the statistical model (data not shown). During the experimental period, the highest milk Se values appeared in the 0.5 Se-Met group, and the lowest milk Se presented in CON and 0.3 Se-Hlan, with intermediate levels in the 0.5 Se-Hlan and 0.3 Se-Met groups. Total Se concentrations in blood and serum are summarized in Figure 2. In blood (Figure 2A), similar total Se values appeared in Se-Met and Se-Hlan supplementation groups, higher than that in CON (P < 0.01). Meanwhile, the highest serum Se appeared in the 0.5 Se-Met group, and the lowest value was in the 0.3 Se-Hlan group, with the other three groups having no significant difference (Figure 2B).

3.4 PCA results

The PCA plots shown in Figure 3 illustrate the variance of measurements affected by Se supplementation (P < 0.05). The first two PCs explained 43% of the total variance. All the samples tended to cluster into three groups (Figure 3A): PC1 reflected the difference between Se-Met and Se-Hlan samples, while PC2 separated CON samples from organic Se samples. The corresponding loading plot lists the explained measurements (Figure 3B). AST, ALT, LDL-C, LDH, and GLU in serum were negatively correlated with PC1 and positively correlated with PC2, while urea, UA, TG in serum, and Se in milk and blood were negatively correlated with PC1 and PC2.

4 DISCUSSION

The influence of Se supplementation in dairy cow diets on milk composition and production in the present study was in line with the results from previous studies (Gong et al., 2014; Juniper et al., 2006), which found that neither organic nor inorganic Se supplementation affected milk composition or production. However, in other reports, fat (Sun et al., 2017), protein (Séboussi et al., 2016), and milk yields were affected by Se sources and dosages.

Most reports do not show a remarkable influence of Se supplementation on the blood biochemical parameters of dairy cows (Phipps et al., 2008), while some researchers have found that the source and amount of Se intake affected plasma metabolites such as BHBA, NEFA, and Na⁺ (Calamari et al., 2010). As indicators of liver function, lower ALT and AST than CON as well as lower AST than Se-Met supplementation appeared in Se-Hlan groups, a result that may reflect a potentially lower metabolic burden of Se-Hlan for the liver (Shaniko et al., 2015; Tsuji et al., 2010). In our study, more GLU and urea appeared in the Se-Met groups than in the Se-Hlan groups. Differences in serum urea among treatments in our study agree with the blood urea nitrogen (BUN) comparison of rats sampled at 1 h when the animals had received an injection of Se-HLan or Se-Met at 1.0 mg Se/kg body weight (Tsuji et al., 2010). The elevated GLU and urea levels for Se-Met intake seem to result from high basic metabolic functioning of the pancreas and liver, as explained by Shi et al. (2017). Although the addition of organic Se (5 g/head/day) for Tabapuã steers reduced UA excretion in urine samples (de Paiva Ferreira et al., 2020), the blood UA of dairy cows was not affected by Se source or dosage (Sun et al., 2019, 2021). Different from the above studies, we found that intake of selenized yeast by cows resulted in higher serum UA than with inorganic Se. Moreover, serum UA tended to decrease with increasing selenized yeast intake in our study. The increasing CREA trend with organic Se level also appeared in our results, and similar CREA levels were presented in all Se treatments, a result that was comparable with a study on mice using different Se sources injected at 0.1 and 1 mg Se/kg body weight. However, an overdose (10 mg Se/kg body weight) of Se-Met caused more serum CREA than that of Se-Hlan (Anan & Ogra, 2013). Considering the comparison of TG between treatments, Se-Met groups had similar TG values with Se-Hlan groups but had higher levels than CON, in accordance with observations in lactating ewes. The addition of inorganic Se for these animals caused lower blood TG than organic Se (Novoselec et al., 2018). Although not reported by other researchers, higher serum LDL-C in Se-Met groups than in Se-Hlan groups occurred in our study. High LDL-C may be associated with high serum Se levels because Se is bound to LDL-C in the blood (Schrauzer, 2000). Regarding LDH, serum measurements of mice injected with organic Se at 1 h observation showed that higher LDH values were presented in the Se-Met group compared to the Se-Hlan group (Tsuji et al., 2010), in line with our observations.

The influence of Se addition for dairy cows on plasma or serum antioxidant indexes was not consistent in previous reports. Some researchers have found that organic Se groups had higher GSH-Px and T-AOC as well as lower MDA compared with inorganic Se groups (Gong et al., 2014; Sun et al., 2017), while other researchers have observed no difference in these indexes between Se treatments (Séboussi et al., 2016). The addition of Se-Hlan in diets for dairy cows showed an MDA drop and a T-AOC increase in one farm observation (Li et al., 2020). In our study, serum GSH-Px and LPO levels showed a trend toward an effect affected by Se-Met and Se-Hlan supplementation. This was not in agreement with a report on broiler chickens that showed similar plasma GSH-Px levels between the two corresponding groups (Celi et al., 2013). As indicators of immune status, antibody IgG, the pro-inflammatory cytokines IL-6, TNF-α, and anti-inflammatory cytokines IL-2 in serum were not influenced by Se treatments in our study, similar to previous reports (Gong et al., 2014).

In our experiment, milk Se level surged rapidly after two weeks of Se-Met supplementation and increased slowly from week 3 to week 8 with some fluctuation, while inorganic Se treatment in CON showed little variation. This was similar to a previous report where the duration of transition for stable milk Se level was 18 days on average after dietary Se supplementation (Calamari et al., 2010). The stable milk Se concentration ranged from 37–53.6, 47–59.7, and 19.7–25.4 μg/kg from week 3 to week 8 for 0.3 Se-Met, 0.5 Se-Met, and CON, respectively, also corresponding to the results of the above research. In our results, the milk Se level for 0.3 Se-Hlan was higher than or similar to that of CON, and the milk Se level for 0.5 Se-Hlan was lower than or similar to that for 0.3 Se-Met. There are two potential reasons for the variation in milk Se among treatments. In ruminants, the absorption and digestibility of organic Se are higher than that of inorganic Se (Shaniko et al., 2015). From the aspect of metabolic pathways, Se-Met can be incorporated into protein synthesis as methionine or converted to selenohomocysteine via biosynthesis of Se-adenosylselenomethionine and then metabolized in a similar trans-selenation pathway as with Se-Hlan (Shaniko et al., 2015; Tsuji et al., 2010). Inorganic Se can be reduced and converted to selenoprotein (Shaniko et al., 2015). Therefore, the intake of Se-Met would lead to the enhancement of protein synthesis and an increase in the levels of selenoprotein, while the intake of Se-Hlan and inorganic Se would only lead to an increase in the levels of selenoprotein. Compared to inorganic Se, intake of organic Se (containing 63% Se-Met) increased the levels of Se-Met and selenocysteine (Se-Cys) in milk to a greater degree, with the Se level in the form of Se-Met being 5–6 times higher than that in the form of Se-Cys (Calamari et al., 2010). Based on our results, 0.5 Se-Met was more effective for the production of Se-enriched milk compared to other treatments.

In a comparison of dairy cows fed varied Se sources, organic Se from yeast or silage increased blood Se more than inorganic Se; different organic Se sources induced similar blood Se levels, while Se sources significantly affected serum Se levels (Séboussi et al., 2016). This trend was in agreement with our study, and our values for blood and serum Se were close to previously reported data (Li et al., 2020; Séboussi et al., 2016). However, there was no difference in blood Se levels between different level organic Se groups in this study, a result that was not in line with previous reports (Calamari et al., 2010). Serum Se was more sensitive to dietary Se supplementation than blood Se, because most of the blood Se is incorporated into red cells during erythropoiesis, and the lifespan of red blood cells for cows is 90–120 days (Calamari et al., 2010). In our observations, serum Se for cows was likely to be affected by the amount of organic Se intake, since the high Se treatment induced higher serum Se than the low Se treatment (Figure 2B). Unlike Se-Hlan, Se-Met from dietary sources is directly involved in serum albumin synthesis as methionine (Cheng et al., 2019), which may explain the higher serum Se level in the Se-Met groups compared to the Se-Hlan groups. Similar serum Se levels in CON and 0.3 Se-Met were presented in our study (Figure 2B), and this was in contrast to previous research (Séboussi et al., 2016). In addition, the lower serum Se in 0.3 Se-Hlan than in 0.3 Se-Met and CON may indicate a lesser impact of Se-Hlan intake on the serum Se fraction. The quantitative differences in serum Se caused by Se sources should be confirmed in further research. Regarding the difference in milk Se and serum Se in response to Se supplementation, milk Se was more sensitive to Se-Met intake for dairy cows (Figures 1, 2B, and 3). This may be associated with the higher Se proportion of Se-Met in milk (about 40–60%) than in serum (about 10–30%) (Awadeh et al., 1998; Calamari et al., 2010). The primary serum Se was incorporated into GSH and selenoprotein P, constant with the response to Se sources (Burk et al., 2001).

Overall, for the efficiency of feed Se sources, Se yeast was more efficient for increasing cows' milk and serum or plasma Se levels than inorganic Se (Youcef & Isabelle, 2016). For the comparison of different types of Se yeast, especially Se-Met vs. Se-Hlan, research on broilers (Celi et al., 2013) and mice (Anan & Ogra, 2013; Tsuji et al., 2010) has been performed, but there have been no previous studies on dairy cows. In our observations, Se-Met resulted in higher milk and serum Se levels than Se-Hlan. This was not in agreement with observations in mice made one hour after administration by bolus of a dose of 1 mg Se/kg body weight, where serum and blood Se levels using Se-Hlan were higher than with Se-Met (Tsuji et al., 2010). Regarding the distribution preference, Se-Met is more efficiently accumulated in the liver and pancreas than Se-Hlan (Anan & Ogra, 2013). A large dose (1 mg Se/kg body weight) bolus administration of Se-Met in mice caused higher levels of LDH, AST, and BUN as well as higher levels of amylase at 1 h compared to controls (without any treatments), while Se-Hlan did not have such effects, suggesting that a large dose of Se-Met induced liver and pancreatic damage (Tsuji et al., 2010). The accumulation of Se-Met for 8 weeks in these organs may affect basic metabolism, and this would cause higher AST, GLU, urea, LDL-C, and LDH than the values presented in the Se-Hlan groups (Table 3 and Figure 3).

In conclusion, our observations suggested that milk and serum Se levels were not affected equally by Se-Met and Se-Hlan. Compared with Se-Hlan, Se-Met was more effective in improving dairy cow serum or milk Se, and it is more likely to be used in Se-enriched milk production. The higher serum AST, GLU, urea, LDL, and LDH observed in Se-Met compared to Se-Hlan seemed to be correlated with a higher hepatic burden for Se-Met metabolization. The exact influence of Se-Met and Se-Hlan on cattle health needs further study.

CONFLICT OF INTEREST STATEMENT

We confirm that there is no conflict of interest with any financial organization in this paper.