Retinol and mineral status in grazing foals during the dry season

The objective of this study was to examine serum retinol and some mineral (zinc, iron, calcium, phosphorus, magnesium, and copper) levels in seven Mangalarga Marchador colts aged 329.48 days, reared on pasture in the city of Montes Claros, MG, Brazil, during the dry season. Equines were evaluated for four periods of 45 days during the total study period of five months. The foals had access to Panicum maximum cv. Tanzania pasture, while water and mineral salt lick were provided ad libitum. Blood samples were taken to determine serum retinol and levels of minerals, along with pasture samples to quantify beta-carotene in the grass. The results indicated that retinol levels varied (2.87 to 1.97 μg/dL) and remained below the standard levels. The levels of zinc, iron, calcium, and phosphorus did not vary significantly with average values of 36.79 μg/dL, 77.32 μg/dL, 10.33 mg/dL, and 9.99 mg/dL, respectively. However, zinc and calcium remained below standard concentrations of 60-120 μg/dL and 10.8-13.5 mg/dL, respectively, since the beginning of the study. On the other hand, copper and magnesium levels decreased over time (97.76 to 77.56 mg/dL and 2.86 to 2.21 μg/dL, respectively), but showed normal levels for horses during the research. Beta-carotene in grass showed a downward trend over time from 38.53 to 0.09 mg/kg of dry matter. The results also indicated a significant relationship between serum retinol and serum copper levels. Vitamin A supplementation is necessary whenever foals are fed pastures during the dry season.


Introduction
Grazing horses derive vitamin A from provitamin A carotenoids present in forages, but the efficiency of this conversion is relatively poor in these animals (Greiwe-Crandell et al., 1997).The most common provitamin A carotenoids in plant-based feedstuffs are β-carotene, αcarotene, and β-cryptoxanthin.The provitamin A carotenoids in major feed ingredients used in livestock diets have not been assessed for almost 50 years (Pickworth et al., 2012).
Beta-carotene is converted into retinol by enzymes in the intestine (Bondi and Sklan, 1984).Iron and zinc are important in the activation of these enzymes and affect the bioconversion of retinol and carotenoids (During et al., 1999;Blomhoff and Blomhoff, 2006).
Serum retinol levels reflect the status of vitamin A in livestock since they are homeostatically controlled from the hepatic reserves (Ross and Harrison, 2007).In addition, the synergistic effect of the minerals in the metabolism of this vitamin is quite intricate (Geor et al., 2013).
Inadequacy of vitamin A, particularly in pastures during the dry season, is expected to compromise the foal growth (Gay et al., 2004) because retinol is involved in several functions such as vision, bone development, reproduction, and integrity of mucosal and epithelial surfaces (Ortega et al., 2010).
The objective of this study was to examine serum retinol and some mineral (zinc, iron, calcium, phosphorus, magnesium, and copper) levels in Mangalarga Marchador grazing foals during the dry season.

Material and Methods
All procedures used in this experiment were approved by the Ethics Committee in Animal Experimentation of Universidade Federal de Minas Gerais (case no.194/08).
The present study was conducted from April to October, in the city of Montes Claros (16°44'06" S and 43°51'42" W) in the state of Minas Gerais, Brazil.The average monthly rainfall level registered during the dry season was 8.75 mm.Seven male Mangalarga Marchador foals aged 329.48±21.15days and weighting 224.42±46.29 kg were included in the experiment.Prior to the trial and every other month until the end of experiments, the foals were treated against ectoparasites with deltamethrin (10 mL diluted into 5 L of water for aspersion of each animal) and against endoparasites with ivermectin (1.6 g for 100 kg of weight) at a strictly maintained dosage according to the manufacturers' recommendation.Foals were kept on five hectares of Panicum maximum cv.Tanzania pasture, while mineral salt (Table 1) and water were provided ad libitum.
Every 45 days for approximately five months, blood and grass samples were collected from the animals and pasture, respectively, to evaluate serum retinol and mineral levels in the animals and beta-carotene in the forage.
Blood sampling was done through jugular puncture using 10-mL vacutainer tubes with no anticoagulant.The tubes were immediately placed into an ice chest protected from light with aluminum foil to prevent the oxidation of retinol.The collected blood was allowed to clot and then was centrifuged at 3000 rpm for 20 min at 4 °C to separate the serum fraction.Serum samples obtained from each foal were then pipetted and placed into amber cryogenic tubes, packed in aluminum foil, and frozen at -20 °C until they were thawed for retinol analysis.Retinol was extracted according to the procedure described by Pesce and Kaplan (1987) and retinol concentration was determined using high-performance liquid chromatography.
The extraction of minerals from blood consisted in diluting 500 μL of serum from the animals in 10-mL flasks containing 0.1% Triton X-100 (Merck Millipore, Darmstadt, Germany).Serum mineral evaluation followed the inductively coupled plasma optical emission spectrometry (ICP-OES) using the technique by Rocha et al. (2009).
Forage samples were randomly and manually collected from different areas of the pasture simulating equine feeding behavior according to Gardner (1986).The grass samples were wrapped in black plastic bags and then cut with scissors in an environment protected from light.The material was packed into several layers of black plastic, wrapped in aluminum foil and frozen at -20 °C for analysis of beta-carotene according to the AOAC (1995).
Retinol concentrations did not show normality or homoscedasticity; therefore, they were analyzed through non-parametric Friedman test.Serum minerals followed a completely randomized block design comparing four treatments (1, 45, 90, and 135 days) in seven blocks (foals), followed by Student-Newman-Keuls test.Betacarotene in Panicum maximum cv.Tanzania grass followed a completely randomized design with four treatments (1, 45, 90, and 135 days) and four replications each, assessed by Student-Newman-Keuls test.Pearson's correlation coefficient was determined between beta-carotene and crude protein contents in grass and serum retinol.All data were analyzed using the software SAEG (Sistema para Análises Estatísticas e Genéticas, version 8.0).To study the association among the variables (minerals, retinol, and beta-carotene), principal-components multivariate analysis of the software INFOSTAT, version 2008, was used.

Results and Discussion
During the dry season, serum retinol concentration of the foals dropped (P<0.05) over the study period (Table 2).
In this research, concentrations of retinol were lower than the standard value of 10 µg/dL reported by Lewis (2000), implying that foals in the present experiment had hypovitaminosis A with minimum liver stock levels and, probably, their hepatic reserves were mobilized to maintain serum retinol levels (Ball, 1998).Foals in this study were probably using their vitamin A reserves before the trial started, as the preceding 30 days had no rain at all, thus reducing the amount of beta-carotene in the pasture (Table 2).
Another relevant factor in the results was the category used in the present experiment, because young animals have little hepatic reserve, which makes them more susceptible to vitamin A deficiency when compared with adult animals.The liver reaches its peak storage of vitamin A only in adulthood (Taipina, 2001).Mineral salt composition (guaranteed analysis kg -1 ): calcium, 140 g; phosphorus, 75 g; sodium, 137 g; magnesium, 14 g; sulfur, 14 g; iron, 1200 mg; manganese, 1200 mg; zinc, 3000 mg; copper, 1250 mg; cobalt, 60 mg; iodine, 130 mg; selenium, 30 mg; and fluorine (max.), 750 mg.Álvarez et al. (2015) reported similar results when evaluating mares reared in a continuous grazing system in an area of silvopastoral system during the spring in Central Spain.Those authors found blood retinol concentration of 6.58 µg/mL, which indicates hypovitaminosis A. In contrast, in Southeastern Brazil, Garcia et al. (2006) found no difference (P>0.05) in serum retinol in pregnant mares kept on pasture during the dry season.
The results also indicated that the concentration of zinc did not vary (P>0.05) during the observed period.However, the values, in general, were suboptimum when compared with the standard values of 60-120 μg/dL according to Wichert et al. (2002).Although mineral salt was offered ad libitum, zinc deficiency was detected in the animals.This may be attributed to levels lower than those mentioned by the manufacturer or, in spite of having free access to the mineral lick, the foals consumed less of it.Rahman et al. (2002) studied different proportions of vitamin A and zinc in humans and came to the conclusion that there was synergism involving these nutrients.Zinc levels affected the synthesis of retinol-binding protein and beta-carotene to retinol conversion.Thus, it can be inferred that, in the present trial, low serum retinol could have been influenced by limited amounts of serum retinol transported to tissues as a result of zinc deficiency or the capacity of the animal to convert pasture beta-carotene into retinol.
Iron levels did not vary (P>0.05) during the study period.From 45 days onward, serum iron levels decreased numerically to values below the standard concentrations for horses (73 to 140 μg/dL) according to the NRC (2007), which indicates deficiency of this mineral.
Calcium and phosphorus concentrations did not vary (P>0.05)either during the observed period.Serum calcium values were always lower than the values of 10.8-13.5 mg/dL suggested by Garcia-Lopez et al. (2001).Low concentrations of calcium may result in poor skeletal development and further lead to problems such as osteopenia in growing animals (Geor et al., 2013).The phosphorus levels were higher than the values reported by Mundim et al. (2004) for equines (4.21±1.2mg/dL).According to the NRC (2007), horses show higher phosphorus levels at birth, which decrease over time until reaching standard concentration as adults.
The results also indicate that serum magnesium concentrations varied (P<0.05) during the research with a decline observed after a third of the study period and values remaining stable until the final collection.In spite of the decline, the levels were well above the values of 1.6-2.0mg/dL mentioned by the NRC ( 2007).These values showed that the foals had no magnesium deficiency.Serum copper levels decreased (P<0.05),but copper concentrations were well within 50-150 μg/dL for horses according to Wichert et al. (2002).
With no consistent rain during the experimental period, the nutritional value of the Panicum maximum cv.Tanzania grass declined as the grasses matured, which may also have influenced beta-carotene concentration.The study also indicated that during the short spell of rainfall volume (21 mm) recorded on the 90th day of the study, the grasses responded rapidly, leading to an enhancement in betacarotene levels, although there was no significant response in serum retinol levels in the foals.This occurrence can be explained by low rainfall, which did not allow homogeneous and efficient budding in the grazing area and, consequently, did not provide sufficient amounts of this nutrient to the animals.In addition, horses have low capacity of converting beta-carotene into retinol with efficiency of only 33% of the amount ingested (Bondi and Sklan, 1984).Insufficient beta-carotene intake associated with low nutritional quality of the forage during the experimental period resulted in loss of serum retinol in foals.A correlation of 0.83 was found between crude protein and serum retinol, which suggests that the animals need to feed on quality forage with adequate crude protein concentrations to maintain retinol levels.
There was positive correlation of 0.81 between crude protein and beta-carotene in grass.This high correlation is because carotenoids protect photosynthetic organisms against potentially harmful photo-oxidative processes and are essential structural components of the photosynthetic antenna and reaction-center complexes.The beta-carotene color is masked by chlorophyll in photosynthetic tissues (Bartley and Scolnik, 1995).
In this research, we observed a relationship of retinol, minerals, and beta-carotene levels using a non-inferential principal-component multivariate analysis (Figure 1).Calcium levels were not considered in the multivariate analysis since the coefficient of variation was quite low at 7.2%.
Serum iron levels inversely varied in relation to betacarotene in grass, i.e., the higher the beta-carotene levels in grass, the lower iron levels were found in serum.
Serum phosphorus levels seemed to be independent of grass quality, but zinc levels were highly correlated with phosphorus levels.Recent studies have indicated that copper levels are correlated with retinol levels.These results are in accordance with the findings of Van De Weyer et al. (2010), who found similar deficient results in beef cows on pastures for this micronutrient.
We found copper as the most associated serum element with retinol, contrary to current literature findings which consider zinc the mineral most associated with retinol.
The reduction of beta-carotene in grass, inflicted by the weather along the dry season, causes a decrease in serum copper, and, consequently, further reduction in serum retinol.
Castenmiller and West (1998) stated that levels of zinc in the body affect the synthesis of retinol-binding protein and beta-carotene conversion into retinol.The correlation between these nutrients were shown by Ahn and Koo (1995) in adult female rats, in which induced zinc deficiency decreased blood levels of retinol.
However, in the present study, the mineral identified as the main component positively correlated with retinol was copper.Similar results were obtained by Root et al. (1999) when studying women in China whose main source of vitamin A was beta-carotene from fresh vegetable diets.Simpson et al. (2004) listed copper-dependent liver proteins in ewes treated with different dietary copper levels.They found an increase in aldehyde dehydrogenase enzyme expression in the oxidation of retinal into retinoic acid.The retinol-binding protein had its activity increased by 1.5-2.0times with higher dietary copper levels.Those authors claimed that aldehyde dehydrogenase binds to free retinal and retinol-binding protein bound to retinal.
Foals in the present study were probably using liver retinol to keep homeostasis.To complete its metabolic pathway, stored retinol must be oxidized into retinal and then into retinoic acid, the latter step being copperdependent, which corroborates the present findings.
Zinc plays an important role in the metabolic pathway of vitamin A, mainly in its transportation to blood vessels.However to have vitamin A acting in epithelial tissues and in cell transcription, the biochemical steps involved require the presence of copper.

Conclusions
The practical recommendation is to supplement foals with vitamin A whenever they are kept on pastures during the dry season.

Table 1 -
Chemical composition of Panicum maximum cv.

Table 2 -
Average serum concentration of retinol and minerals in foals and beta-carotene levels in grass 1Friedman test (P<0.05).Means within columns followed by different letters are different by SNK test (P<0.05).