Exploring the interaction of vitamin A and zinc in finishing cattle

Abstract

Angus-cross steers (n = 95; BW = 473 ± 20 kg) were used to explore the interaction between vitamin A and zinc (Zn) and their influence on performance and carcass quality. Steers were stratified by BW to pens (n = 16; four pens/treatment), each equipped with a GrowSafe bunk (Vytelle, LLC., Lenexa, KS). Treatments were arranged as a 2 × 2 factorial; vitamin A treatments (VA) were: 0 IU supplemental vitamin A/kg DM (0VA) or 2,200 IU supplemental vitamin A/kg DM (2200VA; retinyl acetate, DSM Nutritional Products, Pendergrass, GA); Zn treatments (ZN) were: 0 mg/kg Zn (0ZN) or 120 mg/kg Zn (120ZN; ZnSO₄). Steers were implanted with Synovex Plus (200 mg trenbolone acetate, 28 mg estradiol benzoate; Zoetis, Parsippany, NJ) at trial initiation. Individual BW was collected on days −1, 0, 28, 56, 83, and 84. Blood (serum, plasma) was collected on days 0, 56, and 79. Liver was collected on days 3 and 79. Blood and liver were analyzed for retinol and trace mineral concentrations. Carcass data were collected after a 48-hr chill. Data were analyzed in Proc MIXED or Proc GLIMMIX (SAS 9.4, SAS Inst. Inc., Cary, NC) with fixed effects of ZN and VA and the interaction; blood was analyzed via repeated measures (n = 23–24 per treatment, steer as experimental unit). There were no differences due to ZN, VA, or ZN × VA for live BW, overall ADG, or overall dry matter intake (P ≥ 0.21). Overall G:F was lesser in 0VA compared to 2200VA (VA; P = 0.05). No differences were observed in HCW, marbling, or REA due to ZN, VA, or ZN × VA (P ≥ 0.14). There was a tendency for rib fat to be greater in 0ZN-2200VA compared to other treatments (ZN × VA, P = 0.08). Dressing percentage was greatest in 0ZN-0VA and 120ZN-2200VA and least in 120ZN-0VA and 0ZN-2200VA, with 0ZN-0VA being not different from 120ZN-0VA (ZN × VA, P < 0.01). A greater percentage of carcasses graded average Choice or higher in 0VA compared to 2200VA (VA, P = 0.03). Liver retinol concentrations were lesser in 0VA compared to 2200VA (P = 0.01) and were unaffected by ZN (P ≥ 0.31). Serum retinol in 0VA declined upon removing vitamin A supplementation, 2200VA remained constant and was greater than that of 0VA from day 56 until harvest (VA × Day; P < 0.01). Liver Zn was unaffected by any treatment (P ≥ 0.16). Plasma Zn concentration was affected by ZN × VA (P = 0.02), where it was greatest in 120ZN (0VA and 2200VA), least in 0ZN-0VA, and intermediate in 0ZN-2200VA. Plasma Cu remained constant in 0VA until d 56 then declined prior to harvest, while in 2200VA steers increased by day 56 and remained greater than 0VA until harvest (VA × Day; P = 0.01). Liver Cu was lowest in 120ZN-2200VA compared to all other treatments (ZN × VA; P = 0.04). In finishing beef steers, dietary vitamin A supplementation presents a trade-off between optimal growth performance and carcass merit. Vitamin A supplementation increased plasma Zn concentration, and in this study supplementing both vitamin A and Zn promoted a shift in carcass composition.

Dietary vitamin A and zinc affected circulating concentrations of both, while removing vitamin A from the diet increased carcass merit but reduced feed efficiency.

Introduction

Vitamin A and zinc (Zn) are essential micronutrients in cattle diets (NASEM 2016); however, while research in other species supports an interdependency between these nutrients (Smith et al. 1974; Berzin and Bauman 1987), this has not been explored in feedlot cattle. Optimizing vitamin A and Zn requirements for feedlot cattle could be cost-effective tools to enhance carcass quality and cattle performance.

As reviewed by others (Nguyen et al. 2021), vitamin A has a unique impact on marbling, mediated by retinoic acid—the active metabolite of vitamin A—and its influence on genes associated with myogenesis, adipogenesis, and fibrogenesis (Pyatt et al. 2005; Knutson et al. 2020). The current NASEM (2016) recommendation for vitamin A supplementation for beef feedlot cattle is 2,200 IU/kg DM. Early-life interventions of vitamin A like neonatal injections or increased maternal supplementation during gestation promote adipocyte hyperplasia, increasing marbling potential (Harris et al. 2018; Dean et al. 2024). However, vitamin A supplementation to finishing cattle appears to discourage lipid accumulation within existing cells, and the removal of vitamin A from the finishing diet can enhance final quality grade (Jo et al. 2009).

While vitamin A governs adiposity, Zn is a key micronutrient supporting numerous enzymes involved in protein synthesis, resulting in lean tissue accretion (Nimmanon et al. 2017). NASEM (2016) recommends 30 mg Zn/kg DM, though the common supplementation practice in industry (90 mg Zn/kg DM; Samuelson et al. 2016) is supported by a variety of studies showing positive impacts of Zn on cattle performance (Beeson et al. 1977; Messersmith and Hansen 2024).

The complex physiological interaction between vitamin A and Zn spans absorption to storage to mobilization. For example, retinol binding protein-4 requires Zn for mobilization of vitamin A from the liver (Smith et al. 1974), as does the conversion of retinol to retinal (Sundaresan et al. 1977). Conversely, vitamin A deficiency impairs intestinal Zn absorption by reducing the expression of Zn-binding proteins (Berzin and Bauman 1987). As both micronutrients have potential to impact growth and carcass value, it is important to understand how they interact to influence cattle performance.

Therefore, the objective of this experiment was to evaluate the interaction between supplemental Zn and vitamin A on growth, carcass merit, and trace mineral metabolism in finishing steers. We hypothesized that adding Zn and removing vitamin A would improve growth and carcass quality.

Materials and methods

Procedures and protocols utilized in this experiment were approved by the Iowa State University Institutional Animal Care and Use Committee (IACUC-25-017).

Animals and experimental design

Angus-cross steers (n = 95; BW = 473 ± 20 kg) from a single ranch were used in a study conducted at the Beef Nutrition Farm located in Ames, Iowa from February 2025 to May 2025. Prior to this experiment steers received one of three growing diets (0 [n = 40], 30 [n = 28], or 100 [n = 27] mg supplemental Zn/kg DM; 2200 IU vitamin A/kg DM). Plasma Zn on day −20 was measured on all steers. Steers were allotted to pens (n = 16), blocked by prior Zn dietary treatment/plasma Zn, and stratified by body weight (BW). Each pen was equipped with a GrowSafe Feed bunk (n = 6 steers/pen; Vytelle, LLC.; Lenexa, KS). Each steer was fitted with a unique radio frequency tag in the ear, allowing the GrowSafe bunk to recognize each individual steer and to record individual animal feed disappearance.

A 2 × 2 factorial arrangement of treatments was applied to Zn and vitamin A dietary treatments (n = 4 pens per treatment). Zinc treatments included: 1) no supplemental Zn (0ZN); 2) 120 mg supplemental Zn/kg DM (120ZN; as ZnSO₄). Steers (n = 48) previously on a low supplemental Zn diet and those with the lowest plasma Zn (1.11 mg/L [mean]; 0.89–1.32 mg/L [range]) were assigned to 0ZN treatments. Steers (n = 47) previously on high supplemental Zn diets and with the highest plasma Zn (1.16 mg/L [mean]; 0.99–1.49 mg/L [range]) were assigned to 120ZN treatments. Within 0ZN or 120ZN dietary treatments, plasma Zn was utilized to assign steers to pens ensuring each pen had similar starting plasma Zn concentration. Vitamin A dietary treatments were supplied as retinyl acetate (ROVIMIX A 1000; DSM Nutritional Products; Pendergrass, GA) and included: 1) no supplemental VA (0VA); 2) 2,200 supplemental VA/kg DM (2200VA). All steers received a terminal implant on study day 0 (Synovex Plus; 200 mg trenbolone acetate; 28 mg estradiol benzoate; Zoetis; Parsippany, NJ).

Cattle were fed once daily at 0800 hours, both feed and water were provided for ad libitum intake. Dietary treatments were delivered in a total mixed ration (TMR) using dried distillers grains as the carrier for ZN and VA premixes. All other vitamin and mineral requirements met or exceeded those for finishing beef cattle (NASEM 2016) and the diet is shown in Table 1.

Table 1.

Composition of diets fed.

Ingredient, %DM basis	0ZN-0VA	0ZN-2200VA1	120ZN-0VA2	120ZN-2200VA1,2
Dry rolled corn	60	60	60	60
Corn silage	15	15	15	15
Dried distillers grains	20	15	15	10
Basal premix3,4	5	5	5	5
VA2200 premix3	–	5	–	5
ZN120 premix3	–	–	5	5
Analyzed composition, %DM5
Crude protein	14.42	14.42	14.42	14.42
Neutral detergent fiber	15.68	15.68	15.68	15.68
Ether extract	5.15	5.15	5.15	5.15
NEm, Mcal/kg6	2.09	2.09	2.09	2.09
NEg, Mcal/kg7	1.42	1.42	1.42	1.42
Zn, mg/kg8	32	34	141	145

¹Dietary treatments included supplemental Vitamin A as retinyl acetate (ROVIMIX A 1000; DSM Nutritional Products): 0ZN-2200VA or 120ZN-2200VA.

²Dietary treatments included supplemental Zn as zinc sulfate: 0VA-120ZN or 120ZN-2200VA.

³Treatment premix and basal premix utilized dried distillers grains as the carrier. Basal provided as % DM; limestone (1.5%), Rumensin (0.0135%), and salt (0.31%).

⁴Trace minerals and vitamins provided in the basal per kg of diet DM: 10 mg of Cu (copper sulfate), 20 mg of Mn (manganese sulfate), 0.5 mg of I (calcium iodate), 0.1 mg of Se (sodium selenite), and 0.15 mg of Co (cobalt carbonate), 25 IU of vitamin E.

⁵Based on composited TMR wet chemistry analysis from Dairyland, INC., Arcadia, WI.

⁶Based on NASEM (2016) reported NEm values of feedstuffs.

⁷Based on NASEM (2016) reported NEg values of feedstuffs.

⁸Analyzed values measured by ICP OES (ICP Optima 7000 DV, Perkin Elmer, Waltham, MA).

Steers were slaughtered at a commercial abattoir (National Beef, Tama, IA) via industry-accepted practices on day 85 and hot carcass weight data were collected by trained personnel. Following a 48-h chill, ribeye area (REA), twelfth rib back fat (RF), USDA yield grade (YG; as called by the plant), and marbling scores were collected.

Sample collection

TMR samples were collected weekly and dried in a forced-air oven at 70 °C for 48 h for determination of dry matter (DM). Dry matter intake for each steer was calculated using as-fed intakes as recorded by GrowSafe and applying DM (%) of weekly TMR samples. Individual BW were collected prior to feeding on days −1, 0, 28, 56, 83, and 84, with a 4% pencil shrink applied to all live BW. Average daily gain (ADG) was calculated using BW and days for each period. Feed efficiency (G:F) was calculated by dividing ADG by dry matter intake (DMI) for the respective period.

Jugular blood samples were collected from each steer (n = 95) on days 0, 56, and 79 into vacuum tubes (serum [no additive], potassium EDTA; Becton, Dickinson and Company, Franklin Lakes, NJ). Blood samples on days 0 and 56 were collected prior to feeding. Steers had access to feed before blood collection on day 79 as samples were collected immediately prior to liver biopsy. Whole blood samples were centrifuged at 2,000 × g for fifteen minutes at 4 °C. Serum was protected from light, aliquoted and stored at −80 °C. Plasma was stored at −20 °C. Liver biopsies were collected from a subset of steers (n = 48; 3 per pen) on days 3 and 79 at least 2 h after feeding using methods outlined by Engle and Spears (2000). The subset of steers (n = 12/treatment) were selected such that average BW of the subset was a representation of the average pen BW. Samples were rinsed with phosphate buffered saline before being divided into two aliquots for separate analysis. One aliquot was packaged in tin foil, flash frozen, and stored at −80 °C to maintain retinol integrity. The other aliquot was collected into a 5-mL tube and placed in a cooler with dry ice and was stored at −20 °C.

Analytical procedures

Weekly TMR samples were ground (Retsch Zm100 grinder; Retsch GmBH, Hann, Germany), dried, and composited by treatment for analysis of crude protein (method 990.03; AOAC, 1996), neutral detergent fiber (NDF; method 2002.04; AOAC, 2005), and ether extract (method 920.39; AOAC, 1996) by a commercial laboratory via wet chemistry (Dairyland Laboratories, Inc., Arcadia, WI). Composited TMR samples were acid digested (CMES MARS Xpress, CEM Corporation, Matthews, NC) using trace mineral (TM) grade nitric acid as previously described by Genther-Schroeder et al. (2016) before analysis of Cu, Fe, Mn, and Zn concentrations using inductively coupled plasma—optical emission spectroscopy (ICP-OES; Optima 7000; PerkinElmer, Waltham, MA).

Liver samples stored at −20 °C were dried at 70 °C and acid digested (CMES MARS Xpress, CEM Corporation, Matthews, NC) for analysis of Cu, Fe, Mn, and Zn concentration using ICP-OES with methods previously described (Pogge and Hansen 2013). Instrument accuracy was confirmed using a bovine liver reference sample (National Institutes of Standards and Technology, Gaithersburg, MD). Flash frozen liver samples (day 3, day 79) were analyzed for retinol concentration using i-Check Flouro (Bioanalyt; Teltow, Germany) with methods described by Rients et al. (2026) with slight modifications (µHb Micro Homogenizer, Omni-International, Kennesaw, GA). Sample weight for retinol analysis was between 0.1 and 0.2 g (wet basis). A standard sample provided by Bioanalyt (Teltow, Germany) was run before and after sample measurement run to ensure machine accuracy. A percentage of 21% was applied to convert samples to a dry matter basis. The remaining flash frozen liver from each sample were ground using a mortar and pestle in liquid nitrogen to be used for gene expression analysis for genes related to marbling, retinol-binding protein 4 (RBP4), and trace mineral storage and transport at the final sampling timepoint (day 79; Table S1). Isolation of RNA and cDNA synthesis were completed as described in Rients et al. (2023). Quantity and quality of RNA were measured on a Cytation5 (Agilent Technologies, Winooski, VT) with a Take3Trio (Agilent Technologies) using a nucleic acid quantification protocol. The synthesized cDNA product was stored at −20 °C until quantitative real-time polymerase chain reaction (qPCR) analysis. Quantitative real-time PCR was completed as described by McGill et al. (2016) with ribosomal protein S9 (RPS9) utilized as a reference gene based on prior work describing the stability of this gene in bovine liver under nutritional changes (Janovick-Guretzky et al. 2007; Ostrowska et al. 2014). Stability of the reference gene was confirmed by analyzing RPS9 CT values for ZN, VA, or interactive effects and none were found (P ≥ 0.35), supporting its use for normalization. Conditions for amplification remained the same: (Hold stage) 2 min at 50 °C, 10 min at 95 °C, 40 cycles of 15 s at 95 °C, and 1 min at 60 °C (measure fluorescence step) and a dissociation step of 15 s at 95 °C, 1 min at 60 °C, 15 s at 95 °C and 15 s at 60 °C. Dissociation curves were analyzed for proper product amplification. Relative gene expression is expressed as delta cycle threshold (ΔCT) and was calculated by subtracting the housekeeping gene (RPS9) from the gene of interest. Table S1 provides primers utilized for qPCR.

Plasma samples (n = 95) were analyzed for Cu, Fe, and Zn concentration using methods described by Pogge and Hansen (2013). Instrument accuracy was determined using a human serum sample (UTAK Laboratories, Inc., Valencia, CA). Serum retinol was analyzed using a portable fluorometer according to manufacturer directions (iCheck Fluoro, Bioanalyt, Teltow, Germany). Serum samples used for retinol analysis were thawed on ice and 500 µl was injected into extraction vials (98% hexane, 1% isopropanol, 1% ethanol) and vortexed for 10 seconds. Vials were allowed to settle to complete phase separation and were then inserted into the fluorometer for vitamin A concentration measurement. A standard sample provided by Bioanalyt (Teltow, Germany) was run before and after sample measurement run to ensure machine accuracy. Analysis of circulating metabolites was completed on all steers (n = 95) with serum samples collected on day 56 as steers had already had access to feed prior to blood collection on d 79. Serum glucose (FUJIFILM Wako Diagnostics; intra-assay CV = 4.97%, inter-assay CV = 4.06%) and non-esterified fatty acids (NEFA) (Wako Pure Chemical Industries Ltd., Chuo-ku Oska, Japan; intra-assay CV = 4.64%, inter-assay CV = 8.43%) were assayed in duplicate. Serum samples from day 79 were analyzed for ceruloplasmin concentration using the methods described by Dimitriou et al. (1974).

Statistical analysis

Growth performance, carcass characteristics, liver retinol, liver TM, gene expression and ceruloplasmin data were analyzed as a completely randomized design with a factorial arrangement of treatments using the Proc MIXED procedure of SAS 9.4 (SAS Inst. Inc., Cary, NC). Yield and quality grades were analyzed as a completely randomized factorial design with binomial distributions using Proc GLIMMIX of SAS 9.4 (SAS Inst. Inc.). The model included fixed effects of ZN, VA, and their interaction, with individual steer as the experimental unit (n = 23–24 per treatment). Initial BW (day 0) served as the covariate for all growth performance analysis. Liver retinol and liver TM data for day 79 were analyzed using day 3 values as a covariate. Serum retinol and plasma TM data were analyzed as repeated measures using the MIXED procedure of SAS 9.4 (SAS Inst. Inc.). The model used included fixed effects of ZN, VA, day, and their interaction with day as repeated. The following covariance structures were used based on lowest Akaike’s Information Criterion: serum retinol (CS), plasma Cu (UN), plasma Fe and plasma Zn (CSH). After ear palpitation, it was determined that one steer had lost the implant, thus data from this steer were removed from analysis (0ZN-2200VA). Statistical outliers were determined and removed if data points were beyond three standard deviations from the mean. When the interaction term was significant, means were separated using PDIFF in SAS 9.4. Statistical significance was determined as P ≤ 0.05 and tendencies were determined as 0.05 < P ≤ 0.10.

Results

Growth performance

Live growth performance and CA performance are reported in Table 2. Body weight, ADG, and DMI were not affected by VA, ZN, or the interaction (P ≥ 0.21). However, live G:F was lesser (P = 0.05) and CA:GF tended to be lesser (P = 0.07) in 0VA versus 2200VA.

Table 2.

Effect of dietary vitamin A and Zn on live and carcass adjusted (CA) performance of finishing beef steers.¹

	0ZN		120ZN			P-values
	0VA	2200VA	0VA	2200VA	SEM2	ZN	VA	ZN x VA
Live performance
Day 0 BW, kg	475	475	470	471	4.2	0.25	0.99	0.83
Day 28 BW, kg	534	534	532	533	2.3	0.57	0.87	0.66
Day 56 BW, kg	590	588	587	588	3.5	0.41	0.85	0.35
Day 84 BW, kg	626	631	622	628	4.2	0.44	0.21	0.88
Overall ADG, kg	1.83	1.88	1.78	1.85	0.050	0.45	0.21	0.88
Overall DMI, kg	11.77	11.67	11.75	11.63	0.150	0.84	0.45	0.95
Overall G:F	0.156	0.163	0.152	0.160	0.0036	0.34	0.05	0.88
CA performance3
Final BW, kg	629	626	621	632	4.5	0.81	0.35	0.13
ADG, kg	1.86	1.83	1.77	1.90	0.054	0.84	0.35	0.13
G:F	0.159	0.162	0.152	0.165	0.0042	0.64	0.07	0.18

¹D0 BW served as a covariate in analysis for all growth performance measures other than d 0 BW.

²Highest SEM of any treatment is reported.

³A common dressing percentage (63.8%) was applied to all steers to calculate carcass adjusted final body weight.

Carcass characteristics

Carcass characteristics are reported in Table 3. There was no effect of VA, ZN, or their interaction on HCW, marbling, or REA (P ≥ 0.14). Rib fat tended to be lesser in 0ZN-0VA, 120ZN-0VA, and 120ZN-2200VA versus 0ZN-2200VA (ZN × VA, P = 0.08). Dressing percentage was greatest in 0ZN-0VA and 120ZN-2200VA and least in 120ZN-0VA and 0ZN-2200VA with 0ZN-0VA being not different from 120ZN-0VA (ZN × VA, P < 0.01). Yield and quality grade distribution was not affected by the interaction of ZN and VA (P ≥ 0.15). An increased percentage of carcasses graded Certified Angus Beef (CAB) and USDA Average Choice or above in 0VA versus 2200VA (P ≤ 0.03). Yield grade was not affected by VA (P ≥ 0.39), but supplemental Zn tended (P = 0.07) to increase YG 3 while not affecting YG 4 or 5 (P ≥ 0.17).

Table 3.

Effect of dietary vitamin A and Zn on carcass performance of finishing beef steers.

	0ZN		120ZN			P value
	0VA	2,200VA	0VA	2,200VA	SEM1	ZN	VA	ZN × VA
Carcass traits
HCW, kg	402	400	396	403	3.0	0.81	0.36	0.14
Marbling2	520	522	544	504	51.2	0.88	0.24	0.31
Ribeye area, cm2	91.4	91.7	91.6	92.8	0.58	0.23	0.21	0.40
12th rib back fat, cm	1.50y	1.77x	1.51y	1.44y	0.099	0.10	0.31	0.08
Dressing percentage, %	64.1a,b	63.3c	63.7b,c	64.2a	0.190	0.21	0.48	< 0.01
Yield grade, %3
3	75.0	63.6	79.2	91.3	8.84	0.07	0.67	0.16
4	25.0	26.1	20.8	8.7	9.16	0.17	0.39	0.34
5	0.0	10.3	0.0	0.0	5.878	0.99	0.99	0.99
Quality grade, %3
Select	0.0	0.0	4.2	8.7	5.87	0.97	0.99	0.99
Choice	95.8	86.4	83.3	87.0	7.61	0.31	0.49	0.28
Prime	4.2	13.6	12.5	4.3	7.32	0.97	0.93	0.15
≥ Average Choice	66.7	39.1	54.2	34.8	10.18	0.41	0.03	0.69
CAB	58.3	40.9	66.7	30.4	10.48	0.91	0.01	0.35

¹Highest SEM of any treatment is reported.

²Marbling scores: slight: 300, small: 400, modest: 500, moderate: 600, slightly abundant: 700, abundant: 800.

³Expressed as percentage of the treatment.

^a,b,cWithin a row, means with different superscripts differ (P < 0.05).

^x,y,zWithin a row, means with different superscripts tend to differ (0.05 < P ≤ 0.1).

Retinol

There was no effect of ZN × VA × DAY, ZN × DAY, or ZN × VA for serum retinol (P ≥ 0.46). Serum retinol declined in 0VA across days on feed with a transient increase in 2200VA (VA × Day; P < 0.01; Figure 1A). Overall serum retinol tended to be greater in 120ZN steers than 0ZN steers (ZN, P = 0.10; Figure 1B). Initial and final liver retinol were lesser in 0VA than 2200VA (VA; P = 0.01; Figure 1C). There was no effect of ZN or a ZN × VA interaction for initial or final liver retinol concentrations (P ≥ 0.16).

Figure 1.

A) Effect of vitamin A supplementation (VA) and day of study on serum retinol (µg/L) based on repeated measures analysis of samples collected on days 0, 56, and 79. B) Effect of Zn supplementation (ZN) on overall serum retinol concentration (µg/L). C) Effect of vitamin A supplementation (VA) on liver retinol concentration (µg/g DM) measured on days 3 and 79. Liver retinol concentrations from day 3 were used as a covariate in the analysis of day 79. VA treatments were 0 IU vitamin A/kg DM or 2,200 IU vitamin A/kg DM (supplemented as retinyl acetate), provided for 84 days prior to harvest. ZN treatments included 0 mg Zn/kg DM or 120 mg Zn/kg DM (supplemented as ZnSO₄), provided for 84 days prior to harvest.

Trace mineral

Plasma TM data measured on days 0, 56 and 79 are reported in Table 4. There was no ZN × VA × DAY, ZN × DAY, ZN × VA, or ZN effects on plasma Cu (P ≥ 0.12). Plasma Cu remained steady in 0VA until day 56 and was lower at day 79, while 2200VA plasma Cu increased by day 56 and remained greater than 0VA until harvest (VA × DAY; P = 0.01; Figure 2A). There were no effects on serum ceruloplasmin concentrations at day 79 (P ≥ 0.25). There was a trend for a ZN × VA × DAY effect (P = 0.08; Figure 2B) in plasma Fe where all treatments peaked at day 56, with 120ZN-2200VA tending to be greater than the others. Plasma Fe returned to baseline levels by day 79. There were no ZN × VA × DAY or VA × DAY effects on plasma Zn (P ≥ 0.34). A ZN × DAY effect was observed where 120ZN steers had greater plasma Zn concentrations than 0ZN steers throughout the duration of the trial and both treatments experienced a decline in plasma Zn concentration across days on feed (P < 0.01; Figure 2C). Overall plasma Zn was least in steers receiving 0ZN-0VA and greatest in steers receiving 120ZN (0VA; 2200VA), with 0ZN-2200VA being intermediate (ZN × VA; P = 0.02; Figure 2D).

Table 4.

Effect of dietary vitamin A and zinc on repeated measures analysis of plasma trace mineral concentrations (mg/L) in finishing beef steers.

	0ZN		120ZN			P value
	0VA	2,200VA	0VA	2,200VA	SEM1	ZN	VA	ZN × VA
Overall Cu2	0.83	0.86	0.86	0.90	0.232	0.12	0.13	0.80
Day 0	0.84	0.81	0.87	0.87
Day 56	0.87	0.90	0.87	0.93
Day 79	0.79	0.87	0.84	0.90
Day 79 ceruloplasmin, mg/dL	16.8	18.3	18.9	19.3	1.34	0.25	0.49	0.69
Overall Fe3	1.94	1.99	1.90	2.03	0.547	0.91	0.09	0.45
Day 0	1.85	1.97	1.89	1.90
Day 56	2.07	2.05	1.96	2.19
Day 79	1.89	1.95	1.85	2.01
Overall Zn4	1.11c	1.20b	1.27a	1.30a	0.024	<0.01	0.17	0.02
Day 0	1.16	1.26	1.34	1.32
Day 56	1.10	1.18	1.28	1.32
Day 79	1.08	1.17	1.19	1.25

¹Highest SEM of any treatment is reported.

²ZN × VA × DAY (P = 0.37); ZN × DAY (P = 0.61); VA × DAY (P = 0.01; as shown in Figure 2A).

³ZN × VA × DAY (P = 0.08; as shown in Figure 2B); ZN × DAY (P = 0.88); VA × DAY (P = 0.78).

⁴ZN × VA × DAY (P = 0.34); ZN × DAY (P = 0.01; as shown in Figure 2C); VA × DAY (P = 0.39).

^a,b,c Within a row, means with different superscripts differ (P < 0.05).

Figure 2.

A) Effect of vitamin A supplementation (VA) and day of study on plasma Cu concentrations (mg/L) measured on days 0, 56, and 79. B) Effect of vitamin A supplementation (VA), Zn supplementation (ZN) and day on plasma Fe concentrations (mg/L) measured on days 0, 56, and 79. C) Effect of Zn supplementation (ZN) and day on plasma Zn concentrations (mg/L) measured on days 0, 56, and 79. D) Effect of vitamin A (VA) and Zn (ZN) supplementation on overall plasma Zn concentrations (mg/L). VA treatments were 0 IU vitamin A/kg DM or 2,200 IU vitamin A/kg DM (supplemented as retinyl acetate), provided for 84 days prior to harvest. ZN treatments included 0 mg Zn/kg DM or 120 mg Zn/kg DM (supplemented as ZnSO₄), provided for 84 days prior to harvest.

Liver TM concentrations are reported in Table 5. Liver Cu did not differ on d 3 due to ZN, VA, or their interaction (P ≥ 0.30). However, on day 79, liver Cu was lesser in 120ZN-2200VA compared to others (ZN × VA; P = 0.04). Liver Fe on day 3 was not affected by the interaction of VA and ZN (P = 0.99) but was lesser in 0VA (P = 0.05) vs. 2200VA and tended (P = 0.06) to be lesser in 0ZN versus 120ZN. On day 79, liver Fe was lesser in 0ZN-0VA and 120ZN-2200VA than 120ZN-0VA steers, with 0ZN-2200VA steers being not different from 0ZN-0VA, 120ZN-0VA, or 120ZN-2200VA (ZN × VA, P = 0.04). There were no main effects of ZN or VA on d 79 liver Fe concentrations (P ≥ 0.21). Liver Mn concentrations were unaffected by treatments on day 3 (P ≥ 0.25), but on d 79 0ZN-0VA tended to be greater than all other treatments (ZN × VA; P = 0.06). Liver Zn on day 3 and day 79 was not affected by VA, ZN or the interaction (P ≥ 0.11).

Table 5.

Effect of dietary vitamin A and zinc on liver trace mineral concentration (mg/kg DM) in finishing beef steers.

	0ZN		120ZN			P value
	0VA	2200VA	0VA	2200VA	SEM1	ZN	VA	ZN × VA
Cu
Day 3	314	304	291	289	18.9	0.30	0.76	0.82
Day 792	337a	337a	334a	268b	16.3	0.03	0.04	0.04
Fe
Day 3	115	124	124	133	4.4	0.06	0.05	0.99
Day 792	143b	157a,b	172a	150b	8.9	0.21	0.64	0.04
Mn
Day 3	7.3	6.9	6.7	6.7	0.35	0.25	0.53	0.73
Day 792	7.9x	6.8y	6.6y	6.7y	0.33	0.04	0.12	0.06
Zn
Day 3	113	113	111	123	4.1	0.31	0.11	0.16
Day 792	115	119	125	120	5.6	0.33	0.94	0.43

¹Highest SEM of any treatment is reported.

²Day 3 liver concentration was used as a co-variate for the respective mineral.

^a,b,cWithin a row, means with different superscripts differ (P < 0.05).

^x,yWithin a row, means with different superscripts tend to differ (0.05 < P ≤ 0.1).

Serum metabolites and hepatic gene expression

There were no differences in serum glucose or NEFA concentration at d 56 (P ≥ 0.16; Table 6). No differences in hepatic gene expression of ADHIC, ATP7B, COMMD1, CP, HAMP, MT2A, RBP4, or ZIP8 were seen at day 79 due to ZN, VA, or their interaction (P ≥ 0.11; Table 7).

Table 6.

Effect of dietary Zn and vitamin A on d 56 serum metabolites of finishing beef steers.

	0ZN		120ZN			P value
	0VA	2,200VA	0VA	2,200VA	SEM1	ZN	VA	ZN × VA
Serum glucose, mg/dL	68.9	72.0	67.2	69.6	1.93	0.29	0.16	0.86
Serum NEFA, µEq, L	108.4	107.1	105.6	106.7	5.45	0.77	0.98	0.82

¹Highest SEM of any treatment is reported.

Table 7.

Effect of dietary Zn and vitamin A on d 79 liver gene expression (Δ CT)^1,2 of finishing beef steers.

	0ZN		120ZN			P-Value
	0VA	2200VA	0VA	2200VA	SEM3	ZN	VA	ZN × VA
ADH1C	9.735	9.299	9.775	9.780	0.3150	0.39	0.47	0.46
ATP7B	2.479	2.592	2.381	2.581	0.1269	0.65	0.12	0.72
COMMD1	4.840	4.858	4.794	4.881	0.1492	0.93	0.71	0.81
CP	−2.106	−2.357	−2.399	−2.506	0.2821	0.41	0.51	0.79
HAMP	−0.542	−0.393	−0.769	−0.256	0.3121	0.88	0.27	0.54
MT2A	−0.112	−0.556	−1.013	−0.940	0.4187	0.11	0.64	0.52
RBP4	−0.080	0.141	−0.245	0.136	0.3184	0.78	0.32	0.79
ZIP8	2.325	2.269	2.350	2.330	0.2026	0.82	0.84	0.93

¹Function of each gene can be found in Table S1.

²Delta CT is the CT value of the target gene minus the CT of the housekeeping gene (RPS9).

³Highest SEM of any treatment is reported.

Discussion

Work in other species has established a symbiotic relationship between Zn and vitamin A but our understanding of this relationship is limited in cattle. Zinc is required for vitamin A uptake from the intestine (Jackson et al. 2023), the mobilization of retinol from the liver (Smith et al. 1974), and the conversion of retinol to retinal (Wald 1950). Vitamin A likely plays a less impactful role in Zn status and absorption, but is essential for intestinal epithelial health, the site of Zn absorption, and supports Zn-binding proteins (Miller and Cragle 1965; Berzin and Bauman 1987). This experiment sought to better characterize the interaction between Zn and vitamin A in finishing cattle, specifically examining differences in nutrient status, growth performance, and carcass quality.

All steers had adequate trace mineral status (Cu, Fe, Zn) throughout the duration of the experiment (Puls 1988). Plasma Zn was lesser in 0VA compared with 2200VA, consistent with the work of Berzin and Bauman (1987) who reported serum Zn to be less in chicks fed a diet devoid of vitamin A. Vitamin A devoid chicks also had negative Zn balance (Berzin and Bauman 1987) implying vitamin A is important for Zn absorption. While steers in the present study had biologically normal plasma Zn across treatments, more work is needed to understand how vitamin A influences Zn metabolism in cattle.

Despite receiving 2200 IU vitamin A/kg DM (NASEM 2016) for more than 100 days in their prior growing diet, liver retinol concentration was marginal in all steers at trial initiation in the present experiment (61 µg/g DM; Puls 1994). After eighty-four days, liver retinol of 0VA declined to 25 µg/g DM, a value considered deficient (Puls 1994) while 2200VA maintained liver retinol concentrations. Pickworth et al. (2012) reported similarly marginal liver retinol concentrations in steers at the start of their experiment (∼ 55µg/g DM). In the few studies that measured liver retinol, steers receiving no supplemental vitamin A often reached quite low liver retinol (1–10 µg/g DM). This decline may be more dramatic than that observed in the present study because others had more days without vitamin A (Gorocica-Buenfil et al. 2007a, 2007b, 2007c; Pickworth et al. 2012).

Others have observed an ∼12% improvement in the percentage of carcasses grading average Choice or above when vitamin A supplementation is removed (Gorocica-Buenfil et al. 2007a; Pickworth et al. 2012), consistent with the 20% improvement in 0VA noted in this study. Cattle must meet set criteria (marbling, REA, RF, hide color) to qualify for Certified Angus Beef premiums, and 27% more carcasses received Certified Angus Beef premiums in 0VA compared to 2200VA. In this study, the majority of CAB eligibility is overwhelmingly due to marbling score. The impact of vitamin A on marbling and performance has been heavily researched, but with contradicting results, likely due to the continuous nature of this measurement (Gorocica-Buenfil et al. 2007a; Arnett et al. 2009; Pickworth, et al. 2012; Knutson et al. 2020). Improvements in quality grade are a consistent observation with the removal of vitamin A. An extensive review completed by Nguyen et al. (2021) describes the mechanism behind the often-observed improvements in carcass quality, as retinoic acid negatively influences adipocyte lipid filling. Several factors could be contributing to the effect of vitamin A on carcass quality: cattle type, weaning timeline, dietary management, and general propensity to marble. The steers in the present study were Angus-cross. Others have noted the improvement in marbling due to withdrawal of vitamin A in Angus but not Simmental steers (Knutson et al. 2020) or that genotype may impact response (Ward et al. 2012). Despite differences in retinol status and quality grade in the current study, hepatic mRNA expression of ADH1C, which encodes an enzyme that oxidizes retinol toward retinoic acid, was unaffected by vitamin A. This is not inconsistent with Ward et al. (2012), who linked ADH1C genotype, rather than its hepatic expression, to marbling response. In addition, although less studied, there appears to be no evidence in the literature that increasing vitamin A supplementation beyond NASEM (2016) improves quality grade (Pyatt et al. 2005). Further research examining the effects of breed, weaning strategy, and vitamin A supplementation in finishing cattle is needed.

The impact of removal of vitamin A supplementation on marbling appears to depend upon reduction of hepatic retinol stores to sufficiently decrease circulating retinol that may be converted to retinoic acid and impact lipid metabolism. Despite marginal liver concentrations, all steers in the present study started with very adequate serum retinol concentrations of approximately 380 µg/L and ended with concentrations of 390 or 290 µg/L if receiving 2200VA or 0VA, respectively. In addition, steers receiving 0ZN had overall lesser serum retinol concentrations compared to those receiving 120ZN, demonstrating the role of dietary Zn in maintaining serum retinol concentrations, most likely through RBP4. However, hepatic RBP4 mRNA expression did not differ with ZN, VA, or their interaction, suggesting the influence of dietary Zn on circulating retinol may be exerted through post-transcriptional control of RBP4. While liver retinol concentration was not reported by Pyatt et al. (2005), serum retinol was maintained in steers receiving 2300 or 7200 IU/kg DM (347 vs 355 µg/L), demonstrating it is tightly controlled when dietary consumption is adequate. Pickworth et al. (2012) removed vitamin A supplementation for 184 days prior to harvest and noted an 80 µg/L decline in serum retinol, with a final concentration of 200 µg/L and a thirty-point improvement in marbling score, while the serum retinol concentration of steers receiving 3750 IU/kg DM remained stable at approximately 280 µg/L. Yearling steers in Knutson et al. (2020) were backgrounded on a low vitamin A ration and came into the finishing phase with marginal serum retinol (200 µg/L). Steers receiving 2700 IU/kg DM reached a final serum retinol of 400 µg/L, where steers receiving 723 IU/kg DM reached a final serum retinol of ∼300 µg/L, and expressed a forty-point improvement in marbling score. In 0VA steers, liver retinol concentrations did not deplete to the concentrations where others have seen improvements in marbling score, thus serum retinol concentrations were perhaps not decreased enough to fully express marbling differences (Gorocica-Buenfil et al. 2007a; Pickworth et al. 2012; Knutson et al. 2020). Although dose titration studies are needed to confirm, it appears serum retinol concentrations need to be below 250 µg/L in beef steers to impact marbling scores.

The removal of vitamin A supplementation, or vitamin A supplementation below 2200 IU/kg DM did not induce deficiency symptoms in this experiment or other literature discussed herein. However, steers receiving 0VA tended to have reduced G:F compared to 2200VA (0.15 vs. 0.16). Gorocica-Buenfil et al. (2007b) observed a similar trend where G:F was lesser in steers receiving 0 IU/kg DM. In general, G:F in the present experiment was lesser than in other experiments, where cattle were often yearlings or early-weaned and backgrounded at controlled rates of gain, allowing for compensatory gain later (Pyatt et al. 2005; Gorocica-Buenfil et al. 2007a, 2007b; Gorocica-Buenfil et al. 2008). No other differences in live performance metrics were observed in this experiment or those completed by other researchers (Gorocica-Buenfil et al. 2007a, 2007b; Pickworth et al. 2012).

Interestingly, steers in this experiment receiving both Zn and vitamin A supplementation exhibited the greatest lean carcass characteristics with numerically the highest DP and least RF. The differences observed in DP in this experiment are biologically unclear, warranting further work examining the interaction between vitamin A and Zn. Steers receiving 120ZN tended to have a greater percentage of carcasses receiving a YG3 compared to 0ZN (85.3 vs. 69.3). While other vitamin A research does not report specific Zn supplementation rates, most note adherence to NRC guidelines, thus it is likely they compared vitamin A strategies across cattle receiving at least some supplemental Zn. These studies have not noted differences in RF when VA is removed (Gorocica-Buenfil et al. 2007a, 2007b, 2007c; Pickworth et al. 2012) and Pyatt et al. (2005) observed steers who received 7,200 IU/kg DM actually had greater RF than those receiving 2,300 IU/kg DM. The addition of VA to a diet without supplemental Zn seemed to result in fatter steers in the present study, which could be an artifact of random allocation of steers to diets or could be an interaction of these two nutrients worth further exploration.

Interestingly, biological interactions between Zn, vitamin A, and Cu, like those observed in literature outside of domestic livestock (Sundaresan et al. 1996), were observed in liver and plasma of steers in the present experiment. Steers receiving 2200VA had greater plasma Cu concentrations, perhaps explained by the fact that retinoic acid increases ceruloplasmin—the primary transport protein Cu uses to leave the liver and enter the blood stream (Barber and Cousins 1987; Sundaresan et al. 1996). Perhaps because our shifts in plasma and liver Cu were not dramatic, we observed no differences in hepatic mRNA expression of ceruloplasmin (CP), the Cu-export ATPase ATP7B, the Cu chaperone COMMD1.

The antagonistic relationship between Zn and Cu has been heavily researched, where increased dietary Zn induces greater metallothionein (MT) presence in the intestinal epithelium; but MT can bind both Zn and Cu, resulting in Cu being trapped in the enterocyte rather than in circulation or storage (Hall et al. 1979). Steers belonging to 120ZN-2200VA had the least concentration of liver Cu, hypothetically due to increased binding of Cu to MT in the intestine due to greater Zn presence, reducing the amount of Cu available to be stored in the liver, and greater vitamin A presence stimulating ceruloplasmin production, pulling Cu from the liver to enter circulation. Hepatic MT 2A expression numerically increased due to ZN, but was unaffected by VA or the interaction of ZN and VA, suggesting this is not an explanation for the ZN × VA effect on liver Cu. The changes observed in this experiment demonstrate the importance of considering mineral and vitamin nutritional alterations holistically, rather than independently.

The connection between Fe and vitamin A has been heavily researched in humans and rats, particularly due to vitamin A deficiency associated anemia. Vitamin A deficiency impairs the mobilization of Fe from the liver and therefore reduces hemoglobin formation (Evans 2005). Mejia et al. (1979) and Sijtsma et al. (1993) observed increased liver Fe concentrations during periods of vitamin A deficiency or when vitamin A supplementation was lacking. Steers on the 120ZN-0VA treatment expressed the greatest liver Fe concentration, but steers on the 0ZN-0VA treatment expressed the least liver Fe concentration. Researchers would expect for both treatments receiving no supplemental vitamin A to have greater liver Fe concentrations, leading to further research questions of an interaction between vitamin A, Fe, and Zn. Liver Mn was unexpectedly affected by ZN and VA, being greatest in 0ZN-0VA steers. Liver Mn is a function of ZIP and ZnT transporter movement, which may be affected by Zn. Carmichael et al. (2019) noted 120 mg Zn/kg DM from ZnSO₄ also decreased liver Mn, though they did not observe this effect when inorganic and organic Zn were supplemented together.

Collectively, these data demonstrate reductions in vitamin A supplementation may increase carcass quality, but further research is needed to identify the optimal concentration and timing of supplementation strategy shift to both improve marbling and maintain feed efficiency. Lack of differences in hepatic gene expression due to treatments points toward post-translational effects of these nutrients. In addition, the micronutrient data presented in this article demonstrate that dietary manipulation of one or two micronutrients often influence the absorption, metabolism, and storage of other micronutrients—suggesting work is needed to refine our understanding of these relationships.

Glossary

Abbreviations

ADG

average daily gain

BW

body weight

CA

carcass adjusted

DM

dry matter

DMI

dry matter intake

DP

dressing percentage

EBF

empty body fat

EDTA

ethylenediaminetetraacetic acid

G:F

gain to feed ratio

HCW

hot carcass weight

MT

metallothionein

NEFA

non-esterified fatty acids

REA

rib eye area

RF

12th rib fat thickness

TMR

total mixed ration

TRT

treatment

Zn

zinc

Author contributions

Marissa K. Eekhoff (Conceptualization, Data curation, Formal analysis, Investigation, Project administration, Visualization, Writing—original draft), and Stephanie L. Hansen (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing—review & editing)

Supplementary material

Supplementary data is available at Journal of Animal Science online.

Funding

None declared.

Conflict of interest statement. The authors declare no real or perceived conflicts of interest.