**Evaluation of Resistance to Fescue Toxicosis in Purebred Angus Cattle Utilizing Animal Performance and Cytokine Response**

**Daniel H. Poole 1, \*, Kyle J. Mayberry 1,2 , McKayla Newsome 1 , Rebecca K. Poole 1,3 , Justine M Galliou 1 , Piush Khanal 1 , Matthew H. Poore <sup>1</sup> and Nick V. L. Serão 1,4**


Received: 15 October 2020; Accepted: 10 December 2020; Published: 14 December 2020

**Abstract:** Fescue toxicosis is a multifaceted syndrome common in cattle grazing endophyte-infected tall fescue; however, varying symptomatic responses potentially imply genetic tolerance to the syndrome. It was hypothesized that a subpopulation of animals within a herd would develop tolerance to ergot alkaloid toxicity. Therefore, the goals of this study were to develop selection criteria to identify tolerant and susceptible animals within a herd based on animal performance, and then examine responsive phenotypic and cytokine profiles to fescue toxicosis. Angus cows grazed endophyte-infected tall fescue at two locations for 13 weeks starting in mid-April 2016. Forage measurements were collected to evaluate ergot alkaloid exposure during the study. A post hoc analysis of animal performance was utilized to designate cattle into either tolerant or susceptible groups, and weekly physiological measurements and blood samples were collected to evaluate responses to chronic exposure to endophyte-infected tall fescue. Findings from this study support the proposed fescue toxicosis selection method formulated herein, could accurately distinguish between tolerant and susceptible animals based on the performance parameters in cattle chronically exposed to ergot alkaloids, and provides evidence to warrant additional analysis to examine the impact of ergot alkaloids on immune responsiveness in cattle experiencing fescue toxicosis.

**Keywords:** fescue toxicosis; genetic tolerance; cytokines; cow productivity

**Key Contribution:** Through accurate collection of phenotypic performance parameters in a population of cattle chronically exposed to ergot alkaloids, tolerant animals can be identified and selected to enhance productivity in an environment dominated by toxic tall fescue.

#### **1. Introduction**

Fescue toxicosis, resulting from consumption of ergot alkaloids commonly found in endophyte (*Epichloë coenophiala*)-infected tall fescue (*Lolium arundinaceum* Schreb. Darbysh), significantly impacts livestock health and production globally. Direct exposure to ergot alkaloids from tall fescue and other grass species occurs in production systems that heavily rely on grazing, such as in cow calf and stocker programs in the United States, New Zealand, and Australia; whereas importation of feedstuff has led to ergot alkaloid-induced effects in several Asian countries, including Japan and Korea [1–3]. Due to

the agronomic benefits resulting from symbiosis with the endophyte, including stand persistence and drought tolerance, it is expected that endophyte-infected tall fescue will become the primary forage across more growing zones as the global climate warms. McCulley et al. [4] recently reported that warming air temperatures (+3 ◦C, day and night, year-round) significantly increased concentrations of ergovaline and total ergot alkaloids found in endophyte-infected tall fescue, which may exacerbate fescue toxicosis in future years. Complex plant–fungus–environment interactions exist due to variation in endophyte infection rate, alkaloid production (125 to 5000 µg/Kg), and the animal's voluntary intake [4], all of which can lead to high variation in individual animal responses. Individual variation in response to fescue toxicosis might be, in part, due to host genetics. Galliou et al. [5] tested the association between genotypes from a commercial genetic test for fescue toxicosis and performance in pregnant Angus cows. These authors found significant associations with growth, hair shedding, and calf weaning weight. Altogether, these pose a major challenge to understanding the underlying mechanisms of action responsible for the decreases in livestock productivity. While advancements have been made to better understand the mechanisms associated with fescue toxicosis symptoms, such as focusing on improving growth and reproductive performance of beef cattle grazing endophyte-infected tall fescue, a greater understanding is needed to identify markers for tolerance to fescue toxicosis, which is critical for the sustainability of the global beef industry.

Tall fescue, and more specifically the cultivar Kentucky 31, is the most abundant and economically important cool-season perennial grass in U.S. agriculture. Approximately 8.5 million head of beef cattle reside in southeast or central U.S., where fescue is the predominant forage available, most of which contains the endophyte that causes toxicosis in grazing livestock. While precise data on the rate and composition of ergot alkaloid intake in grazing cattle is limited, the symptoms of this multifaceted disease range from acute outbreaks of gangrenous ergotism to more subtle and chronic decreases in livestock productivity characterized by decreased feed intake and growth performance, compromised reproduction (pregnancy and calving rates), elevated body temperature, and reduced blood flow to the extremities, as reviewed by Strickland et al. [1]. While the gangrenous ergotism is rare and most likely due to acute exposure to very high concentrations of ergot alkaloids, combined with cold temperatures, the decreased livestock health, welfare, and productivity is expected as a result of chronic exposure to ergot alkaloids.

While there have been extensive efforts to develop an innovative and strategic pasture and grazing management system to convert toxic tall fescue to the non-toxic "Novel Endophyte Tall Fescue", many farmers have exercised some caution and hesitation to adopt the technology. This appears to be due to lack of understanding of how to renovate, expense, time, challenging terrain, and lack of feed resources to utilize while the new forage is being established. Essentially, the goal of converting all of the toxic fescue to novel endophyte tall fescue is likely impractical and highly unlikely. Therefore, there has been concurrent interest in identifying ways that producers can select animals based on their tolerance to the adverse effects of fescue toxicosis. As previously stated, animals suffering from fescue toxicosis exhibit decreased growth capability, poor reproductive performance, and have a lack of adequate heat tolerance at least in part due to their retained winter hair coat. Focusing on growth performance selection parameters and the animal's cytokine response to the toxin could be utilized to identify animals that are tolerant to the effects of fescue toxicosis. Therefore, the goals of this study were to develop selection criteria to identify tolerant and susceptible animals based on phenotypic animal performance traits, to evaluate how our criteria matches with the information from a commercial test for fescue toxicosis, and then examine responsive cytokine profiles to identify beef cattle displaying tolerance to fescue toxicosis.

#### **2. Results**

#### *2.1. Changes in Forage Characteristics and Ergot Alkaloid Concentration*

Forage fractions during period 1 (P-1) of the experiment were comprised of 72.5 and 72.9 percent fescue at the Butner Beef Cattle Field Lab (BBCFL) and Upper Piedmont Research Station (UPRS) locations, respectively. The percentage of fescue available to these animals during period 2 (P-2) decreased to 56.0 and 63.3 percent at BBCFL and UPRS, respectively (Table 1). Overall, animals that grazed pastures located at BBCFL were exposed to a lower percentage of fescue compared to UPRS. Additionally, tiller samples to determine the endophyte infection rate identified that the BBCFL pastures had a lower infection rate compared to the UPRS pastures throughout the course of the trial, 80.8 compared to 86.3 percent, respectively (Table 1). As a result of the differences in infection rate and the percent of tall fescue in the sward, ergovaline concentrations differed between these locations (Table 1). The highest concentrations of ergovaline were observed in P-1 for both locations (263.3 µg/kg for BBCFL and 436.7 µg/kg for UPRS) and decreased during P-2 (106.7 µg/kg for BBCFL and 196.7 µg/kg for UPRS) by the end of the trial (Table 1). During P-2 at UPRS, ergosine (333 µg/Kg), ergotamine (175 µg/Kg), ergocornine (53 µg/Kg), ergocriptine (150 µg/Kg), and ergocristine (48 µg/Kg) contributed to total ergot alkaloids. These alkaloids are characteristic of Claviceps purpurea [6], which can commonly infect tall fescue and other grasses. While the concentration of ergovaline decreased during P-2, these concentrations (along with the *C. purpurea* alkaloids at UPRS) were still high enough to elicit many of the negative symptoms associated with fescue toxicosis [7,8]. Lastly, the nutritive value and mineral composition of the tall fescue pastures utilized for this study are presented in Supplementary Table S1. Total digestible nutrients, crude protein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF) were not different between the two locations and did not change over the course of the trial.

#### *2.2. Distribution of T-Snip Genotypes Across FTSM Selected Animals*

The expected proportion of T-Snip genotypes between the two FTSM groups can be see in Table 2. There is a tendency for a different distribution of T-Snip genotypes between EI-TOL and EI-SUS animals (*p* = 0.089). There was a statistical difference (*p* < 0.05) for the proportions of T-Snip genotypes 2 and 3 between EI-TOL and EI-SUS animals. EI-TOL animals had a lower proportion of genotype 2 than EI-SUS animals (0.1 ± 0.07 versus 0.368 ± 0.11), whereas the opposite was found for T-Snip genotype 3 (0.6 ± 0.11 versus 0.316 ± 0.11). There were no differences in the proportion of T-Snip genotypes 1 and 4 between these animals (*p* > 0.05). These results indicated that EI-TOL animals had overall greater frequency of greater T-Snip genotypes than EI-SUS animals.


**Table 1.**Endophyte-infected tall fescue characteristics.

1BBCFL: Butner Beef Cattle Field Laboratory location, Bahama, NC; 2 UPRS: Upper Piedmont Research Station location, Reidsville, NC; 3 P-1: Period 1, using values from weeks one, three, and five; 4 P-2: Period 2, using values from weeks 7, 9 and 11; 5 ES: Entire study, using values from weeks 1, 3, 5, 7, 9, and 11; 6 Phytoscreen© immunochemical endophyte test, Agrinostics Ltd. Co., Watkinsville, GA, USA; 7 HPLC analysis at the University of Missouri Veterinary Medical Diagnostic Laboratory (Columbia, MO;USA) [9]; 8 Denotes not detected via HPLC.


**Table 2.** Expected proportion of T-Snip genotypes <sup>2</sup> between tolerant (TOL) and susceptible (SUS) cows in the endophyte-infected (EI) fescue locations.

<sup>1</sup> FTSM: Fescue toxicosis selection method group. There was an effect (*p* = 0.089) of FTSM on the distribution of T-Snip genotypes; <sup>2</sup> Values of T-Snip (AgBotanica, LCC, Columbia, MO, USA) genotypes range from 1 to 5, with expected greater tolerance to fescue toxicosis in animals with greater values; a,b Expected proportions within a T-Snip lacking the same superscript are statistically different from each other (*p* < 0.05).

#### *2.3. Phenotypic Variables of Animal Performance*

Animal performance is negatively impacted during ergot alkaloid exposure, thus, the focus of this study was to identify tolerant and susceptible animals within an exposed population; however, direct comparisons of performance parameters to animals not exposed to ergot alkaloids demonstrated the severity of exposure throughout the trial. Cattle grazing novel endophyte fescue pastures (EN, non-toxic) had greater body weights compared to animals grazing endophyte-infected (EI, toxic) fescue pastures (*p* < 0.0001; Figure 1). By comparison, animals deemed tolerant (EI-TOL) to ergot alkaloid exposure displayed a positive weight gain compared to animals deemed susceptible (EI-SUS) to ergot alkaloid exposure (Figure 1; Table 3; *p* < 0.05). Additionally, cattle grazing novel endophyte fescue pastures had a greater average daily gain (ADG; 0.703 kg/d) and body condition score (BCS; 6.1) compared to EI-TOL and EI-SUS animals (0.497 and −0.003 kg/d, respectively, for ADG and 5.5 and 5.6, respectively, for BCS; Table 3; *p* < 0.0001). Throughout the course of the study, EN and EI-TOL animals displayed similar hair coat scores (HCS) and were significantly less dense compared to EI-SUS animals (Table 3; *p* < 0.0001). Furthermore, EN animals were better able to shed their hair throughout the trial period compared to EI-TOL and EI-SUS animals (1.99 versus 2.36 and 2.64, respectively; Table 3; *p* < 0.0001). Although within the normal body temperature range for cattle, EN animals were cooler compared to EI-TOL and EI-SUS animals as indicated by rectal temperatures (Table 3; *p* < 0.0001).

a† ab†

a† a†

– –

−

≤ 0.05) ≤

–


**Table 3.** Performance parameters for Angus cows consuming either novel endophyte (EN) or endophyte-infected (EI) fescue from late April to late July 2016.

a–c Within the row, means without a common superscript significantly differ (*p* ≤ 0.05); \* 0.05 < *p* ≤ 0.10 determined a statistical tendency; <sup>1</sup> Values are reported as least square means for the experiment; <sup>2</sup> EN: cattle grazing novel endophyte fescue (nontoxic, *n* = 27); EI-TOL: cattle grazing endophyte-infected fescue deemed tolerant through FTSM (*n* = 20); EI-SUS cattle grazing endophyte-infected fescue deemed susceptible through FTSM (*n* = 20); <sup>3</sup> BCS = body condition score (1–9 scale); <sup>4</sup> HCS = hair coat score (1–5 scale); <sup>5</sup> HSS = hair shedding score (1–5 scale).

The average daily gain was significantly greater in EI-TOL animals during P-1 and over the entire study (ES) compared to EI-SUS animals (Table 4); however, no difference in ADG was observed during P-2 (Table 4), which could be reflective of the change in ergovaline concentrations observed in the fescue pastures between P-1 and P-2 of the trial. In addition to ADG, EI-TOL animals displayed a greater positive change in BCS during P-1, P-2, and throughout the entire study compared to the EI-SUS animals (Table 4; *p* < 0.05). Throughout the course of the study, EI-TOL animals displayed greater change in HCS and HSS in P-1 and P-2 compared to EI-SUS animals (Table 4; *p* < 0.05), with the greatest change occurring in P-1 as the temperature and humidity increased. Thus, EI-TOL animals were able to alter their hair coat density and length earlier in the year when ergot alkaloid concentrations were elevated (P-1), reducing the negative impact of fescue toxicosis on animal performance. Interestingly, no differences were observed when examining the change in HCS or HHS between EI-TOL and EI-SUS animals over the entire study (Table 4; *p* > 0.05). When examining differences in HCS and HSS between P-1 and P-2, there was an increase in hair shedding in the EI-SUS animals in P-2, which could be attributed to the decrease in ergot alkaloid concentration in the fescue pastures, in combination with the fact that elevated temperatures and humidity permitted these animals to display similar hair characteristics to that of the EI-TOL animals. However, no differences were observed in the change in body temperatures during P-1, P-2, or over the entire study (Table 4; *p* > 0.05).


**Table 4.** Performance parameters for Angus cattle grazing endophyte-infected fescue deemed either tolerant (EI-TOL) or susceptible (EI-SUS) through FTSM from late April to late July 2016.

a,b Within the row, means without a common superscript significantly differ (*p* ≤ 0.05); \* *p*-values 0.05 < *p* ≤ 0.10 determined a statistical tendency; <sup>1</sup> P-1: Period 1, using data collected from weeks one, three, and five; <sup>2</sup> P-2: Period 2, using data collected from weeks 7, 9, and 11; <sup>3</sup> ES: Entire study, using data collected from weeks 1, 3, 5, 7, 9, and 11; <sup>4</sup> Values are reported as least square means for the experiment; <sup>5</sup> EI-TOL: cattle grazing endophyte-infected fescue deemed tolerant through FTSM (*n* = 20); EI-SUS cattle grazing endophyte-infected fescue deemed susceptible through FTSM (*n* = 20); <sup>6</sup> ∆BCS = change in body condition score (1–9 scale) over the given time frame; <sup>7</sup> ∆HCS = change in hair coat score (1–5 scale) over the given time frame; <sup>8</sup> ∆HSS = change in hair shedding score (1–5 scale) over the given time frame; <sup>9</sup> ∆RT = change in rectal temperature (◦C) over the given time frame.

#### *2.4. Cytokine Response and Hormone Profiles in FTSM Selected Animals* 1β (CCL4) concentrations were not difference between EI

While performance parameters can be directly selected, changes within the peripheral concentrations of growth factors (FGF1, FGF2, IGF1, and VEGFA); chemokines (CCL2, CCL4, CXCL9, CXCL10); cytokine receptor antagonists (IL36RN, GPRASP-1, NCAM1); anti-inflammatory (IL4 and IL13); and pro-inflammatory (IFNAR1, IFNG, IL1A, IL15, IL2, IL21 and TNF) cytokines, as well as reproductive hormones (prolactin and progesterone), in response to ergot alkaloid exposure would potentially lead to a novel marker for identifying animals resistant to fescue toxicosis.

Of the growth factors, vascular endothelial growth factor A (VEGFA) concentrations were greater in EI-TOL animals compared to EI-SUS animals (6.3 ± 1.1 versus 3.2 ± 1.1 ng/mL, respectively; Figure 2B; *p* = 0.0444). No differences were observed in VEGFA concentrations between locations or over the course of the trial (Table 5; *p* > 0.05). No differences were observed in FGF1, FGF2, or IGF1 concentrations between EI-TOL and EI-SUS animals over the course of the trial (Table 5; *p* > 0.05). Fibroblast growth factor 1, but not FGF2 or IGF1, concentrations tended to be greater in animals at UPRS compared animals at BBCFL (328.6 ± 81.6 versus 62.4 ± 81.6 ng/mL, respectively; Table 5; *p* = 0.0793).

**Figure 2.** EI-TOL animals (*n* = 20) had greater concentrations of GPRASP-1 (**A**) and VEGFA (**B**) in peripheral blood compared to EI-SUS animals (*n* = 20) when chronically exposed to endophyte-infected tall fescue (*p* < 0.05).


**Table 5.**Growth factors, chemokines, and cytokine receptor antagonists.

a,b Within the row, means without a common superscript significantly differ (*p* ≤ 0.05); \* *p*-values 0.05 < *p* ≤ 0.10 determined a statistical tendency; 1 Values are reported as least square means for the experiment; 2 EI-TOL: cattle grazing endophyte-infected fescue deemed tolerant through FTSM (*n* = 20); EI-SUS cattle grazing endophyte-infected fescue deemed susceptible through FTSM (*n* = 20); 3 BBCFL: Butner Beef Cattle Field Laboratory location, Bahama, NC; 4 UPRS: Upper Piedmont Research Station location, Reidsville, NC; 5 Growth factors, chemokines, and cytokine receptors antagonist concentrations reported as ng/mL.

No differences were observed in chemokines, CCL2, CXCL9, or CXCL10 concentrations between EI-TOL and EI-SUS animals over the course of the trial (Table 5; *p* > 0.05). Macrophage inflammatory protein-1β (CCL4) concentrations were not difference between EI-TOL and EI-SUS animals (*p* > 0.05), however, CCL4 concentrations were greater in animals at UPRS compared to animals at BBCFL (161.0 ± 24.0 versus 100.7 ± 24.0 ng/mL, respectively; Table 5; *p* = 0.0793) and increased over the course of the trial (10.7 ± 98.9, 202.6 ± 98.9, and 373.1 ± 98.9, for weeks 1, 7, and 13, respectively; Table 5; *p* = 0.0428). Furthermore, CXCL9 and CXCL10 concentrations tended to be greater in animals at UPRS compared animals at BBCFL (132.5 ± 28.5 versus 58.9 ± 28.5 ng/mL, respectively, for CXCL9 and 73.8 ± 11.0 versus 46.9 ± 11.0 ng/mL, respectively, for CXCL10; Table 5).

G Protein-Coupled Receptor Associated Sorting Protein 1 (GPRASP-1) concentrations were greater in EI-TOL animals compared to EI-SUS animals (4.7 ± 0.2 versus 4.1 ± 0.2 ng/mL, respectively; Figure 2A; *p* = 0.0343). No differences were observed in GPRASP-1 concentrations between locations (Table 5; *p* > 0.05), however, GPRASP-1 concentrations were greatest at the end of P-1, compared to the start and end of the trial (5.0 ± 0.2 versus 4.1 ± 0.2 for weeks 1 and 13, Table 5; *p* = 0.0033). No differences were observed in IL36RN or NCAM1 concentrations between EI-TOL and EI-SUS animals over the course of the trial (Table 5; *p* > 0.05). However, IL36RN and NCAM1 concentrations were greater in animals at UPRS compared animals at BBCFL (333.4 ± 57.4 versus 97.1 ± 57.4 ng/mL, respectively, for IL36RN and 35.5 ± 1.5 versus 31.2 ± 1.5 ng/mL, respectively, for NCAM1; Table 5).

No differences were observed in anti-inflammatory cytokine, IL4, or IL13 concentrations between EI-TOL and EI-SUS animals over the course of the trial (Table 6; *p* > 0.05). However, IL4 and IL13 concentrations tended to be greater in animals at UPRS compared animals at BBCFL (218.8 ± 56.2 versus 83.3 ± 56.2 ng/mL, respectively, for IL4 and 12.5 ± 2.2 versus 6.9 ± 2.2 ng/mL, respectively, for IL13; Table 6). No differences were observed in pro-inflammatory cytokine, IFNG, IL1A, IL15, and IL2 concentrations between EI-TOL and EI-SUS animals between locations or over the course of the trial (Table 6; *p* > 0.05). While no differences were observed in IFNAR1 concentrations between EI-TOL and EI-SUS animals or between locations, IFNAR1 concentrations increased over the course of the trial (41.2 ± 62.4, 165.4 ± 62.4, and 411.9 ± 62.4, for weeks 1, 7, and 13, respectively; Table 6; *p* = 0.0001). Additionally, no differences were observed in IL21 or TNF concentrations between EI-TOL and EI-SUS animals over the course of the trial (Table 6; *p* > 0.05), but concentrations of these cytokines tended to be greater in animals at UPRS compared animals at BBCFL (135.3 ± 24.6 versus 73.7 ± 24.6 ng/mL, respectively, for IL21and 149.4 ± 29.6 versus 67.2 ± 29.6 ng/mL, respectively, for TNF; Table 6).


**Table 6.**Anti-inflammatory and pro-inflammatory cytokines and reproductive hormones.

a,b Within the row, means without a common superscript significantly differ (*p* ≤ 0.05); \* *p*-values 0.05 < *p* ≤ 0.10 determined a statistical tendency; 1 Values are reported as least square means for the experiment; 2 EI-TOL: cattle grazing endophyte-infected fescue deemed tolerant through FTSM (*n* = 20); EI-SUS cattle grazing endophyte-infected fescue deemed susceptible through FTSM (*n* = 20); 3 BBCFL: Butner Beef Cattle Field Laboratory location, Bahama, NC; 4 UPRS: Upper Piedmont Research Station location, Reidsville, NC; 5 Anti-inflammatory and pro-inflammatory cytokines and reproductive hormone concentrations reported as ng/mL.

Of the reproductive hormones examined, prolactin (PRL) concentrations tended to be greater in EI-SUS animals compared to EI-TOL animals (138.4 ± 16.0 versus 98.3 ± 16.0 ng/mL, respectively; Table 6; *p* = 0.0787). Additionally, PRL concentrations tended to be greater in animals at UPRS compared animals at BBCFL (137.6 ± 15.8 versus 99.1 ± 15.8 ng/mL, respectively; Table 6), and was significantly less at the start of the trial compared to weeks 7 and 13 of the trial (64.0 ± 19.4 versus 163.4 ± 19.4 and 127.6 ± 19.4 for weeks 7 and 13, respectively; Table 6; *p* = 0.0019); progesterone (P4) concentrations tended to be greater in EI-TOL animals compared to EI-SUS animals (3.3 ± 0.14 versus 3.0 ± 0.14 ng/mL, respectively; Table 6; *p* = 0.0895). Additionally, P4 concentrations were greater in animals at BBCFL compared animals at UPRS (3.4 ± 1.4 versus 2.8 ± 1.4 ng/mL, respectively; Table 6) and was significantly less at week seven of the trial compared to weeks 1 and 13 of the trial (2.2 ± 0.3 versus 3.5 ± 0.3 and 2.8 ± 0.2 for weeks 1 and 13, respectively; Table 6; *p* = 0.0003).

#### **3. Discussion**

Due to the extensive use of endophyte-infected tall fescue across the globe, the goals of this study were to develop selection criteria to identify tolerant and susceptible animals based on phenotypic animal performance traits, and then examine responsive cytokine profiles to identify beef cattle displaying tolerance to fescue toxicosis. Several studies described breed differences in response to ergot alkaloids, in which breeds such as Senepol and Brahman, which are better able to handle the heat and humidity, outperformed British breeds such as Angus and Hereford [8,10–12] when exposed to toxic tall fescue. These breeds have been used in crossbreeding programs to address fescue toxicosis, but it is unclear if their apparent tolerance is to fescue or adaptation to elevated heat and humidity. However, genetic progress is obtained within the breed, and the identification of genetic variation in animals from the same breed should be exploited to allow for selection for improved response (e.g., tolerance) to fescue toxicosis. Gray et al. [13] examined Angus cattle performance in North Carolina and Mississippi on endophyte-infected fescue pasture. These researchers observed genetic variation for hair coat shedding, indicating that Angus cattle that shed their winter hair coat earlier in the year may be more heat tolerant.

At the genomic level, genes regulating prolactin production have been targeted to identify genetic markers for tolerance to fescue toxicosis. Campbell et al. [14] recently identified a single nucleotide polymorphism (SNP) within the dopamine receptor D2 gene that was associated with variation in calving rates when grazing endophyte-infected tall fescue. In addition, using part of the same data in this study, Galliou et al. [5] showed that a commercial genetic test for fescue toxicosis, T-Snip (AgBotanica, LCC, Columbia, MO, USA), is associated with growth, hair shedding, and calf weaning weight in pregnant Angus cows. With the available T-Snip data, we tested whether the distribution of T-Snip genotypes changed between EI-TOL and EI-SUS animals in our study. According to the company's instructions, T-Snip genotypes range from 1 to 5, with levels of tolerance to fescue toxicosis increasing with the value of the genotype. Although both groups of animals showed T-Snip genotypes 1 to 4, there was a greater proportion of genotype 3 in EI-TOL than in EI-SUS. Similarly, there was a greater proportion of genotype 2 in EI-SUS than in EI-TOL. Therefore, the expected genetic values based on this commercial genetic test was greater in EI-TOL than in EI-SUS. Hence, although the selection of animals in this study was fully based on their phenotypic performance, results from T-Snip genotypes further support that this selection method was able to identify animals with contrasting tolerance to fescue toxicosis. It is important to note that this commercial genetic test does not have a perfect accuracy to predict with phenotypic performance in animals during fescue toxicosis [5]. Therefore, the presence of animals with lower and greater genotype values in EI-TOL and EI-SUS groups, respectively, was expected. Considering these factors, developing a selection method based on the phenotypic performance parameter in cattle from a single breed source in controlled experiments would aid in identifying animals that are potentially tolerant or susceptible to ergot alkaloids.

A major challenge with studying this multifaceted syndrome is that endophyte production of ergot alkaloids varies with season and plant maturity, and contributes to variation in ergot alkaloid intake by grazing cattle. Total ergot alkaloids or ergovaline, the most abundant alkaloid, has been measured in fescue to describe its potential toxicity. Concentrations of ergovaline increase from 250 to 500 µg/Kg in leaf blades and from 500 to 1300 µg/Kg in leaf sheaths from April to May. Seed heads contain the greatest concentration of toxins, and reach concentrations as high as 5000 µg/Kg in June. Ergovaline concentrations decline in August and increase again during fall regrowth [9]. Total ergot alkaloid concentration displays the same seasonal changes as ergovaline [15]. The concentration of total ergot alkaloids and ergovaline declines by 81% to 85%, respectively, from December to March in stockpiled fescue (accumulated during the growing season for grazing during dormancy; [16,17]). There were differences in the percentage of fescue and endophyte infection rate between locations used in this study, which then led to differences in ergovaline concentrations between locations, as well as variation in ergovaline concentration throughout the trial. The increase in ergot alkaloid consumption during period 1 resulted in greater differences in the phenotypic performance parameter, including average daily gain (ADG), body condition scores (BCS), body weight (BW), and rectal temperatures that were evaluated in this study.

The difference in ADG and BCS in response to increased consumption of ergot alkaloids is similar to data reported Thompson et al. [18], which identified a direct negative correlation to an animal's ADG when the pasture infection rate increased. Moreover, comparison of phenotypic performance parameters of animals exposed to ergot alkaloids (EI) to animals not exposed to ergot alkaloids (EN) demonstrated the severity of exposure throughout the trial. Additionally, sufficient body condition has been shown to be one of the most important traits to cattle reproductive efficiency in a cow–calf herd, and decreases as a result of endophyte-infected tall fescue consumption [19].

Aside from animal growth performance, issues with thermotolerance have also frequently been seen in animals suffering from severe cases of fescue toxicosis [20,21]. Many of the symptoms of fescue toxicosis (ergot alkaloid exposure) are amplified during a period of heat stress, and the greatest loss in animal performance and consequently increased production losses occur during the summer months when grazing endophyte infected tall fescue [22]. It has been speculated that heat tolerant (i.e., Bos indicus) breeds of cattle have improved resistance to fescue toxicosis [10,23]. Several studies, including Poole et al. [8], have demonstrated that cattle with adaptions to manage heat stress perform better when exposed to ergot alkaloids and increased environmental temperatures. However, there is conflicting data that report no difference in growth rate [24], hormone concentration [23], or milk production [11] between Bos indicus and Bos taurus cattle grazing endophyte-infected tall fescue. Therefore, it remains unknown if the traits for heat tolerance and tolerance to ergot alkaloids are synonymous.

Correlations have been previously made in which cattle that shed their winter hair coats earlier in the fescue season suffer less from the negative effects of fescue toxicosis [13]. In this study, weekly hair shedding scores (HSS) and hair coat scores (HCS) were taken to evaluate differences based on tolerance designation week to week, as well as initial rate of shedding. Both measures were significantly different between tolerant and susceptible animals grazing the endophyte-infected tall fescue. Both HCS and HSS were similar at the start of the study, and deviation between EI-TOL and EI-SUS animal occurred at week three of the study and continued through week seven (P-1). Interestingly, HSS and HCS re-converged to similar numeric values when ergot alkaloid concentrations decreased, yet ambient temperatures and humidity remained elevated consequentially when prolactin concentrations increased. Decreases in serum prolactin have been associated with fescue toxicity, and are often used as an indicator of ergot alkaloid exposure in cattle [25]. However, increased prolactin concentrations are associated with increasing day lengths and ambient temperatures, and initiate hair shedding [26,27]. Prolactin secretion has been linked with changes in environmental temperature, such that prolactin concentrations are elevated during warmer verses cooler months [28,29], and has been shown to regulate hair shedding [30,31]. It has been speculated that fescue toxicosis-induced hypoprolactinemia prevents shedding of the winter hair coat, therefore, cattle have an elevated body temperature and increased vulnerability to heat stress [32]. In the current study, EI-TOL animals had a

tendency for lower serum prolactin concentrations when compared to EI-SUS animals as determined by the FTSM, thus, it can be assumed that those animals deemed tolerant had sufficient concentrations of serum prolactin to initiate hair shedding. Based on the other physiological measurements, the cattle in this study displayed many other symptoms of fescue toxicosis, and the lack of hypoprolactinemia was unexpected. The interaction of increased THI and exposure to ergot alkaloids impacts the severity of fescue toxicosis (reviewed by one), and the relatively low inclusion rate of the ergot alkaloids in the infected pastures may have altered serum prolactin concentrations. With EI-TOL animals having lower average serum prolactin concentrations than EI-SUS animals, it is hypothesized that these cattle were genetically predisposed to eliciting a shedding response at lower thresholds of prolactin. Using the same animals from this study, Koester et al. [33] showed that EI-TOL and EI-SUS animals have distinct fecal bacterial and fungal communities, further supporting that the proposed FTSM also results in differences in the microbiome of animals under FT stress. This classification using FTSM is further supported by the differences in T-Snip genotypes between EI-TOL and EI-SUS presented in the current study. These hypotheses build from the speculation of Aiken et al. [32] that genetic predispositions exist for necessary prolactin levels to initiate shedding.

More recent studies in various livestock species have reported that immune parameters have moderate to high heritability, as well as high genetic correlation with reproductive performance (lowly heritable traits), indicating that immune parameters can be used as a genetic trait for selecting animals that are resistant to specific diseases [34]. Using the protein array approach, numerous growth factors, chemokines, cytokine receptor antagonists, and anti-inflammatory and pro-inflammatory cytokines were examined during ergot alkaloid exposure. One of the major outcomes of this analysis is that many of the factors examined were significantly different or showed a statistical tendency to be higher at UPRS compared to BBCFL (Tables 5 and 6). While cytokine differences between locations was not the primary focus of this study, these data demonstrate the immune system's responsiveness to varying ergot alkaloid exposure, and warrants further investigation with known concentrations of ergovaline to better understand this interaction. Based on these data, it is hypothesized that cattle at the UPRS location experienced a greater immune response in response to higher ergot alkaloid exposure. A similar report described a hyperactive innate immune response, which may lead to an immuno-compromised animal in stocker steers when chronically exposed to ergovaline [35].

In the current study, vascular endothelial growth factor A (VEGFA) and G Protein-Coupled Receptor Associated Sorting Protein 1 (GPRASP-1) concentrations were greater in EI-TOL animals compared to EI-SUS animals. G Protein-Coupled Receptor Associated Sorting Protein 1 has been associated with downregulation of a variety of G Protein-Coupled receptors, including the D2-dopamine receptor, through lysosomal degradation. The D2-dopamine receptor has been shown to play a direct effect in cattle prolactin secretion, and is also involved in ergot alkaloids exposure's decrease in prolactin secretion [36,37]. Prolactin is decreased when dopamine binds to the D2-dopamine receptor, and the cyclic ring structure of ergovaline closely mimics the ring structure of dopamine, allowing ergovaline to bind the D2-dopamine receptor and thus inhibit prolactin secretion [38,39]. In the current study, EI-TOL had greater GPRASP-1 concentrations, potentially causing the D2-dopamine receptor to be downregulated at a greater rate. This decrease in the prevalence of the D2-dopamine receptor decreased the opportunity for ergovaline to bind and decrease prolactin secretion, giving EI-TOL animals an advantage when grazing endophyte-infected tall fescue. Furthermore, this could contribute to EI-TOL animals having greater sensitivity to prolactin through downregulation of the entire prolactin secretion pathway, resulting in lower biological thresholds needed to initiate the biological roles of the hormone, which was previously discussed.

Vascular endothelial growth factor A is a known vasodilator, and has been shown to increase micro-vascular permeability. In addition, it is known to play a particularly important role in vascular endothelial cells, where higher concentrations increase vascular permeability [40] and could provide protective effects against ergot alkaloid exposure. Vasoconstriction caused by exposure to ergot alkaloids further compounds the issues associated with fescue toxicosis, such as its ability to decrease the animal's capacity for evaporative cooling [41]. Increased VEGFA concentrations and lower rectal temperature observed in the EI-TOL animals provide evidence that the EI-TOL animals selected using the FTSM have a greater ability to avoid heat stress and the vasoconstrictions effects of fescue toxicosis compared to EI-SUS animals. Several studies have examined changes in VEGFA to heat stress, and have focused on tissue and/or cellular responsiveness as opposed to whole animal changes. Vascular endothelial growth factor concentrations in the blood of heat tolerant (Bos Indicus) cattle compared to cattle that are more sensitive to heat stress (Bos Taurus) has not been directly investigated; however, Jyotiranjan et al. [42] and Iqbal et al. [43] reported increases in VEGF gene expression in response to thermal stress in goats and Bos indicus cattle breeds, respectively. Interestingly, of the eight Bos indicus breeds examined, only two breeds (Bhagnari and Lohani) displayed increased VEGF expression, but the authors indicate that this increase in VEGF is associated with adaptation to high altitude as opposed to heat tolerances [43]. Additionally, Jones et al. [44] and Aiken et al. [45] have speculated that decreases in serum progesterone concentrations could be caused by vasoconstriction of blood flow to the ovary in cattle consuming ergot alkaloids. Poole and colleagues [7] confirmed that ergovaline exposure reduced the diameter of the ovarian artery, leading to the functional corpus luteum, thus decreasing circulating progesterone concentrations. In the current study, EI-TOL animals tended to have higher progesterone concentrations compared to EI-SUS animals, which may be linked to greater ovarian artery diameter due to greater concentrations of circulating VEGFA in these animals. Thus, the EI-TOL animal's ability to synthesize higher concentrations of VEGFA may provide the capacity to increase vessel diameter, which would result in substantial mitigation of the vasoconstriction seen with fescue toxicosis. Ultimately, further investigation is needed to evaluate these cytokines and their role in fescue toxicosis, as well as the tolerant animal's ability to mount a stronger immune response to mitigate some of the negative effects of fescue toxicosis.

Taken together, the selection strategy used in this study (i.e., FTSM) was showed to be effective in identifying groups of individuals expressing contrasting responses to ergot alkaloid exposure, i.e., fescue toxicosis [33]. Nonetheless, three items must be further discussed: first, the selection of these animals was based solely on their (adjusted) phenotypic performance. Although the T-Snip data presented in this study support that TOL animals have greater genetic tolerance to ergot alkaloid exposure than SUS animals, the accuracy of identifying individual differences from ergot alkaloid exposure using this test was not 100% [5]. Hence, there is still genetic variation in the animal's response to ergot alkaloid exposure that could be explored. One way of getting ahold of individual genetic variation is through using, for example, the expected progeny differences (EPDs) for the growth rate of these animals as part of the selection criteria. However, there are some limitations on using the current EPDs for this purpose. These EPDs are based on a nationwide genetic evaluation (American Angus Association; http://www.angus.org/), which does not take into consideration the presence of GxE, which seems to be the case for ergot alkaloid exposure and for other stress-related traits [46]. Thus, the use of EPDs for this purpose may bias the selection of animals with greater tolerance to ergot alkaloid exposure. Although no source of genetic information was used in our selection process, it was expected that our approach captured part of the genetic potential of these animals. Assuming a) unbiased estimates for the fixed effects and b) independence between fixed and random effects included in the statistical model used for selection of TOL and SUS animals (see details in Koester at al. [33]), the estimated residuals that were used for classification of TOL and SUS animals should include both genotypic effects (additive and non-additive) and the true residual effects. Thus, assuming at least a moderate narrow-sense heritability for growth rate under ergot alkaloid exposure, it is expected that the estimated residuals include a substantial contribution of additive genetic values. Hence, although the selection process used in this study did not include prior genetic information from these animals, the data presented in this study demonstrate that TOL and SUS animals differed, at some level, in their genetic potential when exposed to ergot alkaloids.

Additionally, another consideration is the potential confounding between the growth rate in the presence or absence of ergot alkaloids. In this study, it was assumed that animals expressed their growth as a function of response to ergot alkaloid exposure. In fact, the difference in results obtained between the two locations with endophyte-infected (toxic) fescue suggests that some minimum concentrations of ergot alkaloids are needed for animals to express a tolerant-related phenotype. In order to obtain accurate information on the relationship between the growth rate in the presence or absence of ergot alkaloids at the genetic level, a much larger sample size, with animals from a substantial number of sires represented across locations with varying concentrations of ergot alkaloids, would be needed to estimate the genetic correlation between the growth rate in the presence or absence of ergot alkaloids. Albeit needed, such a scenario would be difficult to achieve in a timely manner. At the phenotypic level, the use of biomarkers for ergot alkaloid exposure, such as prolactin [25] or those identified in our study (e.g., VEGFA), could provide additional information on the level of ergot alkaloid exposure on animals. This information could be used to evaluate whether a significant physiological response to ergot alkaloid exposure was observed in each animal, potentially allowing for better separation between animals showing a growth rate in the presence or absence of ergot alkaloids.

Lastly, our method was unidimensional, using only growth rate data for the selection of TOL and SUS animals. Hence, the evaluation of a fescue toxicosis selection index using multiple parameters and including other sources of information could be beneficial for more accurate selection of animals for response to ergot alkaloid exposure. Additional sources of information include the use of T-Snip, the biomarkers identified in this population (this study and Koester et al. [33]), genotypes at the dopamine receptor D2 gene locus [14], and more. Opportunities exist for the identification of additional biomarkers based on response-related traits other than growth rate, as used in this study. There is also a need for large-scale genomic studies for the identification of SNPs associated with variation in ergot alkaloid exposure related traits, which could be used to better identify animals with different genetic potential for fescue toxicosis.

#### **4. Conclusions**

Taken together, these data provide support to validate the fescue toxicosis selection method (FTSM) proposed in this study through accurate collection of phenotypic performance parameters in a population of cattle chronically exposed to ergot alkaloids, and provide evidence to warrant additional analysis on the impact of ergot alkaloids on immune responsiveness in cattle that experience fescue toxicosis. Given the increased performance and hair shedding ability, greater hormone regulation and efficiency, and stronger cytokine responses, EI-TOL animals have clear mechanisms available to provide these advantages that can and should be selected for in animals raised in a fescue-dominant environment.

#### **5. Materials and Methods**

#### *5.1. Animal Usage*

This study was conducted at two locations in the piedmont of North Carolina, which were the Butner Beef Cattle Field Laboratory (BBCFL) in Bahama, NC and the Upper Piedmont Research Station (UPRS) in Reidsville, NC. All animal procedures were approved by the North Carolina State University Institutional Animal Care and Use Committee (NCSU IACUC #13-093-A; #17-043-A). Approval for animal use from the University Institutional Animal Care and Use Committee was on 19 August 2013 (#13-093-A) and was renewed and approved on 23 March 2017 (#17-043-A).

#### Animal Management

Cow performance and forage data were collected from late April to late July 2016 (Figure 3) when ergot alkaloid concentrations are greatest. Purebred Angus cows (*n* = 148) that ranged from two to four years of age all grazed endophyte-infected tall fescue during the entire experimental period. Cattle selected for this study were confirmed pregnant at 30 days post artificial insemination via ultrasonography, and were approximately 85 days post-conception at the start of the study. Additionally, cows selected for this study were weaned two weeks prior to the start of the study to remove the effect of lactation on animal performance. Cattle were offered ad libitum water and free choice minerals throughout the duration of the study, in addition to natural shade structures within the pastures. Additionally, a subset of cows (*n* = 27) were maintained under the same conditions on novel endophyte fescue pastures (non-toxic, EN) at the BBCFL to serve as a representative control group. Body weight (BW), body condition scores (BCS, as adapted from Richards et al., [47]), hair shedding score (HSS, as adapted from [13]), hair coat score (HCS, adapted from [48]), rectal temperatures, and jugular blood samples were collected weekly to evaluate the animal's physiological response to ergot alkaloid exposure. Objective scores of BCS, HSS, and HCS were all collected by two trained evaluators and composited for an average score for each animal. Blood samples were collected via jugular venipuncture using 20-gauge needles and sterile 10.0 mL vacutainer tubes that contained no additive (Becton Dickerson, Franklin Lakes, NJ, USA). All blood samples were immediately placed on ice and then transported to the laboratory for processing. Whole blood samples were centrifuged the afternoon following collection each week for 25 min at 1500× *g* at 4 ◦C, then serum was drawn from vacutainer tubes and aliquoted into a glass dram vial and a plastic micro-centrifuge tube and stored at −80 ◦C until progesterone, prolactin, and cytokine analyses were conducted. −

**Figure 3.** Experimental timeline used during cow performance phase of data collection from late-April to late-July 2016.

#### *5.2. Fescue Toxicosis Selection Method (FTSM)*

The classification of animals into tolerance (TOL) and susceptible (SUS) using the proposed Fescue Toxicosis Selection Method (FTSM) is fully described in Koester et al. [33]. In summary, growth data on each animal (Figure 3) was used to estimate the slope of the regression analysis of BW on weeks (average weekly gain; AWG, Figure 4A). This was performed based on three window periods: weeks 1 through 13 (ES), weeks 1 through 7 (P-1), and weeks 7 through 13 (P-2) to assess the effect of the increase in temperature from April to July, availability of forage (see Supplementary Table S1), and exposure of infected tall fescue (Table 1). The residuals from the analysis of AWG in a model including the fixed effects of location, parity, and initial body weight (covariate) were used to identify the window period in which the variation on the data was the largest, indicating the impact of the syndrome [46]. Results from this analysis indicated that data from P-1 resulted in the greatest residual variance [33]. Thus, the forty selected animals, those with the 10 most positive and 10 most negative residuals at each location, representing the TOL and SUS groups, respectively, from P-1 were used to evaluate performance and resistance to fescue toxicosis (Figure 4B).

−

**Figure 4.** Graphical representation of the Fescue Toxicosis Selection Method (FTSM) to select the top 20 tolerant (EI-TOL) and 20 susceptible (EI-SUS) animals for further investigation within the entire population. (**A**) Displays the ranking of growth performance for the entire population (*n* = 149) observed over the 13-week period. (**B**) Displays the deviation in growth performance between the tolerant (*n* = 20) and susceptible (*n* = 20) animals over the 13-week period.

#### *5.3. T-Snip Genotyping*

A detailed description of the genotyping of animals in this study for T-Snip (AgBotanica, LLC, Columbia, MO, USA) can be found in Galliou et al. [5]. In summary, blood cards were collected on each animal and shipped to GeneSeek (Neogen Genomics, Lincoln, NE, USA) for T-Snip genotyping. There were two, one, six, and one animals from EI-TOL at BBCFL for genotypes 1 to 4, respectively. The distribution of these T-Snip genotypes was one, five, three, and one, respectively, for EI-SUS at BBCFL; one, one, six, and two, respectively, for EI-TOL UPRS; and four, two, three, and zero, respectively, for EI-SUS UPRS cows.

#### *5.4. Forage Management*

Cattle at both locations grazed endophyte-infected tall fescue (toxic) pastures throughout the 13-week study. Cattle were rotationally grazed every two weeks at each location to ensure sufficient forage utilization and availability. Composite forage samples were taken from each pasture every two weeks to evaluate the nutrient quality and percentage of forage species available. Forage samples were clipped from approximately 20 locations in each pasture and composited. Fescue tiller samples were collected in November of 2016 to evaluate the pasture endophyte infection rate of the fescue.

Nutrient quality samples were submitted within 24 h of collection for nutrient content, and then the results were averaged by experimental period (North Carolina Department of Agriculture Forage Laboratory, Raleigh, NC; USA, see Supplementary Table S1). From the same composite samples, forage was hand-separated in house by trained technicians to determine the percentage of fescue in relation to other various forage species (Table 3) making up forage dry matter. Collected fescue tiller samples were collected, rinsed, and shipped on ice to determine the pasture infection rate, and the average infection rate is reported by experimental period (Agrinostics Ltd. Co., Watkinsville, GA; USA, Table 3). Subsequent HPLC analysis of the forage samples for ergot alkaloid concentrations (MU Veterinary Medical Diagnostic Lab; [9]) demonstrated the change in ergovaline concentrations between locations and over the course of the grazing period (Table 3).

#### *5.5. Serum Assays and Analysis*

Serum progesterone concentrations (P4) were analyzed on the FTSM, EI-TOL (*n* = 20), and EI-SUS (*n* = 20) animals for both locations at weeks 1, 3, 5, 7, 9, 11 and 13, which represent a bi-weekly measure during the entirety of the collection period. Concentrations were determined by a commercially available radioimmunoassay, the Immuchem Coated Tube Progesterone I125 RIA assay (ICN Pharmaceuticals, Inc., Costa Mesa, CA), as previously described by Lyons et al. [49]. Concentrations are reported in ng per mL, and the interassay and intraassay variation was 7.4% and 4.3%, respectively.

Serum prolactin concentrations (PRLs) were analyzed on the FTSM, EI-TOL (*n* = 20), and EI-SUS (*n* = 20) animals for both locations at weeks 1, 7, and 13, which represent the beginning, midpoint, and end of the collection period. Concentrations were determined by a commercially available Bovine Prolactin ELISA assay (MyBioSource, San Diego, CA, USA), as previously described by Poole et al. [8]. Concentrations were reported in ng per mL, and the interassay and intraassay variation was 11.7% and 4.2%, respectively.

Serum cytokine concentrations were analyzed on the FTSM, EI-TOL (*n* = 20), and EI-SUS (*n* = 20) animals for both locations at weeks 1, 7, and 13, which represent the beginning, midpoint, and end of the collection period. Cytokine concentrations were determined using a commercially available Bovine Cytokine Array (Quantibody® Cytokine Arrays; RayBiotech, Inc., Peachtree Corners, GA, USA). The Q1 array permitted detection of interferon alpha (IFNAR1), interferon gamma (IFNG), interleukin (IL)-13, IL-1A, IL-36Ra, IL-21, chemokine CXCL10, CXCL9, CCL4, and tumor necrosis factor alpha (TNF), whereas the Q2 permitted detection of fibroblast growth factor (FGF)1, FGF1, G Protein-Coupled Receptor Associated Sorting Protein (GPRASP)-1, insulin like growth factor (IGF)-1, IL-2, IL-4, IL-15, CCL2, neural cell adhesion molecule 1 (NCAM1), and vascular endothelial growth factor (VEGFA). All bovine cytokine arrays were conducted according to the manufacturer's recommendations. Following processing, all slides were returned to the manufacturer for laser scanning and data extraction. From the absorbance values provided by the manufacturer, serum concentrations were determined using a software (Q-Analyzer Software for QAB-CYT; RayBiotech, Inc., Peachtree Corners, GA, USA), and concentrations are reported in ng per mL for each individual cytokine.

#### *5.6. Statistical Analysis*

Performance and cytokine data were analyzed using the MIXED procedure of SAS 9.3 [50] with repeated measures. The individual animal was utilized as the experimental unit and the model for body condition scores, positive change in body condition scores, body weight, average daily gain, rectal temperature, hair coat score, hair shedding score, and hormone concentrations, and included Fescue Toxicosis Selection Method outcome (EI-TOL vs. EI- SUS), location (BBCFL vs. UPRS), trial period (P-1, P-2, and ES), and all respective interactions. For cytokine response data analysis, the individual animal was utilized as the experimental unit and the model for each cytokine response included treatment (EI-TOL vs. EI- SUS), location (BBCFL vs. UPRS), and time (Weeks 1, 7, and 13). The distribution of T-Snip genotypes between EI-TOL and EI-SUS was analyzed using a logistic multinomial model using the T-Snip genotype of animals as the response variable and the FTSM group (EI-TOL and EI-SUS) as a fixed effect in the model. This analysis was carried out in R [51]. Results were recorded as least square means ± SEM, where statistical significance was reported at *p* ≤ 0.05 and a statistical tendency at 0.05 ≤ *p* ≤ 0.10.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6651/12/12/796/s1, Table S1. Nutritive Value (on a DM basis) of forage samples.

**Author Contributions:** Conceptualization, D.H.P., N.V.L.S and M.H.P.; methodology, D.H.P., and N.V.L.S.; software, P.K., and N.V.L.S; validation, R.K.P,; formal analysis, K.J.M., R.K.P., and D.H.P.; investigation, K.J.M., M.N., J.M.G., and R.K.P.; resources, D.H.P.; data curation, K.J.M., M.N., and R.K.P.; writing—original draft preparation, K.J.M., and D.H.P.; writing—review and editing, K.J.M., D.H.P., and M.H.P; visualization K.J.M.; supervision, D.H.P., and N.V.L.S.; project administration, D.H.P., and N.V.L.S; funding acquisition, D.H.P., and N.V.L.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the North Carolina Cattlemen's Association, North Carolina Agricultural Foundation, Inc., and Hatch Project 02420

**Acknowledgments:** The authors thank Harrison Dudley, Greg Shaeffer, and the rest of the Butner crew (Butner Beef Cattle Field Laboratory, Bahama, NC), as well as Joseph French and the rest of the UPRS crew (Upper Piedmont Research Station, Reidsville, NC) for their consistent help with animal care. We thank Ruby Monn, Garrett Williams, and Anna Smith for all their help with data collection and laboratory analysis.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Effects of Tall Fescue Endophyte Type and Dopamine Receptor D2 Genotype on Cow-Calf Performance during Late Gestation and Early Lactation**

**Sarah A. Wilbanks 1 , Susan Maggie Justice 1 , Thomas West 1 , James L. Klotz 2 , John G. Andrae 3 and Susan K. Duckett 1, \***


**Abstract:** Grazing endophyte-infected, toxic tall fescue reduces cow/calf production; therefore, this study examines alternate strategies such as use of novel endophyte fescue varieties during late gestation and early lactation or genetic selection of resistant cows. Pregnant cows (n = 75) were randomly assigned to fescue endophyte type: 1) endophyte-infected ergot alkaloid producing tall fescue (E+) or 2) novel endophyte-infected, non-toxic tall fescue (NOV) within maternal (A|A, n = 38 and G|G, n = 37) *DRD2* genotype to examine changes in cow/calf performance and milk production during late gestation and early lactation. Grazing E+ fescue pastures during late gestation reduced cow body weight gain but did not alter calf birth weight compared to NOV. Milk production and calf ADG during the first 30 day of lactation were lower for E+ than NOV. The calving rate was reduced, but not calving interval for E+ cows. The adjusted 205-day weight of calves was lower in those grazing E+ with their dams compared to NOV. There were no interactions between *DRD2* genotype and fescue endophyte type indicating that genotype was not associated with response to E+ fescue in this study. Overall, grazing NOV tall fescue pastures rather than E+ during critical stages of production improved cow gain during late gestation, calving rate, early milk production and calf growth.

**Keywords:** fescue toxicosis; single nucleotide polymorphism; milk production; calf growth

**Key Contribution:** Grazing novel, nontoxic tall fescue pastures during critical stages of cow/calf production can reduce losses associated with reduced cow weight gain during last trimester of gestation, milk production during early lactation and pre-weaning calf growth rates.

#### **1. Introduction**

In the southeastern United States, tall fescue [*Lolium arundinaceum* (Schreb.) Darbysh; *Schedonorus phoenix* (Scop.) Holub] is the dominant cool season, perennial forage available for cow/calf production systems [1–3]. Kallenbach [2] estimates that about 12 million beef cows graze tall fescue in this region with reproductivea and liveweight losses estimated at USD 267 per cow in 2015. Most tall fescue contains an endophyte (*Epichloë coenophialia*) that produces ergot alkaloids, a class of mycotoxins that aid the plant in establishment and drought tolerance but is the causative agent for fescue toxicosis when ingested in livestock [1–7]. Fescue toxicosis is a syndrome that includes vasoconstriction, high body temperature, low heart rate, increased blood pressure, low serum prolactin, agalactia, reduced forage intake, and poor weight gains observed in livestock grazing endophyteinfected, ergot alkaloid producing tall fescue [3,5,8]. Ergovaline is mostly responsible for

**Citation:** Wilbanks, S.A.; Justice, S.M.; West, T.; Klotz, J.L.; Andrae, J.G.; Duckett, S.K. Effects of Tall Fescue Endophyte Type and Dopamine Receptor D2 Genotype on Cow-Calf Performance during Late Gestation and Early Lactation. *Toxins* **2021**, *13*, 195. https://doi.org/10.3390/ toxins13030195

Received: 11 February 2021 Accepted: 5 March 2021 Published: 9 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

vasoconstriction of vasculature in beef cattle consuming endophyte-infected tall fescue [8]. Grazing endophyte-infected tall fescue reduces cow reproductive rates [9–11], lactation and calf weaning weights [3,5,12,13]. Varieties of tall fescue containing novel endophytes are now available that do not produce ergot alkaloids [14] but have a similar agronomic response as E+ [15,16] and are able to persist under grazing conditions [17]. Research has shown that grazing novel tall fescue instead of toxic endophyte-infected (E+) improves calf birth [18] and weaning weights [18,19], and stocker cattle performance [20,21]. Roberts and Andrae [3] suggested that one solution is to renovate and replace toxic tall fescue pastures with novel endophyte varieties; however, producer adoption of the novel varieties has been slow. Others have shown that grazing novel varieties during critical stages of production can improve overall performance and profitability in beef cow/calf production systems [19]. Critical stages in cow/calf production are during the last trimester of gestation when over 75% of fetal growth occurs [22], early lactation when calf growth rates are the highest [23,24], and rebreeding to avoid the negative impact of fescue toxicosis on reproduction [9–11].

Some have suggested there may be a genetic association in the cow population that allows producers to select for cows that are tolerant to the effects of fescue toxicosis [5,25,26]. For animals ingesting ergot alkaloids, a classic response is decreased serum prolactin concentrations [10–13,18,19] and vasoconstriction [3,5,8]. Ergot alkaloids are dopamine agonists that interact with *DRD2* in a competitive manner [27] to alter serum prolactin [28] and vasoconstriction [5–8]. Campbell et al. [29] identified a SNP in dopamine receptor D2 (DRD2) and established a relationship with serum prolactin concentrations in grazing steers and cows. They identified the G|G genotype in steers as having a greater reduction in serum prolactin values than A|A when grazing E+ fescue pastures. They suggested that the A|A genotype may have resistance to fescue toxicosis. Therefore, we chose to evaluate the *DRD2* SNP in cattle to examine if there is any genetic association with response to fescue toxicosis. The objectives of this study were to evaluate: (1) the use of novel, non-ergot alkaloid producing tall fescue (NOV) compared to E+ fescue pastures during the last trimester of gestation (90 days) and early lactation (first 30 days; Figure 1) on cow/calf performance and milk production, and (2) if maternal genotype for the *DRD2* SNP interacted with the response to grazing tall fescue by endophyte type.

**Figure 1.** Timeline for the experimental study. Pregnant cows grazed NOV or E+ pastures during last trimester of gestation (90 days) and early lactation (30 days after calving).

#### **2. Results**

Total ergot alkaloid concentration (year 1) differed by fescue type (*p* < 0.0001) and month (*p* = 0.022) but the interaction was non-significant (*p* = 0.40). Pastures with E+ fescue had higher (*p* < 0.001) total ergot alkaloid levels compared to NOV during the endophyte type treatment period (Table 1). Total ergot alkaloid concentrations were higher (*p* < 0.05) in October and November than January and February (Figure 2A). Hay samples, which were fed in times when forage was limited, also had higher (*p* < 0.001) total ergot alkaloid levels for E+ than NOV (Table 1). In year 2, pasture and hay samples were examined using LC-MS/MS to quantify ergovaline and ergovalinine, which are mostly responsible for vasoconstriction [5,6]. Ergovaline and ergovalinine concentrations differed by endophyte type (*p* < 0.0001) but were non-significant by month (*p* = 0.83) or the interaction between endophyte type and month (*p* = 0.16; Figure 2B). Ergovaline and ergovalinine concentrations were greater (*p* < 0.001) in E+ pastures than NOV during the

months when the endophyte type grazing experiment was conducted. Hay samples from E+ bales contained greater (*p* < 0.001) ergovaline and ergovalinine concentrations compared to NOV.

**Table 1.** Total ergot alkaloid or ergovaline + ergovalinine concentrations by fescue endophyte-infected type, toxic (E+) or novel endophyte infected (NOV) pasture or hay.


a,b Means in the same row differ (*p* < 0.05) by endophyte type.

**Figure 2.** Total ergot alkaloid concentrations (**A**) or ergovaline/ergovalinine (**B**) for novel (NOV) or toxic (E+) endophyteinfected tall fescue pastures by month during time period when endophyte treatments were applied (year 1). Total ergot alkaloids differ (*p* = 0.049) by month but the interaction with fescue endophyte type was non-significant (*p* = 0.40). a,b Means with uncommon superscripts differ (*p* < 0.05). Ergovaline and ergovalinine concentrations did not differ (*p* = 0.47) by month or the interaction (*p* = 0.27) between endophyte type and month.

> The Angus cow herd (n = 227; 3–7 years of age) at the Piedmont Research and Education Center was genotyped for *DRD2* SNP [27]. The cow herd had a distribution of 24% A|A, 56% A|G, and 20% G|G. For this study, we used the homozygote *DRD2* genotypes of A|A (resistant) and G|G (susceptible) to assess if there is any related genetic resistance to fescue toxicosis during late gestation and early lactation. Serum prolactin concentrations were measured during time on endophyte type grazing treatment prior to calving. The two-way interaction between endophyte type and time on fescue was significant (*p* = 0.0088; Figure 3). All other two-way and three-way interactions were nonsignificant (*p* > 0.38). Serum prolactin concentrations were similar at the start of the grazing study (day 0). Serum prolactin concentrations were lower (*p* < 0.05) for E+ compared NOV after 30, 60 and 90 days on fescue treatment. Serum prolactin concentrations increased (*p* < 0.05) during the last 30 days of the grazing study in both NOV and E+; however, concentrations were lower for E+ than NOV. Serum prolactin concentrations were lower (*p* = 0.047) for cows with G|G versus A|A *DRD2* genotypes regardless of endophyte type during the grazing study (Figure 4).

**Figure 3.** Serum prolactin concentrations in cows grazing endophyte-infected ergot alkaloid producing (E+) or novel endophyte-infected (NOV) during the last trimester of gestation by time on treatment. The interaction between endophyte type and time on fescue treatment was significant (*p* = 0.0088). \* Denotes differences between endophyte treatments at that time point (*p* < 0.05).

**Figure 4.** Maternal *DRD2* genotype for serum prolactin concentrations in cows grazing endophyteinfected ergot alkaloid producing (E+) or novel endophyte-infected (NOV) averaged over the last trimester of gestation. The interaction between endophyte type and maternal genotype, maternal genotype and time on fescue, or endophyte type, time on fescue, and maternal genotype were all non-significant (*p* > 0.05). abMeans with uncommon superscripts differ (*p* < 0.05).

Serum non-esterified fatty acids (NEFA) and cholesterol concentrations were also examined during the fescue endophyte grazing treatment. At d 30 of endophyte treatment, the interaction between maternal genotype and fescue endophyte treatment was significant (*p* = 0.017) (Figure 5). Cows with A|A genotype had elevated serum NEFA concentrations when grazing E+ tall fescue compared to NOV. Cows with G|G genotype had similar serum NEFA concentrations when grazing the different endophyte types. At 90 d on treatment, serum NEFA concentrations were higher (*p* = 0.0002) in cows grazing E+ compared NOV regardless of genotype. Serum NEFA concentrations at d 90 were greater (*p* = 0.01) for A|A than G|G genotypes regardless of fescue endophyte treatment. The interaction between genotype and fescue treatment was non-significant (*p* = 0.47) for serum NEFA on d 90. Serum cholesterol concentrations differed (*p* < 0.05) by fescue endophyte type but did not differ by maternal *DRD2* genotype (*p* > 0.37) or interaction (*p* > 0.59) on day 30 and d 90 of fescue endophyte treatment. Serum cholesterol concentrations were greater (*p* = 0.0027) in cows grazing NOV compared to E+ fescue pastures (Figure 6) on day 30 and 90 of fescue endophyte treatment.

**Figure 5.** Serum non-esterified fatty acids (NEFA) in cows grazing endophyte-infected ergot alkaloid producing (E+) or endophyte-infected non-ergot alkaloid producing (NOV) during the last trimester of gestation. The interaction between maternal *DRD2* genotype and fescue endophyte type was significant (*p* = 0.017) on day 30 of fescue endophyte treatment. Endophyte type (*p* = 0.0002) and maternal *DRD2* genotype (*p* = 0.010) were significant but the interaction was non-significant on day 90 of fescue endophyte treatment. a,b Means with uncommon superscripts differ (*p* < 0.05) by endophyte type. c,d Means with uncommon superscripts differ (*p* < 0.05) by maternal genotype.

**Figure 6.** Serum cholesterol concentrations in cows grazing endophyte-infected ergot alkaloid producing (E+) or endophyte-infected non-ergot alkaloid producing (NOV) at 30 and 90 days on fescue endophyte treatments during the last trimester of gestation. All interactions were nonsignificant (*p* > 0.05) and main effects are shown in the figure for each time. a,b Means with uncommon superscripts differ (*p* < 0.05) at each time of sampling.

Cow body weight (BW) was measured over time during the fescue endophyte grazing study. Cow body weight differed by endophyte treatment (*p* = 0.0019) and time on fescue (*p* < 0.001) but the interaction was non-significant (*p* = 0.45). Cow body weight increased over time (*p* = 0.001) during late gestation. Cow BW was lower (*p* = 0.0019) in cows grazing E+ fescue compared to NOV after 30, 60, and 90 of grazing (Figure 7A). There was a trend (*p* = 0.054) for G|G cows to be heavier than A|A cows but there were no significant interactions with genotype and other variables (*p* > 0.34). Average daily gain during the last trimester of gestation differed by endophyte type (*p* = 0.0012) but maternal genotype and all interactions were non-significant (*p* > 0.30). Cows grazing E+ fescue pastures had 30% lower (*p* = 0.0012) average daily gains during the last trimester of gestation compared to NOV (Figure 7B).

**Figure 7.** Changes in cow body weight (BW; **A**) and average daily gain (ADG; **B**) of cows grazing endophyte-infected ergot alkaloid producing (E+) or endophyte-infected non-ergot alkaloid producing (NOV) during last gestation. Interactions between endophyte type, year, and maternal DRD2 genotype were non-significant for body weight gain and average daily gain. \* Denotes differences (*p* < 0.05) in body weight by endophyte treatment. a,b Means with uncommon superscripts differ (*p* < 0.05).

The interaction between endophyte type and maternal *DRD2* genotype was significant (*p* = 0.012) for cow body condition score (BCS). Cows with A|A genotype had lower (*p* < 0.05) BCS when grazing E+ fescue than NOV; however, for G|G cows there was no difference in BCS by fescue endophyte type (Figure 8A). Ultrasound measurements of rump fat thickness and rump muscle depth were obtained at the start and end of the fescue grazing experiment. The change in rump fat thickness did not differ by genotype (*p* = 0.41) or fescue endophyte type (*p* = 0.71) but the change in rump muscle depth was greater (*p* = 0.036) for cows grazing E+ fescue than NOV, regardless of genotype. The interaction between endophyte type and time on fescue was significant for hair coat score (HCS). Hair coat scores did not differ by maternal genotype (*p* = 0.35) or all interactions with maternal genotype were non-significant (*p* > 0.48). Hair coat scores increased (*p* = 0.041) over time in cows grazing E+ fescue and values were greater compared to NOV at day 60 and 90 of treatment (Figure 8B).

**Figure 8.** Body condition score (BCS; **A**) and hair coat scores (HCS; **B**) by fescue endophyte type during the last 90 days of gestation. The interaction between endophyte type and maternal genotype was significant for BCS. a,b Means with uncommon superscripts differ (*p* < 0.05). The interaction between endophyte type and time on fescue was significant for HCS. \* Denotes significance (*p* < 0.05).

Interactions between year, endophyte type and maternal genotype were non-significant (*p* > 0.37) and the main effects are shown in Table 2. Dystocia scores did not differ (*p* > 0.05) by year, fescue endophyte treatment or maternal genotype. Calving interval from year 1 to year 2 did not differ (*p* > 0.05) by fescue endophyte type or maternal genotype. The

calving rate was lower (*p* = 0.012) in cows grazing E+ than NOV. Calf birth weight nor adjusted calf birth weight did not differ (*p* = 0.63) by fescue endophyte treatment. Calf birth weight and adjusted calf birth weights were higher (*p* = 0.018) for A|A than G|G cows and by year with heavier (*p* = 0.055) calves in year 1. Calf sex did not differ (*p* > 0.05) by year, fescue endophyte type or maternal genotype. Cow milk production was estimated using the weigh-suckle-weigh procedure at 30 days after calving. Milk production was 15% lower (*p* = 0.047) in cows grazing E+ fescue compared to NOV. Milk production was higher (*p* = 0.014) in G|G than A|A cows. The interaction between fescue treatment and genotype was non-significant (*p* = 0.99).

**Table 2.** Main effects for year, fescue endophyte type and DRD2 genotype on adjusted calf birth weight and weaning weight of calves. All interactions were non-significant (*p* > 0.05).


<sup>1</sup> Dystocia score: 1 = no difficulty, no assistance required; 2 = minor difficulty, some assistance; 3 = major difficulty, usually mechanical assistance; 4 = caesarian section or other surgery; 5 = abnormal presentation. <sup>2</sup> Adjusted weaning weight was adjusted by cow age and calf sex [30]. Adjusted 205-day weight was the adjusted weaning weight calculated on an equal calf age basis of 205-day. a,b Means in the same row with uncommon superscripts differ (*p* < 0.05) by year. c,d Means in the same row with uncommon superscripts differ (*p* < 0.05) fescue endophyte type. e,f Means in the same row with uncommon superscripts differ (*p* < 0.05) by maternal DRD2 genotype. g,h Means in the same row with uncommon superscripts differ (*p* < 0.10) by fescue endophyte type.

> The interaction between fescue endophyte type and maternal genotype was significant (*p* = 0.036) for calf average daily gain during the first 30 days of lactation. Calf average daily gain (ADG) during the first 30 days of lactation was higher (*p* = 0.036) for calves born to A|A cows grazing NOV than G|G cows grazing NOV or E+ fescue (Figure 9). Calf weaning weight and adjusted calf weaning weights did not differ by fescue endophyte type (*p* > 0.52), genotype (*p* > 0.24) or the interaction (*p* > 0.94). Calf weaning weight and adjusted weaning weights were higher (*p* = 0.021) in year 1 compared to year 2. When weaning weight was calculated on an age adjusted basis, the adjusted 205 days weaning weights tended (*p* = 0.073) to be higher by 8.4 kg for NOV compared E+. Maternal genotype did not alter (*p* > 0.47) adjusted 205 days weaning weight and the interaction with fescue endophyte type was non-significant (*p* = 0.95 & 0.86). Adjusted 205 days ADG of the calves tended to be greater (*p* = 0.072) for NOV than E+.

**Figure 9.** Calf average daily gain during the first 30 days of lactation in cows grazing different fescue endophyte type by maternal DRD2 genotype. The interaction between endophyte type and maternal DRD2 genotype was significant (*p* = 0.036) and simple effects are shown in the figure. a,b Means with uncommon superscripts differ (*p* < 0.05).

#### **3. Discussion**

In the current study, stockpiled tall fescue with either endophyte-infected ergot alkaloid producing (E+) or novel endophyte-infected nontoxic (NOV) pastures were utilized for grazing during the last trimester of gestation (90 days) and early lactation (30 days) in a spring-calving Angus cow herd. In year 1, total ergot alkaloid was examined during the time period when fescue endophyte type treatments were applied (October–March). Total ergot alkaloid concentrations decreased from November and December to January and February. In year 2, a LC/MS analysis was developed to determine specifically ergovaline and ergovalinine levels in forage samples because they have been shown to be the causative agents of fescue toxicosis [5,8]. The concentrations of ergovaline reported in this study are at threshold levels (300–500 ppb) where fescue toxicosis can occur [31]. Kallenbach et al. [32] reported reduced ergovaline concentrations of stockpiled tall fescue from December to March. Curtis and Kallenbach [33] showed that ergovaline concentrations declined 50% in an 84-day period with stockpiling of tall fescue during the winter months. Ergot alkaloid concentration in E+ hay was lower than in the E+ pastures. Similarly, Roberts et al. [34] found ergovaline concentrations in hay are quickly reduced within the first 3 day after mowing and decline gradually during additional months of storage. The levels of ergovaline/ergovalinine in the current study are similar to those reported by Peters et al. [12] and Kallenbach et al. [32] but the total ergot alkaloid values are higher than those published by others for cow grazing E+ fescue [18,19,33].

Pregnant cows grazed E+ or NOV fescue pastures during critical stages of production, last trimester of gestation and the first 30 days of lactation. The interaction between fescue endophyte type and maternal *DRD2* genotype was non-significant and main effects of fescue endophyte type over time and maternal *DRD2* genotype are shown in Figures 3 and 4. Before grazing treatments started (day 0, start of the last trimester of gestation), serum prolactin concentrations were similar. After 30, 60, and 90 days on fescue pastures, serum prolactin concentrations were lower for E+ than NOV. Serum prolactin levels, examined at day 90 of treatment, increased in both E+ and NOV compared to pre-treatment and day 30 and 60 values; however, concentrations were greater for NOV compared to E+. A reduction in serum prolactin concentration is commonly observed in cows grazing E+ fescue pastures [10–13,18,19]. There was a difference in prolactin concentrations among cows of different *DRD2* genotypes but the response to fescue was similar for A|A and G|G genotypes. Campbell et al. [29] identified the *DRD2* SNP suggested that A|A genotype may provide resistance to fescue toxicosis. They hypothesized that spring calving herds should have a higher prevalence of A allele. Our results showed that the A and G allele were about equally distributed in our Angus spring-calving cow herd (0.52 A allele and 0.48 G allele). The majority of our cows were heterozygous (56%) for the *DRD2* SNP with smaller percentages in the homozygous condition (24% A|A, 20% G|G). In this study,

there was no interaction between maternal genotype and fescue endophyte treatment for serum prolactin concentrations indicating a lack of genetic association with prolactin values in cows grazing stockpiled E+ fescue. Others have also examined other single SNPs and found some associations with XK, Kell blood group complex subunit-related family member-4 (XKR4) and serum prolactin levels [35], prolactin gene and calving interval [36], and cytochrome P450 and lower milk production [37]. Selection for a trait like fescue resistance likely involves multiple genes and therefore the use of a single SNP marker may not be as successful [38]. A commercial test, T-Snip™ (AgBotanica, Columbia, MO, USA), is now available that provides a tolerance index for fescue toxicosis. Galliou et al. [39] found a relationship between T-Snip score and cow performance when grazing E+ tall fescue. The benefits of genetic markers for tolerance to fescue toxicosis would be advantageous for tall fescue based grazing systems, but more research is needed in this area.

Serum NEFA and cholesterol concentrations were measured at day 30 and 90 of the grazing endophyte type study to examine changes in metabolism that may be associated with fescue endophyte type or maternal genotype. Non-esterified fatty acids are released into the bloodstream upon lipolysis of adipose triglycerides and elevated levels are associated with negative energy balance in late gestation to early lactation in dairy cattle [40]. Serum NEFA values after 30 days of grazing were higher in A|A cows grazing E+ pastures but unchanged in G|G. At 90 days of treatment, NEFA concentrations were elevated in cows grazing E+ regardless of genotype. Niederecker et al. [41] also reported elevated NEFA values in cows grazing stockpiled tall fescue versus summer baled fescue hay after 56 to 99 days of feeding during late gestation. McArt et al. [40] found that cows with elevated NEFA had longer time to pregnancy, lower milk yield and increased culling rates. Cholesterol concentrations were higher for NOV than E+ at day 30 and 90 of fescue treatment. Others have shown that serum cholesterol content is lower in steers grazing E+ tall fescue pasture versus endophyte-free [42] or novel endophytes [43]. Stuedemann et al. [44] also reported lower plasma cholesterol concentrations in cows grazing E+ fescue and that cholesterol values were negatively associated with rate of N fertilization. The cows grazing high N fertilized E+ pastures had the lowest cholesterol levels and highest incidence of fat necrosis [44]. Cholesterol is the precursor for steroid hormones and higher cholesterol concentrations are associated with shorter intervals from calving to conception and successful pregnancy [45].

It is well documented that grazing of E+ forage results in production losses [1–3,5,25]. Steers grazing ergot alkaloid-producing endophyte-infected tall fescue gained 30% to 70% less on compared to cattle consuming an endophyte-free or novel endophyte tall fescue [3,20,21,46]. In this study, cows grazing E+ fescue had lighter body weights and 30% lower average daily gains during the last trimester of gestation compared to NOV. Others [12,13,18,19] have shown similar results with lower body weight gains in cows grazing E+ fescue compared to endophyte-free or novel endophyte. Reduction in BW gain for cattle consuming toxic fescue has been attributed to a number of things, but most often associated with reduced intake [9,47]. Peters et al. [12] found that forage intake was similar in the initial 28 d even though body weight loss was not. In this study, calf birth weight did not differ between fescue endophyte types even though cow body weight gain was reduced. Similarly, others have also reported no change in calf birth weight when cows grazed E+ fescue prior to calving [19,48,49]. In contrast, others have shown that calves or lambs born to dams exposed to endophyte-infected tall fescue during gestation have reduced birthweights by 5% [18], 15% [50,51] or 36% [52]. Britt et al. [50] found that exposure to ergot alkaloids during late gestation had the greatest impact on fetal development with little to no effect when fed during mid-gestation in sheep. Exposure to ergot alkaloids during late gestation reduced fetal lamb muscle and organ weights resulting in asymmetrical growth, which is indicative of intrauterine growth restriction (IUGR) [53]. These reductions in fetal growth appear related to vasoconstrictive events caused by ingestion of the ergot alkaloids, specifically ergovaline and ergovalinine [54]. In cattle, the placenta continues to increase in weight until near term [55,56] which is

unlike the sheep where placental mass is established by 90 d of gestation [57]. Greenwood et al. [58] suggests that these differences in placental growth may explain why cattle appear less sensitive to under nutrition than sheep. Placental weight and birth weight are highly correlated in beef cattle [59,60]. However, in this study, cow weight gain and body weight were reduced when grazing E+ fescue pastures without any change in calf birth weight. Body condition score was lower for A|A cows grazing E+ than NOV; however, in G|G cows BCS was similar on NOV and E+ but BCS was lower than A|A cows grazing NOV. Hair coat score increased over time in both NOV and E+ but cows grazing E+ fescue had higher HCS on d 60 and 90 than NOV. Gray et al. [61] reported that cows who shed their winter coat before June 1 wean heavier calves. Real-time ultrasound measures of rump fat thickness and gluteus muscle depth at the beginning and end of the study showed that subcutaneous fat thickness was unchanged, but the change muscle depth was greater for E+ than NOV. Ultrasound rump fat and muscle depth measures are useful to monitor changes in body composition in cows grazing toxic tall fescue [62]. These results suggest the cows grazing E+ fescue during the last trimester of rapid fetal growth had reduced body weight gains that lead to mobilization of muscle tissue in order to provide nutrients for fetal growth.

This grazing study was continued through the first 30 days of lactation, which is a period of highest milk production [23,24] when the calf is almost solely dependent on the dam. Grazing E+ fescue during the last trimester and first 30 days of lactation lowered milk production by 15% compared to NOV. Milk production was also greater for G|G genotypes than A|A; however, this was consistent across endophyte treatments. Brown et al. [63] reported a 43% reduction in milk production for Angus cows grazing E+ fescue pastures compared to bermudagrass from d 61 to 200 postpartum. Peters et al. [12] found a 22% reduction in milk production of cows grazing E+ versus endophyte-free tall fescue pastures from birth to weaning. In contrast, Burke et al. [49] did not observe differences in milk production but instead found higher milk fat in cows grazing E+ fescue compared to endophyte-free. Milk production, measured at 70 days of age, did not differ in cows that grazed E+ or NOV in late gestation only [49]. Calf gains during this first 30 days period were lower for E+ than NOV. Feeding toxic endophyte-infected seed to ewes during mid and late gestation reduced milk production by 59 to 82% and reduced pre-weaning growth rates [51], which altered subsequent puberty development and postnatal growth [64]. In addition, cow calving rate was lower in cows that grazed E+ fescue during late gestation and early lactation. Calving interval from year 1 to 2 in this study did not differ among fescue treatments or genotypes. Caldwell et al. [19] also observed reduced calving rates when spring calving cows grazed E+ tall fescue compared to novel endophytes. Calf weaning weights or adjusted weaning weight for cow age and sex did not differ by endophyte type or maternal genotype. However, the adjusted 205-day weight was lower for calves whose dams grazed E+ tall fescue during late gestation and early lactation compared to NOV. Others also found that 205-day weights were lower for calves whose dams grazed E+ fescue from birth to weaning [12] and in spring-calving cows [19].

#### **4. Conclusions**

Grazing E+ fescue pastures during the last trimester of gestation and first 30 day of lactation reduced cow body weight gain but did not alter calf birth weight compared to NOV. Milk production and calf ADG during the first 30 day of lactation were lower for E+ than NOV. Calving rate was reduced but no calving interval for E+ cows. Adjusted 205-day weight of calves was lower in those grazing E+ with their dams compared to NOV. There were no interactions between *DRD2* genotype and fescue endophyte type indicating that genotype was not associated with response to E+ fescue in this study. Overall, grazing E+ tall fescue pastures during late gestation and early lactation appeared to alter the priority of nutrient utilization of the cows to adapt to reductions in gain during the time period of rapid fetal growth. Early milk production was lower for E+, which reduced early calf gains that remained to 205-day of age.

#### **5. Materials and Methods**

All animal experimental procedures were reviewed and approved by the Clemson University Institutional Animal Care and Use Committee (AUP 2017-025) on 22 May 2017. All animal experiments were conducted at the Clemson University Piedmont Research and Education Center (PREC), Pendleton, SC.

#### *5.1. Genotyping*

Hair follicles (5–10) were collected from tail switches of all cows in the Angus cow herd at Piedmont REC (n = 227; 3–7 year of age) and DNA was extracted using Quick Extract™ DNA extraction solution (Epicentre, Madison, WI, USA). The Extracted DNA was shipped to University of Tennessee for *DRD2* SNP genotyping (rs41749780) according to Campbell et al. [29]. Each cow was one of three possible genotypes: A|A, A|G or G|G. Dams with homozygous alleles (A|A and G|G) were used in this study over a two-year period.

#### *5.2. Design*

Cows were pregnancy checked by ultrasound via rectal palpation at day 95 after timed AI by an experienced technician. Pregnant cows (n = 75, 3 to 7 year of age, 608 ± 4.9 kg BW) were randomly assigned to fescue endophyte type: (1) endophyte-infected ergot alkaloid producing tall fescue (E+; Kentucky 31) or (2) novel endophyte-infected nontoxic tall fescue (NOV, Texoma MaxQ II, Pennington Seed, Madison, GA, USA) within maternal (A|A, n = 38 and G|G, n = 37) genotype during the last trimester of pregnancy (90 days) and first 30 days of lactation. Cows remained in the same fescue endophyte type treatment during year 2. Each year, there were two pasture replicates (10 ha pastures/rep; 2 rep/fescue type) per fescue endophyte type. Cows were maintained on non-fescue pastures during the first and second trimester of gestation when not on fescue endophyte treatments. For this study, cows grazed stockpiled fescue pastures with either endophyte-infected, toxic (E+) or endophyte-infected, novel (NOV) tall fescue. Pastures were fertilized 15 days prior to cows being allocated to treatment with 67.7 kg of nitrogen per ha. A strip grazing management practice was implemented through the duration of the project. Paddocks were monitored visually twice weekly and cattle were allotted additional forage when herbage availability dropped below 2000 kg/ha. When all available forage in stockpiled pastures was below 2000 kg/ha, hay of each endophyte treatment type was fed free choice in each paddock. Prior to the start of the grazing experiment each year, tiller samples were taken and sent to Agrinostics (Watkinsville, GA, USA) to determine endophyte infection levels. All pastures used in this study had >90% endophyte infection rates.

#### *5.3. Forage Analysis*

Grab samples of forage were collected during the time period when cattle were grazing the endophyte treatments (late gestation to early lactation) from each paddock when cattle were moved to a new strip of grazing and hay samples collected before feeding. All forage and hay samples were freeze dried and ground using a 1-mm screen Wiley cutting mill (Arthur H. Thomas, Philadelphia, PA, USA). Forage samples were composited on a monthly basis. In year 1, forages and hay were analyzed for total ergot alkaloid infection concentrations (Agrinostics Limited Co., Watkinsville, GA, USA). As ergovaline and its epimer ergovalinine are the most informative ergots involved in vasoconstriction [5,8], we developed an assay to measure concentration of these ergopeptines to obtain more specific information. This assay was not available in year 1 but was developed and utilized in year 2 of this experiment. In year 2, forage and hay samples were analyzed for ergovaline and ergovalinine concentrations by LC-MS/MS at the Multi-User Analytical Laboratory (MUAL) at Clemson University.

For quantification of ergovaline and ergovalinine concentrations, freeze dried forage and hay samples were extracted according to Guo et al. [65]. A tall fescue seed extract was provided by J. L. Klotz and used to establish a calibration range from 0.78 to 100 ng/mL to quantify ergovaline and ergovalinine [66]. Ergotamine tartrate (Sigma Chemical Co.,

St. Louis, MO, USA) was used as an internal standard. Samples were analyzed using a Thermo Orbitrap Fusion™ Tribrid™ Mass Spectrometer (Thermo Fisher, Waltham, MA, USA) with MS1 (ESI positive, 60,000 resolution, scan range 220–1200 m/z) and MS2 (HCD & CID on targeted mass of 25 ergot alkaloids in orbitrap at 15,000 resolution) in the Clemson University Multi-User Analytical Laboratory and Metabolomics Core. The column was Phenomenex Kinetex XB-C18 with 1.7 µm, 100A, 150 × 2.1 mm (Phenomenex, Torrance, CA, USA). Solvent A was 5 mM ammonium bicarbonate and solvent B was 90% acetonitrile with 5 mM ammonium bicarbonate. Quality control checks were performed using instrument blanks, method blanks to determine recovery, and instrument quality control to assess relative standard deviation of 12.5 ppb ergovaline standard. The percent recovery for ergovaline, ergovalinine, and ergotamine were 98, 95, and 100%, respectively. Compound detection was performed using the extracted ion chromatogram for targeted analytes within 2 ppm mass error in Skyline software. Quantification of ergovaline and ergovalinine was performed using linear external calibration curves (0.78 to 100 ng/mL. The method limit of detection for ergovaline/ergovalinine in grass and hay was 11.414 ng per g freeze-dried plant tissue (instrument LOD of 0.78 ppb).

#### *5.4. Blood Samples*

Blood was collected on cows at day 0, 30, 60, and 90 of fescue endophyte treatment in 10 mL vacutainer serum tubes (Covidien Ltd., Monoject; Dublin, Ireland) via the coccygeal vein. All serum samples were allowed to clot and centrifuged at 2000× *g* for 20 min at 4 ◦C to obtain serum. Serum was stored at −20◦C for subsequent prolactin, non-esterified fatty acid (NEFA), and cholesterol analyses. Prolactin concentrations were measured using RIA procedures of Bernard et al. [67]. The intra- and inter-assay coefficients of variation were 5.59 and 5.75 %, respectively. Non-esterified fatty acids on day 30 and 90 were analyzed using NEFA ELISA kit (BiooScientific, Perkin-Elmer, Waltham, MA, USA) with an intraand inter-assay coefficient of variation of 6.42% and 5.56%, respectively. Serum cholesterol was extracted in duplicate using the method of Chauveau-Duriot et al. [68] and samples had an intra- and inter-assay coefficient of variation of 5.29% and 6.00%, respectively. Cholesterol content on day 30 and 90 was analyzed by gas chromotography (Agilent 6850, Santa Clara, CA, USA) using a Restek Rxi-5ms column (15 m, 0.25 mm ID, 0.25 µm, catalog no. 13420; Restek, Bellefonte, PA, USA) according to manufacturer.

#### *5.5. Animals*

Body condition scores (BCS 1–9 with 1 = severely emaciated to 9 = very obese) [69] and hair coat scores (HCS 1–5 with 1 = slick, shiny hair coat to 5 = > 50% of body covered with old, unshed hair) [70] were collected on day 0, 30, 60, and 90 of fescue endophyte treatment. Scores for BCS and HCS were recorded by the same two technicians during the study and average for each cow. Real-time ultrasound measures of rump fat and gluteus medius muscle depth were collected on day 0 and 90 by the same technician using an Aloka 500V and 17.2 cm linear transducer. The image was captured and interpreted using BioSoft Toolbox for beef software (Biotronics, Inc., Ames, IA, USA).

At parturition, calves were processed within 12 h of birth at 800 and 1600. Data collected included date, birth weight, and sex. Calves were ear tagged, tattooed, and males were castrated. Calves were injected intramuscularly with vitamin A (500,000 IU) and D (75,000 IU; VetOne, MWI Veterinary Supply Co., Boise, ID, USA). To estimate milk production, weigh-suckle-weigh technique was utilized at 30 days post-calving [24]. Calves were separated from their dams for 3 h and then allowed to nurse the dams dry. Then, they were separated from the dam again and held for 6 h. After the separation, calves were weighed. The calves were allowed to nurse the dam and then reweighed immediately. The difference in weights between post- and pre-suckling were assumed to represent milk production in that 6 h period and was multiplied by 4 to represent a 24-h period. In April, all cows were synchronized using the fixed-time AI (TAI) 7-day Co-Synch + CIDR protocol for timed AI to a single sire (Connealy Mentor 7374, Select Sires,

registration # 15832714; Spring Hill, TN, USA). At two weeks post TAI, clean up bulls were added and cows were exposed to the clean-up bulls for 70 d. Calves were vaccinated for clostridial agents (One Shot Ultra 7; Zoetis, Parsippany, NJ, USA), bovine rhinotracheitis parainfluenza 3 vaccine (TSV-2, Zoetis), and leptospirosis (Leptoferm-5, Zoetis) in May at 4 month of age. Cows were pregnancy checked by ultrasound via rectal palpation at d 95 after TAI by an experienced technician. Cows that were not pregnant in year 1 were sold. Estimated breeding dates were calculated based on rectal palpation results. Calves were weaned on September 11, 2018 and August 30, 2019 and body weight obtained. Adjusted birth and weaning weights were calculated according to the Beef Improvement Federation Guidelines [30] based on age of the dam and sex of the calf. Adjusted 205-day weights were also calculated [30].

#### *5.6. Statistics*

The univariate procedure of SAS (SAS Inst. Inc., Cary, NC, USA) was used to test all variables for normality. Serum prolactin concentrations were not normally distributed and were log-transformed for data analysis. Ergot alkaloid values were not normally distributed and were square root transformed for statistical analysis. Serum prolactin and ergot alkaloid concentrations results are presented as non-transformed for easier understanding and interpretation. For forage and hay ergot alkaloid data, paddock was the experimental unit. Data were analyzed using mixed procedure with fescue forage type, month, and two-way interaction in the model. For animal data, a 2 × 2 factorial with two-factors of fescue endophyte type (E+ vs. NOV) and maternal DRD2 genotype (A|A vs. G|G) in a split plot design. A mixed model was developed with endophyte type, DRD2 genotype and the 2-way interaction in the model as fixed effects and two different error terms applied. The error term for testing endophyte type was the random effect of paddock within endophyte type by year and the error term for testing DRD2 genotype and its interactions with endophyte type was the random effect of cow within DRD2 and endophyte type combination. The analysis was performed using the GLIMMIX procedure of SAS (SAS Inst. Inc., Cary, NC, USA). For repeated measures analyses, time and all interactions with time were included in the model. Least square means were generated and separated using a Fisher's LSD when the *F*-test was significant for year, fescue endophyte type, maternal genotype and or the interaction. Significance was determined at *p* < 0.05, and trends denoted when *p* > 0.05 to *p* < 0.10. When the interaction was non-significant (*p* > 0.05), main effects of endophyte type or DRD2 genotype are shown in the results. When the interaction was significant (*p* < 0.05), simple effects of endophyte type by DRD2 genotype are shown in the results.

**Author Contributions:** Conceptualization, S.K.D. and J.G.A.; data curation, S.A.W., S.M.J., T.W.; writing—original draft preparation, S.A.W., S.M.J., T.W.; writing—review and editing, S.K.D., J.G.A., J.L.K.; funding acquisition, S.K.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** Technical contribution no. 6684 of the Clemson University Experiment Station. This material is based upon work support by NIFA/USDA under project number SC-1700537.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** Appreciation is expression to the Clemson University Beef Farm staff and AVS 4220 Fetal Fescue Research class for assistance with animal management and sample collection; C. Kojima and F.N. Schrick for DRD2 genotyping and prolactin assays; Clemson University Multi-User Analytical Laboratory and Metabolomics Core for ergot alkaloid analyses.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


*Article*
