https://orcid.org/0000-0002-8085-5302
Figures
Abstract
Gastrointestinal (GI) disease is a major health concern in preweaned dairy calves. The objective of this fixed cohort study was to use RNA isolated from preweaned Holstein and Jersey heifer calf feces to study the molecular adaptations to variable clinical GI disease. The study was conducted on a commercial calf ranch in the western U.S. Enrolled calves were assessed twice daily for variations in demeanor, milk intake, and hydration. Fecal consistency scores were recorded at enrollment (day 1), and on the day (day 10) that a fecal sample was collected for differential gene expression (DGE). Calves with diarrhea on either day were classified as having either uncomplicated, localized GI disease (scours), or systemic GI disease (systemic enteritis). Eighty-four calves’ fecal RNA was evaluated for DGE, of which 33 calves (n = 20 Holstein; n = 13 Jersey) were consistently healthy. The remaining 51 calves (n = 23 Holstein; n = 28 Jersey) experienced varying severity of GI disease during the sampling window. Genes of interest were related to the inflammatory response (i.e., IFNG, NFKB1, NOD2, TLR2, and TLR4) and cell membrane or cytoplasmic transport (i.e., AQP3, FABP2, KRT8 and SLC5A1). Breed-specific findings indicated that AQP3, IFNG, and TLR4 were upregulated in Holsteins with systemic enteritis, whereas KRT8 was downregulated in systemically affected Jerseys. Holsteins did not appear affected by scours aside from a tendency for DGE of toll-like receptors (TLRs) on the day of diarrhea. However, Jersey calves consistently demonstrated a tendency to upregulate IFNG, NFKB1, and TLR4 when affected with either scours or systemic enteritis. These findings were more pronounced in systemically affected Jersey calves and were observed as a delayed response to both scours and systemic enteritis. These findings support previous observations suggesting that Holstein calves may be better equipped than Jersey calves to rapidly fight pathogen invasion.
Citation: McConnel CS, Slanzon GS, Parrish LM, Trombetta SC, Shaw LF, Moore DA, et al. (2022) Transcriptional changes detected in fecal RNA from neonatal dairy calves of different breeds following gastrointestinal disease of varying severity. PLoS ONE 17(12): e0278664. https://doi.org/10.1371/journal.pone.0278664
Editor: Angel Abuelo, Michigan State University, UNITED STATES
Received: January 21, 2022; Accepted: November 22, 2022; Published: December 1, 2022
Copyright: © 2022 McConnel et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This project was supported by Agriculture and Food Research Initiative Competitive Grant no. 2019-68008-29897 from the USDA National Institute of Food and Agriculture, the USDA National Institute of Food and Agriculture, Animal Health & Disease Research Capacity Grant project 1014680, the WSU CVM Caldwell Endowment, and the American Jersey Cattle Association. There was no additional external funding received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Research into dairy heifer health problems tends to simplify disease into preweaned or postweaned gastrointestinal (GI) or respiratory disease, and diagnostic accuracy is often hindered by a lack of detail [1–3]. The diagnosis of GI disease in particular has relied on the observation of diarrhea, but we historically have lacked the knowledge and tools to discriminate between GI disease phenotypes [4, 5]. Although blood biomarkers of inflammation, oxidative stress, metabolism and hydration provide additional insight into overall animal health, these analyses do not describe the actual biological adaptations of the GI tract (GIT) during the onset and progression of disease [6]. Furthermore, standardized methods of classification have been lacking such that what little information is available is often unproductive. As a result, inconsistencies in disease data presentation hinder descriptions, comparisons, and investigations into the overall consequences of disease on heifer health and wellbeing [7]. These misspecifications of disease processes can result in inappropriate therapeutic and preventive actions that impact lifetime wellbeing and put the dairy industry’s reputation at risk for failing to adequately address preventable losses [8].
Efforts to discriminate dairy calf GI disease phenotypes more fully have included transcriptomic assessments of fecal RNA [6]. GI tract epithelial cell exfoliation is associated with the natural turnover process to preserve tissue structure [9], and the daily shedding of these cells into feces provides a source of RNA to evaluate gene expression [10]. This provides a noninvasive sampling method in place of tissue biopsies or postmortem tissue analysis [11]. Importantly, fecal RNA can be differentiated based on the presence of a polyA tail in mammalian RNA that is absent in bacterial RNA. Consequently, Rosa et al. [6] demonstrated differential expression of genes related to cell membrane transport and inflammation in fecal RNA from neonatal Jersey male calves following mild diarrhea. This suggests that RNA isolated from fecal samples provides a potential tool for the evaluation of physiological changes due to GI disease. However, the inflammatory response observed by Rosa et al. in fecal RNA was delayed compared to the pattern of fecal scores and proinflammatory blood biomarkers. Accordingly, further research of transcriptional alterations is warranted to study the intestinal response to inflammation and its variation over time in neonatal dairy calves of different breeds experiencing diverse manifestations of GI disease [12, 13].
Therefore, the current study evaluated transcriptional changes detected in fecal RNA from neonatal dairy calves experiencing varying severity of GI disease with and without therapeutic interventions. We hypothesized that inflammatory and cellular transport signals would be differentially expressed in response to adaptations related to breeds and disease severity and timing. Thus, the objective of this study was to use RNA isolated from fecal samples to study the molecular adaptations to variable clinical GI disease via the expression of select genes involved in inflammation and cellular transport.
Materials and methods
Ethics statement
The research protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Washington State University (ASAF#6414).
Study design and calf enrollment
This fixed cohort study was conducted on a commercial calf ranch in the western U.S. between July 13 and August 6, 2019. The ranch housed approximately 25,000 Holstein, Jersey, and crossbred calves through 200 days of age sourced from multiple dairies. Breed differences were variable within the source dairies ranging from Holstein and Jersey, Holstein or Jersey only, or crossbred from a single source. Calves were fed colostrum at the dairy of birth, transited <2 hours to the ranch, and arrived at the ranch at ≤1 day of age. Transfer of passive immunity was assessed via blood samples obtained from 1-day old calves via jugular venipuncture to measure total serum protein (TSP) using a calibrated refractometer [14].
Animal husbandry
On-farm personnel were responsible for all primary care of the calves including feeding, cleaning, watering, and bedding maintenance. The calves were housed in adjacent individual hutches with no direct contact between calves. Upon entry to the ranch all calves were dehorned using a dehorning paste (Dr. Naylor Dehorning Paste, H.W. Naylor Co. Inc., Morris NY, USA), sprayed with a fly control spray (Ultra-Boss Pour-on Insecticide, Intervet Inc., Merck Animal Health, Omaha NE, USA), and ear notched and tested for bovine viral diarrhea virus (BVDV) using PCR. An enteric clostridial vaccine (Ultrabac CD, Zoetis Inc., Kalamazoo MI, USA) and a vitamin complex with selenium and Vitamin E (MU-SE, Intervet Inc., Merck Animal Health, Omaha NE, USA) also were administered on arrival. An intranasal viral respiratory vaccine (Vista Once SQ, Intervet Inc., Merck Animal Health, Omaha NE, USA) was administered at 2 days of age and a booster was administered at 20 days of age. A 7-way clostridial vaccine (Ultrabac 7, Zoetis Inc., Kalamazoo MI, USA) was administered at 45 days of age. Calves were kept in hutches through approximately 90 days of age.
Calves were fed 2 liters of a custom milk blend in a bottle twice daily from 1 to 15 days of age, and 3 liters of the milk blend thereafter until 52 days of age. From 53 to 60 days of age, calves were fed 2 L once per day and weaned thereafter. The milk blend consisted of pasteurized waste milk together with milk replacer and was targeted for an optimal composition of 13% solids, 22–24% fat, and 28% protein. Under the oversight of a licensed veterinarian following veterinary feed directive guidelines, all calves enrolled in the study and between 5 to 12 days of age received milk medicated with neomycin oxytetracycline per the labeled dose (10 mg/lb body weight/day; Neo-Oxy 100/100 MR, PharmGate, Omaha NE, USA), to combat ongoing GI disease problems. Two liters of customized oral electrolyte solution also were offered after the afternoon feeding during this same risk period from 5 to 12 days of age. Water was available between milk feedings, and a handful of grain mix consisting of pellets, molasses, and whole corn was offered from 3 days of age and steadily increased to approximately 2.25 kg by day 30 with free choice thereafter.
Clinical health assessments and therapeutic interventions
Two on-farm veterinarians oversaw general calf health management and treatment protocols for the ranch. Diagnoses and treatments were based on input from calf health managers’ clinical assessments. Morbidity and mortality records included broad disease diagnoses and associated treatments and were managed on-farm using DairyComp 305 (Valley Agricultural Software, Tulare, CA) with record oversight and health data compilation provided through The HEALTHSUM Syndicate LLC (Sunnyside, WA, USA).
For the purposes of this study, a PhD candidate and WSU veterinarian conducted twice daily clinical evaluations based on a standardized calf health-scoring system [4], prior to and at the time of each a.m. and p.m. milk feeding. Assessments began on the day of arrival through the completion of the follow-up period to 21 days of age. Primary points of consideration included demeanor (bright/alert/responsive vs depressed/dull), milk intake (good appetite; did not finish the milk or did not consume any of what was offered; orogastric intubation), hydration status (mm of ocular recession), and fecal consistency scores (1 = well-formed fecal samples; 2 = semi-formed fecal samples; 3 = loose fecal samples; 4 = watery fecal samples). Due to inconsistencies in fecal appearance within the calf hutches and an inability to observe each calf defecate at the time of all clinical evaluations, fecal consistency scores only were recorded at the time of fecal sampling. Rectal temperature also was recorded for all calves at the time of sampling utilizing a digital thermometer cleaned with alcohol swabs between calves.
Calves with fecal scores of 3 or 4 were diagnosed with diarrhea and classified as having either uncomplicated, localized GI disease (scours), or systemic GI disease (systemic enteritis). Calves with scours had no signs of clinical disease aside from diarrhea. Calves with systemic enteritis had diarrhea along with a behavioral score ≥1 on a matrix incorporating demeanor, mobility, ear positioning, sucking reflex, hydration status based on ocular recession, appetite, and interest in interaction (Table 1). Calves with systemic enteritis were treated medically by farm personnel based on farm protocols incorporating the following alone or in combination: IV fluid therapy (lactated Ringer’s solution), anti-inflammatories (flunixin meglumine or dexamethasone), a commercial yeast product (Celmanax SCP, Arm & Hammer, Church & Dwight Co., Inc., Ewing Township, NJ, USA), a proprietary customized probiotic, kaolin-pectin, or variable IM or SQ antibiotics (ceftiofur crystalline free acid, florfenicol, or oxytetracycline). Calves with fecal scores <3 at the time of sampling were classified as consistently healthy if they did not demonstrate abnormal clinical parameters or behavioral scores during a period of 21 days after arrival or receive any treatment for GI disease, respiratory disease, or other health problems during the sampling period.
Fecal samples for differential gene expression (DGE)
Fecal samples were collected per rectum between July 17 and August 6, 2019 (Fig 1). Heifer calves were randomly enrolled into the sampling scheme and allocated to an initial sampling day (day 1 enrollment sample) between 4 to 12 days of age to have fecal samples (~25–250 g) collected after the morning feeding for microbiome analysis used in a concurrent study [15]. This provided the opportunity to sample healthy calves and calves with previously undiagnosed scours but no other clinical abnormalities. Calves that demonstrated systemic enteritis (i.e., diarrhea plus a behavioral score ≥1) during this same period but prior to a random sample were sampled only once at the time of initial clinical diagnosis irrespective of the randomized sampling scheme.
Nine days after the initial fecal collection each calf was sampled again (day 10 DGE sample; 13–21 days of age) in an effort to align the fecal RNA transcriptomic analysis with previously documented delays in gene expression following evidence of diarrhea [6]. Calves that were initially sampled when healthy or with scours but progressed to systemic enteritis by 12 days of age, were sampled again on the day that systemic enteritis was diagnosed. These calves then were sampled twice more 9 days following both the first (healthy or scours) and second (systemic enteritis) samples to provide 2 day 10 samples. No calf was sampled more than 4 times in total and only one sample ultimately was analyzed for DGE, with preference given to the sample obtained during or following systemic enteritis (e.g., calves with systemic enteritis only were evaluated for disease progression in terms of healthy to systemic, systemic to healthy, or systemic to scours). Fecal samples were collected in sterile Whirl-Pak sampling bags (Cat #14-955-186, Thermo Fisher Scientific, Waltham, MA, USA) and immediately placed in a cooler with ice packs.
RNA extraction
Fecal samples remained on ice until they were processed in an on-farm laboratory within 2–4 hours of collection. Approximately 0.3 g of feces was weighed into a 15 mL conical tube containing 3 mL of RNAlater (Cat #AM7021, Thermo Fisher Scientific, Waltham, MA, USA) and then thoroughly mixed by vortex for 30 sec. Fecal bags and RNAlater tubes containing feces were then frozen onsite at -20°C. Samples were transported on dry ice to the WSU Field Disease Investigation Unit laboratory and stored at -80°C for approximately 3 months prior to processing. Fecal dry matter was performed by weighing out 1 g of feces and drying the samples in an incubator at 37°C for 72 hours. Percent dry matter was calculated by dividing the dry weight by the wet weigh multiplied by 100. Fecal samples in RNAlater were defrosted on ice and vortexed thoroughly for 30 sec at maximum speed. To aid in the removal of RNAlater [16], 3 mL of chilled PBS (Cat #AM1840, MagMAX Total Nucleic Acid Isolation Kit, Life Technologies Corp., Austin, TX, USA) was added and each sample was thoroughly mixed by vortex for 30 sec at maximum speed. To remove solids, samples were then centrifuged at 700 g for 1 min at 4°C. To pellet the cells the supernatant was transferred to a new 15 mL conical tube and centrifuged at maximum speed (4500 g) for 5 min at 4°C. The supernatant was discarded, and the remaining pellet was resuspended in 235 uL of Lysis Buffer provided by the MagMAX kit. The resuspension then was transferred to a bead tube supplied by the MagMax kit. RNA extraction was carried out beginning at the bead beating step in the MagMAX protocol. 90 uL of RNA Storage Solution (Cat #AM7001, Invitrogen, Carlsbad, CA, USA) was used for the final elution step in place of the MagMAX Elution buffer to improve preservation of RNA. RNA was quantified using a Qubit 4 Fluorometer (Cat #33226, Invitrogen, Carlsbad, CA, USA). RNA integrity was determined through fragment analysis using a 5200 Fragment Analyzer System (Agilent, Santa Clara, CA, USA), and 60ul of eluted RNA was stored at -80°C until submission for DGE analysis.
Due to the instability of fecal RNA and Nanostring’s allowance for DV values as a quality measurement, RNA integrity values were not used for quality assessments. Instead, samples with DV200 (percentage of RNA fragments >200 nucleotides) values >50% and concentrations >20 ng/uL were submitted for diagnostic processing at the Primate Diagnostic Services Laboratory (University of Washington, Seattle, WA, USA). Samples with DV200 values <50% and RNA concentrations <35 ng/uL were re-extracted using frozen feces with a Trizol (Cat #15596018, Invitrogen, Carlsbad, CA, USA) extraction method. An RNA Clean & Concentrator Kit (Cat #R1013, Zymo Research, Irvine, CA, USA) was used on all re-extracted samples, and those re-extracted samples with concentrations >60ng/uL were submitted for diagnostic processing as well.
Gene selection and gene expression analysis
Nine candidate genes were selected for this study (Table 2). These genes were related to the inflammatory response (i.e., IFNG, NFKB1, NOD2, TLR2, and TLR4) and cell membrane or cytoplasmic transport (i.e., AQP3, FABP2, KRT8 and SLC5A1). They were selected based on previous investigations into inflammatory pathologies and intestinal mucosa homeostasis [17–19], and gene expression analysis in fecal RNA from nondiarrheic dairy calves [12] and dairy calves undergoing mild diarrhea [6]. The genes glyceraldehyde 3-phosphate dehydrogenase (GAPDH), ribosomal protein S15a (RPS15A) and ribosomal protein 9 (RPS9) were used as internal control genes based on their previous identification as suitable controls [6].
Gene expression analysis of the 9 target genes and 3 housekeeping genes was performed using the Nanostring System (Nanostring Technologies, Seattle, Washington, USA), as previously described [20]. Gene expression was measured using a custom TagSet for the selected genes (S1 Table) on the Nanostring nCounter Analysis System (Nanostring Technologies, Seattle, Washington, USA). NanoString technology uses a digital color‐coded barcode tag with single‐molecule imaging that can detect and count hundreds of unique transcripts per reaction. The geometric mean of the internal control genes was used to normalize the DGE and data retention was accomplished in consultation with NanoString given the novelty of this project’s use of fecal mRNA. Briefly, data selection required average raw counts ≥ 4, > 5 genes with expression above background, and > 20 for the sum of raw counts for all genes within on sample divided by the scaling factor. The analysis of data was performed using nSolver Analysis Software 4.0, including nCounter Advanced Analysis (version 2.0, Nanostring Technologies, Seattle, Washington, USA). More specifically, fold changes and p‐values were calculated using nCounter default settings and a Benjamini-Yekutieli (B-Y) correction for multiple comparisons. The B-Y correction makes the assumption that there may be some biological connection between genes and returns moderately conservative estimates of false discovery rate (FDR). The FDR is the proportion of genes with equal or greater evidence for differential expression (i.e., equal or lower raw p-value) than are expected to be “false discoveries” due to chance.
Sample size
Previous DGE studies have shown that a minimum of 8 animals should be targeted per group (diseased, healthy), with the sample size based on expectations for a multivariate analysis estimating the log fold-change in a target gene’s DGE following a negative binomial distribution [21]. For the purposes of this study the enrollment period was based on the historical prevalence of GI disease of variable severity, mortality rates, average number of calf arrivals per day, and anticipated challenges associated with fecal RNA isolation. Expectations were for quality RNA samples from a minimum of 8 Holstein and 8 Jersey calves within each health status category (consistently healthy, scours, systemic enteritis).
Results
Calves and clinical outcomes
A total of 183 heifer calves were enrolled with RNA extracted from fecal samples and submitted for diagnostic processing at the Primate Diagnostic Services Laboratory using the Nanostring System. Average fecal dry weight percentages for consistently healthy, scouring, or systemically affected calves were 26.4 ± 6.2%, 12.5 ± 5.7%, and 13.9 ± 5.9%, respectively. A total of 99 samples (n = 38 consistently healthy; n = 47 scours; n = 14 systemic enteritis) were excluded from analysis primarily due to insufficient fecal mRNA raw counts or expression above background. Details related to the 84 Holstein and Jersey heifer calves remaining in the analysis are presented in Table 3. Those calves were sourced from 22 separate dairies and all calves were BVDV negative, the average TSP was 6.5 ±0.8 g/dL, and the average ages at enrollment and DGE samples were 7.7 ±2.5 days and 16.7 ±2.5 days, respectively.
Systemic behavioral scores (Table 1) and treatments (flunixin and antibiotics) were assigned or administered during the 10-day period from enrollment to DGE sampling.
A total of 33 calves (n = 20 Holstein; n = 13 Jersey) were consistently healthy. The remaining 51 calves (n = 23 Holstein; n = 28 Jersey) experienced varying severity of GI disease at some point during the sampling window. Average TSP was the same for healthy (6.5 ±0.9 g/dL) and diseased calves (6.5 ±0.8 g/dL); however, the average TSP for Jerseys (6.9 ±0.7 g/dL) was greater than for Holsteins (6.0 ±0.7 g/dL; P < 0.0001). Twenty-three of the diseased calves (n = 8 Holstein; n = 15 Jersey) were scouring at the time of their enrollment sample but healthy 9 days later when sampled for DGE. Two Jersey calves were scouring at both sampling times. Thirteen calves (n = 9 Holstein; n = 4 Jersey) were healthy when enrolled but scouring by the time of their DGE sample.
Thirteen of the 51 calves with GI disease demonstrated systemic enteritis during the sampling period (Table 4). All 13 were treated with flunixin and 8 were treated with antibiotics during the sampling period (Table 3). Seven of the systemic enteritis calves (n = 9 Holstein; n = 4 Jersey) were systemically ill when enrolled (average behavioral score = 8; range 2–17) but progressed to healthy status at least 7 days prior to their DGE sample (systemic to healthy). Two calves (n = 1 Holstein; n = 1 Jersey) were systemically ill when enrolled (behavioral scores 14 and 8, respectively) and scouring 9 days later when sampled for DGE (systemic to scours). Four calves (n = 2 Holstein; n = 2 Jersey) were healthy when enrolled but progressed to systemic enteritis (average behavioral score = 12; range 7–18) when sampled for DGE 9 days later (healthy to systemic).
Behavioral scores are based on the behavioral assessment scoring matrix presented in Table 1.
Differential gene expression
In agreement with a study by Rosa and Osorio [12], raw counts reflecting mRNA abundance in fecal RNA were 8-fold greater on average for KRT8 than the other endogenous genes of interest (S2 Table). NOD2 was not expressed in numerous samples and was removed from further analysis. Log2 fold-changes (ratios) in gene expression were evaluated for diseased calves versus those that were consistently healthy. DGE was evaluated overall and by breed for calves with any GI disease (Table 5), scours or systemic enteritis specifically (Table 6), and scours or systemic enteritis based on the day of diagnosis (Tables 7 and 8 and Figs 2 and 3). These sequential analyses were performed in an effort to assess and highlight the importance of discriminating between breed-specific GI disease phenotypes when evaluating the consequences of disease on calf health and wellbeing. Estimates of false discovery rate using the B-Y correction indicated that TLR4 (B-Y P = 0.01), AQP3 (B-Y P = 0.03), and IFNG (B-Y P = 0.06) were expressed differentially in calves with GI disease at any point during the sampling period (Table 5). It is worth acknowledging that less conservative uncorrected estimates indicated potential differential expression of additional genes including KRT8 (uncorrected P = 0.03), FABP2 (uncorrected P = 0.05), and NFKB1 (uncorrected P = 0.07). When analyzed by breed for calves with GI disease at any point only TLR4 appeared upregulated in Jerseys (B-Y P = 0.07). However, uncorrected estimates suggested the potential for differential expression of AQP3, TLR2, and TLR4 in Holsteins, and FABP2, IFNG, KRT8, and NFKB1 in Jerseys (uncorrected P < 0.1; Table 5).
A more refined analysis based on disease severity indicated that systemic enteritis impacted fecal mRNA DGE more substantially than uncomplicated scours (Table 6). AQP3, IFNG, and TLR4 were upregulated in calves with systemic enteritis (B-Y P ≤ 0.001), and KRT8 appeared downregulated (B-Y P = 0.08). Breed-specific log2 fold-changes indicated that systemically affected Holsteins (n = 6) upregulated AQP3 (1.97 ± 0.90; B-Y P = 0.02), IFNG (1.98 ± 1.00; B-Y P = 0.02), and TLR4 (1.57 ± 1.05; B-Y P = 0.09), whereas Jerseys with systemic enteritis (n = 7) downregulated KRT8 (-0.83 ± -0.50; B-Y P = 0.03). However, uncorrected estimates suggested that systemically affected Jerseys also upregulated AQP3, IFNG, TLR4 along with NFKB1 (uncorrected P < 0.1). Although there were no differences noted in calves with scours using the B-Y correction, uncorrected estimates suggested a similar pattern of DGE in Jerseys with scours (n = 21) as compared to those with systemic enteritis. Scouring Jerseys appeared to upregulate IFNG, TLR4 and NFKB1 and downregulate KRT8 (uncorrected P < 0.1). On the other hand, Holsteins with scours (n = 17) did not appear to be similarly affected and only demonstrated evidence of TLR2 downregulation (uncorrected P = 0.02).
When calves with scours were analyzed based on the day of diagnosis (Table 7 and Figs 2 and 3), only TLR2 was differentially expressed in Holsteins that were healthy at enrollment but diagnosed with scours on day 10 at the time of the DGE sample (n = 9; -1.54 ± -0.79; B-Y P = 0.03). Uncorrected estimates also suggested an upregulation of TLR4 in Holsteins diagnosed with scours on day 10 (uncorrected P = 0.07). Conversely, Jerseys that were scouring at enrollment but healthy on day 10 at the time of the DGE sample only appeared to upregulate TLR4 (n = 15; uncorrected P = 0.04). Similarly, upregulation of IFNG and NFKB1 also was indicated in Jerseys diagnosed with scours on day 1 but healthy on day 10 (uncorrected P < 0.05).
When calves with systemic enteritis were analyzed based on the day of diagnosis (Table 8 and Figs 2 and 3), those that were affected at the time of enrollment but returned to a healthy status ≥7 days prior to the DGE sample (n = 3 Holstein; n = 4 Jersey) upregulated IFNG (B-Y P = 0.01), TLR4 (B-Y P = 0.02), and AQP3 (B-Y P = 0.03) with particularly notable log2 fold-changes (2.33 ± 0.92, 2.18 ± 1.04, and 1.87 ± 1.08, respectively). In addition, uncorrected estimates also suggested upregulation of NFKBI and SLC5AI along with downregulation of FABP2 in those same calves (uncorrected P < 0.1). Although evidence was lacking for overall DGE (B-Y P ≤ 1.0) in the 4 calves that were healthy at the time of enrollment but experienced systemic enteritis on day 10 when sampled for DGE (n = 2 Holstein; n = 2 Jersey), breed-specific heat maps suggested that the 2 Holsteins (Fig 2) might have been more severely affected at that time than the 2 Jerseys (Fig 3).
The various breed-specific differences stimulated further investigation into whether healthy Holstein and Jersey calves might differentially express the genes of interest. Although no genes were differentially expressed based on the B-Y correction, a less conservative estimate (uncorrected p ≤ 0.05) did indicate upregulation of TLR2 and downregulation of KRT8 in healthy Holsteins as compared to healthy Jerseys (Table 9).
Discussion
The findings from this study confirmed our hypothesis that inflammatory and cellular transport signals would be differentially expressed in fecal RNA in response to adaptations related to breeds and GI disease severity and timing. Our results suggested that preweaned dairy calves upregulate the genes AQP3, IFNG, and TLR4 when experiencing undifferentiated GI disease (scours or systemic enteritis; Table 5). A more refined look at GI disease severity and timing indicated that the DGE in fecal RNA was due primarily to systemic enteritis (Table 6), with the inflammatory response and cellular transport signals potentially delayed compared to clinical signs (Table 8) even though all systemically affected calves were treated with flunixin and most (9/13) were treated with antibiotics (Table 3). Additionally, breed-specific findings indicated that upregulation of AQP3, IFNG, and TLR4 in fecal RNA was driven by systemically affected Holsteins, whereas KRT8 was downregulated in Jersey calves with systemic enteritis (Table 6). Furthermore, TLR2 was downregulated in Holstein calves with scours on the day of the DGE fecal sample (Table 7). These findings highlight the importance of discriminating between breed-specific GI disease phenotypes when describing, comparing, and investigating the overall consequences of disease on calf health and wellbeing.
None of the other genes evaluated in this study demonstrated DGE associated with GI disease when using the B-Y correction. This included NOD2 which frequently was left undetected. Unfortunately, the case numbers were small for select DGE assessments of breed-specific impacts of disease severity, timing of disease, treatment effects, and disease resolution or progression. These factors and the inability to track scours duration limited our ability to differentiate expression patterns more robustly. Nonetheless, less conservative estimates of DGE suggested informative breed-specific patterns. Unlike their response to systemic enteritis, Holstein calves did not appear affected by scours aside from evidence of DGE of toll-like receptors (TLRs) on the day of diarrhea (Table 7 and Fig 2). On the other hand, Jersey calves consistently demonstrated a tendency to upregulate IFNG, NFKB1, and TLR4 when affected with either scours or systemic enteritis (Tables 5–7). These findings were more pronounced in systemically affected Jersey calves (Table 6) and were observed as a delayed response to both scours and systemic enteritis (Tables 7 and 8 and Fig 3).
These findings substantiate a previous study suggesting that fecal RNA might be used as a noninvasive tool to investigate intestinal transcriptomic profiling of dairy calves [6]. In that study of 8 neonatal Jersey male calves, upregulation of TLR4 was delayed relative to an increase in fecal scores indicative of mild diarrhea. A similar delay in TLR4 upregulation appeared to be present in the scouring Jersey calves in our study. On the other hand, the apparent upregulation of TLR4 in scouring Holstein calves in our study overlapped with both the occurrence of scours and downregulation of TLR2. Although this TLR2 downregulation may have been influenced by the tendency for higher levels of TLR2 expression in healthy Holstein as compared to Jersey calves (Table 9), a similar downregulation of TLR2 was demonstrated by Rosa et al. [6] suggesting that much of inflammatory response observed in fecal RNA within that study was due to the upregulation of TLR4.
TLRs can detect conserved molecular products of microorganisms and help regulate host adaptive immune responses [22]. Specifically, TLR2 and TLR4 help direct the early innate immune response to microbial challenge [23]. Gram positive lipoteichoic acids (LTA) are recognized by TLR2 whereas TLR4 is responsible for gram negative lipopolysaccharide (LPS) recognition and cell-signaling [24]. Each of these toll-like receptors leads to the expression of distinct inflammatory genes and the induction of various cytokines [23]. For example, an invading pathogen can activate the TLR4 signaling pathway leading to the activation of NFKB1 which promotes the transcription of genes encoding proinflammatory proteins such as IL1B, TNFA, IL6 and IL8 [25]. This may explain the tendency in our study for concomitant upregulation of both TLR4 and NFKB1 in Jersey calves affected with either scours or systemic enteritis. Furthermore, this finding was in agreement with the demonstration by Rosa et al. [6] of NFKB1 upregulation following a similar delay to that observed for TLR4.
In the present study, IFNG was upregulated in response to systemic enteritis across breeds but with greater effect in systemic Holsteins than Jerseys. On the other hand, IFNG only appeared to be upregulated as a delayed response to scours in Jerseys. Although Rosa et al. [6] demonstrated a time-dependent IFNG downregulation in 8 Jersey male calves following mild diarrhea; overall, they showed a pattern of fecal score and pro-inflammatory blood biomarkers (peaking both between 2 and 3 weeks of age) that indicated a pro-inflammatory response during diarrhea in neonatal Jersey calves. It is worth mentioning that a previous study demonstrated that expression of IL23R which stimulates interferon gamma (IFNγ) production by T cells was reduced prior to gradual weaning in Jersey relative to Holstein calves [26]. Regardless of potential breed-specific differences, interferons were originally discovered as agents that interfere with viral replications [27], and IFNγ specifically has well-documented antiviral and immunoregulatory properties [28]. Furthermore, IFNγ has been shown to be both an inducer as well as a regulator of inflammation [29], a duality that confers both resistance and tolerance mechanisms. Although resistance mechanisms are viewed as the primary function of immunity, disease tolerance also is essential in terms of tissue damage control mechanisms that limit the health and fitness costs of infection [30].
The upregulation of AQP3 in our study was primarily a function of systemic disease in Holsteins although there was evidence of an effect in systemically affected Jerseys and across both breeds with scours. Aquaporin-3 is a cell membrane transporter that transports water, glycerol, urea, and other small uncharged solutes across the plasma membrane [31, 32]. It has been shown to be down-regulated during inflammatory bowel disease in humans [33] and rats [34], and mislocalization of Aquaporin-3 within enterocytes was shown to alter the flux of water through enterocytes in mice with bacterial-induced diarrhea [35]. Furthermore, Rosa et al. [6] demonstrated reduced AQP3 expression in the fecal RNA from Jersey male calves experiencing mild diarrhea.
As importantly, Rosa and Osorio [12] demonstrated that mRNA expression of AQP3 tended to be greater in polymorphonuclear leukocytes (PMN) than fecal RNA. Therefore, it is plausible that the fecal RNA samples in our study contained PMN mRNA transcripts and the upregulation of AQP3 and other genes of interest may have been influenced by PMN infiltration into the intestinal lumen. The transmigration of PMNs into the lumen depends on a variety of activities and signaling events with the goal to maintain homeostasis within the intestine. The various signals that direct such migration include the coordinated efforts of cytokines, adhesion molecules, and highly specific chemokines [36].
The infiltration of immune cells may occur due to altered junction connections and impaired intestinal barrier function [37, 38]. Localized and systemic inflammatory conditions are associated with a leaky epithelial barrier and increased intestinal permeability [39]. A leaky intestinal epithelial barrier allows luminal commensal organism components such as LTA and LPS to enter the submucosa and stimulate the immune system leading to the secretion of regulatory cytokines. The interplay of immune cells, LTA, LPS, and cytokines ultimately augments the inflammatory state of the intestine [40]. In fact, it appears that both LTA and LPS can disrupt barrier function by stimulating increased expression of inducible nitric oxide synthase which leads to increased concentrations of nitric oxide [41]. In addition to disrupting tight junction proteins, nitric oxide is implicated in the induction of enterocyte apoptosis and an accelerated rate of intestinal epithelial cell (IEC) shedding potentially exceeding the rate of IEC production and disturbing intestinal homeostasis [42, 43].
This is relevant to the current study in that Rosa et al. [12, 13] demonstrated that even under nondiarrheic conditions, RNA isolated from neonatal dairy calf feces is derived from a considerable number of IECs as evidenced by mRNA expression of KRT8. This held true in our study in that KRT8 mRNA abundance exceeded that from other genes of interest, suggesting that a considerable amount of fecal RNA originated from IECs. Keratins are the largest family of cellular intermediate filament proteins and are predominant in epithelial cells [44]. Specifically, keratin 8 has been associated with mammary epithelial cells and small intestinal mucosa but not immune cells in general or PMNs in particular [12, 45]. Importantly, various lines of evidence suggest that keratins are key regulators of epithelial integrity [46].
The results from our study suggested a tendency for reduced KRT8 expression in fecal RNA from healthy Holstein as compared to Jersey calves. Although not noteworthy in every situation, KRT8 was the only gene demonstrating downregulation in every comparison of diseased versus healthy calves conducted within this study. More specifically, there was a tendency for KRT8 downregulation in Jerseys with scours and substantial evidence of downregulation in systemically affected Jerseys. Studies conducted in humans and mice suggest that keratin 8 is involved in intestinal barrier protection, colonic active ion transport, colorectal hyperplasia, and inflammatory regulation [46]. In fact, indications are that keratin 8 may help suppress NFKB signaling and TLR-dependent inflammation [47], while protecting colonic integrity in support of gut microbiota homeostasis [46]. This aligns with a previous demonstration of associations between the microbiome and the expression of genes regulating the mucosal barrier and innate immunity in neonatal cattle [48]. The interrelatedness and crosstalk between the various components of the GIT (i.e., epithelium, diverse and robust immune arm, and commensal bacteria) cumulatively make the gut the basis for the health and productivity of animals, highlighting that the GIT is responsible for regulating the physiological homeostasis underlying the ability to withstand infectious and noninfectious stressors [49].
The Holstein and Jersey calves in this study were sourced from different dairies. Consequently, there undoubtedly were undocumented differences in dry cow and maternity management impacting calf health. Nonetheless, Jersey calves demonstrated increased levels of TSP relative to Holsteins suggestive of higher maternally derived serum immunoglobulin G concentrations [50]. Even so, it appeared that despite having received enhanced immunological protection from disease, Jersey calves still may have had reduced immune capacity. Holstein calves experiencing systemic enteritis mounted a more robust and potentially more acute DGE response than systemically affected Jerseys to genes that play key roles in the inflammation and cell membrane or cytoplasmic transport. Furthermore, Jersey calves with GI disease had evidence of KRT8 downregulation, which may have adversely affected intestinal barrier protection and inflammatory regulation. These findings align with our previous observations that a greater number of cases of GI disease can progress to necrotizing enteritis with peritonitis in preweaned Jersey as compared to Holstein calves (unpublished data). In addition, Holstein as compared to Jersey calves were less affected by scours with fewer delayed DGE effects. These findings support previous observations by Johnston et al. [26] of breed-specific transcriptional activity suggesting Holstein calves may be better equipped than Jersey calves to rapidly fight pathogen invasion before pathogens can cause disease. That study demonstrated differences between these breeds in the expression of genes involved in immune responses and cell signaling activity. More specifically, Jersey calves demonstrated decreased cellular movement, chemotaxis, and phagocytic functionality relative to Holstein calves [26]. Breed-specific cellular effects may be diverse depending on the pathogens and the nature of inflammation which influence the expression of associated resistance and tolerance mechanisms [51]. Ultimately, a timely and properly controlled immune response is the key to an animal’s resilience based on balancing pathogen elimination (resistance) and tissue damage (tolerance) [52, 53].
Conclusions
This study investigated transcriptional changes in fecal RNA from neonatal Holstein and Jersey calves with GI disease of varying severity. Both breeds demonstrated different levels of transcription for important genes that selectively enhance or alter host innate immune defense mechanisms and modulate pathogen-induced inflammatory responses. These findings indicate divergent susceptibilities and immune capacity that provide insight into individual and population-level markers of disease resilience (resistance and tolerance). Although additional research is required to explain functional mechanisms and bioactivity across dairy populations, defining relationships between molecular regulatory features and phenotypic traits may help reduce disease incidence and improve animal welfare. This is particularly relevant to the preweaned period with its metabolic and physiologic changes and associated increase in disease susceptibility. Ultimately, knowledge of breed-specific immune responses may facilitate targeted prophylactic and therapeutic veterinary interventions, guide improvements in host genetic selection, and clarify the burden of disease on animal wellbeing.
Supporting information
S2 Table. Raw counts for endogenous, housekeeping, negative, and positive probes.
https://doi.org/10.1371/journal.pone.0278664.s002
(XLSX)
Acknowledgments
The authors thank the participating calf ranch and associated personnel for their invaluable assistance with this project.
References
- 1. Gorden PJ, Plummer P. Control, management, and prevention of bovine respiratory disease in dairy calves and cows. Vet Clin North Am Food Anim Pract. 2010;26(2):243–59. pmid:20619182
- 2. Klein-Jöbstl D, Iwersen M, Drillich M. Farm characteristics and calf management practices on dairy farms with and without diarrhea: a case-control study to investigate risk factors for calf diarrhea. J Dairy Sci. 2014;97(8):5110–9. pmid:24881793
- 3.
USDA. Dairy 2007, Heifer Calf Health and Management Practices on U.S. Dairy Operations, 2007. Fort Collins, CO, USA; 2010. Report No.: N550.0110.
- 4. McGuirk SM. Disease management of dairy calves and heifers. Vet Clin North Am Food Anim Pract. 2008;24(1):139–53. pmid:18299036
- 5. Heller MC, Chigerwe M. Diagnosis and treatment of infectious enteritis in neonatal and juvenile ruminants. Vet Clin North Am Food Anim Pract. 2018;34(1):101–17. pmid:29275032
- 6. Rosa F, Busato S, Avaroma FC, Linville K, Trevisi E, Osorio JS, et al. Transcriptional changes detected in fecal RNA of neonatal dairy calves undergoing a mild diarrhea are associated with inflammatory biomarkers. PLoS ONE. 2018;13(1):e0191599. pmid:29373601
- 7. Kelton DF, Lissemore KD, Martin RE. Recommendations for recording and calculating the incidence of selected clinical diseases of dairy cattle. J Dairy Sci. 1998;81(9):2502–9. pmid:9785242
- 8. Compton CWR, Heuer C, Thomsen PT, Carpenter TE, Phyn CVC, McDougall S. Invited Review: A systematic literature review and meta-analysis of mortality and culling in dairy cattle. J Dairy Sci. 2017;100(1):1–16. pmid:28341041
- 9. Bertrand K. Survival of exfoliated epithelial cells: a delicate balance between anoikis and apoptosis. J Biomed Biotechnol. 2011;2011:534139. pmid:22131811
- 10. Donovan SM, Wang M, Monaco MH, Martin CR, Davidson LA, Ivanov I, et al. Noninvasive molecular fingerprinting of host-microbiome interactions in neonates. FEBS Lett. 2014;588(22):4112–9. pmid:25042036
- 11. Kamra A, Kessie G, Chen JH, Kalavapudi S, Shores R, McElroy I, et al. Exfoliated colonic epithelial cells: surrogate targets for evaluation of bioactive food components in cancer prevention. J Nutr. 2005;135(11):2719–22. pmid:16251637
- 12. Rosa F, Osorio JS. Short communication: Comparative gene expression analysis on the enrichment of polymorphonuclear leukocytes and gastrointestinal epithelial cells in fecal RNA from nondiarrheic neonatal dairy calves. J Dairy Sci. 2019;102(8):7464–8. pmid:31178184
- 13. Rosa F, Busato S, Avaroma FC, Mohan R, Carpinelli N, Bionaz M, et al. Short communication: Molecular markers for epithelial cells across gastrointestinal tissues and fecal RNA in preweaning dairy calves. J Dairy Sci. 2021;104(1):1175–82. pmid:33162086
- 14. Deelen SM, Ollivett TL, Haines DM, Leslie KE. Evaluation of a Brix refractometer to estimate serum immunoglobulin G concentration in neonatal dairy calves. J Dairy Sci. 2014;97(6):3838–44. pmid:24704239
- 15. Slanzon GS, Ridenhour BJ, Moore DA, Sischo WM, Parrish LM, Trombetta SC, et al. Fecal microbiome profiles of neonatal dairy calves with varying severities of gastrointestinal disease. PLoS ONE. 2022;17(1):e0262317. pmid:34982792
- 16. Li F, Hullar MA, Beresford SA, Lampe JW. Variation of glucoraphanin metabolism in vivo and ex vivo by human gut bacteria. Br J Nutr. 2011;106(3):408–16. pmid:21342607
- 17. Matsuoka K, Kanai T. The gut microbiota and inflammatory bowel disease. Semin Immunopathol. 2015;37(1):47–55. pmid:25420450
- 18. Al Nabhani Z, Dietrich G, Hugot JP, Barreau F. Nod2: The intestinal gate keeper. PLoS Pathog. 2017;13(3):e1006177. pmid:28253332
- 19. Petnicki-Ocwieja T, Hrncir T, Liu Y-J, Biswas A, Hudcovic T, Tlaskalova-Hogenova H, et al. Nod2 is required for the regulation of commensal microbiota in the intestine. Proc Natl Acad Sci U S A. 2009;106(37):15813. pmid:19805227
- 20. Malkov VA, Serikawa KA, Balantac N, Watters J, Geiss G, Mashadi-Hossein A, et al. Multiplexed measurements of gene signatures in different analytes using the Nanostring nCounter Assay System. BMC Res Notes. 2009;2:80. pmid:19426535
- 21. Wang H, Horbinski C, Wu H, Liu Y, Sheng S, Liu J, et al. NanoStringDiff: a novel statistical method for differential expression analysis based on NanoString nCounter data. Nucleic Acids Res. 2016;44(20):e151-e. pmid:27471031
- 22. Malmuthuge N, Li M, Fries P, Griebel PJ, Guan LL. Regional and age dependent changes in gene expression of Toll-like receptors and key antimicrobial defence molecules throughout the gastrointestinal tract of dairy calves. Vet Immunol Immunopathol. 2012;146(1):18–26. pmid:22321738
- 23. Supajatura V, Ushio H, Nakao A, Akira S, Okumura K, Ra C, et al. Differential responses of mast cell Toll-like receptors 2 and 4 in allergy and innate immunity. J Clin Invest. 2002;109(10):1351–9. pmid:12021251
- 24. Eicher SD, McMunn KA, Hammon HM, Donkin SS. Toll-like receptors 2 and 4, and acute phase cytokine gene expression in dexamethasone and growth hormone treated dairy calves. Vet Immunol Immunopathol. 2004;98(3–4):115–25. pmid:15010221
- 25. Gilbert FB, Cunha P, Jensen K, Glass EJ, Foucras G, Robert-Granié C, et al. Differential response of bovine mammary epithelial cells to Staphylococcus aureus or Escherichia coli agonists of the innate immune system. Vet Res. 2013;44(1):40. https://doi.org/10.1186/1297-9716-44-40
- 26. Johnston D, Earley B, Cormican P, Kenny DA, McCabe MS, Kelly AK, et al. Characterisation of the Whole Blood mRNA Transcriptome in Holstein-Friesian and Jersey Calves in Response to Gradual Weaning. PLoS ONE. 2016;11(8):e0159707. pmid:27479136
- 27. Isaacs A, Lindenmann J. Virus interference. I. The interferon. Proc R Soc Lond B Biol Sci. 1957;147(927):258–67. pmid:13465720
- 28. Schroder K, Hertzog PJ, Ravasi T, Hume DA. Interferon-gamma: an overview of signals, mechanisms and functions. J Leukoc Biol. 2004;75(2):163–89. pmid:14525967
- 29. Zhang J. Yin and yang interplay of IFN-gamma in inflammation and autoimmune disease. J Clin Invest. 2007;117(4):871–3. pmid:17404615
- 30. Martins R, Carlos AR, Braza F, Thompson JA, Bastos-Amador P, Ramos S, et al. Disease Tolerance as an Inherent Component of Immunity. Annu Rev Immunol. 2019;37:405–37. pmid:30673535
- 31. Rojek A, Praetorius J, Frøkiaer J, Nielsen S, Fenton RA. A current view of the mammalian aquaglyceroporins. Annu Rev Physiol. 2008;70:301–27. pmid:17961083
- 32. Verkman AS. Aquaporins in clinical medicine. Annu Rev Med. 2012;63:303–16. pmid:22248325
- 33. Ricanek P, Lunde LK, Frye SA, Støen M, Nygård S, Morth JP, et al. Reduced expression of aquaporins in human intestinal mucosa in early stage inflammatory bowel disease. Clin Exp Gastroenterol. 2015;8:49–67. pmid:25624769
- 34. Zhao G, Li J, Wang J, Shen X, Sun J. Aquaporin 3 and 8 are down-regulated in TNBS-induced rat colitis. Biochem Biophys Res Commun. 2014;443(1):161–6. pmid:24286754
- 35. Guttman JA, Finlay BB. Subcellular alterations that lead to diarrhea during bacterial pathogenesis. Trends Microbiol. 2008;16(11):535–42. pmid:18819802
- 36. Mumy KL, McCormick BA. The role of neutrophils in the event of intestinal inflammation. Curr Opin Pharm. 2009;9(6):697–701. pmid:19854677
- 37. Koch F, Thom U, Albrecht E, Weikard R, Nolte W, Kuhla B, et al. Heat stress directly impairs gut integrity and recruits distinct immune cell populations into the bovine intestine. Proc Natl Acad Sci U S A. 2019;116(21):10333–8. pmid:31064871
- 38. Brazil JC, Lee WY, Kolegraff KN, Nusrat A, Parkos CA, Louis NA. Neutrophil migration across intestinal epithelium: evidence for a role of CD44 in regulating detachment of migrating cells from the luminal surface. J Immunol. 2010;185(11):7026–36. pmid:20974992
- 39. Eckel EF, Ametaj BN. Invited review: Role of bacterial endotoxins in the etiopathogenesis of periparturient diseases of transition dairy cows. J Dairy Sci. 2016;99(8):5967–90. pmid:27209132
- 40. Mani V, Weber TE, Baumgard LH, Gabler NK. Growth and Development Symposium: Endotoxin, inflammation, and intestinal function in livestock. J Anim Sci. 2012;90(5):1452–65. pmid:22247110
- 41. Singh AK, Jiang Y, Gupta S. Effects of bacterial toxins on endothelial tight junction in vitro: a mechanism-based investigation. Toxicol Mech Methods. 2007;17(6):331–47. pmid:20020957
- 42. Potoka DA, Nadler EP, Upperman JS, Ford HR. Role of nitric oxide and peroxynitrite in gut barrier failure. World J Surg. 2002;26(7):806–11. pmid:11948371
- 43. Williams JM, Duckworth CA, Watson AJ, Frey MR, Miguel JC, Burkitt MD, et al. A mouse model of pathological small intestinal epithelial cell apoptosis and shedding induced by systemic administration of lipopolysaccharide. Dis Model Mech. 2013;6(6):1388–99. pmid:24046352
- 44. Iyer SV, Dange PP, Alam H, Sawant SS, Ingle AD, Borges AM, et al. Understanding the role of keratins 8 and 18 in neoplastic potential of breast cancer derived cell lines. PLoS ONE. 2013;8(1):e53532. pmid:23341946
- 45. Moll R, Divo M, Langbein L. The human keratins: biology and pathology. Histochem Cell Biol. 2008;129(6):705–33. pmid:18461349
- 46. Liu C, Liu ED, Meng YX, Dong XM, Bi YL, Wu HW, et al. Keratin 8 reduces colonic permeability and maintains gut microbiota homeostasis, protecting against colitis and colitis-associated tumorigenesis. Oncotarget. 2017;8(57):96774–90. pmid:29228570
- 47. Dong XM, Liu ED, Meng YX, Liu C, Bi YL, Wu HW, et al. Keratin 8 limits TLR-triggered inflammatory responses through inhibiting TRAF6 polyubiquitination. Sci Rep. 2016;6:32710. pmid:27586056
- 48. Malmuthuge N, Griebel PJ, Guan le L. The Gut Microbiome and Its Potential Role in the Development and Function of Newborn Calf Gastrointestinal Tract. Front Vet Sci. 2015;2:36. pmid:26664965
- 49. Kogut MH, Arsenault RJ. Editorial: Gut Health: The New Paradigm in Food Animal Production. Front Vet Sci. 2016;3:71. pmid:27630994
- 50. Ballou MA. Immune responses of Holstein and Jersey calves during the preweaning and immediate postweaned periods when fed varying planes of milk replacer. J Dairy Sci. 2012;95(12):7319–30. pmid:23063147
- 51. Hilchie AL, Wuerth K, Hancock RE. Immune modulation by multifaceted cationic host defense (antimicrobial) peptides. Nat Chem Biol. 2013;9(12):761–8. pmid:24231617
- 52. Medzhitov R, Schneider DS, Soares MP. Disease Tolerance as a Defense Strategy. Science. 2012;335(6071):936–41. pmid:22363001
- 53. Schneider DS, Ayres JS. Two ways to survive infection: what resistance and tolerance can teach us about treating infectious diseases. Nat Rev Immunol. 2008;8(11):889–95. pmid:18927577