You are on page 1of 14

J. Dairy Sci.

103:8601–8614
https://doi.org/10.3168/jds.2020-18417
© 2020, The Authors. Published by Elsevier Inc. and Fass Inc. on behalf of the American Dairy Science Association®.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

The effect of chronic, mild heat stress on metabolic changes of nutrition


and adaptations in rumen papillae of lactating dairy cows
Mehdi Eslamizad,1 Dirk Albrecht,2 and Björn Kuhla1*
1
Institute of Nutritional Physiology “Oskar Kellner,” Leibniz Institute for Farm Animal Biology (FBN), Wilhelm-Stahl-Allee 2, 18196 Dummerstorf,
Germany
2
Institute of Microbiology, Ernst-Moritz-Arndt-University, Felix-Hausdorff-Straße 8, 17487 Greifswald, Germany

ABSTRACT ing substrates for protein glycosylation. In conclusion,


the mild heat stress did not induce barrier dysfunction
Global warming and accompanying high ambi- or inflammatory responses in the rumen epithelium of
ent temperatures reduce feed intake of dairy cows dairy cows, probably because of adaptations in feed
and shift the blood flow from the core of the body intake behavior and defense mechanisms at the tissue
to the periphery. As a result, hypoxia may occur in level.
the digestive tract accompanied by disruption of the Key words: heat stress, rumen, feeding behavior, tight
intestinal barrier, local endotoxemia and inflammation, junctions, toll-like receptor
and altered nutrient absorption. However, whether the
barrier of the rumen, like the intestine, is affected by
INTRODUCTION
ambient heat has not been studied so far. Lactating
Holstein dairy cows were subjected to heat stress at Heat stress is a major concern for dairy cows, espe-
28°C (temperature-humidity index = 76; n = 5) with cially in tropic and subtropic areas, because it compro-
ad libitum feed intake or to thermoneutral conditions mises animal health, welfare, and milk production, all
at 15°C (temperature-humidity index = 60; n = 5) and of which cause significant economic losses (St-Pierre
pair-feeding to heat-stressed animals for a total of 4 et al., 2003). Heat stress is becoming even more of an
d. Gas exchange and feed intake behavior were mea- issue with the continuous elevation of the mean globe
sured in a respiration chamber, and rumen epithelia temperature (IPCC, 2007). The effect of heat stress on
were taken after slaughter. Heat stress significantly performance, behavioral and physiological adaptation,
reduced meal size and whole-body fat oxidation but metabolism, and immune system of cattle has been
increased meal frequency and carbohydrate oxidation. extensively reported. For example, heat-stressed cows
The mRNA expression of toll-like receptor 4 (TLR4) significantly reduce their feed intake to avoid extra heat
and tight junction proteins and the phosphorylation load generated from digestion and metabolic process-
of TLR4 downstream targets (interleukin-1 receptor- ing of nutrients (Baumgard and Rhoads, 2012). Tra-
associated kinase 4, stress-activated protein kinase, ditionally, the decline in milk production during high
p38 mitogen-activated protein kinase, and nuclear environmental temperatures has been attributed to the
factor k-B) in the rumen epithelium were not affected reduction of feed intake (Beede and Collier, 1986; West,
by heat. The proteomics approach revealed increased 1999). However, it was recently stated that the decline
expression of rumen epithelium proteins involved in in feed intake is not fully responsible for the reduction
the AMP-activated protein kinase (AMPK) and insulin in milk yield of heat-stressed dairy cows (Rhoads et al.,
signaling pathways in heat-stressed cows. Also, proteins 2009a; Wheelock et. al., 2010; Baumgard et al., 2011).
involved in chaperone-mediated folding of proteins were By the use of pair-feeding trials, in which ad libitum fed
upregulated, whereas those involved in antioxidant de- heat-stressed cows were compared with counterparts
fense system were downregulated. Further, we found kept under thermoneutral conditions while receiving the
evidence for increased carbohydrate phosphorylation same amount of feed as heat-stressed animals ingested,
accompanied with an increased flux of carbohydrates it was shown that heat stress exerts direct effects on the
through the hexosamine biosynthetic pathway, provid- metabolism and the endocrine systems of cows beyond
the concomitant effect of energy intake restriction.
Those changes may include an increase in basal and
Received February 24, 2020.
Accepted April 18, 2020. stimulated insulin concentrations (O’Brien et al., 2010;
*Corresponding author: b.kuhla@​fbn​-dummerstorf​.de Wheelock et al., 2010) and alterations in the macronu-

8601
Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8602

trient metabolism favoring the utilization of carbohy- latter approach can be useful for the search of welfare
drates and proteins at the expense of fat (Lamp et al., and stress biomarkers in cattle.
2015). Increased carbohydrate utilization is evident by
lower blood glucose levels, greater glucose disposal af-
MATERIALS AND METHODS
ter insulin injection, increased hepatic gluconeogenesis,
augmented depletion of liver glycogen stores, and im- Animals and Treatments
mune cell activation under heat stress (Baumgard and
Rhoads, 2013). However, the biochemical mechanisms All procedures were approved by the ethics com-
accounting for the increased carbohydrate utilization mittee of the state government in Mecklenburg-West
are still not fully understood. For example, it is still Pomerania, Germany (LALLF M-V/TSD/7221.3–
unclear if the absorption of the primary gluconeogenic 1.1–074/12), and the data described herein are part
precursor propionate via the rumen wall is affected by of a recently published study (Vanselow et al., 2016).
heat stress. More specifically, 10 German Holstein dairy cows in
Early studies reported the alteration of the rumen- established second lactation (192 ± 20 DIM) were kept
intestinal epithelium during heat stress. Under condi- under the same normal environmental conditions in a
tions of high ambient temperatures, blood flow is di- freestall barn at the Leibniz Institute for Farm Animal
verted from the viscera to the periphery (Kregel, 2002; Biology (FBN) with daily ad libitum feeding. The ra-
Lambert et al., 2002), leading to a local oxygen and tion consisted on the DM bases of 6.12 kg of corn silage,
nutrient deprivation at the intestinal epithelium and 6.6 kg of grass silage, 0.9 kg of wheat straw, 1.7 kg of
an ultimate reduction of the intestinal barrier function hay, 2.1 kg of concentrate, 1.1 kg of corn meal, 0.8 kg
(Pearce et al., 2013). As a consequence of an increased of extracted canola meal, 0.3 kg of extracted soybean
intestinal epithelial permeability, the transmission of meal, 0.3 kg of wheat, and 0.2 kg of minerals, with a
microbial components from the intestinal tract into the total ME = 10.6 MJ/kg of DM.
circulation may increase (Lim et al., 2007; Pearce et Cows were transported to a climate-controlled cham-
al., 2013), suggesting a central role of the intestine, and ber and kept at thermoneutral conditions [15°C, 64%
potentially the rumen, in endotoxemia and heat stroke relative humidity, and temperature-humidity index
pathophysiology (Leon, 2007; Rhoads et al., 2009b). (THI) = 60] for 6 d to measure ad libitum feed intake.
The intestinal toll-like receptor 4 (TLR4) is activated Thereafter, 5 randomly selected cows were exposed to
in response to injury of the gut (i.e., after damage of the 28°C (52% relative humidity, THI = 76) with ad libitum
intestinal epithelial lining; Abreu, 2010). Recognition of feeding (heat stress group; HS). The ad libitum feed in-
LPS by TLR4 triggers cellular signaling through in- take of 1 HS cow during heat exposure was calculated as
terleukin-1-receptor-associated kinases (IRAK), which percentage of the mean daily intake determined for the
ultimately results in the activation of nuclear factor- 6-d ad libitum feeding period at THI = 60. The remain-
kB (NF-κB) and mitogen-activated protein kinases ing 5 cows were kept continuously at 15°C (THI = 60)
(MAPK). Whether the activation of the TLR4 path- but, with a 1-d time lag, were pair-fed (PF) to allocate
way in the rumen papillae is also involved in heat-stress the same level of feed intake per kilogram of BW as
induced endotoxemia and which molecular mechanisms HS cows consumed. This experimental design allowed
in the rumen epithelium contribute to increased whole- us to differentiate between feed intake and heat-stress
body carbohydrate utilization in heat-stressed dairy effects. Cows were housed in the climate chamber for
cows have not been studied so far. We hypothesized 2.5 d, immediately transferred to a climate-controlled
that (1) heat stress would cause loosening of the tight respiration chamber, and kept for another 36 h under
junctions in the rumen epithelium of dairy cows and the same climatic and feeding conditions, amounting to
that the resulting exposure to luminal bacterial com- a total of 4 d of PF or HS challenge. The stay in the
pounds would activate local inflammatory responses via respiration chamber included a 12-h gas-equilibration
the TLR4 signaling pathway and (2) heat stress affects period, followed by a 24-h gas-exchange measuring pe-
global expression of rumen papillae proteins including riod. In climate and respiration chambers, cows were
those regulating absorption and facilitating whole-body fed at 0700 and 1600 h and milked at 0630 and 1630 h.
carbohydrate utilization. To pursue these hypotheses, In parallel to the milking in respiration chamber, the
we exposed mid-lactation dairy cows for 4 d to either rectal temperature was measured.
continuous heat stress or pair-feeding at thermoneu-
trality, sampled the rumen papillae, and performed Macronutrient Metabolism and Feed Intake Behavior
targeted Western blot and PCR analyses to evaluate
expression of tight junction and TLR4 pathway pro- The dynamics of feed and water intake, gas exchange,
teins and RNA, as well as untargeted proteomics; the and physical activity was recorded for 24 h as previous-
Journal of Dairy Science Vol. 103 No. 9, 2020
Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8603

ly described by (Derno et al., 2009). Briefly, individual (#9252; 1:1,000) and p-SAPK/JNK (#9251; 1:300);
feed intake was measured every 6 min by feed disap- and NF-κB (#4764; 1:500) and p-NF-κB (#3033;
pearance from the feeding bin located on a scale and 1:300; all from Cell Signaling Technology, Cambridge,
connected to an electronic registration device. A meal UK). The membranes were washed 3 times with TBST
was defined when feed intake was greater than 200 g/6 and incubated with horseradish peroxidase-conjugated
min and lasted for more than 2 consecutive measuring goat anti-rabbit secondary antibody (1:10,000; BD
intervals. Water intake was measured by a water meter Pharmingen Inc., San Diego, CA) for 1 h. Finally, blots
connected to an electronic registration device. Physical were washed again 3 times in TBST followed by incuba-
activities were detected by a modified infrared-based tion with chemiluminescent reagent for 1 min (Thermo
motion detector (IS 120, Steinel, Herzebrock-Klarholz, Fisher Scientific) and final exposure to hyperfilms (GE
Germany) by converting movements of the animal to Healthcare, Chalfont St Giles, UK). After detection of
the impulses. Concentrations of CO2 and CH4 in the bands, hyperfilms were scanned and signal intensities
chamber were analyzed by infrared absorption (UNOR were quantified using ImageJ 1.49 software (https:​/​/​
610, Maihak, Hamburg, Germany), and the concentra- imagej​.nih​.gov/​ij/​download​.html). The ratio between
tion of O2 was analyzed paramagnetically (OXYGOR phosphorylated and nonphosphorylated forms of above-
610, Maihak) every 6 min. Subsequently, based on the mentioned proteins was calculated.
amount of O2 consumed and CO2 and CH4 produced,
metabolic heat production, fat oxidation, and carbohy- Quantitative Reverse Transcription-PCR Analysis
drate oxidation were calculated (Eslamizad et al., 2015).
Immediately after the stay in the respiration chamber, Tissue was crushed with a mortar and pestle in liquid
cows were transported to the institute’s slaughterhouse N2 and total RNA was extracted from 65 mg of tissue
(300 m) within 10 min. After captive bolt stunning and powder with TriFast Reagent (Peqlab, Erlangen, Ger-
exsanguination of the cows, tissue samples from the many) and incubated with DNaseI digest (innu PREP,
rumen papillae (~30 cm lateral from the esophageal Analytik Jena, Jena, Germany) according to the manu-
entrance) were obtained within 15 min, snap-frozen in facturer’s instructions. The RNA quality was assessed
liquid nitrogen, and stored at −80°C. using an Agilent 2100 Bioanalyzer (Santa Clara, CA),
yielding RNA integrity number factors between 7.1
Western Blot and 9.1 (mean 8.6). Complementary DNA was synthe-
sized (375 ng of total RNA) using Revert Aid Reverse
Whole proteins from rumen tissue were extracted Transcriptase (Thermo Fisher Scientific, Dreieich,
in a lysis buffer containing 50 mM Tris-HCl (pH 7.8), Germany) and random Hexamer primers (Metabion
1 mM EDTA, 10 mM NaF, 1% IGEPAL CA-630 International, Planegg/Steinkirchen, Germany) and
(Sigma-Aldrich, St. Louis, MO), 0.1% Triton X100, stored at −80°C until use. Intron overspanning primers
0.5% deoxycholic acid, 0.1% SDS, and cocktail protease were designed using Primer3 Plus software (Untergas-
inhibitor (Roche, Basel, Switzerland). Protein concen- ser et al., 2012) or PrimerBLAST (Ye et al., 2012).
trations were measured immediately after extraction Complementary DNA was amplified in triplicate on a
using a Bradford assay kit (Thermo Fisher Scientific, Light Cycler 2.0 (Roche) using specific primers (Sup-
Waltham, MA), and the extracts were stored at −80°C plemental Table S1, https:​/​/​doi​.org/​10​.3168/​jds​.2020​
for later use. Equal protein amounts (50 µg) of boiled -18417) and Light Cycler FAST DNA Master PLUS
samples were separated by SDS-PAGE and electro- SYBR Green I Reaction Mix (Roche). Amplicons were
transferred onto nitrocellulose membranes (Whatman sequenced on an ABI 3130 Genetic Analyzer (Life
Protran BA 83, Dassel, Germany). Membranes were Technologies GmbH, Darmstadt, Germany) and correct
stained with Ponceau S to confirm equal loading. sequence was confirmed. The efficiency of amplifica-
Membranes were blocked for 1 h in Tris-buffered saline- tion was calculated using LinRegPCR software, version
Tween 20 (TBST) solution (20 mM Tris, 0.9% NaCl, 2014.4.1.1 (Ruijter et al., 2013). The gene expression
0.1% Tween 20, pH 7.4) containing 3% BSA before stability of 3 candidate reference genes—CKLF-like
incubation at 4°C with primary antibody in the same MARVEL transmembrane domain-containing protein 6
solution overnight. The following rabbit antibodies and (CMTM6), 60S acidic ribosomal protein P0 (RPLP0),
dilutions were used: interleukin-1 receptor-associated and ELKS/Rab6-interacting/CAST family member 1
kinase 4 (IRAK4; #4363; 1:500) and phosphorylated (ERC1)—was determined in geNorm, and RPLP0 and
(p)-IRAK (p-IRAK4; #11927; 1:300); p38 mitogen- ERC1 were used as references genes for final analysis.
activated protein kinase (p38MAPK; #9212; 1:1,000) Efficiency-corrected data were analyzed as the ratio
and p-p38MAPK (#9211; 1:500); stress-activated between the target gene mRNA abundance and the
protein kinase (SAPK)/c-Jun-N-terminal kinase (JNK) geometric mean of the reference genes using qbase Plus
Journal of Dairy Science Vol. 103 No. 9, 2020
Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8604

software (Biogazelle, Gent, Belgium), and the ratio is list are as follows: repeat count = 1, repeat duration =
presented as relative expression. 30 s, exclusion list size = 500, and exclusion duration =
30. Data files were searched against the National Center
Proteome Analysis for Biotechnology Information Bovine database (www​
.ncbi​.nlm​.nih​.gov/​) using SEQUEST with the common
A total of 50 µg of protein from tissue extracts was contaminant ‘kreatin’ specified. The Sequest search was
loaded on a 1-dimensional 12% SDS-PAGE gel and the carried out considering the following parameters: parent
protein bands were stained with colloidal Coomassie ion mass tolerance of 10 ppm, fragment ion mass toler-
(Roti-Blue, Carl Roth, Germany). Fourteen slices per ance of 0.50 Da, and Met oxidation (+15.99492 Da).
lane (1 lane per animal) were excised using a sterile Each Sequest search included the data from all 14 gel
scalpel. Stained gel slices were transferred into 1.5-mL slices per lane and results loaded into Scaffold software
Eppendorf tubes (Hamburg, Germany) and washed (version 4.8.7, Proteome Software Inc., Portland, OR).
twice with 100 µL of a solution of 50% CH3OH and The Scaffold software was used to validate MS/MS
50% 50 mM NH4HCO3 for 30 min, and once with 100 based peptide and protein identifications. Peptide iden-
µL of 75% CH3CN for 10 min. After drying at 37°C for tifications were accepted if they could be established
20 min, 10 µL of trypsin solution containing 4 µg/mL at greater than 93.0% probability to achieve a false
porcine trypsin (Promega, Madison, WI) was added discovery rate less than 0.1% by the Peptide Prophet
and incubated overnight at 37°C. For extraction, gel algorithm (Keller et al., 2002) with Scaffold delta-mass
bands were covered with 60 µL of 0.1% trifluoroacetic correction. Protein identifications were accepted if
acid in 50% CH3CN and incubated for 30 min under they could be established at greater than 70.0% prob-
shaking. The peptide-containing supernatant was ability to achieve a false discovery rate less than 1.0%
transferred into a clear glass vial and dried at 45°C for and contained at least 2 identified peptides. Protein
100 min in a concentrator (Eppendorf). The dry pep- probabilities were assigned by the Protein Prophet
tides (nonreduced or alkylated) were resuspended in algorithm (Nesvizhskii et al., 2003). Proteins that
10 µL of 50%/49.5%/0.5% (vol/vol/vol) CH3CN/H2O/ contained similar peptides and could not be differenti-
trifluoroacetic acid) and separated and measured online ated based on MS/MS analysis alone were grouped to
by electrospray ionization (ESI)-mass spectrometry us- satisfy the principles of parsimony. Relative quantifi-
ing a Proxeon Easy nLCII- system (Thermo Scientific) cation was performed using the “Normalized NSAF”
coupled to a Thermo Scientific LTQ Orbitrap-XL mass tool, and the t-test feature of the Scaffold software was
spectrometer. A gradient of 0.5%/min (buffer A = used to identify differentially expressed proteins at P
0.1% formic acid in water, Optima LC/MS; buffer B < 0.05. Results were exported as a Microsoft Excel file
= 0.1% formic acid in 99.9% acetonitrile, Optima LC/ (Supplemental Table S2, https:​/​/​doi​.org/​10​.3168/​jds​
MS; Fisher Scientific) was used. Data were acquired on .2020​-18417).
an LTQ Orbitrap-XL mass spectrometer using a 0.1 ×
200 mm column with C18 Aeris Peptide (Phenomenex, Statistics and Bioinformatics Approach
Torrance, CA) at a flow rate of 0.3 µL/min. For MS and
MS/MS analyses, a full survey scan in the Orbitrap-XL For Western blot, PCR, and respiration chamber
with a mass range (m/z 300–2,000) and a Fourier trans- data, differences between HS and PF cows were tested
form (FT) resolution of 30,000 was followed by data- using unpaired t-tests of SAS (SAS Institute Inc., Cary,
dependent fragmentation experiments of the 5 most NC). For proteome analysis, differentially expressed
intense ions. Data were acquired in a data-dependent proteins were submitted to the ClueGo App of the Cy-
“top 5” format, selecting the most abundant precursor toscape software version 3.7.0 (Shannon et al., 2003).
ions from the FTMS scan (mass range 300–2,000 Da). Upregulated and downregulated proteins were assigned
The FTMS scans were acquired with a resolution of to the software as different clusters. The protein list
30,000 and a target value of 1.2 × 106 in the Orbitrap was loaded to the Gene Ontology and Kyoto Encyclo-
analyzer. The ion-trap MS scans were acquired with pedia of Genes and Genomes (KEGG; www​.genome​
unit mass resolution in the LTQ using 3,000 as target .jp/​kegg/​?sess​=​ebfe2ad23e021e38540f798c803dd061)
value, 2 as the default charge state, and a lower inten- database. The ontology selection on the base of biologi-
sity threshold for MS2 of 1,750 counts. The normalized cal processes, immune system activation, and KEGG
collision energy in the collision-induced dissociation was performed by the right-side hypergeometric statis-
was 35 eV and a dynamic exclusion was defined by tic test and its probability was corrected by the Bonfer-
a list size of 500 with exclusion duration of 30 s. The roni stepdown method. The MS proteomics data has
spectra were acquired in the LTQ via collision-induced been uploaded to the ProteomeXchange Consortium
dissociation. The parameters for the dynamic exclusion (http:​/​/​www​.proteomexchange​.org/​) via the PRIDE
Journal of Dairy Science Vol. 103 No. 9, 2020
Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8605

repository with the data set identifier PXD016510 and than PF cows (40.2 ± 0.3°C vs. 38.4 ± 0.1°C, P <
10.6019/PXD016510. 0.05). The HS cows consumed DM at slower rate than
PF cows (P < 0.05). This was associated with more fre-
quent meals, smaller meal sizes, and shorter duration of
RESULTS
meals in HS relative to PF cows (P < 0.05). The mean
Milk Yield and Feed Intake During Ad Libitum time interval between feeding bouts was also shorter in
Feeding at Thermoneutrality HS than PF cows (P < 0.05), while the total time spent
eating or daily water intake was not affected between
Daily milk yield during ad libitum feeding at ther- challenges. However, HS cows had a higher drinking
moneutral conditions (HS: 29.5 ± 1.7 L, PF: 26.0 ± frequency and consumed more water per unit of BW
1.7 L, P = 0.18), as well as during the challenge period (P < 0.05) and tended to oxidize higher amounts of
(HS: 22.1 ± 1.6 L, PF: 22.4 ± 1.9 L, P = 0.88), did carbohydrates per kilogram of DMI than PF cows (P
not differ between PF and HS cows. The percentage = 0.07). According to the feeding regimen, PF cows
decrease in milk yield amounted to 25% for HS and consumed their last meal, which was 2.0 to 17.9 kg of
14% for PF cows. During the ad libitum feeding period OM, 12 to 15 h before slaughter. In contrast, HS cows
at THI = 60, DMI was comparable between HS (17.2 consumed their last meal, which was 0.6 to 11.0 kg, 2.5
± 1.6 kg/d) and PF (16.7 ± 3.6 kg/d) cows (P = 0.05). to 17 h before slaughter.
Additionally, DMI normalized to BW was not different
between groups (HS: 3.20 ± 0.37 kg/d vs. PF: 2.40 ± The TLR4 Pathway and TJ Proteins
0.64 kg/d, P = 0.32). However, BW was greater in PF
than HS cows (Table 1). To examine if there is heat stress–specific activation
of the TLR4 pathway in the rumen papillae, RT-PCR
Feeding Behavior, Macronutrient Oxidation, and Western blot analyses were performed. There was
and Physical Activity During Challenge no significant difference in abundance of TLR4 mRNA
or in phosphorylation of the downstream targets of
During the challenge period, DMI was again greater TLR4 involving IRAK4, p38MAPK, SAPK/JNK, and
in PF than HS cows, but according to the experimental NF-κB (Figure 1), indicating lack of a heat stress–
design, DMI normalized to BW did not differ between specific inflammatory response in the rumen papillae.
HS and PF cows (Table 1). Accordingly, metabolic heat Comparably, the mRNA abundances of tight junction
production was comparable between groups (P = 0.27), proteins, including CLDN3, OCLN, TJP1, TJP2, were
but the mean rectal temperature was greater for HS also not different between HS and PF cows (Figure 2).

Table 1. Dry matter intake, feeding behavior, macronutrient oxidation, and activity of mid-lactation dairy
cows after 4 d of continuous heat stress (HS) compared with pair-fed (PF) cows

Treatment

Item1 HS PF P-value
BW, kg 546 ± 24 688 ± 29 0.01
DMI, kg/d 8.7 ± 1.1 13.0 ± 1.0 0.02
DMI·BW−1·100 1.59 ± 0.16 1.75 ± 0.13 0.49
Meal size, kg of DM·meal−1 0.76 ± 0.12 4.14 ± 1.37 0.03
Meal duration, min·meal−1 28.2 ± 2.8 57.1 ± 9.1 0.01
Meal frequency, d−1 11.6 ± 0.9 4.5 ± 1.9 0.01
Eating rate, kg of DM·min−1 0.032 ± 0.004 0.067 ± 0.017 0.06
Eating time, min·d−1 241.2 ± 18.9 175.5 ± 36.2 0.13
Inter-meal interval, min 96.9 ± 10.6 429.2 ± 133.3 0.03
Water intake, mL·kg of BW−1·d−1 91.7 ± 5.8 69.7 ± 7.0 0.04
Water intake per event, L 2.45 ± 0.17 3.46 ± 0.59 0.11
Drinking frequency, d−1 20.6 ± 0.8 14.3 ± 2.0 0.02
COX, g·kg−0.75·DMI−1·d−1 3.99 ± 0.29 2.97 ± 0.41 0.07
FOX, g·kg−0.75·d−1 10.56 ± 1.26 9.04 ± 0.64 0.36
Activity, 1,000 pulses·d−1 139.6 ± 29.6 336.1 ± 36.6 <0.01
Prandial HPR, kJ·BW−0.75·meal−1 37.9 ± 3.2 192.7 ± 68.7 0.04
mHP, kJ·BW−0.75·d−1 990 ± 26 931 ± 39 0.27
1
COX = carbohydrate oxidation; FOX = fat oxidation; HPR = heat production rise; mHP = metabolic heat
production.

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8606

Figure 1. Effects of 4 d of continuous heat stress (28°C; temperature-humidity index, THI = 76) and thermoneutrality (15°C; THI = 60) on
the activation of the TLR4 signaling pathway in rumen epithelium of dairy cows. The mRNA abundance of TLR4 was normalized to 2 reference
genes, RPLP0 and ERC1 (A). Western blot analysis of the abundance of phosphorylated (p) stress-activated protein kinase (SAPK; B), inter-
leukin-1 receptor-associated kinase 4 (IRAK4; C), p38 mitogen-activated protein kinase (MAPK; D), and nuclear factor k-B (NF-kB; E) were
normalized to the corresponding nonphosphorylated form of each protein. Data are represented as the mean and SE from 5 heat stress (HS) and
5 pair-fed (PF) cows and P-values were from Student’s t-test.

Proteome Analysis comparison using the Scaffold software revealed 154


differentially (P < 0.05) expressed proteins between
A total of 1,086 to 1,642 rumen papillae proteins groups, with 92 upregulated and 62 downregulated
were identified for each animal by MS/MS. The t-test proteins in HS relative to PF cows (Supplemental Table

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8607

Figure 2. The mRNA abundances of rumen tight junction proteins, involving CLDN3 (A), OCLN (B), TJP1 (C), and TJP2 (D) in dairy
cows exposed to continuous heat stress (HS; 28°C; temperature-humidity index, THI = 76) or pair-fed thermoneutral conditions (PF; 15°C; THI
= 60) for 4 d. Data are presented as the mean and SE from 5 HS and 5 PF cows and P-values were from Student’s t-test.

S2, https:​/​/​doi​.org/​10​.3168/​jds​.2020​-18417). Submis- S2). Increased glycolysis, together with increased ex-
sion of the list comprising differentially expressed pro- pression of ATP citrate synthase (ACLY), acyl-CoA
teins to Gene Ontology Biological Process and KEGG synthetase short chain family member 2 (ACSS2), and
pathway analysis extracted 55 proteins, from which dihydrolipoyl dehydrogenase (DLD; Table 2), should
33 activated and deactivated pathways were identified increase acetyl-CoA (Ac-CoA) biosynthesis (Table 2,
based on the P < 0.05 criterion (Figure 3). We did not Figure 3). On the other hand, enzymes of the fatty
find any term or pathway associated with either the acid β-oxidation pathway producing Ac-CoA, such as
immune system or the transport of SCFA or minerals enoyl-CoA hydratase and 3-hydroxy acetyl-CoA dehy-
although some of those proteins were identified (Sup- drogenase, were downregulated under heat stress com-
plemental Table S2). Rather, the majority of differ- pared with isoenergetic conditions at thermoneutrality
entially expressed proteins were found to be involved (Supplemental Table S2). Reduced expression of the
in carbohydrate and protein catabolism, as well as in mitochondrial respiratory chain enzymes cytochrome
the cellular heat-shock response, cellular trafficking, C oxidase (subunit 2) and ATP synthase (subunit B1)
and antioxidant defense. More specifically, enzymes diminish the ATP tone and support the activation
involved in glycolysis and gluconeogenesis, particularly of the 5′ adenosine monophosphate-activated protein
in the phosphorylation and activation of single sugars kinase (AMPK) and the insulin signaling pathways of
such as xylulose kinase, fucokinase, and ribokinase, HS cows via upregulation of calcium-binding protein
were upregulated in HS relative to PF cows (Supple- 39 (CAB39), calcium-binding protein 39-like (CA-
mental Table S2). In addition, the enzyme glutamine:​ B39L), ELAV-like RNA binding protein 1 (ELAVL1),
fructose​-6​-phosphate transaminase 1 (GFPT1), which fructose-bisphosphatase 1 (FBP1), protein phospha-
controls the flux of glucose into the hexosamine path- tase 2 regulatory subunit B α (PPP2R2A), protein
way and thus protein glycosylation, was more highly kinase AMP-activated noncatalytic subunit gamma 1
expressed in HS than PF cows (Supplemental Table (PRKAG1), Ras-related protein Rab-11B (RAB11B),

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8608

Ras-related protein Rab-14 (RAB14), and the adapter tion and protein degradation, chaperone-mediated
molecules crk (CRK), FBP1, flotillin 2 (FLOT2), protein folding, as well as the inhibition of serine and
PRKAG1, AMP-dependent protein kinase type II-α threonine kinase activity in rumen epithelium of HS
(PRKAR2A), and type II-β (PRKAR2B) proteins, relative to PF cows (Figure 3). Vice versa, downregu-
respectively (Table 2). Moreover, upregulation of pro- lation of dynactin subunit 1 (DCTN1), heterogeneous
teins such as cullin 1 through 4 (CUL1, CUL2, CUL3, nuclear ribonucleoprotein U (HNRNPU), nuclear
CUL4A), eukaryotic translation initiation factor 2 mitotic apparatus protein 1 (NUMA1), myosin 5A
subunit 1 (EIF2S1), heat shock protein (HSP) 105 (MYO5A), RAB11B, aminopeptidase N (ANPEP),
kDa (HSPH1), protein OS-9 (OS9), phospholipase A- glutathione S-transferase A5 (GSTA5), glutathione S-
2-activating protein (PLAA), protein transport protein transferase Mu 1 (GSTM1), and P1 (GSTP1) reduces
Sec23A (SEC23A) and 61A1 (SEC61A1), phosducin- mitotic spindle organization, melanosome transport,
like protein 3 (PDCL3), chaperonin containing TCP1 while compromising keratin and pigment metabolism
subunit 4 (CCT4), cysteine and histidine-rich domain- and glutathione metabolism (Table 2, Figure 3). Al-
containing protein 1 (CHORDC1), HSPH1, heat though the regulation of oxidative stress–induced cell
shock protein β-1 (HSPB1), axin interactor (AIDA), death was classified as deactivated in HS compared
CHORDC1, and PRKAR2A support protein process- with PF cows, the process involved increased expres-
ing in the endoplasmic reticulum, protein ubiquitina- sion of HSPH1 and SFPQ, but reduced abundance of

Figure 3. Functional classification of differentially expressed proteins in rumen papillae of dairy cows after 4 d of continuous heat stress based
on Gene Ontology biological process and Kyoto Encyclopedia of Genes and Genomes pathway analyses using ClueGo software (Shannon et al.,
2003; P < 0.05). Thirty-three activated and deactivated pathways (number of nodes) were identified based on the P < 0.05 criterion. Circles in
red represent pathways or processes with upregulated proteins (cluster 1) and blue circles indicate processes with downregulated proteins (cluster
2). The node size refers to the number of differentially regulated proteins involved in each process or pathway, and the intensity of the node color
represents the degree of specificity of the process or pathway to clusters. Gray nodes represent relatively low specificity of regulated processes
because up- and downregulated proteins are involved in these processes. AMPK = AMP-activated protein kinase.

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8609
Table 2. Leading terms of the enriched Gene Ontology biological process or Kyoto Encyclopedia of Genes and Genomes pathway analysis and
differentially expressed proteins involved in the rumen epithelium of dairy cows after 4 d of heat stress compared with pair-fed thermoneutral
counterparts

Pathway or process Count1   Protein name2 P-value3


Activated4      
 Glycolysis/Gluconeogenesis 4 ACSS2, ALDH3A2, DLD, FBP1 2.9 E-3
  Acetyl-CoA biosynthetic process 3 ACLY, ACSS2, DLD 1.0 E-2
  AMP-activated protein kinase signaling 8 CAB39, CAB39L, ELAVL1, FBP1, PPP2R2A, 1.0 E-3
PRKAG1, RAB11B↓, RAB14↓
  Insulin signaling 6 CRK, FBP1, FLOT2, PRKAG1, PRKAR2A, 1.6 E-3
PRKAR2B
  Protein processing in the endoplasmic reticulum 8 CUL1, EIF2S1, HSPH1, OS9, PLAA, SEC23A, 7.7 E-4
SEC61A1, RRBP1↓
  Negative regulation of protein serine/threonine kinase 5 AIDA, CHORDC1, PRKAR2A, PRKAR2B↓, 3.5 E-2
  activity HSPB1↓
  Chaperone-mediated protein folding 4 CCT4, CHORDC1, HSPH1, HSPB1↓ 2.3 E-2
  Protein ubiquitination involved in ubiquitin-mediated 6 CUL1, CUL2, CUL3, CUL4A, OS9, PDCL3 2.0 E-2
   protein catabolic process
Deactivated5      
  Regulation of mitotic spindle organization 4 DCTN1, HNRNPU, NUMA1, VPS4B↑ 4.1 E-3
  Melanosome transport 3 DCTN1, MYO5A, RAB11B 1.0 E-2
  Glutathione metabolism 4 ANPEP, GSTA5, GSTM1, GSTP1 1.7 E-4
  Regulation of oxidative stress-induced cell death 5 HDAC6, HSPB1, PYCR1, HSPH1↑, SFPQ↑ 5.4 E-5
1
The number of proteins involved in the pathway or process.
2
ACLY = ATP citrate synthase; ACSS2 = acyl-CoA synthetase short chain family member 2; DLD = dihydrolipoyl dehydrogenase; AIDA = axin
interactor, dorsalization associated; CHORDC1 = cysteine and histidine-rich domain-containing protein 1; HSPB1 = heat shock protein β-1;
PRKAR2A = cAMP-dependent protein kinase type II-α regulatory subunit; PRKAR2B = cAMP-dependent protein kinase type II-β regulatory
subunit; CCT4 = t-complex protein 1 subunit delta; HSPH1 = heat shock protein 105 kDa; CUL1 = cullin 1; CUL2 = cullin 2; CUL3 = cullin
3; CUL4A = cullin 4A; OS9 = protein OS-9; PDCL3 = phosducin-like protein 3; DCTN1 = dynactin subunit 1; HNRNPU = heterogeneous
nuclear ribonucleoprotein U; NUMA1 = nuclear mitotic apparatus protein 1; VPS4B = vacuolar protein sorting-associated protein 4B; MYO5A
= myosin VA; RAB11B = ras-related protein Rab-11B; CD109 = CD109 molecule; CAB39 = calcium-binding protein 39 (MO25alpha) (protein
Mo25); CAB39L = calcium-binding protein 39 like; ELAVL1 = ELAV-like protein; FBP1 = fructose-1,6-bisphosphatase 1; PPP2R2A = serine/
threonine-protein phosphatase 2A 55 kDa regulatory subunit B, PRKAG1 = 5′-AMP-activated protein kinase subunit gamma-1; EIF2S1 =
eukaryotic translation initiation factor 2 subunit 1 (eukaryotic translation initiation factor 2 subunit α); PLAA = phospholipase A2 activating
protein; SEC23A = protein transport protein SEC23; SEC61A1 = protein transport protein Sec61 subunit α isoform 1; CRK = CRK proto-
oncogene adaptor protein; FLOT2 = flotillin-2; ALDH3A2 = aldehyde dehydrogenase; SFPQ = splicing factor proline- and glutamine-rich;
HDAC6 = histone deacetylase; PYCR1 = pyrroline-5-carboxylate reductase 1, mitochondrial.
3
P-value calculated according to a hypergeometric test and corrected with Bonferroni stepdown using ClueGo software (Shannon et al., 2003).
4
Pathway or processes with corresponding proteins upregulated by heat stress unless indicated by an arrow (↓) representing downregulated
proteins.
5
Pathway or processes with corresponding proteins downregulated by heat stress unless indicated by an arrow (↑) representing upregulated
proteins.

HDAC6, PYCR1, and HSPB1 (Table 1). Accordingly, As designed in the experiment, animals in both
the regulation of oxidative stress–induced cell death is groups consumed same amount of DM per kilogram
not unidirectional, as illustrated in gray in Figure 3. of BW (Table 1). However, the meals (feeding bouts)
were distributed more evenly throughout the day in HS
DISCUSSION compared with PF cows. Accordingly, HS cows ingested
feed more frequently but had reduced duration and
The aim of the present study was to test the hy- size of each meal compared with PF counterparts. The
pothesis that heat stress would affect tight junctions, reduction in feed intake along with reduced physical
inflammatory responses via the TLR4 signaling path- activity appears to be associated with the old concept
way, and global expression of rumen papillae proteins. that feed intake and activity are mechanisms for ther-
We compared HS and PF cows, but did not include ad mal regulation (Brobeck, 1948). Analysis of dynamics
libitum–fed cows at thermal neutral conditions. The of periprandial heat production mainly influenced by
latter group would allow identification of the adapta- diet-induced thermogenesis revealed that the average
tion response from ad libitum intake at thermoneutral- rise during a single meal was less pronounced in HS
ity to reduced feed intake at heat stress; however, the than PF cows. However, daily metabolic heat produc-
chosen experimental design was sufficient to identify tion was comparable between groups, indicating that
heat stress–specific effects at isoenergetic and isonutri- an altered feeding behavior avoids excessive heat load
ent feeding levels. during and after meals. Also, a 4-d heat-stress period

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8610

caused a significant increase in water intake via more the expression of intestinal tight junction proteins are
frequent drinking, most likely to replace the evapora- contradictory in literature. Exposure of caco-2 cells to
tive water loss (Sawka et al., 2001). 39 or 41°C increased occludin expression (Dokladny et
Feeding behavior affects rumen fermentation and thus al., 2008), and pigs exposed to 35°C for 24 h showed
short-chain fatty acid concentrations. Accordingly, the increased expression of claudin 3 and occludin (Pearce
expression of genes associated with short-chain fatty et al., 2013), yet incubation of intestinal epithelial
acid transport and acid-base balance may change with (IEC-6) cells at 42°C for 6 h decreased the expression
feeding; however, we did not observe differences in pro- of the tight junction proteins occludin, ZO1, and E-
tein expression of monocarboxylate transporter 1, for cadherin (Xiao et al., 2013; He et al., 2016). In studies
example, indicating that the different feeding behavior performed on anesthetized primates, exposure to 41°C
of PF and HS cows had no effect on these transporters. did not result in significant changes in the intestinal
In spite of adaptations in feeding behavior and physi- permeability until a rectal temperature of 42°C was
cal activity, rectal temperature was significantly higher reached (Gathiram et al., 1987). The response of tight
in HS than PF cows, indicating that these animals junction protein expression to heat exposure has been
experienced chronic heat stress. Numerous studies have reported to be time and temperature dependent (Xiao
demonstrated that severe heat stress produces a rapid et al., 2013; He et al., 2016). We exposed mid-lactation
increase in intestinal permeability in several species in- dairy cows continuously to mild ambient heat (28°C)
cluding primates, rats, pigs, and cows (Gathiram et al., for a relatively long period of 4 d without analyzing
1988; Hall et al., 2001; Sawka et al., 2001; Koch et al., changes before the 4-d-period. We speculated that
2019). An increased intestinal permeability could po- molecular adaptations occurring in the long term main-
tentially expose immune cells of the intestine to patho- tain rumen epithelial barrier function. It is known that
gen-associated molecular patterns of luminal microbes, HSP interact with tight junction proteins to maintain
and thereby trigger immune responses. In contrast to their structure (Koch et al., 2019). The expression of
the monolayer epithelium of the intestine, the stratified HSPH1 was upregulated (P < 0.05) and that of HSP70
squamous epithelium of the rumen is composed of 4 tended to be higher (P = 0.05) in HS compared with
distinct strata (Steele et al., 2016). Directly in contact PF cows (Supplemental Table S2, https:​/​/​doi​.org/​10​
with the lumen site is the cornified keratinocyte layer, .3168/​jds​.2020​-18417). Particularly, these 2 HSP are
containing many transporters. (e.g., for short-chain implicated in maintaining or protecting intestinal tight
fatty acid absorption). Underneath the keratinocytes junction barrier function (Dokladny et al., 2006; Yang
is the granulosum layer, which is characterized by tight et al., 2007). Proteome analysis also revealed numerous
junctions. Adjacent to the granulosum is the spinosum upregulated proteins significantly enriched in “chap-
layer and the stratum basale, both rich in mitochondria erone mediated protein folding” in HS cows. Another
that facilitate metabolic properties (Steele et al., 2016). mechanism that may contribute to the protection of
Integrity of the 4 strata of the rumen seems to be more tight junction proteins is protein glycosylation. Protein
robust than that in the intestine; however, it can also glycosylation is essential and complementary to the
be damaged, for instance, by pathogenic infections role of classical HSP, and simultaneous appearance of
that result in the intraepithelial invasion of leukocytes both stress-induced glycoproteins and enhanced HSP
and activation of T lymphocytes and macrophages in accumulation after heat stress (i.e., 15–30 min at 45°C)
the basal membrane (Fuertes et al., 2015). We did not is supported by numerous studies (Piper, 1993; Henle
directly measure impairment in terms of permeability et al., 1995). It has been hypothesized that protein gly-
in our study. However, the lack of TLR4 pathway ac- cosylation may have direct antiaggregation functions or
tivation indicated no difference in the rumen epithelial mediates the interaction of proteins with chaperones,
barrier function of HS and PF cows. This conclusion which is particularly important when the cellular ATP
was drawn based on the finding that TLR activation level is low (Henle et al., 1995). Results of our proteome
upon microbial exposure induces secretion of secretory study indicated some degree of protein glycosylation in
IgA into the lumen, produces mucins and antimicrobial rumen papillae of HS cows. The abundance of the rate-
peptides, and stimulates epithelial cell proliferation, limiting enzyme of the hexoseamine pathway, GFPT1,
thereby maintaining intestinal barrier function (John- was more highly expressed in HS cows. This enzyme
ston and Corr, 2016). In addition, analysis of tight catalyzes the transfer of the amine group from gluta-
junction protein expression revealed no significant dif- mine to fructose-6-phosphate to form D-glucoseamine-
ferences between HS and PF cows, indicating no heat 6-phosphate and, with several subsequent enzymatic
stress–specific alteration of the rumen epithelial bar- steps, results in the formation of UDP-GlcNAc, which
rier function. Results of the effects of ambient heat on directly glycosylates proteins (Oki et al., 1999). An-

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8611

other substrate required for glycosylation is Ac-CoA, reduced expression of several enzymes involved in glu-
whose synthesis is also upregulated in HS cows, further tathione metabolism and enzymes regulating cell death
supporting a role of glycosylation of rumen papillae in response to oxidative stress, suggesting that HS cows
proteins to cope with heat stress. Thus, it appears that experiencing mild and chronic heat stress combat the
HS cows develop a protective mechanism involving formation of reactive oxygen species in the rumen epi-
HSP, protein glycosylation, and Ac-CoA biosynthesis thelial cells.
to maintain epithelial barrier function, all preventing We observed increased expression of several proteins
the activation of the TLR4 pathway and subsequent involved in the AMPK signaling pathway in the ru-
inflammation. Moreover, the maintained epithelial bar- men epithelium of HS cows. This agrees with increased
rier function and absence of activated inflammatory phosphorylation of AMPK in skeletal muscle of either
pathways suggest no enhanced immune cell infiltration dairy cows exposed to 28°C for 4 d (Koch et al., 2016)
into papillae tissue. Therefore, the cell population of the or rats challenged for 30 min at 42°C (Goto et al.,
rumen papillae, predominantly consisting of epithelial 2015). Min et al. (2015) also reported higher serum
cells and keratinocytes, is likely not different between AMPK concentrations in dairy cows kept at a THI of
HS and PF cows. 81.7 relative to counterparts kept at a THI of 53.4.
A shift from fat toward carbohydrate oxidation after The AMPK is a central regulator of cellular and or-
heat stress has been reported previously (Baumgard ganismal metabolism in eukaryotes, which is activated
and Rhoads, 2013; Lamp et al., 2015). In most of these when intracellular ATP levels are low (Mihaylova and
studies, each animal was compared with itself in 2 pe- Shaw, 2011). The level of energy intake between HS
riods of thermoneutrality followed by either HS or PF and PF cows was comparable, and therefore should
treatment. For instance, Lamp et al. (2015) reported not account for the activation of AMPK signaling in
that the reduced feed intake of late-pregnant cows di- HS cow. Instead, this effect is seemingly due to direct
minished carbohydrate and increased fat oxidation un- effects of heat stress. Impaired mitochondrial function
der thermoneutral, but not heat-stress conditions, for due to heat stress has been reported in several studies
7 d at 28°C. When whole-body carbohydrate oxidation (White et al., 2012; Huang et al., 2015). This may have
was normalized to metabolic BW and DMI, we found led to ATP deprivation in HS animals beyond the level
that carbohydrate oxidation tended to be higher in HS induced by energy intake reduction. In agreement with
and PF cows. This result confirms increased preference this, increased expression of ATP citrate lyase enzyme
of body tissues for utilizing carbohydrates over fat for and reduced expression of mitochondrial respiratory
energy production in heat-stressed mid-lactation cows. chain enzymes (cytochrome C oxidase and ATP syn-
Moreover, our proteome data suggests that the expres- thase) indicate reduced tricarboxylic acid cycling and
sion of 2 enzymes involved in β-oxidation of fatty acids mitochondrial energy production in HS cows. Evidence
(enoyl-CoA hydratase and 3-hydroxy acyl-CoA dehydro- exists that activated AMPK induces the conservation
genase) was lower in the rumen epithelium of HS than of intracellular ATP levels via multiple downstream
PF cows, which further point to a reduced preference of pathways, including autophagy (Pease et al., 2009; Tsai
fat oxidation under heat-stress condition. Because heat et al., 2016), a process in which regions of cytoplasm,
stress is closely associated with oxidative stress (Oki et insoluble protein aggregates, and damaged mitochon-
al., 1999: Slimen et al., 2014), the shift in fuel oxidation dria are degraded in lysosomes (Hardie, 2011). It seems
is likely an adaptation strategy to reduce formation of that AMPK-mediated autophagy was activated in HS
reactive oxygen species in the mitochondria (Goto et cows, at least in part, to remove damaged mitochondria
al., 2015; Lamp et al., 2015). Heat-induced oxidative and other misfolded proteins. Increased degradation of
stress is largely dependent on the intensity and dura- macromolecules and cellular components is supported
tion of heat stress (Slimen et al., 2014). The effect of by the fact that several proteins involved in ubiquitin-
acute (short and rapid temperature increase) or chronic mediated protein degradation, endoplasmic reticulum
(high ambient temperature over days to weeks) heat protein processing, and serine threonine kinase inhibi-
stress on oxidative stress and the antioxidant defense tion were more highly expressed in HS cows.
system of broiler chicken was reviewed by Akbarian et It is generally known that chronic heat stress in-
al. (2016). The authors described a typical increase in creases the plasma concentration of insulin (O’Brien
the activities of antioxidant enzymes such as catalase, et al., 2010; Wheelock et al., 2010; Min et al., 2015)
glutathione peroxidase, and superoxide dismutase, af- as a result of increased insulin secretion (Itoh et al.,
ter acute heat stress and further showed that chronic 1998). Although we did not measure serum insulin
(long-term) heat stress decreases the abundance and concentration, the highly expressed proteins CRK,
activities of antioxidant enzymes. In agreement with FBP1, FLOT2, PRKAG1, PRKAR2A, and PRKAR2B
the latter study (Akbarian et al., 2016), we observed in HS cows were significantly enriched in the insulin
Journal of Dairy Science Vol. 103 No. 9, 2020
Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8612

signaling pathway, indicating higher insulin action in REFERENCES


HS cows. Several lines of evidence link heat stress to
Abreu, M. T. 2010. Toll-like receptor signalling in the intestinal epi-
endotoxemia (Hall et al., 2001; Lim et al., 2007; Pearce thelium: How bacterial recognition shapes intestinal function. Nat.
et al., 2013) and show that LPS causes a rise in cir- Rev. Immunol. 10:131–144. https:​/​/​doi​.org/​10​.1038/​nri2707.
culating insulin concentration (Waldron et al., 2006; Akbarian, A., J. Michiels, J. Degroote, M. Majdeddin, A. Golian, and
S. De Smet . 2016. Association between heat stress and oxidative
Leon, 2007). Whether increased insulin signaling in stress in poultry; mitochondrial dysfunction and dietary interven-
the rumen epithelium is a result of increased circulat- tions with phytochemicals. J. Anim. Sci. Biotechnol. 7:37. https:​/​
ing insulin concentration due to the leakage in lower /​doi​.org/​10​.1186/​s40104​-016​-0097​-5.
Baumgard, L. H., and R. P. Rhoads. 2012. Ruminant nutrition sym-
parts of the digestive tract needs additional studies. posium: Ruminant production and metabolic responses to heat
However, AMPK phosphorylates the insulin receptor, stress. J. Anim. Sci. 90:1855–1865. https:​/​/​doi​.org/​10​.2527/​jas​
providing a direct link between AMPK and the insulin .2011​-4675.
Baumgard, L. H., and R. P. Rhoads Jr.. 2013. Effects of heat stress
signaling pathway (Chopra et al., 2012). Results of on post-absorptive metabolism and energetics. Annu. Rev. Anim.
an earlier study on rats suggest that even acute heat Biosci. 1:311–337. https:​/​/​doi​.org/​10​.1146/​annurev​-animal​-031412​
stress acts as a rapid stimulator of insulin-independent -103644.
Baumgard, L. H., J. B. Wheelock, S. R. Sanders, C. E. Moore, H.
glucose transport in skeletal muscle, at least in part by B. Green, M. R. Waldron, and R. P. Rhoads. 2011. Postabsorp-
stimulating AMPK (Goto et al., 2015). It seems that tive carbohydrate adaptations to heat stress and monensin supple-
insulin and AMPK signaling pathways act in synergy mentation in lactating Holstein cows. J. Dairy Sci. 94:5620–5633.
https:​/​/​doi​.org/​10​.3168/​jds​.2011​-4462.
to ensure the transport function of the rumen during Beede, D. K., and R. J. Collier. 1986. Potential nutritional strategies
stress conditions. Accordingly, Habegger et al. (2012) for intensively managed cattle during thermal-stress. J. Anim. Sci.
demonstrated that AMPK enhances insulin-dependent 62:543–554. https:​/​/​doi​.org/​10​.2527/​jas1986​.622543x.
Brobeck, J. R. 1948. Food intake as a mechanism of temperature regu-
glucose transporter 4 (GLUT4) regulation, at least in lation. Yale J. Biol. Med. 20:545–552.
myotubes. Chopra, I., H. F. Li, H. Wang, and K. A. Webster. 2012. Phosphory-
lation of the insulin receptor by AMP-activated protein kinase
(AMPK) promotes ligand-independent activation of the insulin
CONCLUSIONS signalling pathway in rodent muscle. Diabetologia 55:783–794.
https:​/​/​doi​.org/​10​.1007/​s00125​-011​-2407​-y.
Behavioral and tissue-level adaptations were success- Derno, M., H. G. Elsner, E. A. Paetow, H. Scholze, and M. Schwei-
gel. 2009. Technical note: A new facility for continuous respiration
fully implemented in dairy cows to reduce heat stress. measurements in lactating cows. J. Dairy Sci. 92:2804–2808. https:​
There was no heat stress–specific change in the integrity /​/​doi​.org/​10​.3168/​jds​.2008​-1839.
of the rumen epithelium as evidenced by comparable Dokladny, K., P. L. Moseley, and T. Y. Ma. 2006. Physiologically
relevant increase in temperature causes an increase in intestinal
expression of tight junction protein and activation of epithelial tight junction permeability. Am. J. Physiol. Gastroin-
TLR4 signaling in HS and PF cows. Further, and in test. Liver Physiol. 290:G204–G212. https:​/​/​doi​.org/​10​.1152/​a jpgi​
contrast to our hypothesis, transport of short-chain .00401​.2005.
Dokladny, K., D. Ye, J. C. Kennedy, P. L. Moseley, and T. Y. Ma.
fatty acids was not specifically affected in HS relative 2008. Cellular and molecular mechanisms of heat stress-induced
to isoenergetically fed counterparts. Increased ambient up-regulation of occludin protein expression: Regulatory role of
heat relative to pair-feeding increased the expression heat shock factor-1. Am. J. Pathol. 172:659–670. https:​/​/​doi​.org/​
10​.2353/​a jpath​.2008​.070522.
of heat shock proteins along with the activation of the Eslamizad, M., O. Lamp, M. Derno, and B. Kuhla. 2015. The control
protein glycosylation machinery, activated AMPK, and of short-term feed intake by metabolic oxidation in late-pregnant
insulin signaling pathways, and increased carbohydrate and early lactating dairy cows exposed to high ambient tem-
peratures. Physiol. Behav. 145:64–70. https:​/​/​doi​.org/​10​.1016/​j​
oxidation at the expense of fat oxidation. These mecha- .physbeh​.2015​.03​.044.
nisms contribute to reduce heat strain and protect ru- Fuertes, M., Y. Manga-González, J. Benavides, M. C. González-Lanza,
men tissue from injury and subsequent inflammation. F. J. Giráldez, M. Mezo, M. González-Warleta, M. Fernández, J.
Regidor-Cerrillo, P. Castaño, M. Royo, L. M. Ortega-Mora, V.
However, whether changes in whole-body macronutrient Pérez, and M. C. Ferreras. 2015. Immunohistochemical study and
oxidation and metabolic pathways in the rumen epithe- mRNA cytokine profile of the local immune response in cattle
lium are caused by endotoxemia eventually induced by naturally infected with Calicophoron daubneyi. Vet. Parasitol.
214:178–183. https:​/​/​doi​.org/​10​.1016/​j​.vetpar​.2015​.10​.012.
the leakage of lower parts of the digestive tract or are Gathiram, P., S. L. Gaffin, J. G. Brock-Utne, and M. T. Wells. 1987.
gross physiological adaptations to heat stress needs to Time course of endotoxemia and cardiovascular changes in heat-
be investigated in future studies. stressed primates. Aviat. Space Environ. Med. 58:1071–1074.
Gathiram, P., M. T. Wells, D. Raidoo, J. G. Brock-Utne, and S. L.
Gaffin. 1988. Portal and systemic plasma lipopolysaccharide con-
centrations in heat-stressed primates. Circ. Shock 25:223–230.
ACKNOWLEDGMENTS Goto, A., T. Egawa, I. Sakon, R. Oshima, K. Ito, Y. Serizawa, K. Se-
kine, S. Tsuda, K. Goto, and T. Hayashi. 2015. Heat stress acutely
The authors thank the Alexander-von-Humboldt activates insulin-independent glucose transport and 5′-AMP-acti-
vated protein kinase prior to an increase in HSP72 protein in rat
Foundation (Bonn, Germany) for financial support. skeletal muscle. Physiol. Rep. 3:e12601. https:​/​/​doi​.org/​10​.14814/​
The authors have not stated any conflicts of interest. phy2​.12601.

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8613
Habegger, K. M., N. J. Hoffman, C. M. Ridenour, J. T. Brozinick, Min, L., J. B. Cheng, B. L. Shi, H. J. Yang, N. Zheng, and J. Q.
and J. S. Elmendorf. 2012. AMPK enhances insulin-stimulated Wang. 2015. Effects of heat stress on serum insulin, adipokines,
GLUT4 regulation via lowering membrane cholesterol. Endocrinol- AMP-activated protein kinase, and heat shock signal molecules in
ogy 153:2130–2141. https:​/​/​doi​.org/​10​.1210/​en​.2011​-2099. dairy cows. J. Zhejiang Univ. Sci. B 16:541–548. https:​/​/​doi​.org/​
Hall, D. M., G. R. Buettner, L. W. Oberley, L. Xu, R. D. Matthes, 10​.1631/​jzus​.B1400341.
and C. V. Gisolfi. 2001. Mechanisms of circulatory and intesti- Nesvizhskii, A. I., A. Keller, E. Kolker, and R. Aebersold. 2003. A
nal barrier dysfunction during whole body hyperthermia. Am. J. statistical model for identifying proteins by tandem mass spec-
Physiol. Heart Circ. Physiol. 280:H509–H521. https:​/​/​doi​.org/​10​ trometry. Anal. Chem. 75:4646–4658. https:​/​/​doi​.org/​10​.1021/​
.1152/​a jpheart​.2001​.280​.2​.H509. ac0341261.
Hardie, D. G. 2011. AMPK and autophagy get connected. EMBO J. O’Brien, M. D., R. P. Rhoads, S. R. Sanders, G. C. Duff, and L. H.
30:634–635. https:​/​/​doi​.org/​10​.1038/​emboj​.2011​.12. Baumgard. 2010. Metabolic adaptations to heat stress in growing
He, S., F. Liu, L. Xu, P. Yin, D. Li, C. Mei, L. Jiang, Y. Ma, and J. Xu. cattle. Domest. Anim. Endocrinol. 38:86–94. https:​/​/​doi​.org/​10​
2016. Protective effects of ferulic acid against heat stress-induced .1016/​j​.domaniend​.2009​.08​.005.
intestinal epithelial barrier dysfunction in vitro and in vivo. PLoS Oki, T., K. Yamazaki, J. Kuromitsu, M. Okada, and I. Tanaka. 1999.
One 11:e0145236. https:​/​/​doi​.org/​10​.1371/​journal​.pone​.0145236. cDNA cloning and mapping of a novel subtype of glutamine: fruc-
Henle, K. J., S. M. Jethmalani, and W. A. Nagle. 1995. Heat stress-in- tose-6-phosphate amidotransferase (GFAT2) in human and mouse.
duced protein glycosylation in mammalian cells. Trends Glycosci. Genomics 57:227–234. https:​/​/​doi​.org/​10​.1006/​geno​.1999​.5785.
Glycotechnol. 7:191–204. https:​/​/​doi​.org/​10​.4052/​tigg​.7​.191. Pearce, S. C., V. Mani, R. L. Boddicker, J. S. Johnson, T. E. Weber, J.
Huang, C., H. Jiao, Z. Song, J. Zhao, X. Wang, and H. Lin. 2015. Heat W. Ross, R. P. Rhoads, L. H. Baumgard, and N. K. Gabler . 2013.
stress impairs mitochondria functions and induces oxidative injury Heat stress reduces intestinal barrier integrity and favors intestinal
in broiler chickens. J. Anim. Sci. 93:2144–2153. https:​/​/​doi​.org/​10​ glucose transport in growing pigs. PLoS One 8:e70215. https:​/​/​doi​
.2527/​jas​.2014​-8739. .org/​10​.1371/​journal​.pone​.0070215.
IPCC. 2007. Climate Change 2007: Contribution of Working Groups I, Pease, S., L. Bouadma, N. Kermarrec, F. Schortgen, B. Régnier, and
II and III to the Fourth Assessment Report of the Intergovernmen- M. Wolff. 2009. Early organ dysfunction course, cooling time and
tal Panel on Climate Change. Core Writing Team, R. K. Pachauri, outcome in classic heatstroke. Intensive Care Med. 35:1454–1458.
and A. Reisinger, ed. IPCC, Geneva, Switzerland. https:​/​/​doi​.org/​10​.1007/​s00134​-009​-1500​-x.
Itoh, F., Y. Obara, M. T. Rose, H. Fuse, and H. Hashimoto. 1998. Piper, P. W. 1993. Molecular events associated with acquisition of
Insulin and glucagon secretion in lactating cows during heat ex- heat tolerance by the yeast Saccharomyces cerevisiae. FEMS Mi-
posure. J. Anim. Sci. 76:2182–2189. https:​/​/​doi​.org/​10​.2527/​1998​ crobiol. Rev. 11:339–355. https:​/​/​doi​.org/​10​.1111/​j​.1574​-6976​
.7682182x. .1993​.tb00005​.x.
Johnston, D. G., and S. C. Corr. 2016. Toll-like receptor signalling Rhoads, M. L., R. P. Rhoads, M. J. VanBaale, R. J. Collier, S. R.
and the control of intestinal barrier function. Methods Mol. Biol. Sanders, W. J. Weber, B. A. Crooker, and L. H. Baumgard. 2009a.
1390:287–300. https:​/​/​doi​.org/​10​.1007/​978​-1​-4939​-3335​-8​_18. Effects of heat stress and plane of nutrition on lactating Holstein
Keller, A., A. I. Nesvizhskii, E. Kolker, and R. Aebersold. 2002. Em- cows: I. Production, metabolism, and aspects of circulating so-
pirical statistical model to estimate the accuracy of peptide iden- matotropin. J. Dairy Sci. 92:1986–1997. https:​/​/​doi​.org/​10​.3168/​
tifications made by MS/MS and database search. Anal. Chem. jds​.2008​-1641.
74:5383–5392. https:​/​/​doi​.org/​10​.1021/​ac025747h. Rhoads, R. P., S. R. Sanders, L. Cole, M. V. Skrzypek, T. H. Elsasser,
Koch, F., U. Thom, E. Albrecht, R. Weikard, W. Nolte, B. Kuhla, G. C. Duff, R. J. Collier, and L. H. Baumgard. 2009b. Effects of
and C. Kuehn. 2019. Heat stress directly impairs gut integrity and heat stress on glucose homeostasis and metabolic response to an
recruits distinct immune cell populations into the bovine intestine. endotoxin challenge in Holstein steers. J. Anim. Sci. 87:78. (Ab-
Proc. Natl. Acad. Sci. USA 116:10333–10338. https:​/​/​doi​.org/​10​ str.)
.1073/​pnas​.1820130116. Ruijter, J. M., M. W. Pfaffl, S. Zhao, A. N. Spiess, G. Boggy, J. Blom,
Koch, F., O. Lamp, M. Eslamizad, J. Weitzel, and B. Kuhla. 2016. R. G. Rutledge, D. Sisti, A. Lievens, K. De Preter, S. Derveaux, J.
Metabolic response to heat stress in late-pregnant and early lacta- Hellemans, and J. Vandesompele. 2013. Evaluation of qPCR curve
tion dairy cows: Implications to liver-muscle crosstalk. PLoS One analysis methods for reliable biomarker discovery: Bias, resolution,
11:e0160912. https:​/​/​doi​.org/​10​.1371/​journal​.pone​.0160912. precision, and implications. Methods 59:32–46. https:​/​/​doi​.org/​10​
Kregel, K. C. 2002. Heat shock proteins: Modifying factors in physi- .1016/​j​.ymeth​.2012​.08​.011.
ological stress responses and acquired thermotolerance. J. Appl. Sawka, M. N., S. J. Montain, and W. A. Latzka. 2001. Hydration ef-
Physiol. 92:2177–2186. https:​/​/​doi​.org/​10​.1152/​japplphysiol​ fects on thermoregulation and performance in the heat. Comp.
.01267​.2001. Biochem. Physiol. A Mol. Integr. Physiol. 128:679–690. https:​/​/​
Lambert, G. P., C. V. Gisolfi, D. J. Berg, P. L. Moseley, L. W. Ober- doi​.org/​10​.1016/​S1095​-6433(01)00274​-4.
ley, and K. C. Kregel. 2002. Selected contribution: Hyperthermia- Shannon, P., A. Markiel, O. Ozier, N. S. Baliga, J. T. Wang, D. Ram-
induced intestinal permeability and the role of oxidative and ni- age, N. Amin, B. Schwikowski, and T. Ideker. 2003. Cytoscape: A
trosative stress. J. Appl. Physiol. 92:1750–1761. https:​/​/​doi​.org/​10​ software environment for integrated models of biomolecular inter-
.1152/​japplphysiol​.00787​.2001. action networks. Genome Res. 13:2498–2504. https:​/​/​doi​.org/​10​
Lamp, O., M. Derno, W. Otten, M. Mielenz, G. Nürnberg, and B. .1101/​gr​.1239303.
Kuhla. 2015. Metabolic heat stress adaption in transition cows: Slimen, I. B., T. Najar, A. Ghram, H. Dabbebi, M. Ben Mrad, and M.
Differences in macronutrient oxidation between late-gestating Abdrabbah. 2014. Reactive oxygen species, heat stress and oxida-
and early-lactating German Holstein dairy cows. PLoS One tive-induced mitochondrial damage. A review. Int. J. Hyperther-
10:e0125264. https:​/​/​doi​.org/​10​.1371/​journal​.pone​.0125264. mia 30:513–523. https:​/​/​doi​.org/​10​.3109/​02656736​.2014​.971446.
Leon, L. R. 2007. Heat stroke and cytokines. Prog. Brain Res. 162:481– St-Pierre, N. R., B. Cobanov, and G. Schnitkey. 2003. Economic losses
524. https:​/​/​doi​.org/​10​.1016/​S0079​-6123(06)62024​-4. from heat stress by US livestock industries. J. Dairy Sci. 86:E52–
Lim, C. L., G. Wilson, L. Brown, J. S. Coombes, and L. T. Mack- E77. https:​/​/​doi​.org/​10​.3168/​jds​.S0022​-0302(03)74040​-5.
innon. 2007. Pre-existing inflammatory state compromises heat Steele, M. A., G. B. Penner, F. Chaucheyras-Durand, and L. L. Guan.
tolerance in rats exposed to heat stress. Am. J. Physiol. Regul. 2016. Development and physiology of the rumen and the lower
Integr. Comp. Physiol. 292:R186–R194. https:​/​/​doi​.org/​10​.1152/​ gut: Targets for improving gut health. J. Dairy Sci. 99:4955–4966.
ajpregu​.00921​.2005. https:​/​/​doi​.org/​10​.3168/​jds​.2015​-10351.
Mihaylova, M. M., and R. J. Shaw. 2011. The AMPK signalling path- Tsai, Y.-C., K.-K. Lam, Y.-J. Peng, Y.-M. Lee, C.-Y. Yang, Y.-J. Tsai,
way coordinates cell growth, autophagy and metabolism. Nat. Cell M.-H. Yen, and P.-Y. Cheng. 2016. Heat shock protein 70 and
Biol. 13:1016–1023. https:​/​/​doi​.org/​10​.1038/​ncb2329. AMP-activated protein kinase contribute to 17-DMAG-dependent

Journal of Dairy Science Vol. 103 No. 9, 2020


Eslamizad et al.: RUMEN PAPILLAE ADAPTATION TO AMBIENT HEAT 8614
protection against heat stroke. J. Cell. Mol. Med. 20:1889–1897. White, M. G., O. Saleh, D. Nonner, E. F. Barrett, C. T. Moraes, and
https:​/​/​doi​.org/​10​.1111/​jcmm​.12881. J. N. Barrett. 2012. Mitochondrial dysfunction induced by heat
Untergasser, A., I. Cutcutache, T. Koressaar, J. Ye, B. C. Faircloth, stress in cultured rat CNS neurons. J. Neurophysiol. 108:2203–
M. Remm, and S. G. Rozen. 2012. Primer3—new capabilities and 2214. https:​/​/​doi​.org/​10​.1152/​jn​.00638​.2011.
interfaces. Nucleic Acids Res. 40:e115. https:​/​/​doi​.org/​10​.1093/​ Xiao, G., F. Yuan, Z. Gu, Z. Liu, Y. Zhang, and L. Su. 2013. Effect
nar/​gks596. of heat stress on intestinal barrier function of human intestinal
Vanselow, J., A. Vernunft, D. Koczan, M. Spitschak, and B. Kuhla. epithelial Caco-2 cells. Med. J. Chin. PLA 38:472–475.
2016. Exposure of lactating dairy cows to acute pre-ovulatory Yang, P. C., S. H. He, and P. Y. Zheng. 2007. Investigation into the
heat stress affects granulosa cell-specific gene expression profiles signal transduction pathway via which heat stress impairs intesti-
in dominant follicles. PLoS One 11:e0160600. https:​/​/​doi​.org/​10​ nal epithelial barrier function. J. Gastroenterol. Hepatol. 22:1823–
.1371/​journal​.pone​.0160600. 1831. https:​/​/​doi​.org/​10​.1111/​j​.1440​-1746​.2006​.04710​.x.
Waldron, M. R., A. E. Kulick, A. W. Bell, and T. R. Overton. 2006. Ye, J., G. Coulouris, I. Zaretskaya, I. Cutcutache, S. Rozen, and T.
Acute experimental mastitis is not causal toward the development L. Madden. 2012. Primer-BLAST: A tool to design target-spe-
of energy-related metabolic disorders in early postpartum dairy cific primers for polymerase chain reaction. BMC Bioinformatics
cows. J. Dairy Sci. 89:596–610. https:​/​/​doi​.org/​10​.3168/​jds​.S0022​ 13:134. https:​/​/​doi​.org/​10​.1186/​1471​-2105​-13​-134.
-0302(06)72123​-3.
West, J. W. 1999. Nutritional strategies for managing the heat-stressed
dairy cow. J. Anim. Sci. 77:21–35. ORCIDS
Wheelock, J. B., R. P. Rhoads, M. J. VanBaale, S. R. Sanders, and L.
H. Baumgard. 2010. Effects of heat stress on energetic metabolism Mehdi Eslamizad https:​/​/​orcid​.org/​0000​-0002​-1071​-3683
in lactating Holstein cows. J. Dairy Sci. 93:644–655. https:​/​/​doi​ Björn Kuhla https:​/​/​orcid​.org/​0000​-0002​-2032​-5502
.org/​10​.3168/​jds​.2009​-2295.

Journal of Dairy Science Vol. 103 No. 9, 2020

You might also like