Heat shock suppresses the innate immune response of bovine endometrial epithelial cells

Abstract

Abstract

With rising global temperatures, it is imperative to determine the impact of heat stress on the physiology of food-producing animals. Dairy cows are susceptible to uterine diseases that reduce fertility. Immune function is important in the development and progression of disease; however, the effect of heat shock on the innate immune capacity of endometrial epithelial cells remains underexplored. Here, we investigated how heat shock alters the innate immune response and mitochondrial respiration of bovine endometrial epithelial cells. Primary endometrial epithelial cells were collected from postpartum cows and cultured in the presence of lipopolysaccharide under thermoneutral (38.5°C) or heat shock (40.0°C) conditions. Time-course and sequential heat shock experiments were conducted to assess gene expression dynamics of HSPA1A, TLR4, CXCL8, IL6, and IL1B. Cell viability was evaluated by MTT assay, and mitochondrial respiration was analyzed using high-resolution respirometry. Heat shock did not affect cell viability or overall mitochondrial respiration but reduced proton leak-related oxygen consumption. Acute heat shock induced HSPA1A expression but suppressed LPS-stimulated CXCL8 and IL6 expression. Expression of TLR4 increased when cells were recovering from heat shock or following sequential heat shock. Sequential heat shock did not affect the expression of pro-inflammatory mediators compared to a single heat shock event. In conclusion, acute heat shock of bovine endometrial epithelial cells reduced the innate immune response to LPS without impacting cell viability or mitochondrial function. Altered endometrial immune function may contribute to the increased incidence of uterine diseases in cows under heat stress conditions.

Lay summary

As global temperatures continue to rise, food-producing animals face increasing challenges to their fertility and productivity. Uterine disease in dairy cows reduces fertility and impacts milk production. Heat stress is known to disrupt immune function, but less is understood about how elevated temperatures affect the cells that line the uterus that respond to invading pathogens. We examined how acute heat exposure affects the immune response of uterine cells to respond to pathogens and found that heat shock reduces their ability to produce key immune molecules needed to fight bacteria. This suggests that heat stress directly weakens the natural defenses of the uterus and helps us understand why cows are vulnerable to reproductive diseases during hot weather. Understanding these effects can guide better management strategies to protect animal health in a warming climate.

Introduction

Global food production faces an escalating challenge from increasingly frequent and intense heat events driven by climate change (IPCC 2023). Heat stress is a major contributor to low fertility in dairy cows (Wolfenson et al. 1988), and intensive genetic selection for milk yield has made modern dairy cattle highly vulnerable to heat stress by increasing feed intake, metabolic heat output, and thermal load (Baumgard & Rhoads 2013, Bernabucci et al. 2014, Loor et al. 2023). Heat stress reduces dry matter intake (feeding) and disrupts metabolic and immune homeostasis, predisposing cows to subfertility and uterine diseases (Das et al. 2016, Gernand et al. 2019, Molinari et al. 2022). Despite these impacts of heat stress, the cellular mechanisms by which elevated temperatures impair immune function and increase disease susceptibility remain poorly understood.

Postpartum uterine diseases are a major health and economic concern for the dairy industry, affecting up to 40% of cows after calving and contributing to reduced milk yield and poor reproductive performance (Sheldon et al. 2019). These diseases arise when pathogenic bacteria, such as Escherichia coli, invade the uterine lumen, triggering inflammation and tissue injury (Bromfield et al. 2015, Sheldon et al. 2019). E. coli possesses the endotoxin lipopolysaccharide (LPS), a major component of the bacterial cell wall and a well-characterized pathogen-associated molecular pattern recognized by the innate immune system. Endometrial epithelial and stromal cells act as key immune sentinels by recognizing pathogen-associated molecular patterns through Toll-like receptors (TLRs), which activate an inflammatory cascade characterized by the production of IL-1β, IL-6, and IL-8 (Cronin et al. 2012, Turner et al. 2014). Although inflammation is necessary for pathogen clearance, excessive or prolonged inflammation can cause tissue damage (Medzhitov 2008, Chovatiya & Medzhitov 2014).

Disease development is governed by interactions between the host, pathogen, and environment (Molinari et al. 2022). Rising environmental temperatures can disturb this equilibrium by promoting pathogen proliferation or weakening host immunity, increasing overall disease susceptibility (Erskine et al. 1988, Hogan et al. 1989, Gernand et al. 2019). Yet, it remains unclear whether the higher incidence and persistence of uterine diseases during hotter months result from increased pathogen load or altered host immune function driven by environmental stress. Heat stress alters endometrial signaling pathways, including prostaglandin and cytokine synthesis (Katagiri et al. 2013, Sakai et al. 2018, 2020, 2022), and in vitro studies indicate an exaggerated inflammatory response in a bovine endometrial epithelial cell line (BEND) exposed to heat shock (Molinari & Bromfield 2023). However, most research on endometrial innate immunity relies on enzymatic or mechanical tissue digestion or long-term culture, methods that can introduce variability in cell viability and isolation efficiency and potentially influence phenotypic and quantitative analysis of cellular responses (Gosens et al. 2002, Skulska et al. 2019, Bondonese et al. 2023). Consequently, the effects of elevated temperatures on the innate immune function of primary endometrial epithelial cells collected directly from healthy cows remain poorly defined. The cytobrush technique overcomes these limitations by enabling non-invasive collection of viable, phenotypically intact epithelial cells from the uterine lumen (Rocha et al. 2022).

The present study aimed to investigate the effects of acute, in vitro heat shock on the innate immune response of primary bovine endometrial epithelial cells collected by non-surgical cytobrush of healthy postpartum cows. The study also evaluated how prior and sequential heat shock influences the innate immune response of endometrial epithelial cells. We hypothesized that acute heat shock disrupts the innate immune function of bovine endometrial epithelial cells and that sequential heat shock would further disrupt responses. Understanding these cellular responses may help explain how heat stress compromises uterine immunity and increases disease susceptibility in dairy cows.

Materials and methods

Endometrial epithelial cell collection and culture

All animal handling and sampling procedures were approved by the Institutional Animal Care and Use Committee of the University of Florida. All cows were housed under cooling conditions consisting of barn shade, fans, and water soakers and were free of clinical signs of disease from calving to the time of cell collection. Endometrial cells were collected on day 21 postpartum following the procedure previously described (Rocha et al. 2022) with minor modifications. Cows were restrained and administered a caudal epidural injection of 60 mg lidocaine hydrochloride 2% (Aspen Veterinary Resources, USA). After disinfecting the perineal area, a sterile cytobrush (endocervical sampling brush; Viamed, USA) mounted on an insemination gun and protected by dual sheaths was passed through the cervix. Once positioned in the uterine body, the brush was exposed and rotated five times against the endometrium to collect luminal epithelial cells. The brush was then placed into 5 mL Hank’s balanced salt solution (HBSS; Cytiva, USA) supplemented with 50 IU/mL penicillin and 50 μg/mL streptomycin (Fisher Scientific, USA) and transported at room temperature to the laboratory for further processing.

The cytobrush-containing tube was vortexed for 10 s, and residual cells were released by scraping the brush against the tube wall with sterile tweezers before discarding it. The cell suspension was centrifuged at 300 g for 5 min at room temperature, and the supernatant was removed. The cell pellet was resuspended in 5 mL pre-warmed red blood cell lysis buffer (1.5 M ammonium chloride, 100 mM sodium bicarbonate, and 10 mM disodium EDTA, dissolved in cell culture-grade water), gently mixed, and incubated for 1 min at room temperature. Isotonicity was restored by adding 10 mL of pre-warmed HBSS supplemented with 50 IU/mL penicillin and 50 μg/mL streptomycin, followed by centrifugation at 300 g for 5 min at room temperature. The final cell pellet was resuspended in 3 mL complete culture medium consisting of RPMI-1640 (Fisher Scientific) supplemented with 10% fetal calf serum, 2 mM L-alanyl-L-glutamine dipeptide (GlutaMAX), 50 IU/mL penicillin, and 50 μg/mL streptomycin. Cell concentration was adjusted to 10⁵ cells/mL, and 500 μL were seeded into 4-well multidishes (Fisher Scientific) and incubated at 38.5°C in a humidified atmosphere of 5% CO₂ for 48 h before further treatment.

Viability and characterization of cultured endometrial epithelial cells

Primary endometrial epithelial cell viability was assessed using the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; Merck, USA) assay as described previously (Tariq et al. 2025). Endometrial cells of individual cows (n = 10) were seeded into 96-well plates (Fisher Scientific). Duplicate wells were exposed to 1 μg/mL ultrapure E. coli O111:B4 LPS (Invivogen, USA) or medium alone under thermoneutral (38.5°C) or heat shock (40.0°C) conditions for 24 h (Fig. 1A). Cells were then incubated with 5 mg/mL MTT for 4 h, washed with pre-warmed Dulbecco's phosphate buffered saline (DPBS), and solubilized with 100 μL dimethyl sulfoxide for 15 min at room temperature in the dark. Optical density (OD₅₄₀) was measured using a microplate reader (BioTek Instruments, USA). Each OD₅₄₀ value was then corrected by subtracting the mean OD₅₄₀ of blank wells that never contained cells. The mean blank-corrected values for each treatment were used for statistical analyses. The same procedure was applied to assess cell viability after sequential heat shock experiment.

Figure 1.

Experimental timelines used to assess the effects of heat shock and LPS on bovine endometrial epithelial cells. Endometrial epithelial cells were isolated using cytobrush on postpartum day 21 of healthy, cooled cows. All cows were free of clinical diseases from calving to day 21 postpartum. Initially, the effects of acute heat shock were evaluated in a time-course experiment (A), in which cells were exposed to shock (40.0°C) or thermoneutral conditions (38.5°C) in combination with 1 μg/mL lipopolysaccharide (LPS) or medium alone during the response period. After that, cells were washed in warmed PBS and allowed to recover in medium alone at 38.5°C for an additional 12 h. Samples were collected at 0, 12, 18, 24, 25, 27, 30, and 36 h relative to the start of the treatment. In addition, cell viability and cellular respiration were evaluated in cells exposed to shock (40.0°C) or thermoneutral conditions (38.5°C) in combination with 1 μg/mL lipopolysaccharide (LPS) or medium alone for 24 h. An additional experiment evaluating sequential heat shock was also conducted (B). Cells were exposed to heat shock (40.0°C) or thermoneutral conditions (38.5°C) in medium alone for 24 h in period 1. Cells were then rested for 12 h in medium alone at 38.5°C before being exposed to heat shock (40.0°C) or thermoneutral conditions (38.5°C) in the presence of 1 μg/mL LPS during period 2. Samples were collected at the end of period 2 for analysis.

To characterize cultured cells, cells were cultured on 8-chamber glass slides for immunocytochemistry. After 24 h of culture with heat shock and LPS (Fig. 1A), cells were fixed in 2% paraformaldehyde for 10 min, permeabilized with 0.1% Triton X-100 for 10 min, and blocked for 30 min with 1% BSA and 10% normal goat serum. Cells were incubated with mouse anti-cytokeratin (1 μg/mL; Fisher Scientific) and rabbit anti-vimentin (1 μg/mL; Fisher Scientific) for 1 h at room temperature or overnight at 4°C, followed by three washes in 0.1% Tween 20 in PBS (v/v). A secondary incubation was performed with anti-mouse Alexa Fluor 594 (4 μg/mL; Invitrogen, USA), anti-rabbit Alexa Fluor 488 (4 μg/mL; Invitrogen), and Hoechst 33342 (1 μg/mL; Invitrogen) for 1 h in the dark. Cells were washed again and mounted under cover glass. Cells were imaged at ×40 magnification on a Zeiss Axioplan 2 microscope (Zeiss, Germany), and images were acquired with a Zeiss Axiocam MRm camera using AxioVision software (v4.8.2; Zeiss) and analyzed using ImageJ (v1.48; National Institutes of Health, USA).

Effect of acute heat shock on innate immune response of endometrial epithelial cells

To evaluate the dynamics of the innate immune response of endometrial epithelial cells exposed to LPS under heat shock and subsequent recovery under thermoneutral conditions, a time-course experiment was performed (Fig. 1A). Endometrial epithelial cells were collected on day 21 postpartum from healthy cows (n = 7) free of clinical diseases from the time of calving until sample collection. Cells were cultured as above and exposed to either ultrapure E. coli O111:B4 LPS (1 μg/mL) or medium alone under thermoneutral (38.5°C) or heat shock (40.0°C) conditions for 24 h. This initial exposure was defined as the response period. After 24 h, cells were washed three times with pre-warmed DPBS and all treatment groups received fresh complete culture medium and were incubated under thermoneutral conditions (38.5°C) for an additional 12 h, which was designated as the recovery period. Cells were collected at 0, 12, 18, 24, 25, 27, 30, and 36 h after the initial treatment. At each time point, total RNA was collected by adding 350 μL of RLT buffer (Qiagen, Germany) to cells and stored at −80°C until further processing.

Effect of sequential heat shock on cell viability and innate immune response of endometrial epithelial cells

To evaluate the effect of sequential heat shock on innate immune function (Fig. 1B), endometrial epithelial cells (n = 10) were first cultured under thermoneutral (38.5°C) or heat shock (40.0°C) conditions for 24 h in medium alone (period 1), followed by washing three times with pre-warmed DPBS. All cells were then incubated in fresh complete medium for 12 h under thermoneutral conditions (resting period). Following the resting period, cells were cultured under either thermoneutral (38.5°C) or heat shock (40.0°C) conditions for 24 h (period 2) with 1 μg/mL ultrapure E. coli O111:B4 LPS. At the end of period 2, cell viability was assessed using the MTT assay (above), or the total RNA was stabilized with 350 μL RLT buffer and stored at −80°C for subsequent RNA extraction and gene expression analysis.

High-resolution respirometry

High-resolution respirometry was performed using an Oroboros Oxygraph-2K (Oroboros Instruments, Innsbruck, Austria) following the manufacturer’s protocol (Fig. 1A). Temperature-controlled 0.5 mL chambers were cleaned and equilibrated with air-saturated RPMI-1640 medium (Fisher Scientific), sealed, and allowed to stabilize for 30 min to correct background oxygen flux using DatLab software (Oroboros Instruments). A modified substrate–uncoupler–inhibitor–titration (SUIT) protocol (O2 ce-pce D004) was used. Endometrial epithelial cells were collected on day 21 postpartum from healthy cows (as above; n = 7). Cells were cultured for 24 h under thermoneutral (38.5°C) or heat shock (40.0°C) conditions and exposed to 1 μg/mL of ultrapure E. coli O111:B4 LPS or medium alone. After treatment, cells were washed with warm PBS, trypsinized, counted, and resuspended in RPMI-1640. Approximately 4.5 × 10⁵ cells were added to each chamber to measure basal respiration. ATP-linked respiration was assessed by addition of oligomycin (2 mM; Sigma-Aldrich, USA). Maximal electron-transfer capacity was obtained by stepwise titration of FCCP (2 mM; Sigma-Aldrich). Residual oxygen consumption was determined following inhibition of complexes I and III with rotenone (0.5 μM; Sigma-Aldrich) and antimycin A (2.5 μM; Sigma-Aldrich), respectively. Oxygen flux was expressed as picomoles O₂ consumed per 10⁵ cells per second.

Effect of heat shock on innate immune response of immortalized bovine endometrial epithelial cells

Both experiments (time course and sequential heat shock) were replicated using immortalized BEND cells (ATCC, USA; CRL-2398) to determine the suitability of using a cell line model for future studies and minimizing animal use. BEND cells were cultured in 75 cm² flasks (Greiner Bio-One, USA) in complete medium containing 40% Ham’s F-12, 40% MEM, 10% fetal bovine serum, 10% horse serum, antibiotics, sodium bicarbonate, and D-valine (Fisher Scientific) as described by Molinari & Bromfield (2023). Cultures were maintained at 38.5°C with 5% CO₂, and the medium was replaced every 48 h. Cells were detached with trypsin–EDTA, washed in complete medium, resuspended at 10⁵ cells/mL, and seeded into 24-well plates (500 μL). After a 24 h equilibration, treatments identical to those applied to primary cells were administered. Gene expression and cell viability were assessed as described above. Mitochondrial respiration and post-sequential heat shock viability were not evaluated in BEND cells.

RNA isolation and real-time reverse transcription-PCR

Total RNA was extracted using the RNeasy Mini Kit (Qiagen) and quantified with a NanoDrop 2000 (Fisher Scientific). Complementary DNA (cDNA) was synthesized using the Verso cDNA Synthesis Kit (Fisher Scientific) with a no-reverse transcriptase control. Primers (Table 1) were designed to span exon–exon boundaries and validated according to Minimum Information for Publication of Quantitative Real-Time PCR Experiments guidelines (R2 > 0.98 and efficiency of 90–110%) (Bustin et al. 2009). Real-time PCR was performed in duplicate using iTaq SYBR Green Supermix (Bio-Rad, USA), 500 nM of each primer, and 2 μL of cDNA on a Bio-Rad CFX Connect system (95°C for 30 s, 40 cycles of 95°C for 5 s, and 60°C for 30 s). A no-template control was used for each primer pair with water replacing cDNA. Relative mRNA expression was calculated using the 2^−ΔCt method, normalized to GAPDH for the time-course experiment and to the geometric mean of ACTB and GAPDH for the sequential heat shock experiment.

Table 1.

Primer sequences used for real-time RT-PCR.

Gene symbol	Primer sequence	Accession number
ACTB	5′ – CAG​AAG​CAC​TCG​TAC​GTG​GG	NM_173979.3
	3′ – TTG​GCC​TTA​GGG​TTC​AGG​G
CXCL8	5′ – GCA​GGT​ATT​TGT​GAA​GAG​AGC​TG	NM_173925.2
	3′ – CAC​AGA​ACA​TGA​GGC​ACT​GAA
GAPDH	5′ – AGG​TCG​GAG​TGA​ACG​GAT​TC	NM_001034034.2
	3′ – ATG​GCG​ACG​ATG​TCC​ACT​TT
HSPA1A	5′ – GAC​AAG​TGC​CAG​GAG​GTG​ATT​T	NM_203322.3
	3′ – CAG​TCT​GCT​GAT​GAT​GGG​GTT​A
IL1B	5′ – CTT​CAT​TGC​CCA​GGT​TTC​TG	NM_174093.1
	3′ – CAG​GTG​TTG​GAT​GCA​GCT​CT
IL6	5′ – ATG​ACT​TCT​GCT​TTC​CCT​ACC​C	NM_173923.2
	3′ – GCT​GCT​TTC​ACA​CTC​ATC​ATT​C
TLR4	5′ – AGC​CAC​GGC​CAT​CCT​CTC​CT	NM_174198.6
	3′ – AGC​TCA​GGT​CCA​GCA​TCT​TGG​T

Statistical analysis

Analyses were performed using SAS statistical software v. 9.4 (SAS Institute, USA) with cow as the experimental unit. Data were analyzed using the general linear mixed model procedure. In experiment 1, cell viability and respirometry were evaluated against the fixed effects of LPS, heat shock (temp), and their interaction, with cow as the random effect. Time-course gene expression data were analyzed as repeated measures against the fixed effects of LPS, heat shock (temp), time, and their interactions. Response and recovery periods were analyzed separately. For the sequential heat shock experiment assessing cell viability and gene expression, the model included the fixed effects of period 1 heat shock, period 2 heat shock, and their interactions, with cow as the random effect. Residual normality was assessed using the Shapiro–Wilk test, and data were transformed when necessary. If data were not normally distributed, the outcome variable was transformed by the Box–Cox transformation. All gene expression data required transformation; however, results are presented as non-transformed values. Results are presented as mean ± standard error of the mean (SEM). Statistical significance was considered when P < 0.05.

Results

Effect of acute heat shock or LPS on viability of endometrial epithelial cells

Immunocytochemistry confirmed that the primary cultures consisted predominantly of epithelial cells, indicated by strong cytokeratin-18 (KRT18) staining with minimal vimentin or double-positive cells (Fig. 2A).

Figure 2.

Characterization and viability of bovine endometrial cell cultures. Endometrial epithelial cells were isolated using cytobrush on postpartum day 21 of healthy, cooled cows (n = 10). All cows were free of clinical diseases from calving to day 21 postpartum. Representative immunocytochemistry micrographs (A) of cultured endometrial cells stained for cytokeratin-18 (KRT18, red), vimentin (VIM, green), and DAPI (blue). The green arrow depicts a vimentin-positive cell, the white arrows depict vimentin and cytokeratin-18 double-positive cells, and the red arrow depicts a cytokeratin-18-positive cell. The scale bar represents 100 μm. Cell viability (B) was assessed by MTT assay after cells were cultured for 24 h under heat shock (40.0°C) or thermoneutral conditions (38.5°C) in combination with 1 μg/mL lipopolysaccharide (LPS) or medium alone. Data were analyzed using the general linear mixed model procedure against the fixed effects of LPS, heat shock (temp), and their interaction, with cow as the random effect, and are presented as mean ± SEM.

Heat shock (40.0°C) for 24 h did not affect endometrial epithelial cell viability compared with thermoneutral conditions (38.5°C; P = 0.652; Fig. 2B). Likewise, LPS exposure did not alter viability relative to medium-alone controls (P = 0.232), and no heat shock × LPS interaction was detected (P = 0.199), indicating that neither LPS nor heat shock is cytotoxic to endometrial epithelial cells.

Acute heat shock increased expression of HSPA1A and TLR4

The expression of HSP1A1 (Fig. 3A) was increased by heat shock during the 24 h response period (P < 0.001) and remained elevated (P = 0.008) during the recovery period when cells were returned to thermoneutral conditions. There was no effect of LPS or the heat shock × LPS interaction on HSP1A1 expression during either culture period. However, duration of culture (P < 0.001) affected HSP1A1 expression in both response and recovery periods.

Figure 3.

Effect of acute heat shock on expression of HSPA1A, TLR4, CXCL8, IL6, and IL1B. Endometrial epithelial cells were isolated using cytobrush on postpartum day 21 of healthy, cooled cows (n = 7) and cultured in vitro. All cows were free of clinical diseases from calving to day 21 postpartum. Cells were challenged with either medium alone or 1 μg/mL lipopolysaccharide (LPS) under heat shock (40.0°C) or thermoneutral conditions 38.5°C) conditions for 24 h (response period). After washing with phosphate-buffered saline, cells were returned to thermoneutral conditions in control medium alone for 12 h (recovery period). Samples were collected at 0, 12, 18, 24, 25, 27, 30, and 36 h after the initial treatment to analyze expression of (A) HSPA1A, (B) TLR4, (C) CXCL8, (D) IL6, and (E) IL1B. Data were analyzed using the general linear mixed model procedure with repeated measurements against the fixed effects of LPS, heat shock (temp), duration of culture (time), and the LPS × heat shock interaction, with cow as the random effect. Data are presented as mean ± SEM relative expression normalized to GAPDH.

Exposure of endometrial epithelial cells to LPS tended (P = 0.096) to increase the expression of TLR4 (Fig. 3B) during the response period, whereas heat shock, duration of culture, and the heat shock × LPS interaction had no effect on TLR4 expression during the response period. Within one hour of cells being returned to thermoneutral conditions during the recovery period, TLR4 expression was increased (P < 0.001) in cells that were exposed to heat shock. The duration of culture in the recovery period affected TLR4 expression (P = 0.001), whereas LPS and the heat shock × LPS interaction had no effect on TLR4 expression during the recovery period. These data show that heat shock induces an immediate increase in HSPA1A that is quickly resolved, and a delayed increase in TLR4 expression occurs only after returning cells to thermoneutral conditions, suggesting disparate mechanisms of heat shock regulation of these two genes.

Heat shock reduces innate immune responses of endometrial epithelial cells

During the 24 h response period, exposure of endometrial epithelial cells to LPS increased (P < 0.001) the expression of CXCL8, IL6, and IL1B (Fig. 3C, D, E). Heat shock reduced the expression of CXCL8 (P < 0.001) and IL6 (P = 0.006), with significant heat shock × LPS interaction for both CXCL8 (P < 0.001) and IL6 (P = 0.033). In contrast, the expression of IL1B was not changed by heat shock (P = 0.111) or the heat shock × LPS interaction (P = 0.149). Duration of culture affected CXCL8 (P < 0.001), IL1B (P = 0.005), and IL6 (P < 0.001).

During the 12 h recovery period, exposure to LPS increased the expression of CXCL8 (P = 0.001), IL1B (P = 0.006), and IL6 (P = 0.021). Heat shock tended to increase the expression of IL1B (P = 0.053) but did not affect the expression of CXCL8 (P = 0.312) or IL6 (P = 0.223). The duration of culture affected both CXCL8 (P < 0.001) and IL6 (P < 0.001) but not IL1B (P = 0.385). There was no heat shock × LPS interaction measured during recovery. Collectively, these data suggest that heat shock reduces the immediate expression of CXCL8 and IL6, while maintaining elevated IL1B expression after heat shock.

Cellular respiration of endometrial epithelial cells is impaired by lipopolysaccharide but not heat shock

To determine the effects of heat shock and LPS on cellular respiration, endometrial epithelial cells were cultured under thermoneutral (38.5°C) or heat shock (40.0°C) conditions with LPS or medium alone for 24 h (Fig. 4). The exposure of cells to LPS reduced basal respiration (P = 0.034), ATP-linked respiration (P = 0.031), maximal respiratory capacity (P < 0.001), and spare capacity (P < 0.001) compared with medium-alone controls. Heat shock or the heat shock × LPS interaction did not affect these parameters. Interestingly, heat shock decreased the oxygen consumption involved in proton leak (P = 0.023), whereas LPS (P = 0.186) and the heat shock × LPS interaction (P = 0.906) had no effect. Non-mitochondrial respiration was unchanged by LPS (P = 0.865), heat shock (P = 0.711), or their interaction (P = 0.865).

Figure 4.

Effect of heat shock and LPS on cellular respiration using high-resolution respirometry. Endometrial epithelial cells were isolated using cytobrush on postpartum day 21 of healthy, cooled cows (n = 7) and cultured in vitro. Cells were challenged with either medium alone or 1 μg/mL lipopolysaccharide (LPS) under heat shock (40.0°C) or thermoneutral conditions (38.5°C) conditions for 24 h. Cells were collected, washed, and evaluated for respirometry variables, including basal respiration (A), ATP-linked respiration (B), proton leak (C), maximal respiration (D), spare respiratory capacity (E), and non-mitochondrial respiration (F). Data were analyzed using the general linear mixed model procedure against the fixed effects of LPS, heat shock (temp), and their interaction, with cow as the random effect, and are presented as mean ± SEM.

Effect of sequential heat shock on endometrial epithelial cell viability and innate immune response

Endometrial epithelial cells were cultured in two periods, the first in medium alone and the second with LPS (Fig. 1B). During both periods, cells were cultured under thermoneutral (38.5°C) or heat shock (40.0°C) conditions in a crossover design with analysis at the end of period 2 (Fig. 5). There was no effect of heat shock in period 1 (P = 0.869), period 2 (P = 0.122), or their interaction (P = 0.834) on cell viability, suggesting that the culture conditions were not cytotoxic.

Figure 5.

Effect of LPS and sequential heat shock on endometrial epithelial cell viability, HSP1A1 expression, and innate immune responses. Endometrial epithelial cells were isolated using cytobrush on postpartum day 21 of healthy, cooled cows (n = 10) and cultured in vitro. All cows were free of clinical diseases from calving to day 21 postpartum. Cells were cultured for 24 h (period 1) under heat shock (40.0°C) or thermoneutral conditions (38.5°C) in medium alone followed by a resting period of 12 h at 38.5°C. Cells were then cultured for a subsequent 24 h (period 2) in the presence of 1 μg/mL LPS under heat shock (40.0°C) or thermoneutral conditions (38.5°C). Cell viability (A) was assessed by MTT assay and measured at an optical density (OD) of 540 nm. Following the second culture period, the expressions of (B) HSPA1A, (C) TLR4, (A) CXCL8, (B) IL6, and (C) IL1B were measured relative to the geometric mean of ACTB and GAPDH. Data were analyzed using the general linear mixed model procedure against the fixed effects of period 1, period 2, and their interaction, with cow as the random effect, and are presented as mean ± SEM. The superscripts represent difference between groups.

Heat shock in period 1 (P = 0.041), period 2 (P < 0.001), and their interaction (P = 0.024) increased HSPA1A expression (Fig. 5B). In particular, heat shock during period 2 induced the greatest increase in HSP1A1 compared to cells exposed to heat shock in period 1 (P < 0.001), consistent with an acute heat shock response. Expression of TLR4 was affected by heat shock in period 1 (P < 0.001) but not in period 2 (P = 0.251) or their interaction (P = 0.551), suggesting a delayed increase in TLR4 expression after heat shock (Fig. 5C).

Heat shock affected CXCL8 expression during both period 1 (P < 0.001) and period 2 (P < 0.001) but not their interaction (Fig. 5D). In particular, cells exposed to heat shock in period 1 exhibited increased CXCL8 expression, whereas heat shock in period 2 decreased CXCL8 expression. Expression of IL6 was decreased by heat shock in period 2 (P < 0.001) but was not affected by heat shock in period 1 (P = 0.954) or the interaction (P = 0.801; Fig. 5E). Expression of IL1B was increased by heat shock in period 1 (P = 0.025) but not affected by heat shock in period 2 (P = 0.613) or their interaction (P = 0.606; Fig. 5F). There was no evidence that sequential heat shock (heat shock in both periods 1 and 2) affected any genes evaluated.

Effect of acute heat shock and LPS on immortalized bovine endometrial epithelial cells

Both acute and sequential heat shock experiments were replicated using immortalized BEND cells to determine the suitability of using a cell line model for future studies (Fig. S1 and S2 (see section on Supplementary materials given at the end of the article)). Acute heat shock affected cell viability (P < 0.001) with heat shock increasing optical density (OD₅₄₀) values in the MTT assay compared with thermoneutral controls (Fig. S1A). Neither LPS exposure (P = 0.289) nor the heat shock × LPS interaction (P = 0.144) affected cell viability. This suggests that acute heat stress increased proliferation of BEND cells, which was not observed in primary endometrial epithelial cells.

During the 24 h response period, expression of HSPA1A, CXCL8, IL6, and IL1B were all affected by LPS, heat shock, and duration of culture, which differed from primary endometrial epithelial cells (Fig. S1B, C, D, E). The expression of CXCL8 and IL6 was also affected by the heat shock × LPS interaction. While the expression of CXCL8 and IL6 was decreased by heat shock in a similar fashion to primary endometrial epithelial cells, heat shock increased the expression of IL1B, which was not affected in primary endometrial epithelial cells.

During the 12 h recovery period (Fig. S1B, C, D, E), expression of HSPA1A, CXCL8, IL6, and IL1B by BEND cells was affected by LPS and duration of culture. Interestingly, heat shock increased the expression of CXCL8, IL6, and IL1B in BEND cells, which was not observed in primary endometrial epithelial cells. Furthermore, the heat shock × LPS interaction affected BEND cell expression of IL6 and IL1B, which was not observed in primary endometrial epithelial cells.

Compared to primary endometrial epithelial cells, sequential heat shock of BEND cells (Fig. S2) affected HSP1A1 and CXCL8 expression in a similar manner during period 1 (P < 0.001) and period 2 (P < 0.001). Conversely, the effect of heat shock on BEND cell IL6 and IL1B expression was opposite that observed in primary endometrial epithelial cells. Heat shock increased the expression of BEND cell IL6 and IL1B during period 1 (P < 0.001) and period 2 (P < 0.001) and affected their interaction (P < 0.03); however, in primary endometrial epithelia cells, heat shock decreased the expression of IL6 and did not affect the expression of IL1B. Furthermore, sequential heat shock of BEND cells exacerbated the expression of both IL6 and IL1B to a greater extent than other treatments, which were not observed in primary endometrial epithelial cells.

Discussion

Heat stress is a major challenge for the dairy industry due to the increased metabolic heat load of high-producing cows that results in subfertility, increased metabolic disorders, and uterine diseases (Das et al. 2016, Fournel et al. 2017, Gernand et al. 2019, Molinari et al. 2022). Although heat stress is known to impair innate immunity, its specific impact on uterine epithelial defense mechanisms remains poorly defined (Dahl et al. 2020, Molinari & Bromfield 2023). Here, we characterized the innate immune function of primary endometrial epithelial cells using time-course and sequential heat shock models. In the time-course experiment, acute heat shock did not affect cell viability but did induce an increase in the expression of HSPA1A, confirming a cellular heat shock response. Similarly, acute heat shock reduced the LPS response of endometrial epithelial cells by reducing CXCL8 and IL6. In contrast, heat shock increased TLR4 expression during the recovery period. Heat shock also reduced oxygen consumption linked to proton leak, while LPS decreased basal, ATP-linked, maximal, and spare respiratory capacity of endometrial epithelial cells. Sequential heat shock did not alter the expression of HSPA1A, CXCL8, IL6, or IL1B compared to a single acute heat shock event. Interestingly, prior heat shock increased TLR4 expression regardless of subsequent incubation conditions, suggesting a carryover effect of heat shock. In contrast, immortalized BEND cells exhibited exaggerated inflammatory responses, with heat shock increasing IL6 and IL1B responses to LPS during the response experiment and amplifying cytokine expression after sequential heat shock. These discrepancies emphasize the limited physiological fidelity of immortalized cell lines and support the use of primary epithelial cells to better represent in vivo uterine immune responses to heat stress. Collectively, these findings demonstrate that heat shock alters the innate immune responsiveness of endometrial epithelial cells to LPS, primarily by suppressing acute inflammatory response while increasing post-stress TLR4 expression.

Bovine immune resilience reflects the balance between resistance, which restricts pathogen growth, and tolerance, which limits tissue damage (Sheldon et al. 2019). Innate immunity provides the uterus with rapid, nonspecific defense against pathogens, where endometrial epithelial cells detect LPS through TLR4 and activate pro-inflammatory signaling to induce cytokines and chemokines, including IL-1β, IL-6, and IL-8 (Cronin et al. 2012, Turner et al. 2014). Increased expression of these immune mediators is linked to uterine diseases, as cows that develop endometritis later have increased endometrial IL1A, IL1B, and IL6 expression early postpartum compared to cows that remain healthy (Herath et al. 2007). TLRs also sense tissue damage, initiating inflammation even without infection (Marsh et al. 2009). In this study, acute heat shock did not alter immediate TLR4 expression but increased expression after heat shock was removed, indicating that prior heat exposure modifies receptor regulation. This observation aligns with previous in vitro and in vivo studies showing that both chronic heat stress (21 days) and acute heat shock (1 h) can upregulate TLR4 expression in immune cells. Increased expression of TLR4 can enhance cellular responsiveness to bacterial components, as observed in pigs that increase IL-6 production during heat stress (Eicher et al. 2004, Zhou et al. 2005, Chen et al. 2006, Ju et al. 2014).

Acute heat shock induced a rapid but transient increase in HSPA1A expression, confirming activation of the heat shock response. The heat shock protein HSP70, encoded by HSPA1A, and other heat shock proteins have been proposed as endogenous activators of TLR4 (Chen et al. 1999, Matzinger 2002, Hentze et al. 2016). Therefore, it is plausible that the increased TLR4 expression observed in this study could be a secondary effect of elevated HSPA1A expression. Moreover, heat shock alone, in the absence of LPS, did not increase the expression of pro-inflammatory mediators CXCL8, IL1B, or IL6, indicating that HSPA1A induction was not sufficient to trigger an innate immune response. Heat shock also reduced LPS-stimulated CXCL8 and IL6 expression despite stable TLR4 expression during heat shock, suggesting suppression of downstream signaling such as NF-κB or MAPK, rather than impaired pathogen recognition (Guzhova et al. 1997). This pattern of TLR4 expression was confirmed in the sequential heat shock experiment, although prior heat shock increased TLR4 expression, heat shock continued to constrain the expression of CXCL8 and IL6 in response to LPS. Notably, prior heat shock increased the CXCL8 response to LPS under thermoneutral conditions, suggesting that heat shock may induce carry-over effects on cells. Here, we measured the gene expression of only three pro-inflammatory mediators, and as such, it is important to consider for future experimentation other mediators of cellular responses that may counteract those observed here, or mediators involved in tissue repair or disease not evaluated here. In addition, it will be imperative to evaluate the protein expression of these various immune modulating factors to determine if protein expression mirrors gene expression as we predict. Based on the data presented, it is worth considering that heat shock exposure may induce cellular alterations that persist after the resolution of heat shock and affect cellular responses under subsequent, physiological conditions.

The findings presented here are similar to previous experiments that describe increased expression of TLR2 and TLR4 in pigs after heat stress (Ju et al. 2014) and reduced IL6 expression in bovine endometrial epithelial cells after heat shock (Sakai et al. 2022). These prior studies suggest a potential inhibitory role of heat shock proteins in regulating cytokine expression. Conversely, our previous work using the immortalized BEND cell line describes increased IL6 and IL1B responses after heat shock. While we replicated this increased inflammatory response in BEND cells in the current study, interestingly, primary endometrial epithelial cells exhibited an opposing response to heat shock by decreasing LPS-induced expression of CXCL8 and IL6. Similar divergent responses between BEND cells and primary endometrial epithelial cells have been reported before and likely reflect differences in stress sensitivity and immune signaling between immortalized and primary cell types (Parent et al. 2002, Godkin et al. 2008). Together, the results here support the argument that primary endometrial epithelial cells provide a more physiologically relevant model than BEND cells to study heat shock-induced modulation of inflammatory signaling. When determining the precise mechanism underlying the suppression of pro-inflammatory gene expression in cells under heat shock conditions, it will be imperative to utilize primary endometrial epithelial cells in place of immortalized BEND cells.

Because immune activation is tightly coupled with cellular bioenergetics, we examined mitochondrial function in cells exposed to heat shock and/or LPS. Here, we demonstrated that LPS reduced basal, ATP-linked, maximal, and spare respiratory capacities, indicating that immune activation imposes a significant energetic burden on endometrial epithelial cells, consistent with the well-established metabolic shift toward glycolysis during innate immune activation (O’Neill & Hardie 2013, Slimen et al. 2014). The reduction in mitochondrial respiration also aligns with prior findings that show LPS induces mitochondrial dysfunction and shifts cellular metabolism toward glycolysis as part of the inflammatory response in macrophages (Tannahill et al. 2013, Mills et al. 2016). Furthermore, stimulation of TLR4 with LPS has been reported to reduce oxygen consumption as an energy-conserving response, reflecting a trade-off between energy production and immune activation (Ganeshan et al. 2019).

Heat shock has also been shown to impair tissue bioenergetics by disrupting mitochondrial biogenesis and oxidative phosphorylation efficiency (Singh 2004, Marquez Acevedo et al. 2025). We initially hypothesized that heat shock would compromise mitochondrial capacity and thereby limit the ability of cells to sustain immune responses to LPS (Bird 2019). Interestingly, heat shock did not impair mitochondrial capacity; instead, it reduced proton leak, suggesting tighter coupling of oxidative phosphorylation (Brand & Nicholls 2011). Reduced proton leak may represent an adaptive cellular strategy to limit heat and reactive oxygen species production (Porter 2001). However, tighter coupling could also limit the dissipation of heat and reactive oxygen, potentially increasing susceptibility to oxidative damage during prolonged heat shock exposure (Brand & Nicholls 2011). Nonetheless, we did not observe any effect of heat shock on basal, ATP-linked, maximum, or spare capacity respiration. It is important to note that mitochondrial respiration was measured after the removal of cells from the heat shock environment and their preparation and analysis was performed under thermoneutral conditions. This short recovery period of ∼30 min may have allowed partial restoration of mitochondrial function, which warrants further investigation under real-time heat shock conditions.

Collectively, these data show that acute heat shock at 40.0°C transiently increases HSPA1A, dampens LPS-induced CXCL8 and IL6 expression, and increases TLR4 expression after the removal of heat shock. Heat shock did not compromise the viability or respiratory capacity of cells, suggesting that inflammatory signaling is more heat-sensitive than cellular metabolism in bovine endometrial epithelial cells. These findings provide novel insight into how heat shock may reduce uterine immune resilience in dairy cows that is related to the increased incidence of uterine disease in heat- stressed cows. These findings highlight the need for broader profiling of cytokine networks involved in disease pathogenesis, including chemokine, pro-inflammatory, anti-inflammatory, and Th17 cytokine networks. Determination of heat-sensitive signaling pathways in the endometrium responsible for immune function and disease development will be helpful in combating heat-related disease development in domestic species.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the work reported.

Funding

This material is based upon work that is supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, under award number 2024-67015-42414. The authors would like to thank the Higher Education Commission of Pakistan for partially funding the PhD program of Arslan Tariq.

Author contribution statement

AT performed experimentation, data analysis, and writing. MS, ZKS, JR, and FY assisted in experimentation and proofreading. JJB acquired funding and performed data analysis, data interpretation, and writing.

Data availability

The data underlying this article will be shared upon reasonable request to the corresponding author.