If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Viability and function dynamics of circulating versus endometrial polymorphonuclear leukocytes in postpartum dairy cows with subclinical or clinical endometritis
Department of Internal Medicine, Reproduction and Population Medicine, Faculty of Veterinary Medicine, Ghent University, Salisburylaan, Merelbeke, 9820, Belgium
Department of Internal Medicine, Reproduction and Population Medicine, Faculty of Veterinary Medicine, Ghent University, Salisburylaan, Merelbeke, 9820, BelgiumLaboratory of Veterinary Physiology and Biochemistry, Department of Veterinary Sciences, University of Antwerp, Universiteitsplein, Wilrijk, 2610, Belgium
Laboratory for Animal Nutrition and Animal Product Quality, Department of Animal Sciences and Aquatic Ecology, Faculty of Bioscience Engineering, Ghent University, Coupure Links, Ghent, 9000, Belgium
Laboratory for Animal Nutrition and Animal Product Quality, Department of Animal Sciences and Aquatic Ecology, Faculty of Bioscience Engineering, Ghent University, Coupure Links, Ghent, 9000, Belgium
Laboratory of Veterinary Physiology and Biochemistry, Department of Veterinary Sciences, University of Antwerp, Universiteitsplein, Wilrijk, 2610, Belgium
Department of Internal Medicine, Reproduction and Population Medicine, Faculty of Veterinary Medicine, Ghent University, Salisburylaan, Merelbeke, 9820, Belgium
We aimed to compare the viability of circulating polymorphonuclear leukocyte (cPMN) and endometrial PMN (ePMN) and their function dynamics in postpartum dairy cows with subclinical (SCE) or clinical endometritis (CE). To do so, blood samples from 38 Holstein cows were collected at −7, 9, 21, and 36 d relative to calving, and endometrial cytology samples from 32 Holstein cows were harvested at 9, 21, and 36 d postpartum. Uterine health status was assessed at 36 d postpartum, and cows were classified as healthy (absence of abnormal vaginal discharge and ≤5% ePMN), SCE (absence of abnormal vaginal discharge and >5% ePMN), or CE (mucopurulent or purulent vaginal discharge and >5% ePMN). Viability (viable, apoptotic, and necrotic) and function parameters phagocytosis (PC), oxidative burst, and intracellular proteolytic degradation were evaluated for cPMN via flow cytometry. For ePMN, only viability and PC were evaluated. The association of cPMN and ePMN viability and functional parameters with reproductive tract health classification were fitted in mixed linear regression models, accounting for repeated measures, sampling day, and interactions of reproductive tract status and day. Cows with CE had a lower proportion of cPMN viability (84.5 ± 2.1%; least squares means ± standard error) and a higher proportion of apoptosis (14.4 ± 2.0%) than healthy (92.4 ± 1.3 and 6.7 ± 1.3%, respectively) or SCE (95.3 ± 2.4 and 3.8 ± 2.3%, respectively) at 9 d postpartum. Interestingly, cPMN intracellular proteolytic degradation was lower [6.2 ± 0.1 median fluorescence intensity (MFI)] in SCE compared with healthy (6.7 ± 0.08 MFI) or CE (6.8 ± 0.1 MFI) at d 9 postpartum. No other differences in cPMN function were found among experimental groups. The proportion of necrotic ePMN was higher for healthy (49.6 ± 5.1%) than SCE (27.4 ± 7.3%) and CE (27.7 ± 7.3%) cows at 36 d postpartum. Also, at 36 d postpartum, the proportion of ePMN performing PC was higher in CE (47.0 ± 8.6%) than in healthy (18.4 ± 7.6%) cows, but did not differ from SCE cows (25.9 ± 8.7%). Results of the present study suggest that cPMN viability and function at 9 d postpartum are associated with the development of uterine disease. Furthermore, ePMN at 36 d postpartum are mostly necrotic in healthy cows but viable and functional in cows with CE, probably due to active uterine inflammation. Remarkably, ePMN in cows with SCE at 36 d postpartum are also mostly viable but seem to display a numerically lower proportion of PC compared with ePMN in CE cows.
Dysregulation of inflammation in the transition period impairs neutrophil function and augments the risk of reproductive tract inflammatory disease in dairy cows (
). The term “dysregulation” accounts for the degree or duration of inflammation that weakens rather than improves health. At the uterine level, it has been shown that a substantial influx of PMN within 2 wk postpartum is associated with better uterine health and fertility (
). Rapid migration of PMN toward the uterine lumen is essential to tip the critical balance between physiological inflammation and repair and pathological inflammation and disease (
). This is because of the potentially pathogenic bacteria such as Bacteroides, Fusobacterium, Porphyromonas, and Trueperella that are ubiquitous in the postpartum uterus of dairy cows (
). Thus, during the first 2 wk after calving, a high number of functionally active PMN is key to maintaining a low relative abundance of pathogenic bacteria in the uterine environment. However, the excessive or persistent presence of dysfunctional PMN in the uterine lumen can induce endometrial cell damage and contribute to disease (
Assessment of the main pathogens associated with clinical and subclinical endometritis in cows by culture and MALDI-TOF mass spectrometry identification.
). It is believed that SCE is a result of metabolic and inflammatory dysfunction that impairs both endometrial PMN (ePMN) function and their ability to undergo apoptosis and necrosis, and ultimately abandon the uterine cavity (
). We also reported that the ePMN in early postpartum healthy cows are remarkably dynamic, showing a decreasing ePMN percentage, a decreasing proportion of apoptotic ePMN versus an increasing proportion of necrotic ePMN, and a decreasing proportion of ePMN performing phagocytosis over time (
). This contrasted with the blood compartment, which only showed a higher number of cPMN in the prepartum period but no changes in viability or function during the peripartum period (
). In the present study, we hypothesized that cPMN and ePMN viability and function are different in cows that develop distinct types of reproductive tract inflammatory disease (CE vs. SCE) versus healthy postpartum dairy cows. Therefore, we aimed to compare the cPMN and ePMN counts and function dynamics in postpartum dairy cows with or without SCE or CE.
MATERIALS AND METHODS
Ethical Statement
For herd A, animal handling and sampling procedures were approved by the ILVO (Flanders Research Institute for Agriculture, Fisheries and Food) Animal Ethics Committee (EC 2018/329, Melle, Belgium). For herd B, all animal handling and sampling procedures were approved by the Ethical Committee of Animal Testing of the University of Antwerp (Antwerp, Belgium) and the Flemish Government, Department of Strategy, International Policy, and Animal Welfare (Brussels, Belgium) with ECD-dossier 2019-44.
Animals and Management
In an observational cohort design, a total of 45 multiparous (parity 2 to 7) Holstein cows, from 2 dairy herds (herd A, n = 26 and herd B, n = 19) in Flanders, Belgium, were enrolled between February and October 2020 in herd A and from April to August 2021 in herd B. Herd A was the experimental dairy farm of ILVO (Melle, East Flanders, Belgium); herd B was a commercial dairy farm (Nevele, East Flanders, Belgium). The per-farm average daily milk yield for herds A and B were 31.5 (fat 4.3% and protein 3.7%) and 30.2 kg/cow (fat 4.2% and protein 3.4%), respectively (test-day recording; Cattle Improvement Co-operative, CRV, Arnhem, the Netherlands).
For both farms, pregnant dry cows were housed in freestall barns and moved to calving pens when imminent calving indicators were observed (e.g., pelvic ligament relaxation, udder distension, and teat filling) or 3 d before the expected calving date. Within 3 d after calving, fresh cows were relocated to a freestall lactating pen and milked twice daily in a parlor.
Not many data are available on the assessment of ePMN viability and function analyzed via flow cytometry, especially not comparing healthy and SCE. Thus, the sample size was calculated for the detection of differences in 10 ± 10 (mean ± SD) in the percentage of PC in cPMN or ePMN between healthy and CE (significance level = 0.05 and power = 80%) as described by
Before sampling, cows were restrained in a headlock stanchion. All samples were collected in the morning (between 0800 and 1000 h). Blood samples were collected 7 to 9 d before expecting calving (or −7 d). Expected calving was calculated as 280 d after the last insemination date. Blood and endometrial samples were collected at 9 ± 1, 21 ± 1, and 36 ± 1 d postpartum. At every sampling day, rectal temperature (using a digital thermometer) and BCS (5-point scale;
) were also measured. Only healthy cows with no fever or clinical signs of any disease other than purulent vaginal discharge were included in the present study.
At −7, 9, 21, and 36 d relative to calving, blood samples were collected from the coccygeal vessels using blood collection needles (20-gauge, BD Vacutainer, PrecisionGlide, Becton Dickinson) and vacuum tubes (BD Vacutainer, Becton Dickinson). For cPMN isolation, vacuum tubes contained 1.5 mL of acid citrate dextrose-A (22 g/L trisodium citrate, 8 g/L citric acid, and 24.5 g/L dextrose); for blood cell count, vacuum tubes contained EDTA (K2E, 0.18%); and for the analysis of nonesterified fatty acids (NEFA) and BHB, vacuum tubes with clot activator (serum separation tube, SST II) were used. Blood for glucose analysis in farm A were collected in vacuum tubes with sodium fluoride (NaF 2.5 mg/mL + potassium oxalate 2 mg/mL). In farm B, blood glucose analysis was done in whole-blood samples using an electronic cowside device (FreeStyle Precision Neo; Abbott Diabetes Care Inc.) immediately following collection. Blood tubes were gently inverted 10 times and transported to the laboratory within 30 min. Blood tubes containing anticoagulants were transported in a refrigerated container.
For endometrial sampling, the perineum of the cow was cleaned with fresh water and iodide soap, dried with paper towels, and disinfected with 70% ethanol. Under rectal guidance, a sterile double-guarded equine cytology brush (9 d postpartum; Minitube) or a histobrush (21 and 36 d postpartum; Puritan) adapted to a stainless-steel AI gun and covered with a plastic sanitary sheath, was manipulated through the cervix, toward the uterine body. The outer guard of the cytobrush device was then pulled back, and the brush was exposed from the inner guard. Under transrectal guidance, the brush was gently pushed against the dorsal wall of the uterine body and rotated 3 times. Once outside the genital tract, the endometrial cytobrush sample was gently rolled onto a clean microscope slide to prepare an endometrial cytology slide, which was then air-dried. Next, the head of the brush was cut with scissors and placed in a 1.5 mL microcentrifuge tube containing 1 mL of RPMI-1640 (Gibco/Thermo Fisher Scientific) supplemented with 0.1% BSA (Sigma-Aldrich) and 0.18% K2EDTA (BD Vacutainer, Becton Dickinson). Samples were transported in a refrigerated container to the laboratory within 30 min.
After endometrial sampling, vaginal discharge was scored using the Metricheck device (Simcro Datamars) and categorized as 0 = clear mucus, 1 = mucus with flecks of pus, 2 = mucopurulent discharge (≤50% pus), and 3 = purulent discharge (>50% pus).
The case disease definition for uterine health status was assessed at 36 d postpartum. Vaginal discharge score ≥2 with endometrial PMN >5% was referred to as CE. Vaginal discharge score ≤1 with endometrial PMN >5% was referred to as SCE. Cows were diagnosed as healthy controls when vaginal score was ≤1 and the endometrial PMN was ≤5% (
Total and Differential cPMN Count, Analysis of Blood Metabolites, and Evaluation of ePMN Slides
Total and differential white blood cell counts were analyzed using the ProCyte Dx hematology analyzer (IDEXX Laboratories Inc.). Blood smear slides were prepared and stained with Diff-Quick (Speedy-Diff complete kit, Clin-Tech Ltd.) to calculate the percentage of band cPMN after counting all cPMN in the marginal zone via light microscopy (Kyowa Optical).
Tubes without anticoagulant and with sodium fluoride were centrifuged at 1,500 × g for 15 min at 21°C, and resultant sera were divided into aliquots and stored at −20°C. Glucose, NEFA, and BHB were analyzed using Gallery Discrete Analyzer (ThermoFisher Scientific) and Randox kits (Randox Laboratories Ltd.). The intra-assay coefficients of variation for all metabolites were <10%. Blood tubes without anticoagulant were not collected at 36 d postpartum in herd A, so no analyses of serum metabolites were performed at this time point.
Endometrial cytology slides were stained with Diff-Quick and mounted with Eukitt (O. Kindler GmbH). All slides were assessed by a trained veterinarian. A total of 300 nucleated cells in randomly selected fields were counted using a bright-field microscope (Kyowa Optical) and the proportion of ePMN to all nucleated cells (ePMN%) was calculated.
cPMN and ePMN Isolation, Viability, and Function
Blood and endometrial PMN isolation were performed as described by
. In short, cPMN were isolated from acid citrate dextrose–mixed blood after Ficoll-Paque PLUS (General Electric Healthcare Bio-Sciences AB) density gradient centrifugation and erythrocyte lysis with water (for 45 s). Isolated cPMN were suspended in a stock solution of RPMI-BSA to a final concentration of 5 × 106 per mL. Endometrial cells were dislodged from the cytobrush by vortexing the microcentrifuge tubes containing the endometrial samples for 1 min. Samples were then washed and erythrocytes were lysed with water. The ePMN suspension was diluted in RPMI-BSA to a final concentration of 5 × 106 ePMN per mL, whereas samples with lower concentrations of ePMN remained undiluted.
After isolation, cPMN and ePMN were immunolabeled and stained as described by
. In short, 1 × 106 PMN were resuspended in 100 µL of flow cytometry buffer (FACS buffer) containing CH138A (final concentration of 10 μg/mL), a bovine granulocyte-specific monoclonal antibody (mouse IgM; Washington State University Monoclonal Antibody Center;
The development and analysis of species specific and cross reactive monoclonal antibodies to leukocyte differentiation antigens and antigens of the major histocompatibility complex for use in the study of the immune system in cattle and other species.
). After incubation (30 min at 4°C in the dark), PMN were washed with FACS buffer and incubated for 15 min at 4°C in the dark in 100 µL of FACS buffer with secondary antibody (goat anti-mouse IgM Alexa-647; final concentration of 4.8 µg/mL, A-21238, Molecular Probes). After incubation and washing, cells were resuspended in 100 µL of an annexin-V + propidium iodide (PI) working solution. This was composed of 1 mL of a Ca2+-rich incubation buffer mixed with 20 µL of Annexin-V-FLUOS (fluorescein; Roche Diagnostics GmbH) and 20 µL of PI (final concentration of 1 µg/mL, Molecular Probes/Invitrogen). After incubation (10 min at room temperature in the dark), cells were put on ice and protected from light until analysis. Autofluorescent controls were included using the same PMN concentration but without addition of antibodies and fluorescent markers.
Functionality tests were performed as described by
. Phagocytosis (PC) tests were performed for cPMN and ePMN. First, a stock solution of zymosan-activated serum was prepared by incubating (60 min at 38.5°C) pooled blood serum from healthy cows with zymosan A from Saccharomyces cerevisiae (final concentration of 10 mg/mL, Sigma-Aldrich). The activated serum was centrifuged at 390 × g for 10 min, aliquoted, and stored at −20°C. Next, 1 × 106 PMN (blood or endometrium) and 50 µL of zymosan-activated serum were suspended in 200 µL of RPMI-BSA. Per 1 × 106 PMN, 1 µL of fluorescently labeled 1-µm beads (FluoSpheres carboxylate, yellow-green (505/515), Life Technologies Corp.) were added. After incubation (30 min at 38.5°C), cells were washed and diluted in RPMI-BSA, put on ice, and protected from light until analysis. Autofluorescent controls were included using the same PMN concentration and zymosan-activated serum, but without addition of the FluoSpheres.
Oxidative burst (OB) and intracellular proteolytic degradation (DQ-ovalbumin, DQ-ova) were only tested for cPMN. For OB, 1 × 106 PMN in 200 µL of RPMI-BSA were supplemented with 2 µL of 2′,7′-dichlorodihydrofluorescein diacetate (H2DCFDA; final concentration of 10 µM; Life Technologies Corp.), and incubated for 15 min at 38.5°C. Next, cells were stimulated with 200 µL of phorbol 12-myristate 13-acetate (Sigma-Aldrich; 200 ng/mL in RPMI-BSA) and incubated again for 15 min at 38.5°C. The autofluorescent controls were not stimulated but supplemented with 200 µL of RPMI-BSA. For DQ-ova, 2 × 106 PMN in 400 µL of RPMI-BSA were supplemented with 10 µL of DQ-ovalbumin (10 μg/mL, Life Technologies Corp.) and incubated for 45 min at 38.5°C. Autofluorescent controls were included using the same PMN concentration but without addition of DQ-ovalbumin. After incubation, all cells (OB and DQ-ova assay) were washed and diluted in RPMI-BSA, put on ice, and protected from light until analysis.
Evaluation of PMN Viability, Function, and Function Capacity
Samples were analyzed using a CytoFLEX 3-laser flow cytometer and CytExpert software v2.0.0.153 (Beckman Coulter Inc.). The recording threshold was set at 10,000 events per sample in a high forward (FSC) × high side scatter (SSC) gate on FSC versus SSC plots, or a measuring time of 120 s (the acquisition velocity was 30–60 µL/min). Cell fluorescence was excited at 488 and 638 nm, and all fluorescent emissions were measured using emission filters as described in Supplemental Table S1 (https://doi.org/10.17605/OSF.IO/4KYXZ;
Viability and function dynamics of circulating versus endometrial polymorphonuclear leukocytes in postpartum dairy cows with subclinical or clinical endometritis.
, the gating strategy was as follows: single-cell populations of PMN were gated on FSC-A versus FSC-H scatter plots to exclude aggregates. Polymorphonuclear leukocytes could be identified morphometrically, based on their characteristic high FSC and high SSC values. This gating strategy was applied for all cPMN assays and for the ePMN PC test (Supplemental Figure S1; https://doi.org/10.17605/OSF.IO/4KYXZ; Bogado Pascotini, 2023). For the ePMN viability assay, PMN were not a priori gated based on morphometry, but all events were included, except the debris fraction on the FSC versus SSC plot. For the cPMN and ePMN viability assays, PMN were additionally gated based on their CH138A × Alexa Fluor 647 positivity. Viable CH138A+ PMN were annexin−/PI−, apoptotic CH138A+ PMN were annexin+/PI−, and necrotic CH138A+ PMN were annexin+/PI+ (Supplemental Figure S2; https://doi.org/10.17605/OSF.IO/4KYXZ; Bogado Pascotini, 2023).
Fluorescence emitted in the PC, DQ-ova, and OB assays were measured with the 525 ± 20 nm filter [fluorescein isothiocyanate (FITC) channel]. The percentage of PMN that performed PC (PPC) was calculated and the difference between the median fluorescence intensity (MFI) of positive samples and the MFI of the respective autofluorescent controls was calculated for PC (MFIPC), OB (MFIOB), and DQ-ova (MFIDQ). Moreover, the PMN function capacity was calculated by multiplying the PPC, MFIPC, MFIOB, and MFIDQ by the cPMN counts (leukogram) or ePMN% of the respective sample times the proportion of viable PMN.
Statistical Analyses
On-farm and laboratory-collected data were exported and merged to Excel (Microsoft Corp.), where data exploration and organization were done using the PivotTables function. The statistical analyses were performed using R (version 4.0.4;
). Blood metabolites, PMN counts, and PMN function parameters were analyzed by Shapiro-Wilk test and, when not normally distributed (P < 0.05), they were transformed (square root or log10). The function lmer of the R package lme4 (
) was used to fit mixed linear regression models. The effect of sampling day, reproductive tract inflammatory disease status (healthy vs. SCE vs. CE), and their interaction were forced into each model to test their effects on blood metabolites, PMN counts, and PMN function parameters. In all models, BCS at enrollment (≤3.5 or ≥3.75) was included as covariate and retained when P < 0.05. All models accounted for repeated measures (day within cow) and the herd origin (A and B) as a random effect. Model residuals were assessed using a scatterplot of the studentized residuals for homoscedasticity, linear predictor for linearity, and a Shapiro-Wilk test for normality. For all transformed and nontransformed variables, the residuals of the models were normally distributed (Shapiro-Wilk P > 0.05). Differences between levels of explanatory variables were assessed with the Tukey post hoc test. Results are expressed as least squares means and standard errors. The Spearman correlation test (r) between each blood metabolite (glucose, NEFA, or BHB) with cPMN and ePMN viability and function parameters and between cPMN and ePMN viability parameters and PC, were calculated. For the above mentioned analyses, R-packages lmerTest (
From the initial pool of 45 cows, 7 cows were excluded due to retained placenta, metritis, displacement of the abomasum, or vaginal score ≥2 with ePMN ≤5%. All 38 cows included were multiparous (parity range 2 to 7); 18 cows had a BCS ≤3.5 (11 from herd A and 7 from herd B) and 20 cows had a BCS ≥3.75 (11 from herd A and 9 from herd B) in their prepartum period. The final sample size for cPMN measurements consisted of 141 blood samples from 38 Holstein cows (22 from herd A and 16 from herd B). For ePMN measurements, 91 endometrial samples from 32 Holstein cows (19 from herd A and 13 from herd B) were included. We were not able to collect or analyze all samples per included cow as originally planned, partly due to SARS-CoV-2 restrictions (cancelled sampling days), incapacity to reach the uterine lumen with the cytobrush, spontaneous in vitro cell aggregation for cPMN samples, or an insufficient number of cells in some ePMN samples.
From the 38 cows for the cPMN measurements, 22 were healthy (14 from herd A and 8 from herd B), 8 were diagnosed with SCE (4 from herd A and 4 from herd B), and 8 were diagnosed with CE (4 from herd A and 4 from herd B). From the 32 cows for the ePMN measurements, 16 were healthy (11 from herd A and 5 from herd B), 8 were diagnosed with SCE (4 from herd A and 4 from herd B), and 8 were diagnosed with CE (4 from herd A and 4 from herd B).
Effect of Uterine Disease on Blood Metabolite Concentrations
The blood concentrations of glucose, NEFA, and BHB from −7 to 21 d relative to calving for each group are shown in Table 1. The concentrations of all blood metabolites changed over time (P < 0.001), irrespective of reproductive tract inflammatory disease status. Compared with healthy cows, CE cows had lower glucose (2.99 ± 0.12 vs 2.34 ± 0.22 mmol/L; P = 0.03) but higher BHB (0.86 ± 0.10 vs 1.53 ± 0.18 mmol/L; P = 0.04) at 9 d postpartum. Interestingly, SCE cows had higher NEFA (0.98 ± 0.09 mmol/L; P = 0.02) at 9 d postpartum and higher BHB (1.29 ± 0.14 mmol/L; P = 0.05) at 21 d postpartum compared with healthy cows (0.56 ± 0.06 mmol/L for NEFA and 0.86 ± 0.09 mmol/ for BHB, respectively). Moreover, at 9 d postpartum, BHB was higher in CE cows (1.53 ± 0.18 mmol/L) than in SCE cows (0.86 ± 0.14; P = 0.02).
Table 1Least squares means (± SE) of blood glucose (mmol/L), nonesterified fatty acids (NEFA, mmol/L), and BHB (mmol/L) in 38 Holstein cows at −7, 9, and 21 d relative to calving
Groups consisted of cows diagnosed healthy (n = 22), with subclinical endometritis (SCE; n = 8) or with clinical endometritis (CE; n = 8) at 36 d postpartum
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
a,b Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
1 Groups consisted of cows diagnosed healthy (n = 22), with subclinical endometritis (SCE; n = 8) or with clinical endometritis (CE; n = 8) at 36 d postpartum
Effect of Uterine Disease on cPMN Counts and Function
The cPMN blood count changed from −7 to 36 d relative to calving (Figure 1A, P < 0.001) but only tended to differ (P = 0.10) between healthy, SCE, and CE cows. Similarly, the proportion of segmented cPMN changed over time (P = 0.0002) but no differences (P = 0.27) were found irrespective of the uterine health status (Supplemental Figure S3; https://doi.org/10.17605/OSF.IO/4KYXZ; Bogado Pascotini, 2023). Cows with CE had lower (P < 0.05) cPMN viability (84.5 ± 2.1%) with a higher (P < 0.04) proportion of apoptosis (14.4 ± 2.0%) than healthy (92.4 ± 1.3 and 6.7 ± 1.3%, respectively) or SCE cows (95.3 ± 2.4 and 3.8 ± 2.3%, respectively) at 9 d postpartum (Figure 1B and C). The proportion of necrotic cPMN (Figure 1D) was low (<1%) and did not differ between groups (P = 0.92). None of the cPMN functions (Table 2) or function capacity parameters (Supplemental Table S2; https://doi.org/10.17605/OSF.IO/4KYXZ; Bogado Pascotini, 2023) changed from −7 to +36 d relative to calving (P > 0.18). Interestingly, cPMN DQ-ova MFI was lower (6.21 ± 0.15 square root MFI; P < 0.05) in SCE than in healthy (6.71 ± 0.08 square root MFI) or CE cows (6.86 ± 0.15 square root MFI) at 9 d postpartum. No other differences in cPMN function or function capacity parameters were found among experimental groups (P > 0.22).
Figure 1Least squares means (± SE) of (A) circulating PMN (cPMN) counts and proportion of (B) viable, (C) apoptotic, and (D) necrotic cPMN in 38 Holstein cows. Groups consisted of cows diagnosed healthy (n = 22), with subclinical endometritis (SCE; n = 8), or with clinical endometritis (CE; n = 8) at 36 d postpartum. Different letters (a, b) indicate P < 0.05.
Groups consisted of cows diagnosed healthy (n = 22), with subclinical endometritis (SCE; n = 8), or with clinical endometritis (CE; n = 8) at 36 d postpartum.
The cPMN function assays were evaluated as the percentage of activated cPMN performing phagocytosis (PC) and the median fluorescence intensity (MFI) for PC, oxidative burst (OB), and endocytic and proteolytic degradation measured via DQ-ovalbumin (DQ-ova).
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
Values for these cPMN function assays were square root transformed.
6.56 ± 0.08
6.64 ± 0.13
6.85 ± 0.13
6.50 ± 0.08
6.33 ± 0.14
6.78 ± 0.13
a,b Means with different superscripts represent significant differences (P < 0.05) among groups (healthy vs. SCE vs. CE) within a row and day relative to calving.
1 Groups consisted of cows diagnosed healthy (n = 22), with subclinical endometritis (SCE; n = 8), or with clinical endometritis (CE; n = 8) at 36 d postpartum.
2 The cPMN function assays were evaluated as the percentage of activated cPMN performing phagocytosis (PC) and the median fluorescence intensity (MFI) for PC, oxidative burst (OB), and endocytic and proteolytic degradation measured via DQ-ovalbumin (DQ-ova).
3 Values for these cPMN function assays were square root transformed.
Effect of Uterine Disease on ePMN Counts and Function
The vaginal discharge score at 36 d postpartum for healthy, SCE, and CE cows was 0.2 ± 0.1 (n = 12 for score 0 and n = 4 for score 1), 0.5 ± 0.2 (n = 4 for score 0 and n = 4 for score 1), and 2.3 ± 0.2 (n = 5 for score 2 and n = 3 for score 3), respectively. The ePMN% in the endometrial cytology samples changed from 9 to 36 d postpartum (P = 0.05), with greater proportions (P < 0.04) of ePMN in CE (53.5 ± 8.0%) than in healthy (13.9 ± 6.1%) and SCE cows (19.3 ± 9.3%) at 21 d postpartum. Lower ePMN% (P < 0.05) were found in healthy (1.7 ± 3.6) than in SCE (20.3 ± 8.0%) or CE cows (26.6 ± 8.0%) at 36 d postpartum (Figure 2A). The proportion of viable, apoptotic, and necrotic ePMN are depicted in Figures 2B to D. The proportion of apoptotic ePMN was the only viability parameter that changed over time (P = 0.03). The proportion of necrotic ePMN was higher (P < 0.03) in healthy (49.6 ± 5.1%) than in SCE (27.4 ± 7.3%) and CE (27.7 ± 7.3%) cows at 36 d postpartum. The proportions of ePMN for PC and the PC function capacity (Figures 3A and B, respectively) were higher at 36 d postpartum (P < 0.03) in CE (47.0 ± 8.6% and 2.4 ± 3.8 log10%, respectively) compared with healthy (18.4 ± 7.6% and 1.0 ± 0.3 log10%, respectively) cows, but did not differ (P > 0.72) in SCE cows (25.9 ± 8.7% and 1.9 ± 0.3 log10%, respectively). The ePMN MFI for PC and the PC function capacity did not differ (P > 0.42) between experimental groups (Supplemental Figure S4A and B; https://doi.org/10.17605/OSF.IO/4KYXZ; Bogado Pascotini, 2023).
Figure 2Least squares means (± SE) of (A) endometrial polymorphonuclear leukocyte-to-all nucleated cell proportion (ePMN%) and the proportion of (B) viable, (C) apoptotic, and (D) necrotic ePMN in 32 Holstein cows. Groups consisted of cows diagnosed healthy (n = 16), with subclinical endometritis (SCE; n = 8), or with clinical endometritis (CE; n = 8) at 36 d postpartum. Different letters (a, b) indicate P < 0.05.
Figure 3Least squares means (± SE) of (A) the proportion of endometrial polymorphonuclear leukocyte (ePMN) phagocytosis (PC) and (B) phagocytosis capacity (ePMN PC × ePMN% × proportion of viable ePMN) in 32 Holstein cows. Groups consisted of cows diagnosed healthy (n = 16), with subclinical endometritis (SCE; n = 8), or with clinical endometritis (CE; n = 8) at 36 d postpartum. Different letters (a, b) indicate P < 0.05.
Correlations Between Blood Metabolite Concentrations and cPMN or ePMN Viability and Function, and Between cPMN and ePMN Viability and Function
None of the correlations between blood metabolite concentrations and cPMN viability and function parameters were significant (P > 0.08), ranging from r = −0.12 to 0.19. Similarly, none of the correlations between blood metabolite concentrations and ePMN viability or PC parameters were significant (P > 0.14), ranging from r = −0.12 to 0.27. The correlations for cPMN and ePMN viability, apoptosis, and necrosis were r = −0.08, 0.13, and −0.10 (P > 0.21), respectively. The correlation between cPMN and ePMN for PPC was r = 0.07 (P = 0.08) and for MFIPC the correlation was r = 0.20 (P = 0.53).
DISCUSSION
The main hypothesis of our study was that cPMN and ePMN viability and function differ, either preceding or at the moment of the diagnosis of SCE or CE at d 36 postpartum, compared with healthy cows. Our results suggest that cPMN viability and function at 9 d postpartum is associated with the pathogenesis of uterine disease. Furthermore, ePMN at 36 d postpartum were mostly necrotic in healthy cows, but they were viable and functional in cows with CE, probably due to active uterine inflammation. Remarkably, ePMN in cows with SCE at 36 d postpartum seemed to display an intermediate phenotype, counting numerically more viable PMN than the healthy cows, but less functional compared with CE cows, as evidenced by the lower proportion of their PC. None of the correlations between blood metabolites and PMN viability and function or between cPMN and ePMN viability and function were significant.
The life cycle of cPMN in transition cows is brief but highly dynamic. After leaving the hematopoietic (i.e., typically the red bone marrow or splenic) compartment, PMN reach the peripheral circulation and remain in the bloodstream for only about 9 h (
). Circulating PMN are continuously replenished; aged PMN (programmed to become apoptotic) are captured in the capillaries of the liver, spleen, or bone marrow, whereas freshly matured PMN are constantly released from their storage pools (
). Inflammation associated with parturition and placental expulsion triggers cPMN to migrate toward the uterus and, in consequence, the cPMN counts transiently decline and the granularity of the nucleus decreases (
). An increase in the proportion of PMN with a nonsegmented nucleus (band cells), the so-called left shift, was also observed between 7 d before and 36 d after parturition in the current data set, independent of the cow's health status (Supplemental Figure S4). Although we did not find differences in cPMN counts in cows diagnosed as healthy, CE, or SCE, the proportion of viable and apoptotic cPMN differed in cows with CE compared with both other experimental groups as soon as 9 d postpartum. It is rare to find high numbers of circulating apoptotic PMN. A previous study found a higher apoptotic potential in cPMN harvested from early lactating than from mid-lactating cows (
). Those authors suggested that apoptosis is induced by various inflammatory mediators (e.g., LPS and tumor necrosis factor-α) that can modulate PMN function (
). A higher percentage of apoptosis might explain the impaired cPMN functions around calving, because apoptotic cPMN have decreased functions (e.g., chemotaxis, PC, and production of myeloperoxidase). This key finding is likely linked with the pathogenesis of CE. However, to date, the proportion of apoptotic cPMN has rarely been evaluated in transition cows.
Robust migration of cPMN to the uterine cavity after calving has been linked with better uterine health (
). However, most high-yielding dairy cows experience increased serum concentrations of NEFA and BHB and decreased glucose within 2 to 3 wk after calving (
). Therefore, cPMN exposed to this altered blood metabolite environment before reaching the uterine lumen may become less functional. As for the present study, at 9 d postpartum we found lower blood glucose and higher BHB in CE compared with healthy cows, and higher NEFA in SCE than in healthy cows. This may explain the differences in cPMN viability and function parameters between cows with uterine disease versus those diagnosed as healthy at the same early time point. Interestingly, in cows with SCE, serum BHB concentrations were also higher but only at 21 d postpartum compared with healthy cows. A previous study showed an altered energy metabolism (i.e., a shift from glucose to high NEFA and BHB) within 3 wk after calving in cows that later were diagnosed with SCE compared with healthy cows (
). Furthermore, large-scale epidemiological studies identified high blood NEFA and BHB in the transition period as risk factors for uterine disease, more specifically for cows suffering from SCE (
). Thus, this changed blood metabolite environment might play a crucial role in cPMN viability and function impairment. Moreover, this may result in an altered innate immune function, triggering the development of uterine disease.
found a significant, negative (r = −0.44) correlation between plasma NEFA and the myeloperoxidase activity of cPMN. In the present study, we attempted to correlate blood glucose, NEFA, and BHB with multiple cPMN and ePMN viability and function parameters. However, we found only weak, nonsignificant correlations. The lack of correlation between energetic metabolites and cPMN and ePMN function parameters in our study is challenging to explain. Possibly, the volatile presence of PMN in the bloodstream conjunctively with their “commuter mode” condition might make them more resilient to biochemical triggers, as do their PMN counterparts in tissues. Clues, risk factors, and modes of action can be inferred from the recent review by
, in which relationships between metabolism and neutrophil function are exhaustively discussed. However, as for the present study, it is important to note that the relatively small sample size or other underlying factors causing uterine disease such as bacterial infection may have played a role.
Polymorphonuclear leukocyte viability and function are traditionally measured in peripheral blood PMN via flow cytometry (
Effect of recombinant bovine granulocyte colony-stimulating factor covalently bound to polyethylene glycol injection on neutrophil number and function in periparturient dairy cows.
). This analytical approach provides valuable insights into innate immune function in transition dairy cows but its usefulness is limited to studying PMN viability and functional capacity in blood. For ePMN, the classic dogma assumes that PMN present in the uterine lumen are already activated (
). In a previous study, we corroborated this by showing that stimulation (with phorbol 12-myristate 13-acetate) of ePMN isolated from the uterus of postpartum cows did not induce a major release of reactive oxygen species, probably because they had already been activated (
). In addition, we previously demonstrated that ePMN in healthy postpartum dairy cows at 9 to 36 d postpartum are highly dynamic, as they can appear in different numbers and proportions of viable, apoptotic, or necrotic PMN, and their phagocytic capacity declines over time (
). The present study now extends these findings by showing that in healthy cows at 36 d postpartum, only one-third of ePMN were viable. Conversely, in cows diagnosed with SCE or CE at the same time point, half of ePMN were viable. This finding is logical for CE because ePMN are supposed to actively fight against uterine pathogens such as T. pyogenes (
). Indeed, ePMN from cows diagnosed with CE had a higher PC than those diagnosed as healthy at 36 d postpartum. Interestingly, the viability of ePMN was as high in cows with SCE as in CE cows at 36 d postpartum, but their PC was numerically lower than in CE cows, similar to the PC of healthy cows. These important results suggest that ePMN in SCE cows are still viable but dysfunctional. Thus, 2 pathways in SCE pathogenesis may be hypothesized; first, SCE may be a product of metabolic and inflammatory dysfunction that impairs the innate immune function and the ability of ePMN to undergo programmed cell death. As a result, ePMN in cows with SCE remain present in the endometrium, impairing reproductive success in the long term, as demonstrated in endometrial samples collected at AI (
). Second, SCE may be an intermediate or transient state between CE and healthy. Future research, including extra sampling time points after 36 d postpartum, might provide a conclusive answer for these prerogatives.
Poor correlations were observed for viability and function parameters between cPMN and ePMN in the current experiment. This may be caused by the diapedesis process of PMN toward the endometrium and the uterine environment, including cytokines and inflammatory mediators, altering the viability or functionality of the PMN. Diapedesis induces apoptosis and decreased PC and OB capacity (
). It is also plausible that distinct PMN phenotypes exist, originating from the bone marrow or PMN changing phenotype instructed in the local tissues (
). This indicates the need to assess both cPMN and ePMN simultaneously in studies involving uterine innate immune function. The evaluation of cPMN and ePMN function capacity (PMN function × PMN counts) did not provide additive evidence in the present study for whether the key driver for uterine health was the PMN quantity, the (individual) PMN function, or a combination of both.
CONCLUSIONS
We showed that PMN differ in specific viability and function parameters in multiparous early postpartum cows with either SCE or CE compared with their healthy counterparts and that these differences are time- and spatial-dependent. Impaired cPMN viability and function were found in SCE and CE at 9 d postpartum. In the uterus, ePMN at 36 d postpartum were predominantly necrotic and hence nonfunctional in healthy cows, they were viable and functional in CE cows, and an intermediate ePMN stage was found in SCE cows (they were mostly viable but less functional with already a reduced PC). No correlations could be found in viability and function parameters between cPMN and ePMN. Future studies should focus on the kinetics of the intriguing immunomodulation and further evaluate both systemic and local PMN phenotypes preceding the diagnosis of reproductive tract inflammatory disease.
ACKNOWLEDGMENTS
This research was funded by the Special Research Fund (BOF) of Ghent University (Merelbeke, East Flanders, Belgium) to Leen Lietaer (BOF17/DOC/269) and BOF of the University of Antwerp (Wilrijk, Belgium; KP 2020; FFB200048) granted to Osvaldo Bogado Pascottini. Bogado Pascottini also received a grant from Fonds voor Wetenschappelijk Onderzoek–Vlaanderen (FWO, Research Foundation, Flanders; Brussels, Belgium; project number 12Y5220N). Stijn Heirbaut received a PhD grant from the Special Research Fund (BOF19/DOC/131) of Ghent University (Merelbeke, Belgium). This research was also partially funded by the government agency Flanders Innovation and Entrepreneurship (VLAIO-Belgium; Brussels, Belgium; LA170830). The authors thank Ann Van Soom (Ghent University, Merelbeke, Belgium) for the use of her laboratory equipment. We also thank Celien Kemel and Marjolein Brack (Ghent University, Merelbeke, Belgium), Ilke Van Hese (Ghent University, Merelbeke, Belgium; ILVO, Melle, Belgium), and the staff of the experimental dairy farm of ILVO (Melle, Belgium) for their help with handling animals and sampling. Special thanks go to Geert and Leen DeWinter (Nevelle, Belgium). The authors have not stated any conflicts of interest.
Viability and function dynamics of circulating versus endometrial polymorphonuclear leukocytes in postpartum dairy cows with subclinical or clinical endometritis.
The development and analysis of species specific and cross reactive monoclonal antibodies to leukocyte differentiation antigens and antigens of the major histocompatibility complex for use in the study of the immune system in cattle and other species.
Effect of recombinant bovine granulocyte colony-stimulating factor covalently bound to polyethylene glycol injection on neutrophil number and function in periparturient dairy cows.
Assessment of the main pathogens associated with clinical and subclinical endometritis in cows by culture and MALDI-TOF mass spectrometry identification.
00127-3&id=gr1.jpg){kind=link}
00127-3&id=gr2.jpg){kind=link}
00127-3&id=gr3.jpg){kind=link}