Novel sampling procedure to characterize bovine subclinical endometritis by uterine secretions and tissue

Theriogenology 141 (2020) 186e196

Contents lists available at ScienceDirect

Theriogenology journal homepage: www.theriojournal.com

Novel sampling procedure to characterize bovine subclinical endometritis by uterine secretions and tissue Anika L. Helfrich a, *, Horst-Dieter Reichenbach b, Marie M. Meyerholz a, € hlich d, Frank Weber a, Holm Zerbe a Heinz-Adolf Schoon c, Georg J. Arnold d, Thomas Fro a

Clinic for Ruminants with Ambulatory and Herd Health Services, Centre for Clinical Veterinary Medicine, LMU Munich, Sonnenstraße 16, D-85764, Oberschleißheim, Germany Institute for Animal Breeding, Bavarian State Research Centre for Agriculture, Prof.-Dürrwaechter-Platz 1, D-85586, Poing, Germany c Institute for Veterinary Pathology, University Leipzig, An Den Tierkliniken 33, D-04103, Leipzig, Germany d Laboratory for Functional Genome Analysis (LAFUGA), Gene Centre - LMU Munich, Feodor-Lynen-Straße, 25, D-81377, Munich, Germany b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 November 2018 Received in revised form 16 August 2019 Accepted 9 September 2019 Available online 16 September 2019

Subclinical endometritis (SE) in cattle is defined as clinically unapparent inflammation of the endometrium. It is reported to impair fertility in affected cows and causes economic loss within the dairy industry. A gold standard for diagnosis of SE has not been set. Uterine cytology and histopathology are both applied, but low agreement between these methods has been described. The objective of the present study was to assess the capability of uterine secretions (US) as a new medium for diagnosis of SE. A novel sampling tool was applied to retrieve US as well as cytological, histological and bacteriological samples of the endometrium after a singular passage through the cervix in 108 dairy cows (43e62 days post-partum [dpp]). To assess the quality of the US samples, a proteome analysis of samples from five healthy donors was performed, demonstrating that in vivo sampling of US was feasible and generated samples suitable for diagnostic purposes. Diagnosis of SE was realized by the combination of clinical, cytological, and histopathological findings. Quantitative analysis of pro- and anti-inflammatory cytokines (interleukin (IL) 1B, IL6, IL8, IL17A, IL10) in US was conducted using AlphaLISA-technology. RNAlater-fixed endometrial biopsies were used for gene expression analysis of the cytokines IL1B, IL6, IL8, IL10 and tumor necrosis factor alpha (TNFa) as well as the prostaglandin-endoperoxide synthase 2 (PTGS2) and the antimicrobial peptide S100A9 by reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR). Cows were assigned to groups according to their uterine health status. A large group of animals (n ¼ 83) displayed no signs of endometritis (E.NEG). Cytological and histopathological examination revealed low agreement; hence, animals with SE were differentiated into SE(cyto) and SE(histo) groups (n ¼ 7 and n ¼ 13, respectively). One animal in group SE(cyto þ histo) as well as four animals with signs of clinical endometritis (CE) were excluded from further analysis. SE(cyto) showed significantly higher median concentrations of IL1B, IL8 and IL17A in US as well as a significantly higher median expression of IL1B, IL8 and IL10 in endometrial biopsies compared to E.NEG. No significant differences were found for IL6 and IL10 in US and IL6, TNFa, PTGS2 and S100A9 in endometrial tissue between these groups. SE(histo) presented no differences concerning the analyzed parameters compared to E.NEG. In conclusion, a method to sample US was successfully established in dairy cows. The cytokines IL1B, IL8 and IL17A are promising candidates in diagnosing cytological endometritis by US. Further assessment of US might contribute to a better understanding of the pathological mechanisms leading to chronic endometrial inflammation and to impaired fertility in affected cows. © 2019 Elsevier Inc. All rights reserved.

Keywords: Uterine secretion Endometrial tissue Inflammatory cytokine PTGS2 S100A9

1. Introduction

* Corresponding author. E-mail address: [email protected] (A.L. Helfrich). https://doi.org/10.1016/j.theriogenology.2019.09.016 0093-691X/© 2019 Elsevier Inc. All rights reserved.

A global decline of the reproductive efficiency in dairy cows has been described within the last five decades [1]. Although more effort concerning efficiency of AI in dairy cattle has been made,

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

conception rates of only 40% are reported [2e4]. Subfertility is the main reason for culling of dairy cows (20%) and is thereby exceeding mastitis and lameness as causes for culling [5]. The resulting economic loss is, amongst others, due to longer calving intervals, higher insemination costs and expenses for restocking of the herds [6]. Subfertility in dairy cows has a multifactorial genesis. The complex interaction between genetic aspects, environmental issues and management strategies has not been fully elucidated so far [1,7,8]. Numerous factors, such as energy homeostasis and immunological status of the animals [6,9], affect ovarian and uterine functions. This might lead to reduced fertility by increasing the number of fertilisation failure and embryonic loss [1,10]. In the puerperal phase, a temporary and controlled inflammation of the endometrium is to some extent physiological and essential for detachment of the placenta, bacterial clearance and involution of the uterus and recovery of the epithelial integrity. These processes are necessary for regeneration of uterine functions [11,12] and the expression ‘postpartum endometrial inflammatory response’ has been proposed by Chapwanya et al. [13]. The regulation of this particular inflammatory response has not been clarified in detail until now [13]. However, the adequate regulation of the uterine microenvironment is crucial for the establishment and maintenance of gestation. Inflammatory changes, such as metritis or endometritis, interfere with the endometrial capacity of conception, implantation and further development of the conceptus [1,14]. For approximately 15 years, bovine subclinical endometritis (SE) has been discussed to play an important role in this context [15]. SE in cows is defined as clinically unapparent inflammation of the endometrium occurring later than 21 days postpartum [16]. According to several authors, impairment of fertility due to SE is detrimental. First service conception rates of affected cows have been reported to be reduced by 18e70% [15,17,18]. Prevalence of SE varies with definition and diagnostic method applied, time post-partum of cows being examined as well as herd specific-factors [19]. Until now, no gold standard for diagnosis of SE has been set [16,20]. Cytological and histopathological examination of the endometrium are both applied as diagnostic devices for SE, but low agreement between these methods has been described [21,22]. Cytology is mainly used because it is comparably easy to conduct, cost-effective and noninvasive [20]. Cytological samples are gained by low-volume uterine lavage (LVL) or cytobrush technique (CB) and the proportion of neutrophils (PMN) in uterine cytology is determined [15,17]. Variable PMN percentage thresholds are applied for cytological diagnosis [23]. The term “cytological endometritis” has been suggested by several authors [24e26]. Endometrial biopsy is considered a more reliable diagnostic method, as it enables the assessment of acute and chronic endometrial alterations including besides the endometrial epithelium also deeper layers like the stratum compactum [23,27e30]. So far, the clinical relevance of these chronic endometrial changes for bovine fertility has not been studied into detail [22,31]. Though the technique is underrepresented in field and research as it is time-consuming and requires a lot of experience by the examiner [31]. Moreover, histological sample quality is often insufficient [21,32] and a detrimental effect of uterine biopsy on subsequent fertility has been reported by some authors [33,34], whereas other authors regard it as a safe technique when adequately applied [35]. Several attempts have been made to develop new diagnostic tools. For example, a ‘cytotape’ has been developed, which allows for cytological sampling of the endometrium at the time of artificial insemination (AI) [36]. However, a diagnostic method for the efficient identification of animals with SE is still lacking. Only the reliable detection of affected animals permits the evaluation of potential therapeutic strategies. The examination of uterine secretions (US) from abattoir uteri generated

187

promising results concerning diagnosis of SE in cattle ex vivo [37]. The objective of the present study was to assess the diagnostic potential of US for diagnosis of SE in dairy cows in vivo. 2. Material and methods 2.1. Experimental animals The study was conducted at a farm of the Bavarian State Research Centre for Agriculture between April 2016 and April 2017. The herd consisted of 188 dairy cows (including dry cows) of different breeds (Simmental, Brown Swiss, Red Holstein, crossbreds) with an average milk yield of 10.404 kg in 2016. Relevant fertility parameters of the herd were 63% first-insemination pregnancy rate, 108 days calving-conception interval and 2.1 AIs per pregnancy. The animals were housed in a free stall barn, fed a total mixed ration (corn and grass silage with individual amounts of concentrate) and had access to fresh water ad libitum. All experimental procedures on animals were carried out in accordance with the European Community Directive 2010/63/EU and were approved by the national authority according to x8 of the German Animal Welfare Act (ROB 55.2-1-54-2532-6-2016). 2.2. Novel sampling tool A novel sampling device was constructed and evaluated on abattoir uteri in a previous project [37]. In the present study, the tool was applied to animals in vivo for the first time. After a singular passage through the cervix, it allows for consecutive collection of US as well as cytological, histological and bacteriological samples of bovine uteri (Fig. 1). A stainless-steel tube (outer diameter 4 mm, wall thickness 0.25 mm extended to 0.75 mm at the fore end, Sawade, Gottmadingen, Germany) with a blunt obturator (MK Medical, EmmingenLiptingen, Germany) served as the working channel. For sampling of US, Merocel sponges (material specification ‘CF 120’, Medtronic, Mystic, USA) were used, which are well characterized in terms of acquiring and releasing different immune markers [38]. After sterilization by gamma radiation (52.6 kGy), they were adhered to a plastic cannula (Simprop, Harsewinkel, Germany) by hot glue. Constant dimensions (2 mm  4 mm x 102.5 mm) and hence weights (0.053 g) of the Merocel sponges were used to enable the ascertainment of individual dilution factors for the US samples by weighing the sponges after sampling. To facilitate the introduction of the flexible Merocel sponge into the rear end of the working channel, a stainless-steel adapter (Sawade, Gottmadingen, Germany) was used. It was imposed on the Merocel sponge in the laboratory and mounted onto the rear end of the working channel at the time of sampling. The modified cytobrush consisted of a steam sterilized commercial interdental brush (DontoDent, Karlsruhe, Germany), which was adhered to a plastic cannula by hot glue. A second plastic cannula (both: Simprop, Harsewinkel, Germany) was used as a protection catheter for the cytobrush. A minimized biopsy forceps (Bema Medical, Stuttgart, Germany) served for collection of biopsy samples of approximately 6 mm  3 mm x 2 mm size for histopathological and microbiological assessment. Prior to each herd visit, the necessary amounts of disposable materials (cytobrushes, Merocel sponges, plastic cannulas) were prepared under a safety cabinet using sterile gloves. The working channel with obturator and the biopsy forceps were reused after steam sterilization. 2.3. Examination and sampling of animals During scheduled visits every two weeks, cows were examined

188

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

Fig. 1. The novel sampling tool consisting of a) working channel with blunt obturator, b) Merocel sponge, c) cytobrush and d) biopsy forceps.

gynaecologically and subsequently sampled after completion of the puerperium. Clinical signs of endometritis were evaluated by both external evaluation and use of a vaginoscope. The oestrus cycle stage was specified by transrectal palpation and ultrasonography €ln, Germany). using a 10 MHz linear array transducer (Esaote, Ko The luteal phase was defined as the presence of a CL > 2 cm, while in follicular phase a CL was absent or in regression, and follicles <2 cm were optional. After thorough cleaning of the vulva with paper towels and water, the working channel with obturator was introduced into the uterus under manual transrectal control. A plastic sleeve (IMV, L'Aigle, France) covering the working channel during insertion into the vagina was perforated when reaching the cervix. The working channel was placed in the uterine horn ipsilateral to the dominant ovarian structure (if present CL, otherwise follicle). After retraction of the obturator, the Merocel sponge was introduced into the uterus and remained in the uterine lumen for 2 min while it was carefully rotated. After retraction, the Merocel sponge was cut off the plastic cannula at a set mark and transferred into a tube with 300 ml extraction buffer (PBS with 100 mg/ml aprotinin). Next, the cytobrush was inserted into the uterus. When reaching the uterine lumen, it was passed through the plastic cannula and rotated by 360 while in contact with the endometrium. It was then retracted into the cannula tip before withdrawal from the working channel. Two cell spreads were prepared from each cytobrush on sterile microscope slides. Subsequently, the cytobrush was rolled over half of a blood agar (Oxoid, Wesel, Germany) for bacteriological examination (BEcyto). Finally, two endometrial biopsies were collected by introducing the biopsy forceps through the working channel into the uterus. One biopsy was placed in 4% formalin for histopathological assessment, and the other one was placed in RNAlater for gene expression analysis and served as a back-up for histological examination in case the formalin-fixed biopsy was not readable. By slight movement of the working channel between the different sampling steps, the sampling of intact endometrial areas was assured. A blood sample was taken from each animal from the tail vein. It served for analysis of plasma concentrations of the cytokines, which were also determined in US. This enabled the arithmetical correction of blood contamination of the US samples.

2.4. Classification of animals In total, 108 animals (53% Simmental, 35% Brown Swiss, sporadic Red Holstein and crossbreds) with an average of three lactations (min: 1, max: 7) were examined and sampled once between 43 and 62 dpp. Their mean current milk yield was 40.8 kg/day (min: 20.6 kg/day, max: 56.3 kg/day). Cows were assigned to groups according to their uterine health status considering results of clinical, cytological and histological examination. Cytological samples were stained by Diff-Quick (Labor þ Technik Eberhard Lehmann GmbH, Berlin, Germany), and the proportion of PMN among 300 nucleated cells was rated under a microscope (Leitz, Stuttgart, Germany) [39]. A threshold of 5% PMN was applied for diagnosis of SE [30]. Processing and histopathological assessment of formalin-fixed endometrial biopsies were performed according to presence, character and degree of inflammatory alterations at the Institute for Veterinary Pathology of the University of Leipzig. The classification scheme according to Heppelmann et al. [40] was applied: Subject to the number and type of inflammatory cells observed, histological endometritis was classified as chronic non-purulent, acute purulent, or chronic purulent and graded as mild, moderate, or severe [40]. In total, 83 animals displayed no signs of endometritis (E.NEG). A group of 21 animals were diagnosed with SE, but cytological and histological results revealed low agreement. This was accounted for by differentiating animals with SE into animals with cytologically and histologically detectable SE (SE(cyto þ histo), n ¼ 1), animals with only cytologically detectable SE (SE(cyto), n ¼ 7) and animals with only histologically detectable SE (SE(histo), n ¼ 13). Four animals showed signs of clinical endometritis (CE, purulent or mucopurulent vaginal discharge and signs of inflammation in cytological and/or histological samples) and together with the only animal with cytologically and histologically detectable SE were excluded from further analysis. According to this group assignment, parameters of US and endometrial expression of inflammatory genes were analyzed comparatively between 103 animals of groups E.NEG (n ¼ 83), SE(cyto) (n ¼ 7) and SE(histo) (n ¼ 13). Their distribution over the luteal and follicular phases was approximately equal (data not shown).

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

2.5. Extraction of uterine secretions Extraction of US from the Merocel sponges was realized by centrifugation referring to the extraction protocol of CrowleyNowick et al. [41], optimized by Rohan et al. [42]. After transferring the Merocel sponges into tubes with 300 ml of extraction buffer (PBS with 100 mg/ml of aprotinin) at the time of sampling, they were cooled on ice and within a maximum of 3 h after sampling were put in the filter of Spin X Centrifuge Tube Filters with a cellulose-acetate-membrane with a 0.45 mm pore size (Costar, Corning, Amsterdam, Netherlands). They were then centrifuged at 14,000g and 20  C for 15 min. The Merocel sponges were transferred into a second filter tube, where another 300 ml of extraction buffer was added, and the centrifugation was repeated under consistent conditions. The two eluates were combined and after determination of their blood contamination were kept at 80  C. The filter of the first filter tube was sampled by a sterile swab, which was rolled on the second half of a blood agar for bacteriological examination (BEfilter). From each animal, individual amounts of US were sampled and extracted by adding 2  300 ml of extraction buffer and centrifugation. An individual dilution factor D was assessed in accordance with Rohan et al. [42]: D ¼ (m(US)þ0.6 g)/ m(US). Hereby, m(US) represents the mass of the sampled US, which is assessed by weighing the filter tube before (m0) and after (m1) adding the wet Merocel sponge with 300 ml of extraction buffer: m(US) ¼ m1-m0-0.3 ge0.053 g. The mass of the Merocel sponges used for US-sampling was 0.053 g. Slight haemorrhages of the mucosa led to contamination of some US samples with blood, which potentially influences the concentrations of the detected cytokines. To correct this effect arithmetically, the blood contamination of the US samples was quantified photometrically. A standard curve of blood contamination was established by adhering 100 ml of whole blood of ten cows to Merocel sponges. This volume was chosen, as it revealed the average amount of US sampled in the precursor project [37]. The Merocel sponges were processed as described above (2  300 ml extraction buffer, centrifugation). The eluates were diluted in 12 steps by adding extraction buffer. By photometrical measurement at 570 nm, oxygenated and deoxygenated haemoglobin were measured likewise [43,44]. The average optical density (OD) of the 10 samples at each dilution stage was used for calculation of a standard curve. The blood contamination of US samples (blood%) was further calculated by its OD. Due to low amounts of obtained US (median 91 mg per animal), laboratory assays providing reliable results despite small sample volumes were required. They are described in the following paragraphs. 2.6. Characterization of US samples by nano-LC-MS/MS based proteome analysis Fifteen microliters of US from five healthy donors was loaded on a sodium dodecyl sulphate (SDS) gel (Novex Tris-Glycine gel 10e20%, Thermo). Proteins were briefly separated (for 10 min at 125 V) and Coomassie stained (Instant Blue, Expedeon). The entire area containing proteins was excised, destained (50% actonitril, 50 mM NH4HCO3) and subjected to in-gel digestion using the following protocol: For protein reduction, the gel slice was incubated in 45 mM DTE/50 mM NH4HCO3 for 30 min at 55  C and subsequent alkylation was performed with 100 mM iodoacaetamide/50 mM NH4HCO3 for 30 min in the dark. In-gel digestion was done using 70 ng Lys-C for 4 h at 37  C and 70 ng Trypsin at 37  C overnight. LC-MS/MS analysis was performed on an Ultimate 3000 (Thermo Scientific, Waltham, MA, USA) nano-chromatography system online coupled to a Q Exactive HFX (Thermo Scientific, Waltham, MA, USA) mass spectrometer. Tryptic peptides diluted in

189

solvent A (0.1% formic acid (FA)) were injected and transferred to a trap column (PepMap 100C18, 100 mm  2 cm, 5 mm particles, Thermo Scientific). Separation was performed at 250 nl/min (Column: PepMap RSLC C18, 75 mm  50 cm, 2 mm particles, Thermo Scientific) with a 160 min gradient from 5% to 25% solvent B (0.1% FA in acetonitrile) followed by a 10 min gradient from 25% to 40% solvent B. MS measurements were performed with a top 15 datadependent method at an ion spray voltage of 2.2 kV. Precursor ions were acquired at a resolution of 60,000 (mass-range: 350e1600), while MS/MS were collected at a resolution of 15,000. For protein identification, MS spectra were searched using the Proteome Discoverer 2.2 in combination with the Bos taurus subset of the SwissProt/TrEMBL database. Protein hits were filtered for high confidence hits (FDR < 1%) and for identifications with at least two individual peptides. Venn diagrams were generated with the help of the InteractiVenn tool [45] and gene ontology (GO) analysis was performed using the PANTHER Classification System [46]. 2.7. Quantification of pro- and anti-inflammatory cytokines Quantitative analysis of pro- and anti-inflammatory bovine interleukins (IL) 1B, IL6, IL8, IL17A and IL10 was conducted using AlphaLISA (amplified luminescent proximity homogeneous assaylinked immunosorbent assay) technology (PerkinElmer, Rodgau, Germany). This method is based on fluorometrically detectable energy transmission between donor and acceptor beads after antibody-mediated binding of the target substance. The test was conducted following the manufacturer's instructions. Donor beads are light-sensitive and were only used in the dark (<100 lux). The protocol for 2 ml sample volume was applied for IL1B, IL6, IL8 and IL17A. For IL10, the 5 ml sample volume protocol was used. The provided standards were pipetted three-fold, the US samples twofold. The measurement was repeated if CV exceeded 10%. The following settings were used at the microplate-reader: Excitation 680 nm, emission 615 nm, excitation time 0.30 s, integration start 0.36 s, integration time 0.40 s. To correct the blood contamination of the US samples arithmetically, the analytes were also quantified in plasma samples. The dilution factor D and the blood contamination blood% of the US samples were accounted for as follows: c(UScalc) ¼ [D$c(US)blood%/100$c(plasma)]/(1-blood%/100). Hereby, c(UScalc) is the concentration of the analyte in undiluted, blood-corrected US samples. D is the dilution factor, c(US) is the measurable concentration of the analyte in diluted and blood contaminated US, c(plasma) is the concentration of the analyte in plasma samples and blood% is the photometrically detected percentage of blood contamination of US samples. If c(US) was below the lower detection limit (LDL) of the assay, the lowest dilution stage of the standard curve was used for c(UScalc). 2.8. Bacteriological examination The two bacteriological samples obtained from each uterus (BEcyto and BEfilter) were cultured on sheep blood agar for 48 h at 37  C. Only plates with at least five colony-forming units were regarded as bacteriologically positive, if a maximum of two different colony morphologies occurred identically in both, BEcyto and BEfilter, or if at least five pure cultures were found in BEcyto only. All other samples with bacteriological growth were considered contaminated. Of bacteriologically positive samples, pure cultures were generated and bacterial species were differentiated using matrix-assisted laser desorption (ionization time-of-flight (MALDITOF) mass spectrometry (Microflex LT and MALDI Biotyper 2.0, Bruker, Billerica, USA).

190

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

2.9. Gene expression analysis RNAlater-fixed endometrial biopsies served for gene expression analysis of the cytokines IL1B, IL6, IL8, IL10, tumor necrosis factor alpha (TNFa), the prostaglandin-endoperoxide synthase 2 (PTGS2) and the antimicrobial peptide S100A9 by reverse transcription quantitative real-time polymerase chain reation (RT-qPCR). Performance and description of RT-qPCR were in accordance with the minimum information for publication of quantitative real-time PCR experiments (MIQE) guidelines [47] respecting the manufacturers’ recommendations. In brief, the endometrial biopsies were homogenised in Innu Speed Lysetubes P containing ceramic beads (Analytik Jena, Jena, Germany) after adding 400 ml of lysis buffer. Subsequently, total RNA was purified by applying the spin column system RNA Mini Kit (Bio&Sell, Feucht, Germany). The RNA quality index (RQI) was assessed by automated electrophoresis in the Experion automated electrophoresis system using an Experion RNA Analysis Kit and RNA StdSens Chips (Bio-Rad, Hercules, USA). It ranges from 1 for completely degraded RNA to 10 for completely intact RNA [48]. For cDNA-synthesis oligo-dT-primer and M-MLV reverse transcriptase H () point mutant (both: Promega, Madison, USA) were used. Oligo-dT primer assured that only mRNA was transcribed. Concentration of cDNA was assessed by the nanophotometer Pearl (Implen, Munich, Germany) and cDNA samples were subsequently diluted to 200 ng/ml. Samples without addition of reverse transcriptase running through the transcription protocol (-RT) served as additional negative controls to detect genomic DNA and contaminations. RT-qPCR was conducted by applying a SensiFAST SYBR No-ROX Kit (Bioline, Luckenwalde, Germany), 1 ml of cDNA template (200 ng) and 0.2 mM of each primer (biomers.net, Ulm, Germany and eurofins Genomics, Ebersberg, Germany) in a total reaction mixture volume of 20 ml. Primer sequences used for gene amplification are depicted in Table 1. The master mix was prepared at an isolated laboratory bench. Standards and no-template-controls (NTC) were applied in triplicate and samples in duplicate. The reaction mixtures were established in transparent 96 well-plates (Frame Star, 4titude, Berlin, Germany) and underwent a two-step protocol in a TOptical thermocycler (Analytik Jena, Jena, Germany), which started with 2 min of incubation at 95  C, followed by 40 cycles of 5 s denaturation at 95  C and 15 s annealing and extension at 60  C. For PTGS2, a three-step protocol was applied (40 cycles: 5 s at 95  C, 10 s at 60  C, 15 s at 72  C). Fluorescence detection took place after each annealing and extension step. Cq-values were acquired by the TOptical software. Eventually a melting curve was generated by heating the PCR-products from 60  C to 95  C (DT ¼ 0.5  C) to verify amplification of specific cDNA. Quality

parameters of each qPCR-run included coefficient of determination and efficiency of the standard curve (reference: > 0.985 and 90e110%, respectively), at least one NTC negative, target-specific melting temperature. Absolute quantification of specific mRNA in endometrial samples (copies/ml) was accomplished by adding standard curves as previously described [49]. Briefly, serial dilutions of the specific cDNA subclone with defined concentrations (106, 105, 104, 103, 102 copies/ml) were added to each qPCR-plate. Samples were reevaluated, if SD was >0.7 cycles. RT-qPCR: reverse transcription quantitative real-time PCR; for: forward-primer; rev: reverse-primer; bp: base pairs. 2.10. Statistical analysis Data were managed in Microsoft Excel. Statistical analyses were conducted using the programming language R, version 3.3.2 [53]. The Shapiro-Wilk-test revealed that the results of AlphaLISA and RT-qPCR were not normally distributed. Significant differences between the groups E.NEG, SE(cyto) and SE(histo) were therefore determined using the non-parametric Kruskal-Wallis-test. The level of significance was set at P < 0.05. For pairwise group comparisons the Mann-Whitney-U-test was applied as a post-hoc test with Bonferroni adjustment of level of significance (P < 0.05/ 3 ¼ 0.0167). Dilution- and blood-corrected cytokine concentrations of US and mRNA contents of endometrial biopsies are presented in Tukey-boxplots, where the box includes 50% of the data and the horizontal line in the box presents the median values. Outliers and extreme values were included in all statistical analyses. The relationship between interleukin concentration in US and endometrial gene expression was analyzed by Spearman's correlation analysis for IL1B, IL6, IL8 and IL10. To rule out effects of outliers on statistical results and data interpretation, this analysis was also conducted excluding outliers (values below and above the box and whiskers plot). As no differences in significance of correlations were found, the inclusion of outliers to the data set is considered acceptable. Only significant differences are mentioned in the results section and are described by the fold differences between the median values. Additional results are illustrated in the boxplots and the supplementary materials. 3. Results 3.1. Applicability of sampling method and sample quality The novel endometrial sampling tool was applicable in all presented animals. Trueperella pyogenes was found in two animals with CE. One animal with cytologically detectable SE revealed

Table 1 Primer sequences for RT-qPCR and resulting amplicon length. Gene

Primer sequence (5‘ / 3‘)

Amplicon length

Reference

Accession No.

IL1B

for TTC TCT CCA GCC AAC CTT CAT T rev ATC TGC AGC TGG ATG TTT CCA T for GGA GGA AAA GGA CGG ATG CT rev TCT GCG ATC TTT TGC TTC AGG AT for CCT CTT GTT CAA TAT GAC TTC CA rev GGC CCA CTC TCA ATA ACT CTC for TGA CTT TAA GGG TTA CCT GGG TT rev GCT TCT CCC CCA GTG AGT TC for CTTCTGCCTGCTGCACTTCG rev GAGTTGATGTCGGCTACAACG for CTCTTCCTCCTGTGCCTGAT rev CTGAGTATCTTTGACTGTGGGAG for GGCTAGGGCACTATGACAC rev GGCCACCAGCATAATGAAC

198 bp

[50]

M35589

195 bp

[37]

NM_173923.2

189 bp

[51]

NM_173925

131 bp

[37]

NM_174088

156 bp

[51]

NM_173966

359 bp

[52]

AF031698

179 bp

[49]

NM_001046328

IL6 IL8 IL10 TNFa PTGS2 S100A9

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

growth of Streptococcus uberis and Escherichia coli in endometrial samples. In the uterus of one cow without endometritis Streptococcus pluranimalium was found. No anaerobic culture was performed and therefore endometrial colonization with relevant anaerobic bacteria (e. g. Fusobacterium necrophorum, Prevotella species [16]) may have been missed. In total, 94% of all cytological samples and 96% of all histological samples were evaluable. Due to redundancy in sample recovery, at least one endometrial smear and one biopsy sample (RNAlaterfixed samples served as back-up) could be evaluated in each animal. Eight RNAlater-fixed endometrial biopsies were not available for gene expression analysis, as they were either used for evaluation of histopathological interpretability of RNAlater-fixed samples (n ¼ 4) or applied as histological back-up, in case the formalin-fixed biopsy was not readable (n ¼ 4). The average RNA-concentration of the remaining RNAlater-fixed biopsies was 325.7 ng/ml and the average cDNA-concentration was 817.0 ng/ml prior to dilution to 200 ng/ml. The median RQI was 8.7 (Q1: 8, Q3: 9.5), which reveals very good RNA-quality. Two samples (2%) did not meet quality criteria (RQI > 6, RNA-concentration > 50 ng/ml) and therefore were excluded from further microbiological evaluation. No cDNA of IL1B, IL6, IL8, TNFa and S100A9 was detectable in samples transcribed without addition of RT (-RT). Due to insufficient sample material, this evaluation was not conducted for PTGS2. For IL10 the -RTanalysis repeatedly showed high Cq-values and a distinct IL10specific peak in the melting curve. The reason for this could not be clarified within the present study. Therefore, the IL10-results of RT-qPCR analysis are to be considered conditionally. A median of 91 mg US (Q1: 50 mg, Q3: 124 mg) per cow were sampled, resulting in a median dilution factor of 7.6 (Q1: 5.8, Q3: 12.9). Median photometrically determined blood contamination of the US samples was 5.9% (Q1: 2.4%, Q3: 13.8%). In total, 11% of US samples were excluded from further analysis due to a dilution factor of >50 (n ¼ 2) or a blood contamination of >33% (n ¼ 9), as arithmetical correction was not considered reasonable. To further characterize the US collected with the new sampling device, a proteome profile of five samples using nano-LC-MS/MS was generated. In total, 2,648 proteins were identified, of which only 4% were detected in only one sample, but 73% in all samples (Fig. 2) demonstrating the reliability of the sample collection procedure. The GO analysis of the identified proteins revealed that the US

191

contains proteins covering a broad variety of functional classes. An overview of the results is given in the supplementary material. Several of the identified proteins were related to defence against pathogens e.g., a variety of complement proteins, S100 proteins A8 and A9 as well as proteins related to interferon/interleukin signalling, such as the interferon regulatory factors IRF3 and IRF6, interleukin 18 and the interleukin enhancer-binding factors ILF2 and ILF3. To allow a meta-analysis of the data, the entire list of identified proteins is available in the supplementary material. 3.2. Cytokine concentrations in uterine secretions US samples of 92 cows were included in the determination of cytokine concentrations by AlphaLISA. Singular concentration values, which were <0 after correction of blood contamination, were excluded from statistical evaluation. The 92 animals showed the following uterine health status: n(E.NEG) ¼ 72, n(SE(cyto)) ¼ 7, n(SE(histo)) ¼ 13. Median concentrations of IL1B, IL8 and IL17A in US were higher in SE(cyto) compared to E.NEG (P < 0.0167, 268-fold, 5-fold, 3-fold, respectively). Furthermore, SE(cyto) revealed higher concentrations of IL8 in US compared to SE(histo) (P < 0.0167, 4-fold). Concerning uterine health status no differences were found in IL6 concentration in US. E.NEG and SE(histo) revealed no differences for the analyzed interleukins in US (Fig. 3). The statistical analysis was also conducted using the cytokine concentrations in US, which were only corrected by dilution factor but not by blood contamination. Analogue results were found (data not shown). 3.3. Endometrial expression of inflammatory genes Endometrial biopsies of 93 animals were included in gene expression analysis. The animals showed the following uterine health status: n(E.NEG) ¼ 73, n(SE(cyto)) ¼ 7, n(SE(histo)) ¼ 13. Compared to E.NEG, SE(cyto) showed a higher median expression of IL1B, IL8 and IL10 in endometrial biopsies (P < 0.0167, 3-fold, 18-fold, 2-fold, respectively). Median expression of IL1B and IL8 was higher in SE(cyto) compared to SE(histo) (P < 0.0167, 3-fold, 19fold, respectively). However, IL6, TNFa, PTGS2 and S100A9 were not differentially expressed depending on the uterine health status (data of the latter three not shown), and no differences were found between E.NEG and SE(histo) (Fig. 4). By Spearman's correlation analysis a weak positive relationship was found for IL8 concentration in US and IL8 gene expression in endometrial biopsies (P < 0.05, rho ¼ 0.33). For IL1B, IL6 and IL10, no significant correlation was found. 4. Discussion 4.1. Successful application of the novel sampling tool in vivo

Fig. 2. Venn diagram of US proteins identified in samples from 5 different healthy donors. Numbers in the diagram indicate number of identified proteins.

The novel sampling tool, developed and evaluated ex vivo on isolated abattoir uteri by Hillmer [37], was successfully applied in vivo. After a singular passage through the cervix, it allows for consecutive collection of US and cytological, histopathological and bacteriological samples of the bovine uterus. The sampling of US with Merocel sponge material in vivo was shown to be a noninvasive, well-tolerated technique and generated samples of high quality for further analyses. There was no indication that the sampling technique including two biopsy samples did affect fertility of the respective animals. This result is in accordance with those of Chapwanya et al. [35], concluding that when applied properly, bovine endometrial biopsy is a safe technique. Nevertheless other authors have described a detrimental effect of

192

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

Fig. 3. Interleukin concentrations in uterine secretions subject to uterine health status. Singular concentration values, which were < 0 after correction of blood contamination, were excluded from statistical evaluation and are respected by the number of animals presented. c(UScalc): dilution- and blood-corrected concentrations of the analyte in uterine secretions; n: number of animals; IL: interleukin; E.NEG: animals without endometritis; SE(cyto): animals with only cytologically detectable subclinical endometritis; SE(histo): animals with only histologically detectable subclinical endometritis; ‘*’: significant (P < 0.0167 with Bonferroni adjustment).

endometrial biopsy sampling on bovine fertility parameters, which potentially accounts for its low acceptance in clinical as well as in scientific context [33,34]. It is possible that the reported differences are due to the application of biopsy forceps with differing sizes. In the present study, a minimized biopsy forceps was applied, which may explain the unaffected fertility parameters of the sampled animals.

4.2. Blood contamination of US samples is negligible The contamination of US samples with blood occurred frequently and was due to unavoidable slight mucosal haemorrhages resulting from introducing the working channel into the uterus. A photometrical method was established to quantify the percentage of blood contamination of US samples by means of a standard curve and to arithmetically correct the interleukin concentrations in US for blood contamination. Thus blood contamination, which was not visible to the naked eye, could be quantified and corrected arithmetically. The statistical analysis of interleukin concentrations in US concerning uterine health revealed comparable results if an arithmetical correction of blood contamination

was applied or not. This could be the result of low blood contamination of US samples in this study (median 5.9%), which did not influence interleukin concentrations in US. Another possible explanation is that in cases of endometritis, the analytes examined in this study occur in much higher concentrations in US, resulting in negligible blood contamination. To the authors’ knowledge, this is the first study respecting and quantifying the blood contamination of US samples, while other authors have disregarded its effect [54,55]. In future studies, arithmetical correction of blood contamination of US is supposed to not be essential concerning the interleukins analyzed here. Still, this is not generally the case: if the analyte concentration is higher in blood compared to US, an influence on the results is to be expected (e.g., for TNFa, data not shown). However, it is recommended to exclude samples with high percentages of blood contamination (>33%) from analysis.

4.3. Low agreement between cytological and histological examination In the present study, low agreement between cytologically and histologically detected SE was found (SE(cyto) vs. SE(histo)). Only

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

193

Fig. 4. Gene expression in endometrial biopsies subject to uterine health status. IL: interleukin; E.NEG: animals without endometritis; SE (cyto): animals with only cytologically detectable subclinical endometritis; SE (histo): animals with only histologically detectable subclinical endometritis; ‘*’: significant (P < 0.0167 with Bonferroni adjustment); n(E.NEG) ¼ 73, n(SE(cyto)) ¼ 7, n(SE(histo)) ¼ 13.

one animal showed signs of inflammation in both cytological and histological samples, whereas seven and 13 animals revealed characteristics of inflammation only in cytological respectively histological samples. Several studies revealed low sensitivity and very high specificity of cytological examination if histological analysis was compared to a reference method [22,56,57], meaning that cytologically diagnosed SE can mostly be detected via histopathology, but histologically detected inflammation cannot always be detected via cytology. Differing inflammatory processes of the endometrium are mainly discussed as an explanation: cytological examination considers the content of PMN and therefore detects acute or ‘active‘ [58] inflammation, in which PMN gather at the endometrial surface, the epithelium. Chronic changes, like e.g., lymphoplasmacellular infiltration of the endometrium or vascular congestion, take place in deeper layers of the endometrium and can be exclusively detected by histopathological examination [32,56,58]. However, the relevance of these chronic alterations of the endometrium with regard to subfertility of dairy cows remains unclear [22,31,34,59]. Fewer studies concluded a low sensitivity of histopathological examination in comparison to the reference method cytology [21,56], meaning that acute inflammation presented by PMN can usually be detected by both methods. Meira et al. [21] found for approximately 60% of the animals with cytologically detected SE false-negative results in histo-pathological examination, but did not hypothesise any reasons. One explanation for the findings in the present study is the use of a modified biopsy tool. The respective biopsy samples are comparatively small and represent mainly superficial layers of the endometrium. Although the readability of the small-sized biopsy samples in comparison to bigger samples was evaluated and considered appropriate in a preliminary study [37], inflammatory processes might not be characterized sufficiently. Additionally, as in some samples luminal epithelium was damaged or even lacking completely, superficially located PMN were detected cytologically but were not included in the histological evaluation. Considering the broad consensus of cytological and AlphaLISA results concerning SE-diagnosis, the cytological

evaluation of the endometrium seems favourable in terms of SE diagnosis compared to the evaluation of the small endometrial biopsies, which were recovered in the present study. Under the described conditions, no effect of cytologically nor histologically detected SE on fertility parameters was detected. 4.4. Uterine secretions might represent a promising medium for diagnosis of subclinical endometritis With this project, a first step was made to evaluate the use of US for diagnosis of SE. The sampling and characterization of US, which was applied for the first time ex vivo by Hillmer [37], represent a promising tool to diagnose SE in dairy cows, but require further investigation. It is based on the idea that US is a fluidal film covering the endometrium and is formed by secretion of uterine glands as well as endometrial epithelium, stroma and immune cells. Its composition is supposed to represent the health status of the entire endometrium. The proteome analysis demonstrated that US contain a plethora of proteins with a broad variety of functions, among them proteins related to inflammation and immune defence. Furthermore, the number of identified proteins is reasonably high, compared to other studies addressing the proteome of uterine luminal fluid in humans [60,61], ewes [62] and cattle [63,64]. Together with the high overlap of the identified proteins in all samples, this demonstrates the suitability as well as the reliability of the Merocel swab method. While the examination of cytokine concentrations in uterine lavage samples [54,55] and of cervico-vaginal mucus [65] of cows with and without clinical or subclinical endometritis have been conducted and published elsewhere, to the authors’ knowledge this is the first study evaluating the concentration of cytokines in pure US in vivo. Using the methods explained above, it was possible to quantify individual dilution factors of US for adequate comparison of the samples. In contrast to uterine flushing samples, in which the actual amount of US and the dilution effects throughout the flushing medium are neglected [54,55], this novel method implies the possibility to arithmetically detect the absolute quantity of analytes in pure US.

194

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

Another advantage of the novel tool compared to uterine flushing is, that the Merocel swab method enabled sample collection in all cows, whereas Kasimanickam et al. [66] reported a failure to retrieve fluids by uterine lavage in 17% of cows. As assays must work with small sample volumes and still produce reliable results, AlphaLISA-technology was used in this study. The abundance of cytokines in US differed significantly between animals with cytologically detectable SE and animals without endometritis: In SE(cyto), the median concentration of proinflammatory cytokines IL1B, IL8 and IL17A in US was significantly higher. These inflammatory modulators are crucial in effectively starting and enhancing the local immune response by attracting and activating circulating leucocytes (mostly PMN and macrophages), forming endothelial adhesion molecules and increasing vascular permeability [67,68]. IL1B promotes the secretion of prostaglandins, chemokines (e.g., IL8) and other interleukins (e.g., IL6). An upregulated endometrial expression of IL1B has repeatedly been found in cows with endometritis [69e72]. Kim et al. [54] found no influence of cytological endometritis on IL1B-levels in uterine flushings, but neglected the described dilution factor of uterine secretions by the flushing medium and applied a different PMN-threshold (18%) compared to the present study. The chemokine IL8 is regarded as the strongest chemotactic factor for PMN [73,74]. Its chemoattractive potential in the bovine uterus has been documented in vivo [75]. Other groups found its upregulation in endometritic cows in product and gene expression levels [54,70,72]. A synchronous upregulation of IL1B- and IL8-mRNA abundance has also been found in several studies [70,76,77] and might be due to the fact that IL1B potentially promotes IL8secretion [73]. The critical role of IL17A in the regulation and enhancement of the immune response was only detected approximately 25 years ago [78,79]. Its relevance in bovine subclinical endometritis has not been explored in detail, but some results indicate that it is involved in endometrial immune defence [80]. To our knowledge, the present work is the first analysing IL17A in cows with SE on product level. In contrast to these results, US of animals with only histologically detectable SE and animals with healthy uteri revealed cytokine concentrations of comparable levels. These findings indicate that merely histologically detectable, chronic inflammation of the endometrium does not modulate the uterine environment concerning the investigated analytes. Until now, the prospective value of these chronic inflammatory alterations of the endometrium for bovine reproductive performance has not been confirmed [22,31]. Thresholds of cytokine concentrations in US to differentiate between animals with and without cytologically detectable SE as well as sensitivity and specificity of the analysis have yet to be established in a follow-up study including a suitable number of animals. Additionally, it should be evaluated whether thresholds independent from the time point of examination in relation to calving are applicable. The detected differences in US cytokine concentrations could form the basis for future research projects focusing on the development of a cow-side-test for diagnosis of SE by means of US. The outstanding advantage of US analysis compared to cytological analysis is the impartial quantification of analytes which are relevant for inflammation, whereas cytological results depend on the experience of the person conducting the analysis. The results of US analysis were confirmed on endometrial gene expression level. The pro-inflammatory mediators IL1B and IL8 revealed a higher median mRNA-abundance in SE(cyto) compared to E.NEG and SE(histo). Additionally, the median expression of the anti-inflammatory cytokine IL10 was higher in the endometrium of SE(cyto) compared to E.NEG, whereas no effect was found on product level. In addition to post-transcriptional regulation

accounting for this divergence [81], an effect of the high Cq-values in the -RT-analysis of IL10 in this study has to be considered. IL10 is regarded as the major anti-inflammatory cytokine confining the immune response and controlling tissue damage due to excessive immune reactions. Hillmer [37] and Brodzki et al. [55] considered the anti-inflammatory presence cohesive to subclinical endometritis as a reason for the establishment or maintenance of chronic inflammation. This might prevent the establishment of an adequate postpartum endometrial inflammatory response [13], which is necessary for bacterial clearance, uterine involution and the reestablishment of endometrial integrity [11,12]. 4.5. Uterine secretions are a promising medium for further research on SE in cattle Analyzing US in cattle permits a detailed view into the characteristics of the uterine milieu including potentially existing inflammatory processes as well as its bacteriological status. In future, this could be used to clarify patho-physiological mechanisms leading to subfertility in dairy cows on molecular level in holistic approaches. The use of US could represent an important part of uterine proteomic analyses and could lead to further identification of biomarkers with relevance for pathogenesis and chronification of SE. Moreover, differentiation among cows with SE according to clusters of regulated analytes might be promising. Furthermore, sampling of US could be used to evaluate the mechanisms and effectiveness of new prophylactic or therapeutic strategies. The combined characterization of US and uterine bacteriological status by the novel sampling device might enable identification of SE cows requiring treatment: self-cure rates of 40%e66% have been reported for SE detected between four and five weeks post-partum [19,82,83]. However, some cows with differing characteristics might depend more on treatment to reinstall uterine functions within the expected time. Besides, analysis of US might contribute to illuminate the mechanisms leading to impaired fertility in SEaffected cows. 5. Conclusions The sampling and evaluation of uterine secretions is considered a promising method for characterization of subclinical endometritis in cattle. The cytokines IL1B, IL8 and IL17A are promising diagnostic candidates for cytologically detectable SE in US. To determine thresholds of cytokine concentrations in US for diagnostic purposes, a large-scale study including an adequate number of SE animals is necessary. The examination of US might serve as an alternative to cytological diagnostic techniques and implies the outstanding advantage to provide impartial results detached from the examiner's experience in an assay which is simple and fast to perform. As it provides a numerical value which cannot be influenced by the researcher, diagnostic bias can be reduced. It enables objective information of comparable quality and accuracy among different examiners and groups, which is not obtained by endometrial cytology. The new diagnostic tool offers the possibility to compare cytological as well as histological findings with characterization of US. In the long term, this could be helpful for the development of treatment strategies or prophylactic routes to face bovine subclinical endometritis. Avenues for further research include the determination of thresholds of cytokine concentrations in US for SE diagnosis as well as sensitivity and specificity of the analysis and the identification of additional biomarkers. Holistic strategies, including proteomic analyses of US, are to be considered to investigate the pathophysiological mechanisms leading to bovine subclinical endometritis and subsequent impairment of fertility.

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

Declaration of interest None. Acknowledgements This work was supported by the Association for Bioeconomy Research (FBF). The authors' sincere gratitude is expressed to the staff of the LVFZ Achselschwang, esp. Robert Schwaiger, and Sandra Kirsch for their patient support and fruitful collaboration in animal sampling. The authors acknowledge Dr. Georg Wolf of the Institute for Infectious Diseases and Zoonoses at LMU Munich, for enabling bacteriological differentiation by MALDI-TOF mass spectrometry and Dr. Anna Rieger for assistance in statistical inquiries. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.theriogenology.2019.09.016. References [1] Walsh SW, Williams EJ, Evans ACO. A review of the causes of poor fertility in high milk producing dairy cows. Anim Reprod Sci 2011;123:127e38. [2] Butler WR. Effect of protein nutrition on ovarian and uterine physiology in dairy cattle. J Dairy Sci 1998;81:2533e9. [3] Lucy MC. Reproductive loss in high-producing dairy cattle: where will it end? J Dairy Sci 2001;84:1277e93. [4] Rethmeier J, Wenzlau M, Wagner M, Wiedemann S, Bachmann L. Fertility parameters in German dairy herds - determined by herd size and milk yield? Reprod Domest Anim 2018;53(Suppl. 1):33. [5] Parkinson TJ. Subfertility and Infertility - infertility and subfertility in the cow: structural and functional abnormalities, management deficiencies and nonspecific infections. In: Noakes D, Parker J, England G, editors. Veterinary reproduction and obstetrics. Elsevier Saunders; 2009. p. 393e475. [6] Pryce JE, Royal MD, Garnsworthy PC, Mao IL. Fertility in the high-producing dairy cow. Livest Sci 2004;86:125e35. [7] Roche JF. The effect of nutritional management of the dairy cow on reproductive efficiency. Anim Reprod Sci 2006;96:282e96. [8] LeBlanc SJ. Assessing the association of the level of milk production with reproductive performance in dairy cattle. J Reprod Dev 2010;56:1e7. [9] Abdelli A, Raboisson D, Kaidi R, Ibrahim B, Kalem A, Iguer-Ouada M. Elevated non-esterified fatty acid and b-hydroxybutyrate in transition dairy cows and their association with reproductive performance and disorders: a metaanalysis. Theriogenology 2017;93:99e104. [10] Long S. Pregnancy and Parturition - abnormal development of the conceptus and its consequences. In: Noakes D, Parker J, England G, editors. Veterinary reproduction and obstetrics. Elsevier Saunders; 2009. p. 123e45. [11] Sheldon IM, Cronin JG, Healey GD, Gabler C, Heuwieser W, Streyl D, et al. Innate immunity and inflammation of the bovine female reproductive tract in health and disease. Reproduction 2014;148:41e51. [12] LeBlanc SJ. Reproductive tract inflammatory disease in postpartum dairy cows. Animal 2014;8:54e63. [13] Chapwanya A, Meade KG, Foley C, Narciandi F, Evans ACO, Doherty ML, et al. The postpartum endometrial inflammatory response: a normal physiological event with potential implications for bovine fertility. Reprod Fertil Dev 2012;24:1028e39. [14] Gilbert RO. The effects of endometritis on the establishment of pregnancy in cattle. Reprod Fertil Dev 2011;24:252e7. [15] Kasimanickam R, Duffield TF, Foster RA, Gartley CJ, Leslie KE, Walton JS, et al. Endometrial cytology and ultrasonography for the detection of subclinical endometritis in postpartum dairy cows. Theriogenology 2004;62:9e23. [16] Sheldon IM, Lewis GS, LeBlanc SJ, Gilbert RO. Defining postpartum uterine disease in cattle. Theriogenology 2006;65:1516e30. [17] Gilbert RO, Shin ST, Guard CL, Erb HN, Frajblat M. Prevalence of endometritis and its effects on reproductive performance of dairy cows. Theriogenology 2005;64:1879e88. [18] Barlund CS, Carruthers TD, Waldner CL, Palmer CW. A comparison of diagnostic techniques for postpartum endometritis in dairy cattle. Theriogenology 2008;69:714e23.  ski W, Podhalicz-Dzie˛ gielewska M, Zdun  czyk S, Janowski T. The diag[19] Baran nosis and prevalence of subclinical endometritis in cows evaluated by different cytologic thresholds. Theriogenology 2012;78:1939e47. [20] De Boer MW, LeBlanc SJ, Dubuc J, Meier S, Heuwieser W, Arlt S, et al. Invited review: systematic review of diagnostic tests for reproductive-tract infection and inflammation in dairy cows. J Dairy Sci 2014;97:3983e99. [21] Meira EBS, Henriques LCS, Sa LRM, Gregory L. Comparison of ultrasonography and histopathology for the diagnosis of endometritis in Holstein-Friesian

195

cows. J Dairy Sci 2012;95:6969e73. [22] Madoz LV, Giuliodori MJ, Migliorisi AL, Jaureguiberry M, de la Sota RL. Endometrial cytology, biopsy, and bacteriology for the diagnosis of subclinical endometritis in grazing dairy cows. J Dairy Sci 2014;97:195e201. [23] Wagener K, Gabler C, Drillich M. A review of the ongoing discussion about definition, diagnosis and pathomechanism of subclinical endometritis in dairy cows. Theriogenology 2017;94:21e30. [24] Dubuc J, Duffield TF, Leslie KE, Walton JS, LeBlanc SJ. Definitions and diagnosis of postpartum endometritis in dairy cows. J Dairy Sci 2010;93:5225e33. [25] Denis-Robichaud J, Dubuc J. Determination of optimal diagnostic criteria for purulent vaginal discharge and cytological endometritis in dairy cows. J Dairy Sci 2015;98:6848e55. [26] Pascottini OB, Hostens M, Sys P, Vercauteren P, Opsomer G. Cytological endometritis at artificial insemination in dairy cows: prevalence and effect on pregnancy outcome. J Dairy Sci 2017;100:588e97. [27] Bondurant RH. Inflammation in the bovine female reproductive tract. J Anim Sci 1999;77:101. [28] Cengiz M, Altun S, Comakli S, Ozkaraca M, Cengiz S, Hayirli A. Sensitivity and specificity of endometrial cytopathological diagnosis in slaughtered cows. Reprod Domest Anim 2016;51(Suppl. 2):82. [29] Chapwanya A, Meade KG, Doherty ML, Callanan JJ, Mee JF, O'Farrelly C. Histopathological and molecular evaluation of Holstein-Friesian cows postpartum: toward an improved understanding of uterine innate immunity. Theriogenology 2009;71:1396e407. €ntzke J, Drillich M, De la Sota RL. [30] Madoz LV, Giuliodori MJ, Jaureguiberry M, Plo The relationship between endometrial cytology during estrous cycle and cutoff points for the diagnosis of subclinical endometritis in grazing dairy cows. J Dairy Sci 2013;96:4333e9. [31] Pascottini OB, Opsomer G. Postpartum uterine diseases in dairy cows: a review with emphasis on subclinical endometritis. Vlaams Diergeneeskundig Tijdschrift 2016;85:378e85. [32] Bonnett BN, Miller RB, Etherington WG, Martin SW, Johnson WH. Endometrial biopsy in Holstein-Friesian dairy cows. I. Technique, histological criteria and results. Can J Vet Res 1991;55:155e61. [33] Etherington WG, Martin SW, Bonnett B, Johnson WH, Miller RB, Savage NC, et al. Reproductive performance of dairy cows following treatment with cloprostenol 26 and/or 40 days postpartum: a field trial. Theriogenology 1988;29:565e75. [34] Bonnett BN, Martin SW, Meek AH. Associations of clinical findings, bacteriological and histological results of endometrial biopsy with reproductive performance of postpartum dairy cows. Prev Vet Med 1993;15:205e20. [35] Chapwanya A, Meade KG, Narciandi F, Stanley P, Mee J, Doherty ML, et al. Endometrial biopsy: a valuable clinical and research tool in bovine reproduction. Theriogenology 2010;73:988e94. [36] Pascottini OB, Dini P, Hostens M, Ducatelle R, Opsomer G. A novel cytologic sampling technique to diagnose subclinical endometritis and comparison of staining methods for endometrial cytology samples in dairy cows. Theriogenology 2015;84:1438e46. [37] Hillmer M. Uterussekret zur Diagnostik und Pathogeneseforschung der chronischen Endometritis beim Rind. Diss. med. vet. Ludwig-MaximiliansUniversit€ at München; 2018. [38] Castle PE, Rodriguez AC, Bowman FP, Herrero R, Schiffman M, Bratti MC, et al. Comparison of ophthalmic sponges for measurements of immune markers from cervical secretions. Clin Diagn Lab Immunol 2004;11:399e405. [39] Melcher Y, Prunner I, Drillich M. Degree of variation and reproducibility of different methods for the diagnosis of subclinical endometritis. Theriogenology 2014;82:57e63. € mmling A, Ulbrich S, Weinert M, Piechotta M, [40] Heppelmann M, Bro Wrenzycki C, et al. Effect of suppression of postpartum ovulation on endometrial inflammation in dairy cows. Theriogenology 2015;84:155e62. [41] Crowley-Nowick PA, Bell MC, Brockwell R, Edwards RP, Chen S, Partridge EE, et al. Rectal immunization for induction of specific antibody in the genital tract of women. J Clin Immunol 1997;17:370e9. [42] Rohan LC, Edwards RP, Kelly LA, Colenello KA, Bowman FP, CrowleyNowick PA. Optimization of the weck-Cel collection method for quantitation of cytokines in mucosal secretions. Clin Diagn Lab Immunol 2000;7:45e8. [43] Zijlstra WG, Buursma A. Spectrophotometry of hemoglobin: absorption spectra of bovine oxyhemoglobin, deoxyhemoglobin, carboxyhemoglobin, and methemoglobin. Comp Biochem Physiol B Biochem Mol Biol 1997;118: 743e9. [44] Sheth SA, Yanamadala V, Eskandar EN. Intraoperative human functional brain mapping using optical intrinsic signal imaging. In: Chaudhary V, editor. Advances in brain imaging. InTech; 2012. p. 77e100. [45] Heberle H, Meirelles GV, da Silva FR, Telles GP, Minghim R. InteractiVenn: a web-based tool for the analysis of sets through Venn diagrams. BMC Bioinf 2015;16:169. [46] Mi H, Muruganujan A, Casagrande JT, Thomas PD. Large-scale gene function analysis with the PANTHER classification system. Nat Protoc 2013;8:1551. [47] Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, et al. The MIQE guidelines: minimum information for publication of quantitative realtime PCR experiments. Clin Chem 2009;55:611e22. [48] Fleige S, Pfaffl MW. RNA integrity and the effect on the real-time qRT-PCR performance. Mol Asp Med 2006;27:126e39. [49] Lind M, Sipka AS, Schuberth HJ, Blutke A, Wanke R, Sauter-Louis C, et al. Location-specific expression of chemokines, TNF-a and S100 proteins in a teat

196

A.L. Helfrich et al. / Theriogenology 141 (2020) 186e196

explant model. Innate Immun 2015;21:322e31. [50] Neuvians TP, Schams D, Berisha B, Pfaffl MW. Involvement of proinflammatory cytokines, mediators of inflammation, and basic fibroblast growth factor in prostaglandin F2-induced luteolysis in bovine corpus luteum. Biol Reprod 2004;70:473e80. [51] Yang W, Zerbe H, Petzl W, Brunner RM, Günther J, Draing C, et al. Bovine TLR2 and TLR4 properly transduce signals from Staphylococcus aureus and E. coli, but S. aureus fails to both activate NF-kB in mammary epithelial cells and to quickly induce TNFa and interleukin-8 (CXCL8) expression in the udder. Mol Immunol 2008;45:1385e97. [52] Odau S, Gabler C, Holder C, Einspanier R. Differential expression of cyclooxygenase 1 and cyclooxygenase 2 in the bovine oviduct. J Endocrinol 2006;191:263e74. [53] R Core Team R. A language and environment for statistical computing. In: R foundation for statistical computing; 2016. Vienna, Austria. [54] Kim I-H, Kang H-G, Jeong J-K, Hur T-Y, Jung Y-H. Inflammatory cytokine concentrations in uterine flush and serum samples from dairy cows with clinical or subclinical endometritis. Theriogenology 2014;82:427e32. [55] Brodzki P, Kostro K, Brodzki A, Wawron W, Marczuk J. Inflammatory cytokines and acute-phase proteins concentrations in the peripheral blood and uterus of cows that developed endometritis during early postpartum. Theriogenology 2015;84:11e8. €nderungen beim [56] Merbach S. Prognostische bewertung endometrialer vera €t Leipzig; 2011. rind. Diss. med. vet. Universita [57] Pascottini OB, Hostens M, Dini P, Vandepitte J, Ducatelle R, Opsomer G. Comparison between cytology and histopathology to evaluate subclinical endometritis in dairy cows. Theriogenology 2016;86:1550e6. [58] Pascottini OB, Hostens M, Dini P, Vandepitte J, Ducatelle R, Opsomer G. Distribution of inflammation and association between active and chronic alterations within the endometrium of dairy cows. Reprod Domest Anim 2016;51: 751e7. [59] Rodenbusch S. Makroskopische und histopathologische Untersuchungen am Genitaltrakt sub- und infertiler weiblicher Rinder im klinischen Kontext unter besonderer Berücksichtigung der Endometriumbiopsie. Diss. med. vet. Uni€t Leipzig; 2011. versita [60] Soleilhavoup C, Riou C, Tsikis G, Labas V, Harichaux G, Kohnke P, et al. Proteomes of the female genital tract during the oestrous cycle. Mol Cell Proteom 2016;15:93e108. [61] Casado-Vela J, Rodriguez-Suarez E, Iloro I, Ametzazurra A, Alkorta N, GarcíaVelasco JA, et al. Comprehensive proteomic analysis of human endometrial fluid aspirate. J Proteome Res 2009;8:4622e32. [62] Koch JM, Ramadoss J, Magness RR. Proteomic profile of uterine luminal fluid from early pregnant ewes. J Proteome Res 2010;9:3878e85. [63] Forde N, McGettigan PA, Mehta JP, O'Hara L, Mamo S, Bazer FW, et al. Proteomic analysis of uterine fluid during the pre-implantation period of pregnancy in cattle. Reproduction 2014;147:575e87. [64] Faulkner S, Elia G, O'Boyle P, Dunn M, Morris D. Composition of the bovine uterine proteome is associated with stage of cycle and concentration of systemic progesterone. Proteomics 2013;13:3333e53. [65] Adnane M, Chapwanya A, Kaidi R, Meade KG, O'Farrelly C. Profiling inflammatory biomarkers in cervico-vaginal mucus (CVM) postpartum: potential early indicators of bovine clinical endometritis? Theriogenology 2017;103: 117e22. [66] Kasimanickam R, Duffield TF, Foster RA, Gartley CJ, Leslie KE, Walton JS, et al. A comparison of the cytobrush and uterine lavage techniques to evaluate

[67] [68] [69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78]

[79]

[80]

[81] [82]

[83]

endometrial cytology in clinically normal postpartum dairy cows. Can Vet J 2005;46:255e9. Singh J, Murray RD, Mshelia G, Woldehiwet Z. The immune status of the bovine uterus during the peripartum period. Vet J 2008;175:301e9. Holdsworth SR, Gan P-Y. Cytokines: names and numbers you should care about. Clin J Am Soc Nephrol 2015;10:2243e54. Kasimanickam R, Kasimanickam V, Kastelic JP. Mucin 1 and cytokines mRNA in endometrium of dairy cows with postpartum uterine disease or repeat breeding. Theriogenology 2014;81:952e8. Gabler C, Fischer C, Drillich M, Einspanier R, Heuwieser W. Time-dependent mRNA expression of selected pro-inflammatory factors in the endometrium of primiparous cows postpartum. Reprod Biol Endocrinol 2010;8:152. Wagener K, Pothmann H, Prunner I, Peter S, Erber R, Aurich C, et al. Endometrial mRNA expression of selected pro-inflammatory factors and mucins in repeat breeder cows with and without subclinical endometritis. Theriogenology 2017;90:237e44. ~o KN, Santos NR, Galva ~o JS, Gilbert RO. Association between endomeGalva tritis and endometrial cytokine expression in postpartum Holstein cows. Theriogenology 2011;76:290e9. Wuyts A, Struyf S, Gijsbers K, Schutyser E, Put W, Conings R, et al. The CXC chemokine GCP-2/CXCL6 is predominantly induced in mesenchymal cells by interleukin-1b and is down-regulated by interferon-g: comparison with interleukin-8/CXCL8. Lab Investig 2003;83:23e34. Romagnani P, Lasagni L, Annunziato F, Serio M, Romagnani S. CXC chemokines: the regulatory link between inflammation and angiogenesis. Trends Immunol 2004;25:201e9. Zerbe H, Schuberth HJ, Engelke F, Frank J, Klug E, Leibold W. Development and comparison of in vivo and in vitro models for endometritis in cows and mares. Theriogenology 2003;60:209e23. Fischer C, Drillich M, Odau S, Heuwieser W, Einspanier R, Gabler C. Selected pro-inflammatory factor transcripts in bovine endometrial epithelial cells are regulated during the oestrous cycle and elevated in case of subclinical or clinical endometritis. Reprod Fertil Dev 2010;22:818e29. Peter S, Michel G, Hahn A, Ibrahim M, Lübke-Becker A, Jung M, et al. Puerperal influence of bovine uterine health status on the mRNA expression of proinflammatory factors. J Physiol Pharmacol 2015;66:449e62. Rouvier E, Luciani MF, Mattei MG, Denizot F, Golstein P. CTLA-8, cloned from an activated T cell, bearing AU-rich messenger RNA instability sequences, and homologous to a herpesvirus saimiri gene. J Immunol 1993;150:5445e56. Yao Z, Fanslow WC, Seldin MF, Rousseau A-M, Painter SL, Comeau MR, et al. Herpesvirus Saimiri encodes a new cytokine, IL-17, which binds to a novel cytokine receptor. Immunity 1995;3:811e21. Johnson H, Torres CG, Carvallo F, Duchens M, Peralta OA. Endometrial expression of selected transcripts in postpartum of primiparous Holstein cows with clinical and subclinical endometritis. Anim Reprod Sci 2015;156:34e9. Day DA, Tuite MF. Post-transcriptional gene regulatory mechanisms in eukaryotes: an overview. J Endocrinol 1998;157:361e71. Kasimanickam R, Kasimanickam VR, Olsen JR, Jeffress EJ, Moore DA, Kastelic JP. Associations among serum pro- and anti-inflammatory cytokines, metabolic mediators, body condition, and uterine disease in postpartum dairy cows. Reprod Biol Endocrinol 2013;11:103. Dubuc J, Duffield TF, Leslie KE, Walton JS, LeBlanc SJ. Randomized clinical trial of antibiotic and prostaglandin treatments for uterine health and reproductive performance in dairy cows. J Dairy Sci 2011;94:1325e38.