An immune/inflammatory challenge induced by bacterial infection is suspected to affect the reproductive process in females. It was reported that pelvic inflammatory disease and metritis are important causes of serious disease and infertility in humans and domestic animals. Each year in the United States, more than 1 million women experience an episode of acute pelvic inflammatory disease and more than 100,000 become infertile (Williams et al., 2008). Moreover, it is increasing evidence that uterine disease also affects ovarian function. The most numerous pathogenic bacteria in the bovine uterus is Escherichia coli and its presence is specifically associated with ovarian dysfunction (Williams et al., 2007). The effects of E. coli seems to be mediated directly through the bacterial endotoxin - lipopolysaccharide (LPS), or indirectly through the inflammatory mediators associated with E. coli infection including pro-inflammatory cytokines which concentration in the blood is highly elevated during urine infection (Williams et al., 2008). It was also described that the systemic inflammation induced by intravenous injection (i.v.) of LPS significantly decreases the concentration of luteinizing hormone (LH) (Daniel et al. 2003; Herman et al. 2010) and could even lead to the disruption of ovarian cyclicity (Peter et al. 1989). It is worth mentioning that bacterial endotoxins including LPS are commonly used as in vivo model of inflammation because they induce pathophysiological responses throughout the organism which are very similar to a real bacterial infection (Tilders et al. 1994). At least partially, the endotoxins may affect the LH secretion directly at the level of anterior pituitary (AP) (Coleman et al. 2003; Haziak et al. 2013). However, it is believed that the major anti-luteinizing effect of inflammation is focused on the suppression of gonadotropin-releasing hormone (GnRH) secretion in the hypothalamus as LH synthesis and release are dependent on the pattern of this decapeptide pulsatile secretion. It was previously reported that pro-inflammatory cytokines which penetrate the hypothalamic area during inflammation can alter the GnRH pulse amplitude, duration, and frequency and can disturb the female ovulatory cycle (Rivier and Vale, 1990; Kalra et al. 1998; Karsch et al. 2002). These cytokines such as interleukin (IL) -1??, IL-6 and tumor necrosis factor (TNF) ?? may affect the GnRH secretion directly acting via their corresponding receptors occurring widespread in the hypothalamic area (McMahan et al. 1991; Gadient and Otten, 1993; Vitkovic et al. 2000). The results of previous studies showed that IL-1?? is a potent negative regulator of GnRH secretion (Kang et al. 2000; Herman et al. 2012). IL-1?? could directly modulate GnRH synthesis and release because in vitro study performed on the immortalized GnRH neurons showed the expression of its type one receptor directly on these cells (Igaz et al. 2006). However, IL-1?? and other pro-inflammatory cytokines may also modulate the GnRH secretion indirectly via prostaglandins (PGs) ' dependent pathway; they are all responsible for stimulation of PGs synthesis in the hypothalamus during sepsis (Netea et al. 2000). The study performed on rats showed that pharmacological inhibition of cyclooxygenase (COX) -1 and 2 required for PGs production prevented the IL-1 and TNF??-induced inhibition of GnRH/LH secretion (Rivest and Rivier, 1993; Yoo et al. 1997). These results suggest that the reduction of PGs synthesis in the brain could reverse the negative effect of inflammation on the reproduction process at the level of central nervous system (CNS).
The aim of the present study has been to determinate the effect of the i.v. injection of meloxicam, a selective COX-2 inhibitor, on GnRH/LH secretion in the anestrous ewes during an immune/inflammatory challenge induced by repeated LPS treatment.
Materials and methods
The studies were performed on adult, 3-year-old Polish mountain sheep in the anestrous season (April'May). To determine the anestrous state, all ewes were bled by jugular venipuncture every 48 h for 10 days, and concentrations of progesterone in serum were used to identify cyclic and anestrous ewes. The anestrous state was determined on the basis of low progesterone level in blood plasma which never exceeded 1 ng/ml during the 10-day-period. All animals were in good condition. The body condition of all individuals was estimated at three on a five-point scale. The animals were maintained indoors in individual pens and exposed to natural daylight. The ewes were well adapted to the experimental conditions; and always had visual contact with neighbouring ewes, even during the experimental period, to prevent stress due to social isolation. The animals were fed a constant diet of commercial concentrates with hay and water available ad libitum.
All procedures on animals were performed with the consent of the Local Ethics Committee of the Warsaw Agriculture University.
Venous catheters were implanted into the jugular vein on the day prior to the experiment. Ewes (n = 24) were randomly divided into four experimental groups: Group I - control (n = 6); Group II ' meloxicam treated (n = 6); Group III ' LPS-treated (n = 6); group IV ' LPS- and meloxicam- treated (n = 6). In treated animals, the immune stress was induced by the i.v. administration of LPS from Escherichia coli 055:B5 (Sigma, St. Louis, MO, USA) dissolved in saline (0.9% w'v NaCl) (Baxter, Deer'eld, IL, USA) at a concentration of 10 mg/L into the jugular vein (400 ng/kg). For six days, the animals received single injection of LPS or saline at 8 a.m. At the day 7 of the experiment, 30 min prior to saline/LPS treatment ewes from group II and IV received i.v. injection of meloxicam (500 ??g/kg) (Boehringer Ingelheim GmbH, Germany) respectively. The effectiveness of meloxicam treatment in LPS treated animals was verified by body temperature measurement. In all meloxicam-treated animals completely suppression of LPS induced fever was observed (data not shown). At the same time, the animals from control and LPS treated groups received 2 ml of saline injection. At the day 7 of the experiment, jugular blood samples (5 ml) were taken for measurement of the peripheral hormone concentration at 10 min intervals beginning 2 hrs before the i.v. administration of LPS or an equivalent volume of saline injection and continuing for 3 hrs. Blood samples were collected into heparinized tubes and immediately centrifuged for 10 min at 1000 ?? g at 4 ??C. Plasma was stored at -80 ??C until assayed.
After the end of blood collection, the animals were euthanized immediately and the brains were rapidly removed from the skulls. Next, the AP and hypothalamic structures were dissected. The block of brain encompassing hypothalamic structures were sectioned sagittally and dissected from both sides into four parts, i.e., the preoptic area (POA), anterior hypothalamus (AHA), medial basal hypothalamus (MBH) and median eminence (ME), according to stereotaxic atlas of the sheep brain (Welento et al. 1969). Landmarks were the mammillary body, median eminence and optic chiasm. The depths of the cuts were 2 to 2.5 mm for MBH and 2.5 to 3 mm for AHA and POA. All tissues were frozen immediately after collection in liquid nitrogen and were stored at -80 ??C.
Radioimmunoassay of plasma hormones
Radioimmunoassay for LH
The plasma LH concentration was assayed with a double-antibody RIA using anti-ovine-LH and anti-rabbit-'-globulin antisera and ovine standard (teri.oLH, Tucker Endocrine Research Institute), according to Stupnicki and Madej (1976). The assay sensitivity was 0.3 ng/ml and the intra- and inter-assay coefficients of variation were 8% and 11.5%, respectively.
Radioimmunoassay for FSH
The concentration of FSH was determined by double antibody radioimmunoassay (RIA) using anti ovine-FSH (teri.anti-oFSH) and anti-rabbit-'-globulin antisera, according to L'Hermite et al. (1972). The anti-FSH, as well as the FSH standard (teri. oFSH-and teri. FSH ig), was kindly supplied by Dr. L.E. Reichert Jr. (Tucker Endocrine Research Institute LLC, Atlanta, Georgia, USA). The assay sensitivity was 1.5 ng/ml and the intra- and inter-assay coefficients of variation were 3.5% and 11.0%, respectively.
Radioimmunoassay for prolactin
The plasma prolactin concentration was assayed by a radioimmunoassay double-antibody method, using specific anti-ovine prolactin and anti-rabbit-??-globulin antisera according to Woli??ska et al. (1977). The prolactin standard for iodination was obtained according to the method described by Kochman and Kochman (1977). The assay sensitivity for prolactin was 2 ng/ml, and the intra- and inter-assay coefficients of variation were 9% and 12%, respectively.
Radioimmunoassay for cortisol
The cortisol concentrations were determined by radioimmunoassay (RIA) according to Kokot and Stupnicki (1985), using rabbit anti-cortisol antisera (R/75) and an HPLC-grade cortisol standard (SIGMA, USA). The assay sensitivity was 1 ng/ml and the intra- and interassay coefficients of variation for cortisol were 9% and 12%, respectively.
ELISA assay for the GnRH concentration in the POA
The concentrations of GnRH in the POA were determined with a commercial GnRH ELISA kit (BlueGene Biotech CO., LTD. China) dedicated for sheep. The tissues were homogenized in 400 ??l of phosphate buffered saline (0.02 M). Then homogenates were subjected to two freeze-thaw cycles to further break the cell membranes. After that, the homogenates were centrifugated for 15 min at 1500 x g in 4 ??C. The supernatants were aliquoted and stored until assay in ' 80 ??C. All steps in the assays were performed according to the manufacturer's instructions. The incubation of plates and absorbance measurement at 450 nm were performed using a VersaMax reader (Molecular Devices LLC., Sunnyvale, California, United States). The assay sensitivity was 1.0 pg/ml. The values of GnRH concentration were normalized to total protein content in each sample assayed using Bradford method.
Determining the relative gene expression
Total RNA from the AP tissues was isolated using NucleoSpin?? RNA II Kit (MACHEREY-NAGEL Gmbh & Co; D??ren, Germany) according to a manufacturer's instruction. The purity and concentration of isolated RNA were spectrophotometrically quantified by measuring the optical density at 230, 260 and 280 nm in a NanoDrop 1000 instrument (Thermo Fisher Scientific Inc., Waltham, USA). The RNA integrity was verified by electrophoresis using 1 % agarose gel stained with ethidium bromide. Maxima' First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Fisher Scientific Inc., Waltham, USA) was used to prepare cDNA synthesis. As a starting material for this PCR synthesis 2 ??g of total RNA was used.
Real-time RT-PCR was carried out using HOT FIREPol EvaGreen?? qPCR Mix Plus (Solis BioDyne, Tartu, Estonia) components and HPLC-grade oligonucleotide primers synthesised by Genomed (Poland). Speci'c primers for determining the expression of housekeeping genes and the genes of interest were designed using Primer 3 software (Table 1). One tube contained: 4 ??l PCR Master Mix (5x), 14 ??l RNase-free water, 1 ??l primers (0.5 ??l each, working concentration was 0.25 ??M) and 1 ??l cDNA template. The tubes were run on the Rotor-Gene 6000 (Qiagen, Duesseldorf, Germany). The following protocol was used: 95??C in 15 min for activating Hot Star DNA polymerase and finally the PCR including 30 cycles at 95??C in 10 sec for denaturation, 60??C in 20 sec for annealing, and 72??C in 10 sec for extension. After the cycles, a final melting curve analysis under continuous fluorescence measurements was performed to confirm the specificity of the amplification.
Relative gene expression was calculated using the comparative quantification option of Rotor Gene 6000 software 1.7. (Qiagen, Duesseldorf, Germany). The second differential maximum method (Rasmussen, 2001) was used in this analysis to calculate reaction efficiencies and a set percentage of the maximum fluorescence value to calculate the beginning of the exponential phase. To compensate a variation in cDNA concentrations and the PCR efficiency between tubes, an endogenous control gene was assayed in each sample and used for normalization. Initially, three housekeeping genes: GAPDH, ??-actin, PPIC were tested. The BestKeeper was used to determine the most stable housekeeping gene, for normalizing genes of interest expression. The BestKeeper based on the pair-wise correlation analysis of all pairs of candidate genes (Pfaffl et al. 2004) and calculates variations of all reference genes (SD (?? Ct)). GAPDH was chosen as the best endogenous control gene. They had the lowest SD (?? Ct) value and a good correlation coefficient with the remaining analyzed housekeeping genes.
Statistical analysis of data
The results of hormones concentration are presented as the mean ?? S.E.M. All experiments consisted of a baseline period when no treatment was given (2 to 0.5 h before) and a period after treatment (1 to 3 h after). To identify treatment effects, the mean values for the baseline and treatment periods were obtained. Next, the data were analyzed using a one-way ANOVA to identify significant interactions for all parameters between the control and experimental groups. ANOVAs for the hormones parameters excluded data during the first hour after LPS or saline treatment to allow time for treatments to take effect. When a significant treatment by time interaction was observed, a post hoc analysis was conducted to identify treatment effects the Tukey's test was used to compare pre- compared with post-treatment values. Statistical significance was defined as P < 0.001.
During data analysis of hormone concentration significant differences in the level of prolactin between individuals assigned for meloxicam and LPS treated group were found at the day 7 of the experiment before last LPS injection. Based on this observation, the group of animals treated with meloxicam and LPS were divided into two subgroups of three individuals each in treated groups. Three individuals characterized by the high level of circulating prolactin which did not show changes in LH release after meloxicam treatment were assigned to Meloxicam I group (n=3). Three other individuals characterized by the level of prolactin not different from control group which characterized increased LH release after meloxicam treatment were assigned to Meloxicam II group (n = 3).
The results of GnRH concentration were analyzed using a one-way ANOVA to identify significant interactions for all parameters between the control and experimental groups. A post hoc analysis based on the Tukey's test was conducted to identify treatment effects. The results are presented as the mean value ?? S.E.M Statistical significance was defined as P < 0.01.
The results of relative gene expression are presented as the mean value ?? S.E.M. The significance of the differences in the levels of GnRH between the experimental groups was assessed by the Mann-Whitney U test. Statistical significance was defined as P < 0.05.
Effect of meloxicam and repeated LPS treatment on LH, FSH, prolactin and cortisol release
Lasting 7 days administration of LPS significantly suppressed (P < 0.001) LH release. However, no additional effect of subsequent LPS treatment on LH release suppression was determined at the day 7 of the experiment. The effect of meloxicam on LH release among endotoxin treatment animals was differentiated. In three individuals meloxicam successfully (P < 0.001) abolished the inhibitory effect of LPS. On the other hand, in three other individuals the meloxicam treatment did not affect LH release. No effect of both LPS and meloxicam on FSH release was observed (Fig. 1). There was no effect of repeated LPS treatment on the cortisol concentration assayed before endotoxin injection at the day 7 of the experiment (Fig. 2). However, significant (P < 0.001) reductions of cortisol release were assayed in all LPS treated groups in comparison with the cortisol level in these animals taken before LPS injection. In three meloxicam treated individuals, the reduction of cortisol secretion was significant (P < 0.001) even compared to the control animals (Fig. 2). In the animals under immune stress prolactin release was not affected. However, it is worth noting that significant (P < 0.001) differences in the prolactin level were determined among meloxicam ' treated groups (Fig. 2). It should be stated that no differences in release of all analyzed hormones were found between the control and meloxicam control groups.
Effect of meloxicam and repeated LPS treatment administration on GnRH synthesis in the POA
LPS-treatment decreased (P < 0.05) the concentration of GnRH compared to both control groups. Meloxicam treatment restored the GnRH level to control value in three individuals which also showed raising the level of LH. On the other hand, no effects of meloxicam treatment were observed in three other LPS-treated ewes (Fig.3).
Effect of meloxicam and LPS on GnRH gene expression in the hypothalamus
Lasting 7 days LPS treatment reduced the gene expression of GnRH in all analyzed hypothalamic structures. The effect of meloxicam treatment on GnRH gene expression in the hypothalamic area was differentiated. Meloxicam treatment significantly (P < 0.05) abolished the suppressory effect of LPS-induced inflammation on GnRH gene expression in the POA. On the other hand, no effect of meloxicam on GnRH mRNA level was found in the MBH, AHA and ME (Table 2).
Effect of meloxicam and LPS on COX-1 and COX-2 gene expression in the hypothalamus
LPS-treatment strongly (P < 0.05) increased COX-2 gene expression in all analyzed hypothalamic structures. Meloxicam injection lowered the level of COX-2 mRNA in the POA, AHA, MBH and ME in LPS treated animals. No effects of LPS and meloxicam on COX-1 gene expression were found in all analyzed hypothalamic structures (Table 3).
Effect of meloxicam and LPS on LH??, FSH??, GnRH-R and prolactin gene expression in the anterior pituitary
It was found that meloxicam reduced (P < 0.05) inhibitory effect of prolonged LPS treatment on LH?? and GnRH-R gene expression in the anterior pituitaries collected from the same individuals which were characterized by increased LH release. No changes were observed in the FSH?? transcription. Although, no effect of LPS on prolactin gene expression was determined in the AP in most ewes, three individuals who did not show changes in LH release after meloxicam treatment were characterized by significantly (P < 0.05) higher prolactin gene expression comparing with the other animals (Table 4). It is worth mentioning that no differences in these genes expression were found between the control and meloxicam control groups.
The results of our study showed that repeated LPS treatment strongly suppressed the activity of hypothalamic-pituitary unit in anestrous ewes. Lasting 7 days LPS-treatment, reduced GnRH synthesis in the hypothalamus at the transcriptional and post-transcriptional levels as well as LH?? gene expression and release from the AP. However, the FSH secretion remained unaffected. These effects of prolonged LPS treatment are similar to the results of our previous studies which showed that even single LPS injection significantly inhibits the hypothalamic-pituitary-gonadal (HPG) axis at the level of the CNS (Herman et al. 2010; Herman and Tomaszewska-Zaremba, 2010). This also supports the previous report about suppressive effect of bacterial endotoxin on the LH secretion in sheep. Coleman et al. (1993) showed that intravenous LPS administration reduced the plasma concentration of LH and number of LH pulses but did not affect pulse amplitude in the castrated ram. In ewes the effect of LPS treatment seems to be dependent on the level of ovarian steroids. Endotoxin significantly delayed the time an experimentally induced LH surge in ovariectomized ewes, but did not alter the amplitude, duration, or prevent the occurrence of the LH surge (Battaglia et al. 1999). This delaying effect of LPS on the LH surge was dependent upon the moment when endotoxin was introduced in relation to the onset of the oestradiol signal. When endotoxin was administered early in the initial period of oestrogen sensitivity, it completely blocked the LH surge in most ewes, but when endotoxin was administered after the period of oestrogen sensitivity, the LH remained unaffected (Battaglia et al. 1999). Peripheral administration of endotoxin invariably interrupted the preovulatory oestradiol rise and delayed or blocked the LH surges in the cycling ewes. Endotoxin inhibition of the preovulatory rise in oestradiol appeared to be influenced partly by the suppression of LH pulsatility (Battaglia et al. 2000).
Although, both single and repeated endotoxin treatment suppress the HPG axis activity, the mechanisms through which those pathological states affect the reproduction process appeared to differ. In the inhibition of GnRH/LH secretion occurring during acute inflammation stress and activation of the hypothalamic-pituitary-adrenal axis play an important role. The studies carried out on sheep showed that LPS injection strongly stimulates cortisol release (Debus et al. 2002; Herman et al. 2010). Cortisol is known as a potent inhibitor of the reproductive process in sheep, especially reducing the LH release (Debus et al. 2002). However, the suppressive effect of cortisol on LH release is dependent on the reproductive status of ewes. The ovarian steroids, particularly estradiol, enable the cortisol suppression of LH pulse frequency in sheep. On the other hand, cortisol-dependent inhibition of LH release is minimal in the ovariectomized ewes devoid of gonadal steroids (Oakley et al. 2009a, 2009b). Moreover, the other HPA axis components, such as corticotropin-releasing hormone and arginine vasopressin were also described as inhibitors of the GnRH/LH secretion in sheep (Battaglia et al. 1998). The results of our study showed that repeated LPS treatment did not result in the activation of the HPA axis, in contrast to immune stress induced by single LPS injection. Therefore, the role of cortisol in the suppression of the HPG axis during prolonged endotoxin treatment seems to be marginal. Despite this fact, it should be noted that not only the lack of cortisol stimulation after LPS injection, but a decrease in the circulating level of this hormone was determined in animals treated with endotoxin. Moreover, in three meloxicam treated individuals this reduction in cortisol release was even significant to the intact animals. In case of meloxicam treated ewes, this lowering of cortisol release could be explained by analgesic action of the drug. However, it is more possible that observed correlations result from animal anxiety in the period before administration of the endotoxin. It is indicated by slightly and not statistically higher levels of cortisol in these animals at the beginning of the experience.
Our results showed that prolonged LPS treatment increased COX-2 gene expression in all analyzed hypothalamic structures when COX-1 transcription stayed unaffected. We also determined that meloxicam treatment did not affected COX-1 transcription in the hypothalamus. On the other hand, meloxicam reduced COX-2 gene expression in all analyzed hypothalamic structures. However, the level of COX-2 gene expression in the POA, AHA and MBH was still significantly elevated compared to its control level. Only in the ME, meloxicam generally restored COX-2 transcription to the control level. The varied effects of meloxicam on COX-1 and COX-2 gene expression did not surprise because meloxicam is known as a selective COX-2 inhibitor. Cruz et al. (2008) showed that meloxicam inhibited not only COX-2 activity and prostaglandin E2 (PGE2) synthesis but also COX-2 gene expression. Generally, it has been accepted that COX-1 is a housekeeping enzyme responsible for modulating physiological events and is present in most tissues including stomach, kidney and platelets, whereas COX-2 is highly induced in various cells by pro-inflammatory stimuli, mitogens and cytokines (Vane and Botting, 2001). However, recent investigations have demonstrated that the roles of COX-1 and COX-2 are oversimplified. It was showed that COX-2 is present under nonpathological conditions in tissues such as the brain and spinal cord, playing an important role in the maintenance of physiological homeostasis (Martinez et al. 2002). Previous findings have also shown that selective inhibition of COX-2 only partially reduces the level of PGs at the site of inflammation in comparison with nonselective COX inhibitors, which reduce PGs to undetectable levels (Anderson et al. 1996). This suggests that COX-1 may also contribute to the pool of PG at the site of inflammation. However, COX-2 seems to be dominant but not an exclusive source of PG formation during inflammation (Huntjens et al. 2005). Determined in our study reduction of COX-2 gene expression in the hypothalamus suggests that meloxicam decreased inflammatory induced PGs synthesis in the hypothalamic area. This could have a profound effect on the suppression of the HPG axis, because centrally acting prostaglandins are considered as potent down regulators of GnRH/LH secretion (Harris et al. 2000).
We also proved that the suppressory effect of repeated LPS treatment on the hypothalamic-pituitary unit could be abolished by a single meloxicam injection. In all ewes, the pharmacological blockade of COX-2 by meloxicam effectively reversed the negative effect of an immune/inflammatory challenges on GnRH transcription in the POA, which is a key structure for GnRH synthesis because more than half of GnRH neurons population have their pericarions located in this structure (Caldani et al. 1988). These results support the previous study which showed that flurbiprofen, a non-selective COX inhibitor, reverses the suppressory effect of endotoxin. This suggests that PGs mediate the suppressive effects of this immune/inflammatory challenge on pulsatile GnRH and LH secretion in ewes (Harris et al. 2000). Moreover, the results of the study carried out on rat also showed that pro-inflammatory cytokines such as IL-1?? affect the GnRH secretion mainly indirectly via PGs-dependent pathway and suggested the primary role of PGs in the inhibition of the hypothalamic-pituitary unit during inflammation (Rivest and Rivier, 1993). On the other hand, Breen et al. (Breen et al. 2004) provided scientific evidences undermining the pivotal role of PGs in the endotoxin-induced suppression of pulsatile GnRH/LH secretion in ewes. They showed that, although flurbiprofen abolished endotoxin-induced fever, which is a centrally generated, prostaglandin-mediated response, it failed to reverse blockade of the LH surge. This indicates that endotoxin blocks the LH surge centrally, suppressing GnRH secretion via a mechanism not requiring PGs. It should be also noted that in our work some results concerning the effect of meloxicam treatment on LH secretion are not conclusive. Unblocking of GnRH transcription in the POA after meloxicam treatment did not simply result in incensement of GnRH synthesis and LH secretion in all ewes. In three animals, meloxicam restored GnRH synthesis and LH secretion to the control level; however in three other individuals no effect of meloxicam treatment on GnRH and LH secretion was observed. These ambiguous effects may simply result from limited time of the experiment but also from different level of prolactin in these two groups of animals. The group of individuals which did not react on meloxicam treatment was characterized by higher prolactin secretion. It is worth mentioning that this group of ewes did not exhibit elevated level of prolactin at the day 1 of the experiment, before first LPS injection (data not showed). The different effect of LPS induced inflammation on prolactin secretion in some individuals could result from different adaptation of the animals to endotoxin. The high level of prolactin in the blood is usually associated with the acute inflammation (Bondanelli et al. 2008; Herman et al. 2010). However, the elevated level of prolactin is not obvious marker of inflammation because during chronic inflammatory state the secretion of this hormone is rather unregulated (Bondanelli et al. 2008; Scotland et al. 2011). This suggests that three ewes characterized by the high level of the prolactin did not develop adaptive immune system as observed in chronic inflammation. This different character of inflammation ongoing in the ewes with low and high concentration of prolactin in the blood could at least partially explain different effectiveness of meloxicam treatment in these two groups of LPS-stimulated animals. It suggests that meloxicam is more effective in restoring the proper GnRH/LH secretion during chronic but not acute inflammation. The high concentration of circulating prolactin may directly influence on the reproductive process which at least partially could explain diverse effect of meloxicam treatment found in our study. Prolactin may affect GnRH secretion in the hypothalamus as well as the pituitary responsiveness to GnRH stimulation. During physiological states associated with elevated prolactin blood concentration (i.e., pregnancy, pseudopregnancy, postpartum and lactation), the LH concentration is always decreased (Milenkovi?? et al. 1994). There are scientific evidences suggesting that prolactin suppresses the LH release via inhibition of GnRH secretion in the hypothalamus. The results of in vitro study carried out on the immortalized GnRH neurons showed that prolactin acting via its corresponding receptors inhibits GnRH release and possibly GnRH gene expression in these cells (Milenkovi?? et al. 1994). The study performed on mice showed the evidence on involvement of prolactin in the regulation of GnRH neurons activity. Grattan et al. (2007) described that 7 days of intracerebroventricular prolactin administration potently suppressed serum LH levels in ovariectomized, estrogen-treated mice. They proved that the ability of central prolactin to reduce LH secretion was mediated via prolactin receptors which transcript expression was determined directly in GnRH neurons. The thesis about the importance of hypothalamic action of prolactin supports the reports showing that circulating prolactin crosses the blood-cerebrospinal fluid barrier thanks to the existence of a prolactin receptor-mediated mechanism for its transport in the choroid plexus (Walsh et al. 1987, 1990). Prolactin could also directly modulate the pituitary responsiveness to GnRH stimulation. The potency of prolactin to suppress the LH secretion was described in ex vivo study carried out on pituitary explants from ovariectomized rats which showed that in anterior pituitaries exposed to elevated levels of prolactin, LH secretion and pituitary responsiveness to GnRH could be impaired (Cheung, 1983). It was also found that artificially induced hyperprolactinemia decreased the number of LH plaque-forming cells in ovariectomized rats (Sortino and Wise, 1989). Moreover, the in vivo and in vitro studies showed that in hyperprolactinemic rats the GnRH induced LH response is reduced (Smith, 1978, 1982). This reduced reaction of pituitaries from hyperprolactinemic rats for GnRH stimulation could be a result of expression of GnRHR in this tissue. The studies carried out on male rats showed that chronic prolactin administration suppressed the pulsatile LH secretion, pituitary GnRH receptor content and pituitary responsiveness to GnRH (Fox et al. 1987). This supports our results which showed that the group of meloxicam treated animals which did not show increasing LH secretion was characterized by decreased GnRHR gene expression, and this reduction was even significantly stronger than observed in a group of LPS treated animals.
In conclusion, the results of the present study show that the LPS-induced decrease in GnRH and LH secretion may be reduced by the COX-2 inhibitors, possibly resulting from the suppression of PGs synthesis in the hypothalamic area and attenuation of the stress response. However, the effectiveness of these inhibitors may be reduced during inflammatory states characterized by the high level of circulating prolactin.