Introduction

Pituitary adenylate cyclase-activating polypeptide (PACAP) is a multifaceted neuroendocrinological mediator, which was originally isolated from the hypothalamus1,2, but it turned out that it is produced by a variety of tissues and cell types, including neurons and immune cells3. The role of PACAP is established in different homeostasis processes such as nociception, inflammation, energy balance, and thermoregulation4,5,6,7.

Systemic inflammation is often induced in experimental models by the administration of bacterial lipopolysaccharide (LPS). High doses of LPS (several mg/kg) cause serious, often lethal systemic inflammation, which is typically accompanied by a decrease in deep body temperature (Tb)8. In such severe forms of systemic inflammation, PACAP exerted a protective effect in mice and dogs9,10. In addition, the protective role of PACAP signaling was also demonstrated in LPS-induced shock with mice genetically lacking the specific PACAP receptor PAC111. An interaction between PACAP signaling and pro-inflammatory cytokines, such as interleukin (IL)-6 and tumor necrosis factor (TNF)-α, was also shown in such experimental models9,11,12.

In contrast to severe systemic inflammation, mild-to-moderate systemic inflammatory response can be triggered by lower doses of LPS (often in the µg/kg range), which is characterized by fever. The febrile response is mediated by pyrogenic cytokines (IL-1, -6, and TNF-α) and by the activation of the arachidonic acid (AA) pathway, wherein cyclooxygenase (COX) enzyme activity leads to production of prostaglandin E2 (PGE2)13,14. In the median preoptic nucleus (MnPO), PGE2 activates thermoregulatory neurons, which leads to the rise of deep Tb via increased thermogenesis and skin vasoconstriction8. In addition to this well-established molecular pathway, neuroendocrinological mediators, for example, substance P and cholecystokinin signaling, have also been identified as contributors to the fever response15,16. Interestingly, the intracerebral administration of PACAP to rats resulted in increased oxygen consumption (thermogenesis) and skin vasoconstriction in rats17, which is similar to the thermoeffector pattern of febrigenesis. Moreover, the PACAP-induced increase in Tb could be attenuated by COX enzyme inhibition18, indicating an interaction between PACAP signaling and the COX pathway. However, the potential role of PACAP in the development of the LPS-induced fever response has not been investigated despite its complex function in homeostasis, which entails both pro- and anti-inflammatory roles11,19,20,21.

In the present work, we studied how the genetic ablation of PACAP influences the LPS-induced early and later phases of the fever response by comparing mice, which had the Pacap gene homozygously either present (Pacap+/+) or absent (Pacap-/-) due to a targeted disruption22. In thermophysiological experiments, we recorded changes in deep Tb in response to LPS. To identify the involved molecular mechanisms, we measured serum cytokine levels, as well as tissue COX-2 mRNA expression in this animal model.

Results

LPS-induced changes in the thermoregulatory response of Pacap +/+ and Pacap -/- mice

To compare the fever response between Pacap+/+ and Pacap-/- mice, we administered LPS [120 µg/kg, intraperitoneally (i.p.)] or saline to the mice of both genotypes. The infusion of saline did not cause any effect on deep Tb in either genotype, as expected based on our previous study16. On the contrary, LPS-treated mice developed fever as compared to their saline-treated counterparts (Fig. 1). In Pacap+/+ mice LPS caused a characteristic fever response: their deep Tb started to increase at 40 min, plateaued (~ 39.1 °C) between 100 and 170 min postinfusion, then it gradually decreased, but remained elevated compared to saline treatment throughout the experiment (p < 0.05 at 50–360 min). These findings are in agreement with those reported in genetically unmodified mice in previous studies16,23. However, in Pacap-/- mice, the LPS-induced fever response was less pronounced than in their Pacap+/+ littermates: their Tb reached plateau at 60 min (~ 38.9 °C), but started to decrease earlier (at 140 min) and returned to the level of saline-treated mice already at 200 min post-LPS administration (Fig. 1). Statistically, the intergenotype difference between the LPS-treated groups was significant at 160–250 min and 290–340 min postinfusion (p < 0.05). Importantly, the Tb of the LPS-treated Pacap-/- mice was markedly (0.5–0.8 °C) lower than that of Pacap+/+ mice starting from 160 min post-LPS infusion until the end of the experiment (p < 0.05).

Fig. 1
figure 1

Changes in deep (colonic) Tb in response to i.p. administration of LPS (120 µg/kg) or saline in Pacap+/+ and Pacap-/- mice. *p < 0.05, **p < 0.01 intergenotype difference in LPS-treated mice; #p < 0.05, ##p < 0.01 difference between treatments in Pacap-/- mice; $p < 0.05, $$p < 0.01, $$$p < 0.001 difference between treatments in Pacap+/+ mice.

Full size image

LPS-induced changes in COX-2 mRNA expression in the lungs, liver, and brain of Pacap +/+ and Pacap -/- mice

Next, we compared the LPS-induced COX-2 expression in peripheral LPS-processing organs (i.e., lungs and liver) and in the brain between Pacap+/+ and Pacap-/- mice. For that, lung, liver, and brain samples were collected at 210 min post-LPS infusion, i.e., at the time point at which the biggest difference in Tb was found between the genotypes (Fig. 1). In Pacap+/+ mice, LPS induced a significant increase in COX-2 expression in the lungs (Fig. 2A), liver (Fig. 2B), and brain (Fig. 2C), consistent with findings from previous studies in genetically unmodified rodents16,24,25. In Pacap-/- mice, the LPS-induced increase in COX-2 expression was also significant compared to saline treatment in the lungs and the brain, but not in the liver (Fig. 2). Importantly, we found a significant difference in COX-2 mRNA expression between LPS-treated Pacap+/+ and Pacap-/- mice: in Pacap-/- mice, the LPS-induced COX-2 mRNA expression was significantly reduced in the liver (Fig. 2B), whereas it was more pronouncedly elevated in the brain compared to Pacap+/+ mice (Fig. 2C).

Fig. 2
figure 2

Relative COX-2 gene expression in the lungs (A), liver (B), and brain (C) of Pacap+/+ and Pacap-/- mice after infusion of LPS (120 µg/kg) or saline. Tissue samples were collected at 210 min postinfusion. Number of animals in the corresponding groups are indicated in the figure. Significant differences are marked as *p < 0.05, **p < 0.01, ***p < 0.001.

Full size image

LPS-induced changes in serum cytokine levels of Pacap +/+ and Pacap -/- mice

The serum concentrations of the pyrogenic cytokines IL-1α and -β were significantly increased in Pacap+/+ mice in response to LPS compared to their saline-treated counterparts (p < 0.001 for both cytokines), whereas in Pacap-/- mice LPS did not cause a significant change in their levels compared to saline treatment (Fig. 3A and B). The intergenotype difference between the LPS-treated groups was also significant for both IL-1α and -β (p < 0.05 for both cytokines). In cases of IL-6 and TNF-α, the administration of LPS resulted in a marked (p < 0.001) rise in both genotypes, while saline had no meaningful effect (Fig. 3C and D). The levels of the anti-inflammatory IL-10 were higher in LPS-treated than in saline-treated mice in both genotypes (p < 0.001), however the rise was significantly attenuated in Pacap-/- mice compared to their Pacap+/+ littermates (p < 0.05) (Fig. 3E).

Fig. 3
figure 3

Serum cytokine concentrations in Pacap+/+ and Pacap-/- mice. Serum IL-1α (A), IL-1β (B), TNF-α (C), IL-6 (D), and IL-10 (E) concentrations in Pacap+/+ and Pacap-/- mice in response to LPS (120 µg/kg) or saline. Blood samples were collected at 210 min postinfusion. Number of animals in the corresponding groups are indicated in the figure. Significant differences are marked as *p < 0.05, ***p < 0.001.

Full size image

Discussion

The present study provides novel evidence demonstrating that the fever response to bacterial endotoxin is significantly attenuated in Pacap-/- mice. We also found that the LPS-induced COX-2 mRNA upregulation was reduced in the liver, whereas increased in the brain of Pacap-/- mice compared to their Pacap+/+ littermates. Furthermore, the blood levels of the pyrogenic cytokines IL-1α and -β, as well as the level of the anti-inflammatory cytokine IL-10 were decreased, while levels of the pro-inflammatory IL-6 and TNF-α tended to be lower in the absence of PACAP in response to LPS.

Systemic inflammation involves a balance between pro- and anti-inflammatory processes26: the systemic inflammatory response (SIR) is counteracted by a compensatory anti-inflammatory (CAR) response27. Systemic inflammation is often associated with changes in deep Tb, most commonly with fever that is a thermoregulatory response of the host organism to facilitate the elimination of the pathogens8. In experimental models, systemic inflammation-associated fever is often induced with the administration of the bacterial endotoxin LPS.

PGE2 production in peripheral organs such as the lungs and liver contribute to the initial phases of LPS-induced fever, while later phases are maintained by production of PGE2 within the brain8,28,29. It should be also mentioned that by using genetically modified mice, Shionoya et al.30 showed that PGE2 synthesis in brain endothelial cells is responsible for all phases of fever, which challenges the importance of peripheral organs in febrigenesis. In the early phase, LPS via binding to toll-like receptor 4 (TLR4), activates the AA pathway involving the upregulation of COX-2, as well as the increased production of pro-inflammatory (pyrogenic) cytokines such as IL-1, -6, and TNF-α via nuclear translocation of nuclear factor kappa B (NF-κB). Within the brain, PGE2 binds to prostaglandin E type 3 (EP3) receptor expressed on thermoregulatory neurons in the MnPO of the hypothalamus, which in turn leads to the elevation of Tb8,29,31. While the aforementioned canonical mechanisms of fever are well established (for reviews, see8,29,32), novel mediators, which can influence the activities of the classical pathways are often discovered15,16,33,34. A schematic of the described mechanisms of febrigenesis, also including how they can be influenced by PACAP based on the findings of the current study, are presented in Fig. 4.

Fig. 4
figure 4

Schematic representation of febrigenesis and the potential influence of PACAP. For explanation, see text. POA, preoptic area of the hypothalamus.

Full size image

It has been demonstrated by several research groups that PACAP plays an important role in thermoregulation and contributes to the defense of energy homeostasis (for reviews, see4,35). It was also suggested that PACAP contributes to the LPS-induced systemic inflammation in dogs10, mice9,11,21, and rats12. However, it has remained unclear until now, whether and how PACAP may play a role in the development of the LPS-induced elevation of deep Tb. Here, we showed that LPS-induced fever was attenuated in the absence of PACAP starting from ~ 120 min post-LPS infusion, which clearly indicates that PACAP is required for the febrile response. This finding is in harmony with our previous results about the fever-like thermoregulatory response to PACAP administration in rats and the decreased basal Tb and metabolic rate in Pacap-/- mice presumably caused by enhanced suppression of thermogenesis by neurons in the medial preoptic area17.

To explore the molecular mechanism in the background of the attenuated fever response in the absence of PACAP, we studied the expression of COX-2 in peripheral LPS-processing organs (i.e., lungs and liver) and in the brain. We showed that in the absence of PACAP the LPS-induced upregulation of COX-2 was attenuated in the liver, whereas it was augmented in the brain. The reduced hepatic COX-2 upregulation can well explain the attenuated fever response to LPS in the absence of PACAP (see Fig. 4). In our previous study, we found a similarly reduced fever and attenuated COX-2 upregulation in peripheral organs in response to LPS in the genetic absence of the neurokinin-1 receptor16, which is a key receptor in substance P signaling36. Interestingly, a close interaction between PACAP and substance P pathways in inflammatory processes has been demonstrated in multiple studies37,38,39. Hence, it is plausible to assume that PACAP and substance P interact within the pyrogenic pathways of febrigenesis. It should be also mentioned that PACAP was recently shown to stimulate neuroendocrine and behavioral stress responses through corticotropin-releasing factor-containing neurons in the hypothalamus40. From human studies and animal experiments it is also known that LPS stimulates the hypothalamic–pituitary–adrenal axis, which is thought to be an important adaptive immunoregulatory response41. Since corticotropin-releasing factor plays an important role in the regulation of COX-2 under inflammatory conditions42, it may be assumed that an interaction between PACAP and the hypothalamic–pituitary–adrenal axis could have also contributed to our findings.

The augmented COX-2 upregulation induced by LPS in the brain of the Pacap-/- mice was an unexpected finding. Different studies indicated that PACAP receptors are widely expressed throughout the cranial vasculature (for review, see43). In the central nervous system, the major source of PGE2 is the brain endothelium8,29,32. It was shown that PACAP inhibits the COX pathway of rat cerebral microvessels44, which may present a direct interaction between PACAP and COX-2 in brain endothelial cells to explain the higher LPS-induced COX-2 expression in the Pacap-/- mice observed in our study. As an alternative explanation, microglial cells are among the main source of inflammatory mediators in the brain. It has been suggested that PACAP may exert its effects in part by acting directly on microglial cells as a microglia-deactivating factor45,46, which may lead to enhanced production of pro-inflammatory mediators, including PGE2, in the absence of PACAP. Since increased COX-2 expression is typically associated with higher PGE2 levels, a more pronounced febrile response could be expected. It is possible, however, that in the absence of PACAP downstream mediators in the AA cascade do not function properly, which could explain a compensatory, yet still ineffective, overexpression of COX-2 in the brain. In support of this scenario, it was shown that the majority of EP3 receptor-expressing neurons in the MnPO, which play a key role in the mediation of LPS-induced fever, express PACAP47. However, in that study it remained unanswered whether the release of PACAP is necessary to induce LPS fever. Based on our findings, it is plausible that the malfunction of these neurons in the absence of PACAP could contribute to the attenuated fever response even when COX-2 expression was higher. However, the discovery of the exact mechanism of the PACAP-COX-2 interaction in the brain was beyond the scope of the present study, thus remains subject of future research.

Pro-inflammatory cytokines, like IL-1α, -β, IL-6, and TNF-α, are crucial for triggering the immune response against pathogens. Nevertheless, anti-inflammatory cytokines, such as IL-10, also have an important role in limiting inflammation and in preventing tissue damage of the host48,49. Since pro- and anti-inflammatory cytokines play a pivotal role in the LPS-induced fever signaling pathway, we measured serum concentrations of these cytokines to assess whether the LPS-induced cytokine production is suppressed in the absence of PACAP. We found that in Pacap-/- mice the LPS-induced elevation of IL-1α, -β, and IL-10 were significantly reduced, while that of IL-6 and TNF-α tended to be lower than in control mice. Since IL-1, -6, and TNF-α are well-documented pyrogenic cytokines (for review, see8,32), the smaller increase in their blood levels in response to LPS can explain the attenuation of the fever response in the absence of PACAP. It should be also noted that Mota and Madden50 showed that circulating IL-1β can elicit fever by activating brain circuits in the absence of PGE2 production, which raises the possibility for a role of PACAP in COX-independent febrigenesis.

The decreased levels of IL-10 in response to LPS in the absence of PACAP was an unexpected finding, since IL-10 has an antipyretic role in LPS fever51,52. However, taking into consideration the mitigated elevations of COX-2, pyrogenic cytokine levels, and deep Tb, which are constituents of SIR, the smaller increase in anti-inflammatory mediators as part of the CAR can be expected. Thus, our findings suggest that PACAP is required for the production of the canonical pyrogenic cytokines in fever, but it also contributes to the regulation of the anti-inflammatory response, thereby maintaining the SIR-CAR balance in systemic inflammation. When PACAP is absent, the alteration in the pro- and anti-inflammatory cytokine levels as well as in the activity of AA cascade leads to the dysregulation of the SIR-CAR balance leading to a weakened fever response (Fig. 4).

Limitations of our study should be also mentioned. Fever signaling was examined only at a single time point in the current experiments due to the explorative nature of the study. It would be also interesting to see temporal changes of COX-2 expression and cytokine levels in the Pacap-/- mice, since the kinetics of enzymes and cytokines incorporate time-dependent changes of their synthesis, metabolism, and elimination, which can show variations during systemic inflammation. Our results can serve as an encouraging basis for designing future studies, which aim to discover the kinetics of the inflammatory mediators in the absence of PACAP. Further, the use of constitutive PACAP knockout mice did not allow us to functionally analyze the site(s) of action of PACAP. An elegant method to discover the sites of action would be to use organ or tissue specific modulation of PACAP expression. For example, by using conditional PACAP deletion in their recent study, Bakalar et al.53 could distinguish between neurotransmitter and non-neurotransmitter functions of PACAP, including thermoregulatory. They showed that restriction of PACAP deletion to the hypothalamus prevents the thermoregulatory changes observed in constitutive Pacap-/- mice, however hypothalamic PACAP-expressing neurons did not require PACAP for fasting-induced hypothermia. Similar animal models could also advance our knowledge regarding the specific site of action of PACAP in systemic inflammation-associated fever, but that was beyond the scope of the present study.

In conclusion, the modulation of COX-2 expression, as well as pro- and anti-inflammatory cytokine concentrations, by PACAP is required for the development of the fever response associated with systemic inflammation.

Materials and methods

Animals

Experiments were conducted in 42 Pacap+/+ and Pacap-/- adult mice of both sexes. As in earlier studies17,54,55, the mice were obtained from the Animal Facility of Medical School, University of Pecs. Generation by a gene-targeting technique, maintenance, and backcrossing of PACAP deficient mice on a CD1 background has been described previously22,56. Animals were housed in temperature-controlled rooms on a 12 h light–dark cycle. Standard rodent chow and tap water were available ad libitum. At the time of the experiments, the mice weighed 37.2 ± 0.8 g.

For thermophysiological experiments mice were extensively handled (5 min per day for 8 days) and then habituated to the experimental setup as follows. Each mouse was adapted to staying in a wire-mesh confiner every day for gradually increasing time intervals (8 training sessions, 1–4 h each), as in previous studies57,58. The cylindrical confiner restricted the animal from turning around but permitted limited forward and backward movement. Small animals are readily adaptable to restraint since after habituation they respond to it with neither stress fever59 nor other signs of stress60,61,62.

All procedures were conducted under protocols approved by the Institutional Animal Use and Care Committee of the University of Pecs (registration no.: BA02/2000-6/2018) and were in accordance with the directives of the National Ethical Council for Animal Research and those of the European Communities Council. The study is reported in compliance with ARRIVE guidelines.

Anesthesia and perioperative care

Mice were anesthetized via i.p. injection of a ketamine-xylazine cocktail (81.7 and 9.3 mg/kg, respectively) and received antibiotic prophylaxis with gentamicin (6 mg/kg) intramuscularly. To prevent intra- and postoperative hypothermia, mice were kept on a heating pad (model V500DVstat; PECO Services Ltd., Brough, UK) during the surgery, and then they were allowed to recover from anesthesia in a temperature-controlled chamber (model BJPX-B400II; Biobase; Jinan, China) set to an ambient temperature of 28 °C. Mice were allowed to recover for 4 days before the experiment.

Intraperitoneal catheter implantation

To minimize stress during substance administration in the experiment, a polyethylene (PE)-50 catheter filled with pyrogen-free saline was implanted into the peritoneal cavity of each mouse, similarly as in previous studies57,58. In brief, through a small midline incision on the abdominal wall, the internal end of the catheter was fixed to the left side of the abdominal wall with a suture, while the external end of the catheter was tunneled under the skin to the nape, where it was exteriorized and heat-sealed. The surgical wound was sutured in layers. The catheter was flushed with 0.1 ml of saline on the day after the surgery and every other day thereafter.

Experimental setup

The mice were placed in the thermocouple thermometry setup in cylindrical confiners and equipped with copper-constantan thermocouples (Omega Engineering, Stamford, CT, USA) to measure colonic temperature, a form of Tb. The colonic thermocouple was inserted beyond the anal sphincter (3 cm deep); fixed to the base of the tail with adhesive tape; and plugged into a data logger device (Cole-Palmer, Vernon Hills, IL, USA) connected to a computer. Animals in their confiners were then placed into a temperature-controlled incubator (model MIDI F230S; PL Maschine Ltd., Tarnok, Hungary) or into a biochemistry incubator (model BJPX-Newark; Biobase; Jinan, China), in which the ambient temperature was set to 33 °C, which was thermoneutral for mice in these setups. In restrained mice in our setups, the thermoneutral zone is 31–33 °C16,57, while 30 °C is slightly below the zone63. The pre-implanted i.p. catheter was connected to a PE-50 extension, which was prefilled with the substance of interest and connected to a syringe placed in an infusion pump (model 975; Harvard Apparatus Inc., Holliston, MA, USA). The thermocouple thermometry setup has been extensively used by our group and has allowed us to study the dynamics of the fever response in rats and mice15,16,64.

Substance administration

LPS from Escherichia coli 0111:B4 was purchased from Sigma-Aldrich (St. Louis, MO, USA). A stock solution of LPS (5 mg/ml) in pyrogen-free saline was stored at −20 °C. On the day of the experiment, the stock was diluted to a final concentration of 36 µg/ml. The diluted LPS solution or saline (for control animals) was infused (26 µl/min for 4 min) through the extension of the i.p. catheter to deliver LPS at a final dose of 120 µg/kg. Deep Tb was monitored for 6 h after the infusion. Administration of the substances was carried out between 10:00 a.m. and 11:30 a.m. in the experiments. After the experiment, the animals were euthanized with sodium pentobarbital (100 mg/kg, i.p.).

Molecular biology

Tissue harvesting

On the day of the experiment, each mouse was placed in a confiner and left to acclimate in the incubator for ~ 2 h and then infused with LPS or saline as in the thermophysiological experiments. The time point for tissue harvesting was chosen based on the Tb curves. We collected blood and tissues at the time when we observed the biggest difference in deep Tb between the LPS-treated Pacap-/- and Pacap+/+ mice. At the 210-min time point after infusion, the mice were anesthetized with ketamine-xylazine cocktail through the i.p. catheter. Blood samples were collected from the left ventricle, transferred to an Eppendorf tube, and were allowed to clot for 25 min before centrifugation at 10000 rpm for 10 min, in a refrigerated centrifuge at 4 °C. Serum fractions were collected, pooled, and stored at −80 °C until use in the assay. For collection of lung, liver, and brain tissue samples for RT-qPCR, each mouse was transcardially perfused with 0.1 M phosphate-buffered saline. Samples of the liver and the lung were collected rapidly and snap frozen in liquid nitrogen. Then the entire brain was removed, the hypothalamus was dissected and frozen. All tissue samples were stored at −80 °C.

Measurement of serum IL-1α, IL-1β, IL-6, IL-10, and TNF-α cytokine concentrations

Luminex xMAP technology was used to determine the protein concentrations of IL-1α, IL-1β, IL-6, IL-10, and TNF-α cytokines performing Milliplex Mouse High Sensitivity T Cell Magnetic Bead Panel (catalog number: MHSTCMAG-70K, Merck KGaA, Darmstadt, Germany) according to the manufacturer’s instructions. Briefly, all samples were thawed and tested undiluted in a blind-fashion and in duplicate. All reagents of the kit were brought to room temperature before use. 50 µl volume of each sample, standard, and control was added to a 96-well plate (provided with the kit) containing 25 µl mix of capture antibody coated bead sets, each internally color-coded with fluorescent dyes. Following 16 h of incubation, biotinylated detection antibody and streptavidin-PE were added to the plate after 60 and 30 min incubation steps, respectively. After the last washing step, 150 µl drive fluid was added to the wells, the plate was incubated for an additional 5 min on a shaker and immediately read on the Luminex MAGPIX instrument. Luminex xPonent 4.2 software was used for data acquisition. Five-PL regression curve were generated to plot the standard curves for all analyte by the Belysa v1.1 (Merck Millipore, Darmstadt, Germany) software calculating with bead median fluorescence intensity values. Results are given in pg/ml.

RNA isolation and quantitative real-time polymerase chain reaction (RT-qPCR)

As in our previous studies63,65, total RNA was extracted from the lungs, liver, and whole hypothalamus of treated and untreated Pacap+/+ and Pacap-/- mice. The extraction was performed using TRI Reagent (Molecular Research Center Inc., Cincinnati, OH) in combination with the Direct-Zol RNA Isolation Miniprep Kit (Zymo Research, Irvine, CA), following the manufacturer’s protocols. RNA samples were treated with deoxyribonuclease I (Zymo Research, Irvine, CA) and quantified by Jenway Genova Nano Micro-Volume Spectrophotometer (Thermo Fisher Scientific, Budapest, Hungary). 500 ng purified RNA was reverse transcribed into cDNA using Tetro cDNA Synthesis Kit (Meridian Bioscience, Memphis, Tennessee, USA). Real-time qPCR was conducted with Quantstudio 5 System using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a reference gene. Each reaction contained 20 ng cDNA, 1 × Luminaris HiGreen Low ROX qPCR Master Mix (Thermo Fisher Scientific), and 0.3 µM from each primer. RT-qPCR cycle conditions were as follows: 95 °C for 10 min, followed by 40 cycles of 95 °C for 30 s, 60 °C for 30 s, then 72 °C for 1 min. The following primer pairs were used to amplify the target loci: GAPDH sense: 5’-TTCACCACCATGGAGAAG-3’ and antisense: 5’-GGCATGGACTGTGGTCATGA-3’. COX-2 sense: 5’-GGGTTGCTGGGGGAAGAAA-3’ and antisense: 5’- CTCTGCTCTGGTCAATGGAGG-3’. All reactions were carried out in triplicate, and the mean value of the threshold cycles (Ct) was used for the determination of mRNA expression levels. The relative gene expression ratios were calculated according to the comparative ΔΔCt method using samples of untreated animals as calibrator. Measurements included a dissociation curve analysis to verify the amplification specificity. During the gene expression calculations, primer efficiencies were considered66.

Data processing and analysis

Data on deep Tb, serum cytokine levels, and COX-2 expression were compared by two-way ANOVA, followed by Fisher’s LSD post hoc tests, as in our previous studies17,63. For statistical analysis, Sigmaplot 11.0 (Systat Software, San Jose, CA, USA) software was used. Differences were considered statistically significant when p < 0.05. All data are presented as mean ± SEM.