Changes of c-Fos expression and NADPH-d activity in claustrum induced by chronic muscle inflammation in cat (a preliminary study)

An investigation of the central mechanisms underlying muscle inflammation and musculoskeletal pain is an important step to find means for the prevention or treatment of muscle inflammation. One of the insufficiently studied brain structures involved in the transmission of nociceptive information is the claustrum (CL). Therefore, the aim of the study was to reveal changes in the patterns of Fos-immunoreactivity and NADPHdiaphoreactivity in the nucleus claustrum (CL) and additionally in the ventral putamen (Pu) during chronic inflammation of m. gastrocnemius-soleus in cat, induced by intramuscular injection of complete Freund's adjuvant (CFA). Immunohistochemical and histochemical techniques were used to detect Fos-immunoreactive (Fos-ir) and NADPH-diaphorase reactive (NADPH-dr) neurons within studied structures. It was revealed that nine days after CFA-induced muscle inflammation the level of Fosimmunoreactivity and NADPH-d reactivity within the CL and in the ventral part of Pu increased two-fold in comparison with the control. Because the CL is reciprocally connected with many structures of the brain cortex and subcortical structures, all these structures can be pathways of transmission of nociceptive information, nevertheless, it can be assume that the central amygdala nucleus may make the main nociceptive contribution to the activation of neurons within the CL. It is known that CL is mutually related to Pu, but it was not possible to assess their mutual influence in this study. The results of the study of the Fos-ir neurons distribution in CL and Pu under conditions of long-term muscles inflammation indicate the active involvement of the mentioned structures in the formation of adaptive reactions. The increase in the number of neurons with NADPH-d reactivity in CL and Pu indicates that NO-signals play a significant role in the formation and amplification of the response to painful impulses from inflamed muscles. In addition, further research is needed to accurately identify all possible nociceptive inputs to the CL and to separate the emotional, motor, auditory, and visual components that may accompany nociceptive processes.


Introduction
It is well known that inflammatory muscle diseases lead to a significant decrease in a human's life quality [31]. Musculoskeletal pain or myalgia is the result of muscle inflammation, which leads to a significant modulation of motor activity [18,20,30]. In this regard, it is a considerable interest to study the effect of inflammatory processes (occurring in the skeletal muscles) on the brain structures. To study central mechanisms underlying musculoskeletal pain, electrical, mechanical, or chemical excitation of highthreshold group III and IV muscle afferents has been widely explored in tests on humans and animals. Intramuscularly injected agents, such as capsaicin, carrageenan, potassium chloride, hypertonic saline, or others, were used to induce muscle pain and, thereby, to excite thin muscle afferents [30]. Peripheral and central mechanisms underlying chronic pain are still poorly understood; this type of pain is difficult to ameliorate. The induction of persistent inflammatory pain by intramuscular (i.m.) injections of complete Freund's adjuvant (CFA) could be a convenient approach in examination of brain structures involved in muscle nociception [1,30,33].
One of the understudied structures of the brain is the claustrum. It is known that claustrum receives input from almost all regions of cortex and projects back to almost all regions of cortex. Although there are enough studies of this structure, including studies on human under pain conditions, nevertheless, the function of the claustrum remains unclear [4]. Morphological investigations are one of the appropriate ways to study inflammatory pain processes at the CNS level. The best known morphological techniques are the detection of c-Fos protein in the nuclei of activated neurons and detection of nerve cells containing nitric oxide synthase.
The expression of immediate early genes, especially c-fos, is widely used for mapping functional pathways, including the circuitry that underlies the transmission of nociceptive information in the CNS. The proto-oncogene c-fos is expressed in neurons in response to various stimuli, and its c-Fos-protein product can be easily detected with immunohistochemical techniques [12,23]. Therefore, the immunohistochemical method of c-fos expression (as a marker of neuronal activation) to determine which neural elements are activated. Another method of investigation is the identification of nitric oxide synthase (NO-synthase or NOS-containing neurons) within the spinal and brain structures. The neuronal variant NADPH-diaphorase is known to correspond to a form of nitric oxide synthase [28,32]. NO-synthase acts as a marker of late changes in gene expression and effects on various physiological and pathophysiological functions of the central and peripheral nervous system [7,14].
Thus, the aim of the study was to reveal changes in patterns of Fos-immunoreactivity and NADPH-diaphorase reactivity within the nucleus claustrum (CL) and additionally in the ventral part of Putamen (Pu) under condition of chronic inflammation of the m. gastrocnemius-soleus (GS) (which is most frequently involved in daily life) induced by i.m. injection of CFA.

Materials and methods
Cats weighing 2.4 to 3.8 kg were randomly divided into 2 groups. 1st group -control intact animals (n=4). 2nd groups -adjuvant-injected cats (n=4). The animals were purchased from a state-controlled animal farm through the common animal facility of the Bogomolets Institute of Physiology (Kyiv, Ukraine). The experimental animals were kept in an air-filtered, humidity -(55±5 %) and temperaturecontrolled (20-22 °C) room with filtered air. The present study was approved by the Bioethics Committee of the Institute and performed in accordance with the European Union Directive of 22 September 2010 (2010/63/EU) for the protection of animals used for scientific purposes.
Inflammation was induced by infiltrating the GS muscles with 1 ml of the CFA modified with Mycobacterium butyricum (Calbiochem, USA) dissolved in 1 ml of Ringer's solution. Injections of CFA solution were performed under full inhalation anesthesia into the left GS muscles (three injections of 0.2 ml into each head). Cats were placed in cages with soft bedding and allowed to recover for 9 days [30].
Nine days after the induction of inflammation of GS muscles, all animals were deeply anesthetized with sodium pentobarbital (75 mg/kg, i.p., Sigma, USA) and perfused through the ascending aorta with a 0.9 % physiological saline (500 ml) followed by a fixative solution (1500 ml) containing 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Samples of the lumbar and sacral spinal cord were quickly removed; blocks were postfixed in the same fixative overnight and cryoprotected in phosphate-buffered sucrose at 4 °C for 48 h. Part of the brain was cut into frontal frozen 40-µm-thick sections. Brain sections were collected in wells with cold phosphate-buffered saline (0.01 M phosphate buffer containing 0.9 % NaCl), to be processed immunohistochemically and histochemically. The Fos immunoreactivity was detected according to a standard avidin-biotin-peroxidase technique [16] using a rabbit polyclonal antibody against c-Fos protein 1:2000 (Ab-5, Oncogene Research, USA) and a commercial kit 1:200 (ABC Kit, PK 4001, Vector Laboratories, USA). Fosimmunoreactive (Fos-ir) neurons were observed as units with black nuclei due to nickel-intensified 3,3'diaminobenzidine tetrahydrochloride (Sigma, USA), staining. To evaluate the possibility of double labeling of Fos immunoreactivity and NADPH-diaphorase reactivity in neurons, half of the immunostained sections were additionally incubated in 0.1 M phosphate buffer (pH 7.4) containing 0.3 % Triton X-100, 0.2 mg/ml nitroblue tetrazolium (Sigma, USA), and 0.5 mg/ml β-NADPH tetrasodium salt (Sigma, USA) at 37 °C for 30 to 60 min [32]. NADPH-diaphorase reactive (NADPH-dr) neurons were detected as light-blue cells with unstained nuclei. The location of Fos-ir and NADPH-dr cells were controlled according to the stereotaxic atlas [21].
The density of Fos-ir neurons and NADPH-dr cells were counted bilaterally in the cat brain sections (in all animals) at the levels A+11/12 (within 400 x 400 µm 2 areas) and analyzed. Only high intensely stained NADPH-dr neurons were used for statistical analysis. The Shapiro-Wilk test was used to test the normality of the data distribution. The obtained values are expressed as means ±SEM. Intergroup differences were evaluated by two-way ANOVA followed by the Bonferroni's multiple comparison test. Differences of the mean values with p<0.05 were considered as significant. Data analysis was performed using software "Origin 8.5" (OriginLab Corporation, USA).

Results
In the intact cats, the mean density of Fos-ir neurons within CL were 6.8±0.8 and 7.2±1.3 (on the left and right side of the section, respectively). Additionally, the labeled neurons were counted in adjacent to CL ventral part of Pu. The density such cells were 13.6±2.1 and 10.2±3.3 units, respectively. Note, difference between amount of activated neurons on left and right sides within these nuclei was not statistically significant (p>0.05) (Figs. 1 and 2). In comparison with control animals, increase in the mean density of Fos-ir cells was registered in CL and Pu in CFAinjected cats. Maximal density of Fos-ir neurons was recorded within CL and Pu on contralateral side of the brain section (in relation to the CFA-injection side, p<0.05). Therefore, mean density of labelled cells were 16.1±1.1 and 20.5±1.6 in CL, and 30.3±4.5 and 42.3±5.8 in ventral part of Pu (on the ipsi-and contralateral side, respectively) ( Figs. 1 and 2).
NADPH-d reactivity within CL and Pu were also registered. In the control animals, mean density within these nuclei on left and right sides were 5.1±0.8 and 4.2±2.1 in CL, and 10.8±0.7 and 12.8±1.9 in Pu. Values obtained in CL and Pu were not statistically significant between left and right section side for each nucleus (p>0.05) (Figs. 1  and 3). Compared with the control, the level of NADPH-d reactivity was significantly higher (p<0.05) in the CFA-injected cats. Thus, mean density within studied nuclei on ipsi-and contralateral side were 8.4±0.9 and 9.7±0.8 in CL, and 11.9±0.7 and 20.5±1.5 in Pu, respectively (Figs. 1  and 3).
It should be noted that in comparison with the control, the average density of labeled Fos-ir as well as NADPH-dr neurons in the contralateral CL and Pu increased by twofold. Also, note that the Fos-ir and NADPH-dr neurons were intermingled within studied nuclei, but double-labeled cells were not revealed.

Discussion
The results of the study demonstrate that 9 days after CFA-induced muscle inflammation the level of Fosimmunoreactivity and NADPH-diaphorase reactivity within the CL and in the ventral part of Pu increased two-fold  It is known that changes in neuronal activity within the CNS structures are mainly initiated by the inflow of nociceptive signals from high-threshold muscle afferents [5]. Moreover, nociceptive information from inflamed muscles can be transmitted not only through the spinothalamic tract, but also through direct and indirect pathways to the hypothalamus and limbic structures [8,9,22]. Also it was shown that intraplantar application of CFA induced expression of microglial and astrocytic markers and proinflammatory cytokines IL1b and IL-6 (involved in inflammation and immune reaction) at the spinal level, brainstem, and forebrain (cerebral hemispheres, limbic system, thalamus, hypothalamus, and corpus callosum) [24].
According to the H. Sherk and K. Shima et al. [26,27] the claustrum is a telencephalic structure characterized as a subcortical nucleus connected reciprocally with the cerebral cortex. The connections between the CL and the neocortex are well organized. Areas of cortex within one lobe that are interconnected with cortico-cortical connections share afferent and efferent projection zones in the CL, suggesting their interactions by way of convergent inputs. In addition, K. Shima et al. showed that neural activity in the supplementary motor area and precentral motor cortex both projected to the CL [27]. Numerous neuroanatomical studies in the cat and monkey have revealed widespread connections between the claustrum and most allocortical and neocortical regions. Thus, the CL is interconnected with the frontal lobe including motor cortex, prefrontal cortex and cingulate cortex, visual cortical regions in the occipital lobe, temporal and temporopolar cortices, parietooccipital and posterior parietal cortex, the frontoparietal operculum, somatosensory areas, prepiriform olfactory cortex and the entorhinal cortex. The CL is also connected to the hippocampus, amygdala, and caudate nucleus. [4]. In addition, significant influence on neuronal activation under nociceptive stimulation is pain anticipation (medial frontal cortex, cerebellum), attention to pain (anterior cingulate cortex, dorsolateral prefrontal cortex), emotional aspects of pain (CL closely connected to amygdala) and motor control [10]. In our study, apparently, nociceptive information could be transmitted to the CL via the most of the above-mentioned structures. However, it is quite a difficult task to accurately identify all possible nociceptive inputs to the CL and to separate the emotional, motor, as well as auditory and visual components, which requires a lot of additional research.
As mentioned above, significant changes in c-Fos expression levels were also observed in putamen. The Pu, together with the caudate nucleus, makes up the striatum, a major site of cortical and subcortical input into the basal ganglia. Although the putamen is frequently activated during pain, its role during pain has often been assumed to be related to motor processing [29]. However, it was shown that the putamen may contribute to the Fig. 2. The mean density ± SEM of Fos-ir neurons in claustrum and ventral part of putamen at the levels A+11/12 (within 400 x 400 µm 2 areas in the cat brain sections) in control animals (unilateraly) and on ipsil-and contralateral side of the cats brain after injection of complete Freund's adjuvant. Asterisks indicate signi?cant differences in the density of labeled cells between those in control animals and cats after injection of complete Freund's adjuvant on contralateral side; sign # -control mean density vs. ipsilateral mean density of adjuvant injected cats. Fig. 3. The mean density ± SEM of NADPH-dr neurons in claustrum and ventral part of putamen at the levels A+11/12 (within 400 x 400 µm 2 areas in the cat brain sections) in control animals (unilateraly) and on ipsil-and contralateral side of the cats brain after injection of complete Freund's adjuvant. Asterisks indicate significant differences in the density of labeled cells between those in control animals and cats after injection of complete Freund's adjuvant on contralateral side; sign # -control mean density vs. ipsilateral mean density of adjuvant injected cats. processing of sensory aspects of pain. The striatum is rich in opioid receptors and contains nociceptive neurons responsive to graded noxious stimuli [2,29]. Although the Pu is mutually related to the CL [17], it is quite problematic to assess their mutual influence.
In our study, NADPH-dr/NOS-containing neurons in CL and Pu were found to be twice larger in comparison with the controls. Earlier it was shown that NO is involved in the development of both acute and chronic inflammation, and different NOS inhibitors demonstrate an anti-inflammatory action [6]. An NO-mediated increase/decrease in the cerebral neuronal activity results probably from activation/ inhibition of different ion channels in the neuronal membranes [25,30]. D.V. Hinova-Palova et al. showed that there are two populations of NADPH-dr neurons in the CL. One population, consisting of large and medium-sized positive neurons, represents projection neurons, while the other population of small positive neurons corresponds to local circuit neurons [13] and only the densely stained cells were GABA immunoreactive [11]. In our experiments, it was found that a large number of NADPH-dr neurons are localized in Pu, especially in the ventral part. It was shown that NO-generating interneurons are key modulators of neuronal activity in Pu in rats and are intensely innervated by glutamatergic and dopaminergic-afferent projections from the substantia nigra [15], as well as that NO regulates signaling cascades in intermediate chains of the system central nucleus amygdala-substantia nigra-putamenthalamus [19].

Conclusions
1. The results of the study of the Fos-ir neurons distribution in CL and Pu under conditions of long-term muscles inflammation indicate the active involvement of the mentioned structures in the formation of adaptive reactions.
2. The increase in the number of neurons with NADPHd-positivity in CL and Pu indicates that NO-signals play a significant role in the formation and amplification of the response to painful impulses from inflamed muscles.
3. Further research is needed to accurately identify all possible nociceptive inputs to the CL and to separate the emotional, motor, auditory, and visual components that may accompany nociceptive processes.