c-fos gene expression is increased in the paraventricular hypothalamic nucleus of Sprague-Dawley rats With visceral pain induced by acetic acid without detectable changes of corticotrophin-releasing factor mRNA: A quantitative approach with an image analysis system
The immediate early gene c-fos is one of the most studied genes in the CNS as a marker for neuronal activation (Edwards et al.,1999; Abraham and Kovacs,2000; Konkle and Bielajew,2004). It is thus generally believed that activation of a neuronal system with c-fos expression in the brain is a hallmark for reflecting the functional status of a discrete brain structure. The c-fos belongs to the family of immediate-early transcription factor genes that are believed to function in coupling short-term signals elicited by extracellular to long-term changes in cellular phenotype by orchestrating changes in target gene expression (Curran and Morgan,1995). However, the expression of c-Fos, which is normally low, can be increased by a number of pharmacological, physiological, and behavioral manipulations (Morgan and Curran,1989; Herrera and Robertson,1996). Therefore, the measurement of c-Fos protein levels, the product of c-fos gene activation, has been used as a marker for activated neurons (Hoffman et al.,1993; Edwards et al.,1999; Abraham and Kovacs,2000). In recent years, c-Fos protein and c-fos mRNA have been widely used as markers for the impact of drug treatments (Young et al.,1991; Kreuter et al.,2004), experimental procedures (Worley et al.,1993; Del Bel et al.,1998), and environmental changes such as light (Rusak et al.,1992), ambient temperature (Bratincsak and Palkovits,2004), and novelty/stress (Emmert and Herman,1999; Amico et al.,2004). The c-fos gene can also be induced by pain (Snowball et al.,2000; Nakagawa et al.,2003). Visceral or somatovisceral pain is modeled in animals, as assessed by the writhing test (abdominal contractions) after an intraperitoneal injection of acetic acid (Snowball et al.,2000; Nakagawa et al.,2003; Sinniger et al.,2004). However, pain is often one of the concerns in our daily life (Cervero and Laird,1999); more specifically, visceral pain is a common form of pain associated with disease processes (Cervero and Larid,1999). Conceivably, to better understand visceral or somatovisceral pain in animal models is important for the advancement of pain treatment and management in patients (Cervero and Laird,1999; Joshi and Gebhart,2000).
It is clear that c-fos gene expression is one of the most important CNS parameters in the field of neuroscience research. However, the traditional method of studying c-fos gene expression by examining c-Fos-positive neurons is a labor-intensive task (Snowball et al.,2000; Nakagawa et al.,2003; Sinniger et al.,2004). Recently, the approach with the optical fractionator probe in conjunction with stereo-investigator software (MicroBrightField) on a PC system connected to a microscope interfaced with digital camera has been used for counting c-Fos-positive neurons (West et al.,1991). In fact, counting c-Fos-positive neurons with the optical fractionator probe is very time-consuming (Sari and Zhou,2004). Here we explored a way that can effectively assess c-Fos immunoreactivity through the use of a combination of immunohistochemistry and quantitative morphometric image analysis. The c-Fos immunoreactivity in a discrete brain region is equivalent to the signals of c-fos mRNA at the initial stage of their induction (Zangenehpour and Chaudhuri,2002).
For this study, we hypothesized that c-Fos immunoreactivity in a discrete brain region is compatible to the signals of c-fos mRNA at the initial stage of their induction and therefore can be used to reflect the number of c-Fos-positive neurons in the same region. This hypothesis was tested with a visceral pain model on rats induced by acetic acid, as the latter is known to activate the c-fos gene in the CNS (Snowball et al.,2000; Nakagawa et al.,2003; Sinniger et al.,2004). This study suggests that both c-Fos immunoreactivity (as detected by regular immunohistochemistry) and c-fos mRNA (as detected by in situ hybridization) can be efficiently measured using a computer-assisted image analysis system (Hwang et al.,2004b,2005; Suzuki et al.,2004). It is more effective, more objective, and less time-consuming than the manual counting of c-Fos-positive cells in discrete brain regions. Furthermore, the same approach used to detect c-Fos-positive neurons in the paraventricular hypothalamic nucleus (PVN) induced by acetic acid was also applied to determine if corticotrophin-releasing factor (CRF) mRNA was activated at the same window of the time frame.
MATERIALS AND METHODS
Animals and Experimental Treatments
Young adult male Sprague-Dawley rats (weighing from 164 to 181 g for the c-Fos study, and from 187 to 209 g for the CRF study) were purchased from Harlan-Sprague Dawley (Indianapolis, IN) and housed at the Laboratory Animal Center of the Indiana University School of Medicine. All of the experimental animals were kept at a constant ambient temperature and under a steady light/dark cycle with free access to food and water. Animals were acclimatized for 7 days to the housing cage before the experimental treatment. There were three groups of animals: the control animals received the intraperitoneal (i.p.) injection of saline (0.9% NaCl; 1 ml); the sham-control animals received the needle insertion without saline; and for the experimental group, each conscious rat received an i.p. injection of 2% acetic acid (AA in 1 ml saline) to induce the visceral pain according to Nakagawa et al. (2003). Each rat was returned to the home cage. Then, 60 min after the treatment, animals were anesthetized and perfused with the fixative discussed below. The experimental protocols were approved by the institution animal care and use committee.
For studying c-Fos immunohistochemistry and c-fos mRNA in situ hybridization, rats were anesthetized with ketamine cocktail (0.1 ml/100 g body weight, i.p.), perfused transcardially with 50 ml PBS, and followed by 300 ml of a fixative 60 min after the experimental treatment (Nakagawa et al.,2003). The fixative contained 4% formaldehyde (from Fisher) and 0.2% picric acid in PBS (pH 7.4). After perfusion, the brains were quickly removed and postfixed in the same fixative for 1 hr. Tissue samples were then rinsed in PBS and placed in 20% sucrose in PBS for cryoprotection at 4°C for 2 days. They were then frozen with pulverized dry ice powder and stored at −74°C until use. Serial 30 μm thick sections of the PVN were cut transversely with a Leica cryostat. One set of tissue sections were directly collected on gelatin-coated slides for the c-fos mRNA study, while their immediately adjacent tissue sections were cut but collected sequentially in a small vial containing PBS for the c-Fos immunohistochemistry, which appeared to have stronger staining in 30 μm section than that in 16 μm sections. On the other hand, for studying CRF mRNA, rats from another set of the experiment were also decapitated 60 min after the experimental procedures. Brains were removed, trimmed, and frozen quickly using dry ice powder. The PVN and central nucleus of the amygdala (CeA) were, however, cut into 16 μm frozen sections (Hwang et al.,2004b,2005; Suzuki et al.,2004) and collected on gelatin-coated slides. Tissue sections on slides were fixed with 4% formaldehyde in PBS as we did previously (Hwang et al.,2004b,2005; Suzuki et al.,2004).
In Situ Hybridization for c-Fos mRNA and CRF mRNA
In situ hybridization for c-fos mRNA was conducted in a similar way as we studied different mRNAs previously (Suzuki et al.,2004; Hwang et al.,2005). Briefly, the 51-mer antisense oligodeoxynucleotide probe, labeled with [33P]-dATP, of the following sequence was used to label the c-fos mRNA: 5′-GGGATAAAGTTGGCACTAGAGACGGACAGATCTGCGCAAAAGTCCTGTGTG-3′ as used by Lillrank et al. (1996). In situ hybridization, rinses, and autoradiography were carried out we did previously (Hwang et al., 2004, 2005; Suzuki et al., 2004). After vacuum-drying overnight, tissue slides were lined up on cardboards, apposed to BioMax MR films (Kodak, Rochester, NY), and processed for autoradiographic exposure in light-tight cassettes to produce autoradiograms. Those shown in Figures 2 and 4 were images taken from the autoradiograms on the films under an Olympus microscope at the same power for taking pictures for c-Fos immunoreactive images for the match and comparison. The c-fos mRNA signals were detected on BioMax MR films after 7 days of autoradiographic exposure. For the CRF mRNA, its signals in the PVN and CeA were detected using antisense 60-mer oligodeoxynucleotide probe of CRF cDNA with the following sequence: 5′-GTAAATCTCCATCAGTTTCCTGTTGCTGTGAGCTTGCTGAGCTAACTGCTCTGCCCTGGC, whose specificity had been described in our previous reports (Hwang et al.,2004b,2005).
For c-Fos immunohistochemistry, tissue sections collected in the vials were first washed in three changes of 0.1 M PBS (pH 7.4) and then placed in 0.1 M PBS containing 2% hydrogen peroxide for 20 min to reduce the endogenous peroxidase activity. Following three rinses in PBS, sections were incubated in the blocking medium containing 5% normal goat serum and 0.3% Triton X-100 for 30 min to block the nonspecific binding. After three washes in PBS, the sections were incubated in the rabbit polyclonal anti-c-Fos antiserum (Santa Cruz Biotechnology, Santa Cruz, CA) at the dilution of 1:1,000 with the blocking medium for 24 hr at room temperature. After the incubation in primary antibody, sections were incubated with a biotinylated goat antirabbit IgG (1:400) at room temperature for 30 min, followed by the standard avidin-biotin complex (ABC) procedure (Vector Laboratories, Burlingame, CA) with 0.05% diaminobenzidine (DAB) as a substrate for the horseradish peroxidase (HRP).
Quantitative Image Analysis
The signals for c-fos mRNA, CRF mRNA, and c-Fos immunostaining were measured using the Dell Optiplex GX 260 computer system and AIS v6.0 software (Imaging Research, Ontario, Canada), similar to what we did previously (Hwang et al.,2004a,2004b,2005). A digital camera mounted on the Olympus microscope was used to image sections in bright field, which were then displayed on a high-resolution monitor. All images (expressed in relative optic density, or ROD) were captured on the same day in order to maintain uniform settings adjusted at the beginning of capturing. The quantification of c-Fos immunostaining for each PVN section was carried out by montaging 10 rectangles (1,000 μm2 per rectangle), which covered all c-Fos immunostaining in a PVN profile for each section. However, the nonspecific background signals for each section were determined by measuring the ROD of an area lateral to the fornix of its own tissue section. The specific signals for each of 10 rectangles were obtained by subtracting the nonspecific background signals from total signals of densitometric readings (Kunkler and Hwang,1995). The sum of these 10 specific signals of c-Fos immunoreactivity (ir) was used to represent one PVN section. Four to five PVN sections per rat were examined, and three to four rats per group were studied. The general approach was similar to that reported by Chang et al. (2000).
However, the total signals of c-fos mRNA in the PVN were obtained by encircling the entire PVN profile with c-fos mRNA autoradiogram detected on the films, and the specific signals for c-fos mRNA were obtained by subtracting its own tissue nonspecific background signals, as we did previously for other mRNA studies (Hwang et al.,2004a,2004b,2005). Data collected between the experimental versus control groups were subjected to the unpaired two-tailed Student's t-test. Statistical differences were considered significant if P < 0.05.
This study reaffirmed the general concept that c-Fos immunoreactivity and c-fos mRNA examined were localized within the same neuronal system in their immediately adjacent PVN sections (Fig. 1). It is also clear that c-fos mRNA signals were much more intense than c-Fos protein signals from the same animal (Fig. 1). The results of this study indicated that there was an increase of c-fos mRNA expression in the PVN in response to visceral pain when induced by acetic acid (Fig. 2, bottom), as compared with those by saline (Fig. 2, middle) or sham-control treatment (Fig. 2, top). Similarly, c-Fos-ir in the PVN was also significantly increased in animals treated with 2% AA (Fig. 3), as compared with that in the control/sham-animals. There was no statistical difference in c-Fos-ir in the PVN between the sham-control and saline-treated groups (Fig. 3).
Although there was little c-Fos-ir in the PVN of the sham-control or saline-treated animals (Fig. 2A and B), c-fos mRNA signals were detectable in the PVN (Fig. 4A and B). This suggested that there was a basal level of the transcription of c-fos gene for c-fos mRNA, but not for its translation to c-Fos protein.
The adjacent PVN tissue sections of three groups of animals described above for c-Fos-ir were processed for the detection of c-fos mRNA. C-fos mRNA expression in the PVN of the experimental group of animals treated with 2% AA was qualitatively enhanced (Fig. 4C) when compared with the sham-control (Fig. 4A) and saline-treated animals (Fig. 4B). Acetic acid treatment significantly increased c-fos mRNA expression in the PVN (Fig. 5, right column), as compared to the sham-control and saline-treated groups. There was no statistical difference in c-fos mRNA level in the PVN between the control and sham-groups of animals (Fig. 5, left two columns).
Furthermore, CRF mRNA expression in the PVN and CeA was also studied. Surprisingly, we did not observe any detectable difference of CRF mRNA either in the PVN or in the CeA between acetic acid-treated animals and the control animals (saline-treated group and sham-group; Table 1).
Table 1. CRF mRNA (relative optical density/unit area) in the paraven-tricular hypothalamic nucleus (PVN) and central nucleus of the amygdala (CeA) of Sprague-Dawley rats treated with 2% acetic acid or saline treatment, or received sham injections. Each group contained five animals (n = 5). Data were expressed in Mean ± SEM
CRF mRNA expression was not statistically different among groups either in the PVN or in the CeA.
0.31 ± 0.07
0.30 ± 0.09
0.33 ± 0.04
0.15 ± 0.02
0.15 ± 0.01
0.18 ± 0.08
This study has provided evidence that both c-Fos protein and c-fos mRNA can be quantitatively measured in the same discrete brain region like the PVN. To our knowledge, we have shown for the first time of its kind that increases of both c-fos mRNA and c-Fos-ir in their immediately adjacent PVN sections of rats can be reliably measured using an image analysis system with an anatomical resolution. This new approach is clearly more efficient and manageable than the traditional manual method of counting c-Fos-positive cells or neurons (Hitzemann and Hitzemann,1997; Ryabinin et al.,1997; Kreuter et al.,2004). Results from this study are consistent with the work reported both by Sinniger et al. (2004), who have illustrated increased c-fos mRNA in the PVN, and by Snowball et al. (2000), who have shown increased number of c-Fos-positive neurons in the PVN after an AA treatment, suggesting the validity of our approach. However, it is important to mention that c-Fos induction in the PVN may only be obvious at their initial stage of induction (within 1 hr), since c-Fos induction is transient in nature (Zangenehpour and Chaudhuri,2002). The c-Fos-ir was assessed effectively using a regular immunohistochemistry in conjunction with a computer-assisted image analysis system (Hwang et al.,2004b,2005; Suzuki et al.,2004). In brief, the quantitation of c-Fos-ir and c-fos mRNA signals, and CRF mRNA signals induced by the AA treatment, is effective.
Intraperitoneal Injection of Acetic Acid for Inducing Visceral Pain
The writhing test is characterized by abdominal contractions and is an effective and straightforward method to induce visceral pain without the need for surgical procedures, as with gastric distention (Traub et al.,1996). The latter has its obvious advantage for studying the visceral pain mechanism in a specific visceral region such as the stomach. On the other hand, visceral pain induced by an i.p. injection of acetic acid is by nature more systemic, mimicking situations such as food poisoning or disseminated cancerous growth within the abdominal cavity. This study confirms that the utilization of acetic acid is a practical approach to examine central structures associated with visceral or somatovisceral pain.
PVN as a Key Structure for Visceral Pain
In this study, we examined whether the PVN and CeA in the forebrain were involved in the induction of visceral pain. Since both the PVN and CeA contain the neuronal systems that produces CRF, CRF mRNA signals in the PVN and CeA were studied in addition to c-Fos immunohistochemistry and c-fos mRNA in the PVN. This study is unique in terms of the examination of both c-Fos-ir and c-fos mRNA in adjacent cryostat sections of the PVN from the same Sprague-Dawley rats following an i.p. injection of 2% AA. Studying adjacent tissue sections of the same animals for two parameters has several advantages: a reduction of the animal use, and a reduction of individual variability between two parameters. More importantly, they provide quantitative data with anatomical resolution. The latter is particularly critical when an estimation of the parameter within the subdivisions of a discrete brain region, as evident from our previous studies (Hwang et al.,2004a,2004b; Suzuki et al.,2004).
This study has demonstrated that, 1 hr after the i.p. injection of acetic acid according to Nakagawa et al. (2003), there is a robust expression of c-Fos-ir and c-fos mRNA occurring in the PVN, suggesting that it is a key brain structure involved in the visceral pain mechanism. Interestingly, the number of c-Fos-positive neurons in the PVN is also increased after noxious stimulation to the lower back muscle (Ohtori et al.,1985). With this premise, it was an unexpected finding that CRF mRNA signals in the PVN from another set of Sprague-Dawley rats receiving the same acetic acid treatment were not significantly different from those of the control rats. Similarly, CRF mRNA signals in the CeA are not significantly different between the experimental and control groups. However, it is unknown if CRF mRNA signals will be increased in the PVN and CeA of experimental animals when an intronic probe is used.
In addition, it is unclear from this study what neurons are activated during visceral pain to have c-fos gene expressed in the PVN. But it is well known that the PVN consists of the magnocellular and parvocellular divisions (Sawchenko and Swanson,1985), and that the PVN is involved in visceral pain processing (Snowball et al.,2000). More specifically, neurons in the parvocellular PVN are heavily activated to express c-Fos mRNA 1 hr after acetic acid treatment (Sinniger et al.,2004). Anatomically, the parvocellular PVN contains a variety of neurons that can synthesize CRF, neurotensin, enkephalin, vasoactive intestinal polypeptide, thyrotropin-releasing factor (Cessatelli et al.,1999; Raptis et al.,2004; Sarker et al.,2004), cocaine- and amphetamine-regulated transcript (CART) (Raptis et al.,2004; Fekete et al.,2005), and vasopressin (Jiang et al.,2004). However, according to Sinniger et al. (2004), c-fos mRNA induced by AA seems to be in CRF-immunopositive neurons. It is unknown if the c-Fos gene is also expressed in other peptidergic neurons in the PVN in response to the pain. However, one of the most plausible explanations to our observation with no difference of CRF mRNA signals in the PVN between the control and experimental groups is perhaps due to the potential that CRF gene transcription for CRF neurons in the PVN in response to AA is far behind the c-fos gene expression. In other words, 1 hr after the AA treatment applied for this study, the CRF mRNA in the PVN may has not been upregulated yet. This is consistent with the fact that c-fos is one of the cellular immediate early genes well known for its early expression during neuronal activations (Morgan and Curran,1989).
Both c-Fos Immunoreactivity and c-fos mRNA as Excellent Markers
A tandem approach to measure both c-Fos-ir and c-fos mRNA in the PVN of rats following a treatment is practical, as two parameters are better than one with c-Fos immunoreactivity alone. A consistent expression of c-fos mRNA and c-Fos-ir in response to an experimental procedure may be essential, as shown in this study that there is an increase of both c-Fos-ir and c-fos mRNA in the PVN of rats 60 min after the AA treatment. This is parallel to the study showing that c-fos mRNA is proceeded by c-Fos protein expression in rat visual cortex at their early stage of gene expression (Zangenehpour and Chaudhuri,2002). Thus, a concomitant assessment of these two parameters would be a better choice as an indicator for the neuronal activation, even though the use of c-Fos protein detection with immunohistochemistry is popular (Traub et al.,1996; Hitzemann and Hitzemann,1997; Ryabinin et al.,1997; Bonaz et al.,2000; Snowball et al.,2000). This is perhaps due to the fact that c-Fos antibodies are commercially available for allowing immunohistochemical localizations of c-Fos protein.
c-Fos Immunohistochemistry in Relative Optical Density With Image Analysis System as a Proper Approach for Assessing Neuronal Activation
An approach to the utilization of c-Fos antibody to examine c-Fos-positive neurons is one of the most commonly used methods in neuroscience (Ryabinin et al.,1997; Angeles-Castellanos et al.,2005; Lo and Chen,2005), in which c-Fos-positive neurons are manually counted with the aid of computer software. Although the technique is valid, such manual counting protocols are very time- and resource-intensive (Sundquist and Nisenbaum,2005). To circumvent this problem, the number of c-Fos-positive neurons was estimated by assessing the ROD of c-Fos-ir in the discrete brain region using an image analysis system as above.
Interestingly, this study has shown that c-Fos-ir (as expressed by ROD to reflect the number of the c-Fos-positive neurons in the PVN) is significantly increased 60 min after the AA treatment. Furthermore, the increased c-Fos-ir in the PVN of rats treated with 2% AA is parallel to the increased c-fos mRNA in their ROD (Figs. 3 and 5). Such an expression of c-fos mRNA in ROD is compatible to that by Emmert and Herman (1999) and Morgan et al. (1992). The result is also comparable to other published works (Bonaz et al.,2000; Snowball et al.,2000), which have indicated that the same AA treatment can increase the number of c-Fos-positive neurons in the PVN using the manual counting methods. Conceivably, the latter is obviously more subjective than the parametric number derived from the computer-assisted image analysis system we used (Hwang et a.l,2004b,2005; Suzuki et al.,2004). Collectively, the new practical approach for assessing the ROD for c-Fos-ir is consistent with the literature. Therefore, the ROD approach for c-Fos-ir is valid for studying neuronal activation using an image analysis system to measure c-Fos-ir in cryostat tissue sections after an application of the traditional immunohistochemistry protocol. To this end on tissue sections, it is critical to mention that vibratome tissue sections, commonly used for immunohistochemistry studies, are not suitable for this quantitative evaluation due to their great variability in thickness.
Our result showing an increase of c-Fos-ir with quantitative immunohistochemistry (Mausset-Bonnefont et al.,2003; Kretz et al.,2004) is technically similar, but not the same, to quantitative cytochemical investigation into the cytochrome oxidase activity by measuring its DAB reaction density (Shumake et al.,2002; Bertoni-Freddari et al.,2004). In fact, Mize (1994) has long advocated for the use of this quantitative immunohistochemistry using an image analysis system (Mize et al.,1988). More specifically, the ROD derived from the antibody and tissue antigen immunoreaction to form the HRP-DAB reaction product in a discrete brain region can be quantitatively measured (Mize et al.,1988; Di Stefano et al.,2001; Mausset-Bonnefont et al.,2003; Kretz et al.,2004) to assess the functional status and experimental or pathological conditions. Lastly, it is interesting to point out that the quantitative immunohistochemistry described in this report is a method equivalent to Western blotting but with an anatomical resolution, since its principle is not different from the traditional Western protocol.
In conclusion, we have used quantitative immunohistochemistry in conjunction with a computer-assisted image analysis system to demonstrate that neurons in the parvocellular division of the PVN are activated to express its significantly increased c-Fos immunoreactivity in a visceral pain model, as induced by an i.p. injection of 2% acetic acid. It is important to mention that the increased c-Fos immunoreactivity in a quantitative unit (in ROD) is parallel to the increased expression of c-fos mRNA in their immediately adjacent sections. Clearly, this c-Fos study has suggested that PVN neurons are involved in the visceral pain process. However, future efforts need to examine further the chemical natures of these parvocellular neurons, since an identification of the chemical nature of these neurons may provide further insights for the development of a pharmacotherapeutic strategy for the management of visceral pain.
The authors are grateful to Dr. Harry June for allowing the use of his AIS image analysis system, Tad S. Hwang for reading the draft, and Dr. Wayne S. Hwang for editing the manuscript. Supported by National Institute on Alcohol and Alcoholism grant AA13521 (to B.H.H. and Z.-H.G.) and a postdoctoral fellowship from the Department of Anatomy and Cell Biology, School of Medicine, Indiana University (to R.S.).