Induction of Heme Oxygenase-1 Is Involved in Carbon Monoxide-Mediated Central Cardiovascular Regulation

Heme oxygenase (HO) is the rate-limiting enzyme responsible for the catabolism of heme and subsequent production of carbon monoxide (CO), biliverdin, and iron (Maines, 1997; Maines et al., 1998). To date, three isoforms of HO have been identified. HO-1, a 32-kDa protein, is induced by heme and numerous oxidative stressors that are enriched in spleen and liver. HO-2, a 36-kDa protein, is present abundantly in the brain and testis as a constitutive enzyme (Maines, 1997; McCoubrey et al., 1997). HO-3 is a 33-kDa protein that shares approximately 90% homology with HO-2 but exhibits a relatively low catalytic activity (McCoubrey et al., 1997). HO and its metabolic products have been implicated in the regulation of numerous biological processes (Maines, 1997; Maines et al., 1998). Among them, CO derived from HO activity has been shown to function as a neurotransmitter (Verma et al., 1993), a vasodilator (Furchgott and Jothianandan, 1991), and an endogenous modulator of the nitric oxide (NO)-cGMP signaling in brain (Ingi et al., 1996).

HO-1 has been demonstrated to provide cytoprotection in various in vitro and in vivo systems. The activation of HO-1 gene has been considered to be an adaptive cellular defense mechanism (Poss and Tonegawa, 1997). In addition, either acute (Sacerdoti et al., 1989) or chronic (Escalante et al., 1991) administration of an inducer of HO-1 to spontaneously hypertensive rats led to a normalization of blood pressure. Other inducers of HO-1 or HO substrates have also been shown to decrease blood pressure in hypertensive rats (Levere et al., 1990; Martasek et al., 1991; Johnson et al., 1996). Moreover, treatment of normal (Johnson et al., 1995) or endotoxemic (Yet et al., 1997) rats with inhibitors of HO (metalloporphyrins) has been shown to produce an increase in systemic arterial pressure. Because biliverdin itself has not been associated with the regulation of blood pressure (Johnson et al., 1995), these studies provided evidence that CO via the HO activity may contribute to the regulation of blood pressure.

HO is widely expressed in brain and is responsible for the CO-generating ability of the brain, including brainstem (Ewing and Maines, 1991; Maines et al., 1993). In the central nervous system, the nucleus tractus solitarii (NTS) is the site where afferent fibers arising from the arterial and cardiopulmonary baroreceptors make the first central synapse and thus play an important role in the integration of autonomic control of cardiovascular system (Reis, 1984). It has been reported that CO formed within the NTS subserves a vasodepressor mechanism that is tonically active in awake rats (Johnson et al., 1997). Furthermore, we have shown that unilateral microinjection of hemin or hematin, a heme molecule cleaved by HO to yield CO, into the NTS produce dose-related depressor and bradycardic effects (Lo et al., 2000; Lin et al., 2003). On the other hand, either systemic administration (Johnson et al., 1997) or direct microinjection of HO inhibitor zinc deuteroporphyrin 2,4-bis glycol into the NTS attenuates the baroreceptor reflex (Lo et al., 2000). Taken together, these findings suggested that CO within the NTS might play an important role in the regulation of cardiovascular function. However, it is not clear whether HO actually exists in the NTS.

The present study investigated whether hemin injection into the NTS induced different HO isoform (HO-1 and HO-2) expression in addition to their in situ localizations after induction. Our results showed that both HO-1 and HO-2 exist in the NTS, but only the HO-1 level is significantly elevated upon hemin microinjection. These results suggested that HO-1 might play a role in central cardiovascular regulation.

Materials and Methods

Materials. Experimental drugs such as urethane and l-glutamate were purchased from Sigma Chemical Co. (St. Louis, MO). ZnPPIX was obtained from Tocris Cookson Ltd. (Bristol, UK). ZnPPIX was dissolved in 50 mM Na₂CO₃, pH 8.8 to 9.4, immediately before use. Hemin (1 nmol/60 nl; MP Biomedicals, Irvine, CA) was dissolved in 30% 0.1 N NaOH, pH 8.6 to 9. All other drugs were dissolved in normal saline on the day of the experiment.

Animal Procedures and Tissue Preparation. Male Sprague-Dawley rats (250-350 g) were obtained from National Science Council Animal Facility and housed in the animal room of Kaohsiung Veterans General Hospital (Kaohsiung, Taiwan). The rats were kept in individual cages in a room in which lighting was controlled (12 h on/12 h off), and temperature was maintained at 23-24°C. The rats were given Purina Laboratory Chow (Purina, St. Louis, MO) and tap water ad libitum.

All animal protocols have been approved by the Research Animal Facility Committee at Kaohsiung Veterans General Hospital. Humane treatment is administered at all times. Rats were anesthetized with urethane (1.0 g/kg i.p. and 300 mg/kg i.v.) if necessary. The preparation of animals for intra-NTS microinjection and the methods used in the localization of the NTS have been described previously (Tseng et al., 1996). For microinjections into the NTS, a glass cannula was filled with l-glutamate (0.154 nmol/60 nl) to functionally identify the NTS. A specific decrease in blood pressure (BP) and heart rate (HR) (≥-35 mm Hg and -50 bpm) was demonstrated after microinjection of l-glutamate in the NTS. After that, BP and HR were observed through microinjection of hemin (1 nmol/60 nl) before and 10 min after intra-NTS administration with HO inhibitor ZnPPIX (1 nmol/60 nl) or vehicle alone (50 mM Na₂CO₃, 60 nl).

After completion of the experiments, rats were perfused intracardially with saline followed sequentially by a solution of 4% formaldehyde. Paraffin embedding was performed. Five-micrometer sections of the brainstem were stained with H&E. The proper placement of the pipette tip in the NTS was verified by histological examination. Maps and coordinates (from bregma) are taken from the atlas of Paxinos and Watson (1986). For total RNA and protein extractions, the NTS was removed immediately after the completion of the experiments without formaldehyde perfusion and cryopreserved for further extraction.

Western Blot Analysis. At the time points 10 min or 4 h after microinjection of hemin or pretreatment with ZnPPIX into the NTS, the NTS tissue was separated carefully under the microscopy examination from individual rats. In brief, tissues on both sides of the dorsomedial part of the medulla oblongata at the level of NTS (1 mm rostral or caudal from the obex) were collected by micropunches made with a stainless steel bore (1 mm i.d.) (Chan et al., 2004). Thereafter, these individual NTS tissues were used for western blot. Western blot was used to determine the HO-1 and HO-2 protein expression levels in the NTS. The procedures were as described previously with minor modifications (Huang et al., 2004). In brief, total protein were prepared by homogenized NTS in lysis buffer containing 20 mM imidazole-HCl, pH 6.8, 100 mM KCl, 2 mM MgCl, 20 mM EGTA, pH 7.0, 300 mM sucrose, 1 mM NaF, 1 mM sodium-vanadate, 1 mM sodium molybetadate, 0.2% Triton X-100, and proteinase inhibitor cocktail (Roche Diagnostics, Indianapolis, IN) for 1 h at 4°C. Equal amounts (30 μg/sample assessed by bovine serum albumin protein assay; Pierce Chemical Co., Rockford, IL) of protein were separated on 12.5% SDS-Tris glycine gel electrophoresis and transferred to a nitrocellulose membrane (GE Healthcare, Little Chalfont, Buckinghamshire, UK). The membrane was blocked with 5% nonfat milk in Tris-buffered saline/Tween 20 buffer (10 mM Tris, pH 7.5, 100 mM NaCl, 0.1% Tween 20) and incubated with rabbit anti-rat HO-1 polyclonal antibody (1:4000; StressGen, Victoria, BC, Canada), rabbit anti-rat HO-2 polyclonal antibody (1:4000; Stress-Gen) (Ndisang et al., 2002), and mouse anti-α tubulin antibody (1:10,000; Sigma) (Huang et al., 2004) in Tris-buffered saline/Tween 20 with bovine serum albumin and incubated for 1 h at room temperature. Peroxidase conjugated anti-mouse or rabbit antibody (1: 5000 with HO-1, 1:15,000 with HO-2, and 1:10,000 with α-tubulin; Jackson ImmunoResearch Laboratories Inc., West Grove, PA) was then applied. The membrane was developed using ECL-Plus protein detection kit (GE Healthcare). The protein expression levels were then determined using NIH Image 1.61 software (National Institutes of Health, Bethesda, MD).

Immunohistochemistry Analysis. Immunohistochemical staining was performed according to procedures as described previously (Huang et al., 2004). In brief, rat brain was fixed with 4% formaldehyde. Paraffin-embedded serial sections were cut at 5-μm thickness. The sections were deparaffinized, biotin blocked (biotin blocking system; DakoCytomation Ltd., Ely, Cambridgeshire, UK), microwaved (0.01 M citric buffer, pH 6.4), quenched (3% H₂O₂/methanol), blocked (3% goat serum), and incubated in IHC-specific rabbit anti-rat HO-1 antibody, rabbit anti-rat HO-2 antibody (1:500, StressGen), rabbit anti-GFAP antibody (1:500; DakoCytomation Ltd.), or mouse anti-NeuN antibody (1:500; Chemicon International, Temecula, CA) at 4°C overnight. After incubating the sections with biotinylated secondary antibody (1:200; Vector Laboratories, Burlingame, CA) for 1 h and then AB complex (1:100) for 30 min at room temperature. The sections were visualized using diaminobenzidine substrate for 5 min (Vector Laboratories) and counterstained with hematoxylin. The sections were then photographed with an Olympus microscope equipped with a CCD imaging system (Olympus, Tokyo, Japan).

Quantitative Real-Time RT-PCR. Total RNA from NTS before and after hemin treatment was extracted using TRI reagent according to the manufacturer's protocols (Molecular Research Center, Cincinnati, OH). Five micrograms of total RNA was used to synthesize first strand cDNA using 50 U of StrataScript reverse transcriptase and 500 ng of oligo(dT) primer (Stratagene, La Jolla, CA). Reverse transcription products (1/20) were used as template for quantitative real-time RT-PCR in LightCycler technology (Roche Diagnostics) using a SYBR green assay. PCR reaction was performed in 50 μl of SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) containing 10 μM forward primers and reverse primers and approximately 30 ng of cDNA. Amplification and detection were performed by one cycle of 95°C for 10 min, 40 cycles of 95°C for 15 s, 62°C for 20 s, and 72°C for 15 s. After completion, a final melting curve was performed by denaturation at 95°C for 15 s and then was recorded by cooling to 60°C and then heating slowly until 95°C for 20 min according to the dissociation protocol of ABI7700 instrument. The primer sequences for HO-1: forward primer, 5′-AGCTCTATCGTGCTCGC-3′; and reverse primer, 5′-GTGTTCCTCTGTCAGCAGT-3′, which amplified 110 base pairs of HO-1 cDNA fragment. The β-actin mRNA level was determined using: forward primer, 5′-TCACCCACACTGTGCCCATCTACGA-3′; and reverse primer, 5′-CAGCGGAAC CGCTCATTGCCAATGG-3′, which amplified a 295-base pair β-actin cDNA fragment.

Statistical Analysis. All data were expressed as mean ± S.E.M. A paired Student's t test (before and after pretreatments), unpaired Student's t test (for control and study group comparisons), or repeated-measures analysis of variance followed by Dunnett's test for significant differences were applied to compare group differences. Differences with P < 0.05 were considered significant.

Results

Cardiovascular Effects of Hemin in the NTS. In agreement with our previous finding, intra-NTS microinjection of hemin (1 nmol), a heme molecule cleaved by HO to yield CO, resulted in hypotension and bradycardia (Lo et al., 2000; Lin et al., 2003). The administration of vehicle (30% 0.1 N NaOH) itself elicited only slight cardiovascular effects, and the pattern of response was different from that of hemin. After pretreatment with an HO inhibitor, ZnPPIX (1 nmol), for 10 min, the depressor and bradycardiac response to hemin were attenuated significantly (from -45 ± 2 mm Hg and -64 ± 7 bpm to -26 ± 3 mm Hg and -28 ± 5 bpm, respectively, P < 0.05, paired Student's t test) (Fig. 1, A and B). Pretreatment with the vehicle of ZnPPIX did not affect the cardiovascular responses to hemin (from -43 ± 3 mm Hg and -59 ± 7 bpm to -41 ± 3 mm Hg and -65 ± 6 bpm, respectively, P > 0.05, paired Student's t test). To exclude the possibility that the hypotension and bradycardia were caused by biliverdin or iron, we tested the cardiovascular effects of biliverdin and iron by the NTS microinjection. Unlike hemin, neither of them had cardiovascular effect when microinjection of biliverdin or ion even at the high dosage of 1 mol/60 nl into NTS. These findings suggested that CO in the NTS might play an important role in the regulation of cardiovascular function.

HO-1 Protein Induced by Hemin Microinjection into the NTS. The roles of HO in central cardiovascular control remained undetermined. Two types of heme oxygenases were thought to be responsible for the HO functions in controlling cardiovascular effects: the inducible HO-1 and the constitutive HO-2. To determine the roles of these two enzymes in central cardiovascular control, we determined whether HO-1 and HO-2 protein expression were induced in the NTS after microinjection of hemin. Western blot analysis revealed a significant increase in HO-1 protein expression (Fig. 2, A and B, 1.77 ± 0.07-fold, P < 0.05, n = 6) after hemin microinjection into the NTS compared with that of control (Fig. 2A, lane 4 versus 3). To identify the successful placement of the NTS, the l-glutamate response was taken as the control group. Furthermore, pretreatment with ZnPPIX attenuated hemin-induced increase in HO-1 expression (Fig. 2A, lane 5; 1.18- ± 0.17-fold, P < 0.05). In contrast, HO-2 protein expression showed no significant changes after hemin injection or ZnPPIX pretreatment compared with control (Fig. 2C, 1.19- ± 0.27- and 0.83- ± 0.08-fold, respectively). These results suggested the HO-1 elevation might play a role in hemin-mediated central cardiovascular effects.

Hemin-Induced HO-1 in Situ Protein Expression in the NTS. We then determined the in situ protein expression levels and localizations of HO-1 and HO-2 after microinjection of hemin and pretreatment with ZnPPIX. Figures 3 and 4 showed immunohistochemical analysis of HO-1 and HO-2 expression in the NTS, respectively. Relatively low to mild levels of HO-1 expression were observed in very few cells in the NTS of the control group (l-glutamate injection only) (Fig. 3, A and B). However, intensive immunostaining of the HO-1 in many cells with different morphology was noted after injecting hemin into the NTS (Fig. 3, D and G). Qualitative and quantitative analyses of HO-1 protein expression were performed after hemin microinjection and ZnPPIX pretreatment according to the methods described under Materials and Methods. Figure 3E showed significant induction of HO-1 protein in the right hemisphere of NTS after hemin injection compared with the control (l-glutamate injection only) in the left hemisphere (26.7 ± 0.9% versus 10.3 ± 0.5%, n = 32, P < 0.01). The potency of hemin-induced HO-1 expression did not differ in the anatomical localization of NTS because equal induction of HO-1 was found after injecting hemin into either the left or right hemisphere of NTS (Fig. 3, C and D versus F and G). Pretreatment of ZnPPIX significantly inhibited the induction of HO-1 in the NTS compared with control (l-glutamate injection only) (Fig. 3J; 18.6 ± 0.7% versus 26.9 ± 0.9%, n = 32, P < 0.01). HO-2 protein expression was then examined after hemin injection with or without pretreatment of ZnPPIX. Figure 4 demonstrated in situ HO-2 expression in NTS after hemin injection and pretreatment of ZnPPIX. Quantitative analysis showed no significant changes of HO-2-positive cells in the NTS after hemin injection into right hemisphere of NTS compared with control (l-glutamate injection only) in the left hemisphere (Fig. 4E; 35.4 ± 1.2% versus 36.3 ± 1.0%, n = 32). Pretreatment with ZnPPIX did not significantly affect the HO-2 expression in the right hemisphere of NTS compared with that of control (l-glutamate injection only) in the left hemisphere (Fig. 4J; 32.9 ± 0.9% versus 35.0 ± 0.9%, n = 32). It is surprising that the HO inhibitor blocked the hemin-induced HO-1 induction. This suggests that an HO product, carbon monoxide, biliverdin, or iron, is responsible for the induction. The immunohistochemistry experiments were performed to investigate whether biliverdin or FeCl₂ solution elevates the HO-1 protein level in the NTS. Neither biliverdin nor FeCl₂ solution can elevate the HO-1 protein level in the NTS (data not shown). Taken together, we can exclude the possibility that biliverdin or iron is responsible for the HO-1 induction upon microinjection of hemin into the NTS. CO cannot be directly microinjected into the NTS, but by exclusion, CO is highly suspected to be responsible for the HO-1 elevation.

Hemin Microinjection-Induced HO-1 Protein Expression in NTS Neurons and Glial Cells. While examining the in situ protein expression of HO-1 in the NTS after hemin injection, we observed at least two major cell types stained positive for HO-1 antibody. One cell type was with smaller nucleus with multiple fine protrusions (Fig. 5C, arrowhead). The other cell type was with larger nucleus with flat and polygonal morphology (Fig. 5C, arrow). These cells were further confirmed to be of glial and neuronal origin by staining positive for GFAP and NeuN antibody, respectively (Fig. 5, D and H). The cells in the NTS that were found stained positive for HO-2 antibody were primarily large cells with scanty cytoplasm with flat and polygonal morphology (Fig. 5G, arrow). They were further identified to be of neuronal origin because they stained positive with NeuN antibody. These results indicated that the protein expression of HO-1 in NTS neurons and glial cells might be involved in central controls of cardiovascular effects.

Induction of HO-1 mRNA Expression in the NTS after Hemin Microinjection. We further determined whether HO-1 mRNA expression levels were altered after microinjection of hemin into the NTS. The quantitative real-time PCR was performed to examine the level of HO mRNA. The results showed that elevated HO-1 mRNA level was observed in the 10 min (26% increase) and 4 h (121% increase) after hemin microinjection (Fig. 6). These results indicate that HO-1 protein induction in the NTS after hemin injection may be mediated through transcriptional modification at the RNA level.

Discussion

In the present study, we demonstrated that microinjection of hemin, a precursor for CO production, into the NTS induced depressor and bradycardic effects. These effects of hemin required HO because prior administration of an HO inhibitor (ZnPPIX) significantly suppressed the cardiovascular effects of intra-NTS microinjection with hemin (Fig. 1, A and B). These results were similar to our previous findings (Lo et al., 2000; Lin et al., 2003). It has been shown that administration of an HO inducer such as heme-l-arginate or heme-l-lysinate causes a marked decrease in BP in spontaneously hypertensive rats (Johnson et al., 1996). Intracardiac injection of high-titer retrovirus encoding HO-1 gene into spontaneously hypertensive rats has been shown to decrease blood pressure (Sabaawy et al., 2001). In addition, HO inhibitors have been shown to increase BP and peripheral resistance and suggest that endogenous CO subserves a tonic vasodepressor function (Johnson et al., 1995). Direct micro-injection into the NTS of heme oxygenase inhibitor, zinc deuteroporphyrin 2,4-bis glycol or ZnPPIX, attenuates the baroreceptor reflex (Johnson et al., 1997; Lo et al., 2000). It has been indicated that CO apparently acts in the NTS, promoting changes in glutamatergic neurotransmission and lowering blood pressure (Johnson et al., 1999). We have reported that metabotropic glutamate receptors may couple to the activation of HO via the liberation of CO to participate in central cardiovascular regulation (Lo et al., 2002). Taken together with our results that neither biliverdin nor iron, two products from hemin, showed cardiovascular effect when microinjected into NTS, these findings suggest that the HO/CO system may be involved in the regulation of various neural and cardiovascular functions.

The NTS is a nexus of nerve endings for cardiovascular chemo- and baroreceptor to regulate cardiovascular functions (Spyer, 1981; Reis, 1984). To date, no systematic study has been carried out to correlate the actual expression levels of HO-1 and HO-2 protein in the NTS. In our study, we examined the pattern of expression of HO-1 and HO-2 in the NTS by Western blot and immunohistochemical analysis. Our results showed the expression of HO-1 protein was significantly enhanced after hemin microinjection into the NTS in comparison with that of control. However, hemin treatment did not increase the expression of HO-2 protein in the NTS. On the other hand, HO-2 protein expression levels were not different between hemin treatment and ZnPPIX pretreatment groups (Fig. 2). Moreover, immunohistochemical staining also showed quantitative increase in HO-1 protein expression after hemin treatment. Pretreatment with ZnPPIX reduced hemin-induced HO-1 protein expression in the NTS. The expression of HO-2 protein was not different among the hemin-treated groups and not affected by ZnPPIX pretreatment (Figs. 3 and 4). The quantitative real-time PCR was used to examine the level of HO mRNA (Fig. 6), and the results showed that HO-1 mRNA level was elevated in the 10 min and 4 h after hemin microinjection. These results supported our results that HO-1 protein level increased after hemin injection and indicated that HO-1 protein induction may be mediated through transcriptional modification at the RNA level. Together, these results implied that hemin induced the expression of HO-1, leading to the increased HO activity, and enhanced CO production might be contributed to the central cardiovascular regulation.

Under normal conditions, HO-1 is present in the whole brain at the limit of detection by radioimmunoassay (Sun et al., 1990) and Western blot analysis (Trakshel et al., 1988). However, it has been shown that under normal conditions, the HO-1 isozyme is expressed in very few select neuronal and non-neuronal cell populations (Ewing and Maines, 1991; Ewing et al., 1992). It is usually considered that induction of HO-1 protects cells from oxidative stress and has an important role in antioxidant defense responses (Applegate et al., 1991; Chen et al., 2000). Given that prior HO-1 induction protects against insults in some systems, possible neuroprotective therapies based on inducing overexpression of HO-1 have been suggested (Panahian et al., 1999). In addition, it has been demonstrated that either acute (Sacerdoti et al., 1989) or chronic (Escalante et al., 1991) administration of an inducer of HO-1 (stannous chloride) to spontaneously hypertensive rats led to a normalization of blood pressure. Other inducers of HO-1 or HO substrates have also shown to decrease blood pressure in hypertensive rats (Levere et al., 1990; Martasek et al., 1991; Johnson et al., 1996). In the present study, we have demonstrated that CO via HO-mediated hemin metabolism can significantly regulate the central cardiovascular effects. Likewise, we have previously reported that the NO was involved in the NTS regulation of blood pressure and that NO synthase inhibitor attenuated baroreflex activation (Lo et al., 1996; Tseng et al., 1996). Another report has shown that CO shares some of the chemical and biological properties of NO (Marks et al., 1991). Endogenous CO production could lead to cGMP synthesis through activation of guanylyl cyclase (Marks et al., 1991; Verma et al., 1993). The enzyme HO could act as a source of CO in neurons. Two forms of this enzyme are found in the brain. HO-1 normally shows only a limited distribution, but its synthesis can be selectively increased in certain neurons and glial cells through activation of heat shock elements by diverse stimuli. Conversely, HO-2 is widely distributed in the brain under all conditions (Maines et al., 1998).

Furthermore, we observed the in situ protein expression of HO-1 in the NTS after hemin microinjection, we observed at least two major cell types stained positive for HO-1 antibody. These cells were further confirmed to be of glial and neuronal origin by staining positive for GFAP and NeuN antibody, respectively. HO-1 protein expression was predominantly noted in neurons of the NTS, and glial cells were sparsely distributed in the NTS. However, HO-2-positive cells were found localized mostly in neurons (Fig. 5). It has been suggested that excitotoxin-induced toxicity is partially mediated by oxidative stress (MacGregor et al., 1996; Matsuoka et al., 1998), and the glial cells in which HO-1 protein was induced may be resistant to oxidative stress. Overexpression of HO-1 may contribute to the resistance of glial cells to oxidative stress. Moreover, a recent report has suggested that overexpression of HO-1 might exert cytoprotective and antiapoptotic effects on glial cells (Munoz et al., 2005). At the same time, glial cells overexpressing HO-1 may be involved in repairing the lesioned area and reconstruction of the neuronal functions and/or scavenging dead neurons and/or terminals. HO-1 overexpression in NTS neurons might also elicit additional protection toward stress induced by hemin. The actual roles of HO-1-expressing glial cells and neurons in NTS-mediated central control of cardiovascular effects remains to be determined.

HO-2 was found concentrated in the brain and testes, accounting for the great majority of HO activity in the brain (Maines, 1997). Evidence has suggested that HO-2, localized to selective neuronal populations (Verma et al., 1993; Maines, 1997), plays a major role in neuromodulatory activities, with CO participating as a putative neurotransmitter. Thus, HO-2 mRNA and protein are selectively concentrated in discrete neuronal populations, although most, if not all, neurons do possess HO-2. The abundant expression of HO-2 has been associated with neuronal protection against oxidative stress injury by quenching free radicals (Dore et al., 1999). In our experiments, we found HO-2 expression in NTS neurons. Even though we did not detect any significant changes of HO-2 expression after hemin injection or pretreatment with ZnPPIX, the possibility of increased enzyme activity due to post-translational modification cannot be excluded. A similar finding has been shown previously by Ndisang et al. (2002). In their study, the expression of HO-2 protein was not found different among all animal groups tested and not affected by hemin treatment. Although we have shown that hemin induces HO-1 expression that suggests HO-1 might be responsible for generation of CO in the NTS, it is still possible that the more abundant HO-2 also plays a role in CO generation in the NTS. The phosphorylation and activation of HO-2 remains unclear and will be investigated in the future to distinguish the roles of HO-1 and HO-2 in the NTS.

The HO-CO system has already been proven to be a potential regulator of various neural and cardiovascular functions. Results from this study show that HO-1 and HO-2 expression occurred in the NTS, and we provide a broad array of evidence that HO-1, rather than HO-2, is implicated in the NTS effect. Furthermore, we are the first group to demonstrate the enhanced HO-1 protein expression in situ in the NTS after hemin injection. We further observed that at least two major cell types stained positive for HO-1 antibody. We also suggested that microinjection of hemin into the NTS may induce the activation of HO-1 via the liberation of CO to participate in central cardiovascular regulation. The identification of the actual roles of HO-1 expression in glial cells and neurons in the NTS will be a matter of future research.