The heme oxygenase isozymes, HO-1 and HO-2, oxidatively cleave the heme molecule to produce biliverdin and the gaseous messenger, CO. The cleavage results in the release of iron, a regulator of transferrin, ferritin, and nitric oxide (NO) synthase gene expression. Biliverdin reductase (BVR) then catalyzes the reduction of biliverdin, generating the potent intracellular antioxidant, bilirubin. We report an age-related decrease in HO-1 and HO-2 expression present in select brain regions including the hippocampus and the substantia nigra, that are involved in the high order cognitive processes of learning and memory. The age-related loss of monoxide-producing potential in select regions of the brain was not specific to the HO system but was also observed in neuronal NO-generating system. Furthermore, compared to 2-month old rats, the ability of aged brain tissue to respond to hypoxic/hyperthermia was compromised at both the protein and the transcription levels as judged by attenuated induction of HO-1 immunoreactive protein and its 1.8 Kb transcript. Neotrofin™ (AIT), a cognitive-enhancing and neuroprotective drug, caused a robust increase in HO-1 immunoreactive protein in select neuronal regions and increased the expression of HO-2 transcripts. The potential interplay between regulation of HO-2 gene expression and the serum levels of the adrenal steroids is discussed. We suggest the search for therapeutic agents that reverse the decline and aberrant stress response of HO enzymes may lead to effective treatment regimens for age-associated neuronal deficits.

Biliverdin reductase catalyzes the reduction of biliverdin, the product of heme oxygenase (HO) activity, to bilirubin. The reductase is unique among all enzymes characterized to date in being dual pH/cofactor-dependent. Until now the enzyme was assumed to be a noninducible cytosolic protein. This report, for the first time, demonstrates induction and nuclear localization of reductase in rat kidney in response to HO-1 inducers: bacterial lipopolysaccharide (LPS) and bromobenzene. The study also demonstrates that nuclear localization requires an intact nuclear localization signal and is responsive to cGMP. Specifically 16 h after treatment of rats (i.p.) with LPS (5 mg/kg), there was an increase in nuclear biliverdin reductase as determined by immunostaining, Western blotting, and activity analysis. Induction and nuclear localization of the reductase in kidney was also observed in bromobenzene-treated rats (2 mmol/kg, s.c., 24 h). The reductase message levels, however, were not increased in response to either treatment, suggesting post-transcriptional activation of the reductase by LPS and bromobenzene. The mechanism of nuclear transport of the reductase was examined using HeLa cells transfected with the hemagglutinin-tagged reductase construct. When cells were treated with 8-BrcGMP the protein translocated into the nucleus. Mutation of the putative nuclear localization signal domain of the reductase blocked nuclear transport of the protein. We suggest the significance of nuclear localization of the reductase may relate to: 1) chain-breaking antioxidant activity of bilirubin; 2) inhibition of superoxide formation by bilirubin; and 3) modulation of the signal transduction pathways.

Balloon angioplasty revolutionized interventional cardiology as a nonsurgical procedure to clear a diseased artery of atherosclerotic blockage. Despite its procedural reliability, angioplasty's long-term outcome can be compromised by restenosis, the recurrence of arterial blockage in response to balloon-induced vascular trauma. Restenosis constitutes an important unmet medical need whose pathogenesis has yet to be understood fully and remains to be solved therapeutically. The radical biomediator, nitric oxide (NO), is a natural modulator of several processes contributing to postangioplasty restenosis. An arterial NO deficiency has been implicated in the establishment and progression of restenosis. Efforts to address the restenosis problem have included trials evaluating a wide range of NO-based interventions for their potential to inhibit balloon-induced arterial occlusion. All types of NO-based interventions yet investigated benefit at least one aspect of balloon injury to a naive vessel in a laboratory animal without inducing significant side effects. The extent to which this positive, albeit largely descriptive, body of experimental data can be translated into the clinic remains to be determined. Further insight into the pathogenesis of restenosis and the molecular mechanisms by which NO regulates vascular homeostasis would help bridge this gap. At present, NO supplementation represents a unique and potentially powerful approach to help control restenosis, either alone or as a pharmaceutical adjunct to a vascular device.

Increasing evidence suggests that S-nitrosothiols (thionitrites) might represent naturally occurring nitric oxide surrogates and function as intermediates in nitrogen monoxide metabolism. A facile, sensitive, and selective micromethod has been developed and validated for quantification of S-nitrosothiols as their mercury-displaceable nitrogen monoxide content. In this method, brief (5-min), room-temperature pretreatment of S-nitrosothiol with a molar excess of aqueous mercuric chloride was used to liberate into solution, quantitatively, the nitrogen monoxide moiety, which rapidly and quantitatively converted to its stable solution end-product, nitrite. Solution nitrite was reduced back to nitric oxide with vanadium(III), and the nitric oxide was detected by gas-phase chemiluminescence after reaction with ozone in a commercial nitric oxide analyzer. A linear relationship was observed between S-nitrosothiol-bound nitrogen monoxide and ozone-chemiluminescent detector response over a wide range (16.3-3500 pmol) of nitric oxide, as generated by reaction of vanadium(III) with either nitrite standard or mercury-treated S-nitrosothiol. Assay response was quantitatively identical for equivalent amounts of nitrite and S-nitrosothiol-bound nitrogen monoxide. The method displayed 96% selectivity for nitrite vs. nitrate and negligible (<2%) interference by nitrosated compounds bearing nitrogen monoxide moieties bound to either nitrogen or carbon. The lower limits of quantitative sensitivity and qualitative detection were below 50 and 20 pmol S-nitrosothiol-bound nitrogen monoxide-equivalents, respectively. The intraday and interday coefficients of variation did not exceed 8%. This technique has been applied to quantify structurally diverse natural and synthetic S-nitrosothiols with quantitative recovery from complex biological samples such as culture media and plasma at levels of nitrogen monoxide-equivalents undetectable by the popular Saville colorimetric method.

Heme oxygenase (HO) proteins are members of the HSP30 family and consist of 2 isozymes identified to date, termed HO-1 and HO-2 [1,2]. Separate genes encode the isozymes [3] and protein products which are immunochemically distinct, share less than 50% similarity at the amino acid sequence level. Each form, however, shows greater than 90% similarity among species, including human and the rat (reviewed in ref. [4]). Furthermore, these isozymes function in a well-defined role to carry out oxidation of the heme molecule (Fe-protoporphyrin IX) in concert with NADPH-cytochrome P450 reductase. The oxidation of heme is isomer specific and results in the formation of bile pigments, carbon monoxide, and iron. The heme molecule constitutes the prosthetic moiety of hemoproteins, such as hemoglobin, myoglobin, catalase, soluble guanylate cyclase, cytochrome b(5), cytochromes P450 and NO synthase.HO-1 also known as heat shock protein (HSP) 32 is encoded by a gene which is exquisitely stress-responsive and a host of stimuli that mediate oxidative stress cause induction of the protein both in vivo [4-7] and in vitro [8,9]. The HO-2 form shows a unique pattern of regulation from that of HO-1. HO-2 is a constitutive protein and its expression is not affected by the inducers of HO-1 tested to date [10]; rather, the only known regulator of HO-2 yet identified is adrenal glucocorticoids [11]. The two isozymes display vast differences in tissue distribution and under normal conditions HO-1 is present in the whole brain at the limit of immunodetection [6,12] and is discreetly localized in select neuronal populations [13,14]. HO-1 protein (similar to 32 kDa) and its similar to 1.8 kb transcript are increased [6,13-16], however, in response to stressful stimuli primarily in non-neuronal cell populations.The heme oxygenase system serves in both a catabolic and anabolic capacity in the cell. In the former capacity, it down-regulates cellular heme and hemoprotein levels. And, as such it inactivates the most effective catalyst for formation of free radicals, the heme molecule [17,18]. In its anabolic role, as noted above, heme oxygenase produces bile pigments, carbon monoxide, and iron, all of which are biologically active: bile pigments function as antioxidants [19]; the carbon monoxide generated by HO activity has been correlated with the generation of cGMP [7,20-24]; and iron regulates expression of various genes, including that of HO-1 itself [25], as well as transferrin receptors, ferritin, and NO synthase [26-28].We used rabbit anti-rat HO-2 polyclonal antibody and HO-2 cDNA to localize HO-2 immunoreactive protein and the 1.3(-) and 1.9 kb homologous transcripts, respectively, in rodent brain as visualized by histochemical staining procedures. These protocols provide the first detailed description of methodologies successfully used to define the pattern of HO-2 expression at the transcriptional and translational levels in the adult rat brain [14,29] and glucocorticoid-treated newborn rats [30]. The procedures described herein have the virtue of being non-radioactive, as well as applicability to the systemic organs, such as the cardiovascular system [7,23] and the male reproductive organs [31-33]. Visualization of cellular HO-2 expression aids in assessment of potential sites of carbon monoxide, iron, and bilirubin production within the nervous system.

Nitric oxide (NO.) and NO. donors incite NAD- [i.e., mono(ADP-ribosylation)] and NADH-dependent posttranslational protein modifications by an as yet unknown mechanism. A route of pyridine nucleotide-dependent, NO.-stimulated protein modification has recently been hypothesized [S. Dimmeler, and B. Brune, (1992) Eur. J. Biochem. 210, 305-310; J. S. Stamler (1994) Cell 78, 931-936]. An essential feature of this proposed mechanism is NADH nitrosation, for a nitroso-NADH adduct is considered to be a key reactant in the generation of pyridine nucleotide-modified protein. To evaluate at the molecular level the ability of NADH to act as a nitrosation substrate, the potential effects of NO., the nitrosothiols S-nitrosoglutathione and S-nitrosocysteine, the nitrosating agent tert-butyl-nitrite, and the NO. metabolite peroxynitrite on the molecular and functional (i.e., hydride-transfer) properties of NADH have been directly assessed at physiological pH. Exposure of NADH to NO. or nitrosothiol altered neither the hydride-transfer capability of the pyridine nucleotide nor its ultraviolet spectrum in ways suggestive of NADH nitrosation. As determined by NMR spectroscopy, NADH was refractory to the well-recognized nitrosating agent tert-butyl nitrite. Consequently, it appears that NADH is an unfavorable substrate for nitrosation under physiological conditions. These data are inconsistent with the proposal that NO. or a NO.-derived nitrosating agent interacts with NADH to generate the nitroso-NADH hypothesized to be essential to NO.-stimulated, pyridine nucleotide-dependent protein modification. Peroxynitrite, a possible source of nitrosating compounds, readily oxidized NADH to NAD, but demonstrated no potential to form a nitroso-NADH adduct. The facility with which NADH is oxidized to NAD has implications for peroxynitrite-mediated tissue damage.

Although nitrosothiols have been suggested to act as regulators of cell (patho)physiology, little is known about the pharmacology of nitrosylated proteins as nitric oxide (NO.) congeners. We describe the molecular consequences of nitrosylating bovine serum albumin (BSA) at multiple specific sites and demonstrate that the product S-nitrosoproteins exert NO.-like activity. The content of nucleophilic nitrosylation sites (i.e., free sulfhydryl groups) in native BSA was increased by either reduction with dithiothreitol or thiolation with N-acetylhomocysteine. Fourteen moles of nitrogen monoxide (NO)/mol BSA equivalent were then selectively positioned on either the endogenous sulfhydryl groups of reduced BSA or the homocysteine moieties of thiolated BSA, respectively. Each resulting S-nitrosoprotein adduct was an oligomeric mixture across the >2000 kDa to approximately 66 kDa molecular mass range. The BSA-derived S-nitrosoproteins were immunoreactive with antibodies against native BSA but evidenced compromised long-chain fatty acid binding. Both types of BSA-derived S-nitrosoproteins suppressed human coronary artery smooth muscle cell proliferation to a similar degree (IC50 approximately 70 microM NO. equivalents) and were significantly more effective antiproliferative agents than a standard NO. donor, DETA NONOate. Antiproliferative bioactivity reflected the NO functionalities carried by each protein, but was independent of molecular mass of the nitrosylated BSA adducts. These data exemplify the rational design and characterization of protein-based S-nitrosothiols as NO. congeners and suggest that such agents could have therapeutic potential as NO delivery systems.

By using in situ hybridization and immunohistochemistry, the distribution patterns of heme oxygenase (HO)-1 (HSP32) transcript and protein were studied, and their response to thermal stress was examined. And, by using an HO-1 cDNA probe and polyclonal antibody, the levels of HO-1 mRNA and protein in normal and heat-shocked testis were quantified, The digoxigenin-labeled probe detected a strong signal for HO-1 transcript in Leydig cells, and in the Sertoli cells, spermatogonia, primary spermatocytes, and spermatids of the seminiferous tubules. In all cell types, the transcript was predominantly concentrated in the nucleus in a defined pattern. Thermal stress (42 degrees C, 20 min) did not change the cell population pattern of HO-1 transcript expression; however, it did cause distortion of the nuclear pattern and diffusion of the transcript signal in cells, Hyperthermic treatment of rats resulted in a modest (2- to 2.8-fold), time-dependent, and sustained (1-16 h) increase in testicular 1.8-kb HO-1 mRNA. Immunohistochemical analysis of normal and heat shock patterns of testicular HO-1 expression showed robust staining of Sertoli and Leydig cells after heat shock; in normal tissue, immunoreactivity was low in these cell populations. As with the transcript distribution, hyperthermia did not affect the pattern of HO-1 immunoreactivity, and the protein was not detected in spermatogenic cells under control or stress conditions. In the Leydig cells, hyperthermia led to a more than 8-fold increase in the intensity of cytoplasmic staining for HO-1 protein. Consistent with the selective expression of HO-1, the level of the single HO-1 immunoreactive protein (similar to 32 kDa) detected in total testis microsomes showed a modest (1.5-fold) increase 6 h after heat shock. Data are consistent with the suggestion that differential distribution of HO-1 protein in the germ cell line and Sertoli cells reflects differential HO-1 mRNA processing in these cell types. The increase may be essential for the catalysis of the heme moiety of denatured hemoproteins such as cytochrome P450 and hemoglobin heme and hence may protect against heme-catalyzed free radical formation. We propose that induction of HO-1 protein in Sertoli and Leydig cells may function to protect the spermatogenic cells under conditions of thermal stress.

Activity of the stress protein, heme oxygenase-1 (hsp32; HO-1), produces carbon monoxide (CO), the potential messenger molecule for excitatory N-methyl-D-aspartate receptor-mediated events, in the hippocampus. Long-term stress caused by elevated adrenocorticoids induces pathological changes in CA1-CA3 neurons of the hippocampus; the adrenal hormones also exacerbate damage from stress. In rats chronically treated with corticosterone, we examined expression of HO-1 and its response to thermal stress in the hippocampus. An unprecedented appearance of scattered immunoreactive astrocytes marked the molecular layer of the hippocampus in corticosterone-treated rats. Steroid treatment showed no discernible effect on whole-brain HO-1 mRNA. When these rats were subjected to hyperthermia, neurons in the CA1-CA3 area, including pyramidal cells, exhibited intense immunoreactivity for the oxygenase and a pronounced increase (approximately 10-fold) in number. HO-1 is essentially undetectable in this area when rats are exposed to chronic corticosterone alone or thermal stress by itself, or in control rats. In contrast, similar analysis of hilar neurons showed no apparent effect on either the number or relative intensity of HO-1-immunostained cells after treatment. Corticosterone treatment also intensified the stress response of cerebellum, including Purkinje cells and Bergmann glia in the molecular layer. In brain, despite a pronounced reduction in NO synthase activity in corticosterone-treated and/or heat-stressed animals, the level of cyclic GMP was not significantly reduced. These observations are consistent with the hypothesis that responsiveness to environmental stress of CA1-CA3 neurons brought about by chronic elevation in circulating adrenocorticoids results in an increased excitatory neuronal activity and eventual hippocampal degeneration. Moreover, these findings yield further support for a role of CO in the production of cyclic GMP in the brain.

The heme oxygenase isozymes, HO-1 and HO-2, oxidatively cleave the heme molecule to produce antioxidants, the bile pigments, the gaseous cellular messenger, CO, and iron, a regulator of transferrin, ferritin, and nitric oxide synthase gene expression. HO-1 (hsp32) is a stress-inducible enzyme, whereas HO-2 is constitutively expressed at high levels in the testes and brain. In the present study, using immunohistochemical and in situ hybridization techniques, we report for the first time the cellular distribution of HO-1 and HO-2 in the testes of normal and heat-shocked rats and define a cell-specific expression of the isozymes and a stage-specific expression of HO-2 in the organ. In normal tissue, HO-1 was present at low levels in the Sertoli cells and could not be detected in germ or Leydig cells. HO-2, on the other hand, was most prominently expressed in residual bodies and was not detected in spermatogonia. Modest levels of HO-2 were observed in spermatocytes, spermatids, and select Leydig cells. In contrast, prominent expression of HO-2 messenger RNAs (mRNAs) was detected by in situ hybridization in spermatogonia, as well as spermatocytes, spermatids, and residual bodies of the seminiferous epithelium. The expression pattern of HO-2 protein and transcript in testes of heat-stressed (42 C; 20 min) rats did not differ from that in the control animals, whereas the expression pattern of HO-1 differed from that in the controls, in which distinct populations of Leydig and Sertoli cells displayed intense immunoreactivity. Thermal stress also resulted in an increase (2.8-fold) in the testicular HO-1 mRNA level within 1 h after treatment, followed by a significant increase (32%) in total microsomal heme oxygenase activity 6 h after treatment. Notably, this increase followed a significant depression (36%) in enzyme activity, which was detected 1 h after hyperthermia. The disparity between HO-2 mRNA and protein distribution clearly indicates cell-specific differences in the translational efficiency of HO-2 transcripts. It appears that HO-2 mRNA translation is linked to the maturation and expression of a factor(s) that regulates this process. This, in turn, appears to coincide with sperm development. HO-1 activity, on the other hand, which has a transcriptional component to its regulation, may have a role in maintenance of the conditions required for spermatogenesis.

The antioxidant enzyme superoxide dismutase (EC 1.15.1.1) (SOD) catalyzes the conversion of superoxide anion radical (O2.−) to hydrogen peroxide and molecular oxygen. SOD helps prevent tissue damage by O2.−and its metabolites, and augmentation of tissue SOD is a useful therapeutic strategy in certain diseases having an oxidative-injury component. Routine application of direct SOD assays is not technically facile, since the short half-life of the O2.−substrate and its free radical nature necessitate specialized analytical equipment to detect and measure O2.−chemically. Consequently, indirect SOD assays which monitor some change in an indicator substance reacting with O2.−are routinely used, particularly for biological samples. Limitations of indirect test systems utilizing heme-based indicators for the presence of O2.−and/or enzymatic O2.−generators led us to develop a SOD microassay based on spectrophotometric assessment of O2.−-mediated nitro blue tetrazolium reduction by an aerobic mixture of NADH and phenazine methosulfate, which produces superoxide chemically at nonacidic pH (Rao,Free Radical Biol. Med.7, 513–519, 1989). The proposed SOD assay system is formatted for use in an automated 96-well microplate reader and has the virtues of a nonheme indicator, a nonenzymatic O2.−source, physiological pH, and economy of time and materials. The assay has been applied to measure purified and tissue SOD (Cu,Zn- and Mn-types) activity as well as O2.−turnover by small-molecule “SOD mimetics.”

Biliverdin reductase regulates heme oxygenase activity by removing the inhibitory product of the oxygenase activity, biliverdin; and reducing it to bilirubin. The other products of the oxygenase are carbon monoxide and Fe. To date, biliverdin reductase remains unique among all enzymes described by using 2 different cofactors (NADPH and NADH) at different pH ranges. The present study reports on the developmentally regulated changes in the pattern of protein expression and the level of biliverdin reductase transcript in rat brain. Biliverdin reductase activity of the brain cytosol with both NADPH (pH 8.7) and NADH (pH 6.7) exhibited developmental changes with the activity increasing after birth, reaching an adult level by day 28 postpartum. When analyzed by Western blotting the immunoreactive protein detected increased as the animal matured (day 1 to 28 postparturition). Northern blot hybridization of RNA isolated from rat brain revealed the presence of approximately 1.5 kb biliverdin reductase transcript at all stages of development ranging from 1 day post partum to 20 months. The level of the transcript was developmentally regulated and a gradual increase ( approximately 4-fold) was observed from day 1 after birth to adulthood and was maintained in 20 month old animals. Cellular localization, using immunohistochemical technique, revealed age-related pattern of expression of the reductase in select regions such as the cortex, substantia nigra, hippocampus and in the cerebellum; the changes, however, did not follow the same pattern. To elaborate, in the cortex, the reductase expression increased when 7-day-old animals were compared with young adults (2 months old) and then declined in the 20-month-old animals. In the substantia nigra the level of reductase expression progressively declined with age when 7-day-old neonate, 2- and 20-month-old animals were compared. In the hippocampus, a distinct reductase-expressing cell population residing between CA1 and the dentate gyrus was observed in the 7-day-old animals; these cells were not detected in the adults (2 or 20 months old). In the cerebellum, the expression of the reductase reflected the developmental organization of this region. We postulate that age-dependent increase of the brain reductase at the transcript and protein levels in the course of maturation serves to control heme oxygenase activity which also displays a developmental pattern in the organ. As such, the reductase modulates generation of biologically active heme degradation products; bilirubin, carbon monoxide and Fe.

Presently we have investigated the carbon monoxide generating capacity of the cardiovascular system under normal and stress conditions by examining the microsomal heme oxygenase system at the transcript, protein and activity levels; and have assessed response of heart nitric oxide (NO) synthase activity and cyclic GMP levels to stress. Heme oxygenase (HO) isozymes, HO-1 (HSP32) and HO-2, catalyze the rate limiting step in the only known pathway in eukaryotes for the generation of the potential cellular message, carbon monoxide, and the antioxidant, bilirubin. We show expression of HO-1 and HO-2 at both the transcription and protein levels under normal conditions in the heart and descending aorta, and demonstrate the sensitivity of only the HO-1 isozyme to heat stress in these tissues. The ratio of the two HO-2 homologous transcripts (approximately 1.9 and 1.3 Kb) present in the atrium, ventricles and descending aorta and their levels were not altered by hyperthermia (42 degrees C, 20 min) when measured 1 or 6 hr after treatment. In contrast, hyperthermia caused a rapid, robust and coordinate increase of approximately 10- to 32-fold in the approximately 1.8-Kb HO-1 mRNA in these tissues when measured 1-hr post-treatment. Hyperthermia also caused a significant increase in both HO-1 protein and heme degradation capacity in the heart. Furthermore, the induction of HO-1 protein in the heart was accompanied by a significant elevation in tissue cyclic GMP level first detected 1-hr post-treatment and was sustained 6 hr after heat shock.(ABSTRACT TRUNCATED AT 250 WORDS)

Presently we describe, for the first time, induction of microsomal heme oxygenase-1 (HO-1) mRNA and protein in response to ischemia/reperfusion and therefore define HO-1 as stress protein in the kidney. Specifically, Northern blot analysis of kidneys of rats subjected to bilateral ischemia for 30 min revealed an increase of 8- to 10-fold in the level of 1.8 Kb HO-1 mRNA 6 hr after reperfusion. The increase in transcript level was maintained when assessed after 24 hr. The levels of 1.3 and 1.9 Kb transcripts for the second isozyme of HO, HO-2, were decreased at both time points. The increase in HO-1 mRNA was reflected in HO-1 protein level, as judged by Western blot analysis and at the level of activity as judged by the rate of bilirubin formation. An absence of change in adrenal HO-1 mRNA level subsequent to renal ischemia/reperfusion suggested that the induction of kidney HO-1 did not reflect a generalized response of the rat organs to stress; rather, it was a target organ specific response. Moreover, in kidneys subjected to ischemia 6 and 24 hr after reperfusion, significant increases in the cellular content of heme were observed; heme is a known inducer of HO-1 synthesis. Ischemia/reperfusion also adversely affected concentration of cytochrome P-450 in both mitochondrial and the microsomal fractions of the kidney. We suggest that increase in tissue heme levels may be a significant factor in damage caused by ischemia/reperfusion to renal tissue, whereby the metalloporphyrin promotes oxygen-free radical formation.(ABSTRACT TRUNCATED AT 250 WORDS)

The purpose of these studies was to determine whether the total immunoreactive alpha-inhibin protein concentration in seminal plasma correlated with serum gonadotropin levels or semen characteristics and to identify the forms of alpha-inhibin present in seminal plasma. Thirty-eight serum samples from men being evaluated for infertility were selected for study based on their serum hormone profiles and semen parameters. Serum LH and testosterone levels were normal, but FSH levels ranged from normal to hypergonadotropic (> 20 IU/L). Most semen parameters were within normal ranges, but germ cell numbers ranged from normal to azoospermic. Thus, seminal plasma from these men provided a unique opportunity to examine the antigenic forms of alpha-inhibin in individuals in whom strong correlations between inhibin and FSH levels might be predicted because of the observed ranges of FSH levels and germ cell numbers. Seminal plasma alpha-inhibin was characterized by RIA or Western blotting, using an antiserum directed against the N-terminal of the alpha-subunit of mature [32,000 mol wt (M(r))] inhibin. The antiserum recognized the alpha-subunit of dimeric inhibin as well as free alpha-inhibin and alpha-inhibin precursor proteins. Total immunoreactive alpha-inhibin ranged from 8.21-43.99 nmol/L in seminal plasma. However, alpha-inhibin levels were not statistically correlated with serum FSH levels or any of the measured semen parameters (including germ cell number). In contrast, the immunoreactive alpha-inhibin concentration in seminal plasma was negatively correlated (P < 0.01) with the serum LH level. Western blot analyses revealed that multiple forms of immunoreactive alpha-inhibin are present in seminal plasma. The majority of immunoreactivity was associated with monomeric proteins (ranging from 58,000-95,000 M(r)) that were larger than the alpha-subunit (21,000 M(r)) predicted for mature dimeric human inhibin (32,000 M(r)). The relative amounts of individual forms of immunoreactive alpha-inhibin varied among the patients studied, but could not be correlated with other serum or seminal parameters measured. Our observations demonstrate that various monomeric alpha-inhibin proteins are present in human seminal plasma. It is unlikely that these proteins alone or combined with inhibin beta-subunit proteins have identical biological activities. Thus, until assays specific for each of the various forms of immunoreactive alpha-inhibin are developed, their role as well as that of inhibin in the endocrine or local modulation of testicular function cannot be deduced from RIA data alone.

Two heme oxygenase (HO) isozymes-HO-1, which is a heat shock protein (HSP32), and HO-2-catalyze the isomer-specific production of biliverdin IXalpha and carbon monoxide. The latter has the potential of functioning as a neurotransmitter, whereas the reduced form of biliverdin, bilirubin, has potent antioxidant activity. Formation of bilirubin is catalyzed by biliverdin reductase (BVR). The reductase is a unique enzyme in being dual pyridine nucleotide and dual pH dependent. Here, we show that the reductase is resistant to thermal stress at both the protein and message level. We further demonstrate that the reductase is coexpressed in cells that display HO-1 and/or HO-2 under normal conditions, as well as in regions and cell types that have the potential to express heat shock-inducible HO-1 protein. Exposure of male rats to 42-degrees-C for 20 min did not decrease brain BVR activity, but caused a slight increase in NADPH- and NADH-dependent activities at 1 and 6 h following hyperthermia. High levels of the approximately 1.5-kb BVR mRNA were detected in control brain; it too displayed thermal tolerance. Similarly, the pattern of multiplicity of net charge variants of the enzyme purified from brain of heat-shocked rats did not differ from the control pattern. Immunochemical localization of BVR protein in normal brain correlated well with the presence of HO-1 and/or HO-2 throughout the forebrain, diencephalon, cerebellum, and brainstem regions. There were select neuronal and nonneuronal cells in the substantia nigra and cerebellum that did express the reductase under normal conditions, wherein no HO isozymes could be detected. The same population, however, responded to heat shock by an intense increase in the level of HO-1. We postulate that the constitutive presence of the reductase in this cell population and the overall thermal stability of the enzyme represent a safeguard mechanism in the brain for the prompt conversion of biliverdin to bilirubin under conditions when oxidation of the heme moiety of denatured hemoproteins by HO-1 is accelerated.

Abstract: In mammalian systems, the heme oxygenase (HO) isozymes HO‐1 (HSP32) and HO‐2 oxidatively cleave the heme molecule to produce bile pigments and carbon monoxide. Although HO‐1 is inducible by various chemicals in systemic organs and cell culture systems, this communication reports for the first time the induction of this stress protein and its transcript by a chemical in the brain. In addition, this study demonstrates expression of HO‐1 in select populations of cells in the brain in response to GSH depletion. Specifically, treatment of adult rats with diethyl maleate (DEM; 4.7 mmol/kg) caused a pronounced decrease in brain GSH content within 1 h. GSH levels remained significantly depressed for at least 24 h postinjection. Northern blot analysis of brain poly(A)+ mRNA following DEM treatment revealed on the average a sixfold increase in the 1.8‐kb HO‐1 mRNA level compared with that of controls; concomitant with this change was a decrease in GSH levels. Total brain HO activity was not significantly altered along with the increase in HO‐1 mRNA level. The increase in transcription of HO‐1 was a direct response to GSH depletion, as judged by the observation that treatment of neonatal rats with L‐buthionine‐(S,R)‐sulfoximine (BSO) (3 mmol/kg, twice daily, for 2 days), a selective inhibitor of GSH synthesis, caused a marked depression in total brain GSH level and a concomitant increase in brain 1.8‐kb HO‐1 mRNA content. The magnitude of the increase was up to ∼ 11.5‐fold that of the control level, as evidenced by northern blot analysis. In contrast, the level of two homologous HO‐2 transcripts (1.3 and 1.9 kb) did not increase in response to either DEM or BSO treatment. Analysis of brain HO‐1‐immunoreactive protein following DEM treatment for 9 h indicated induction of HO‐1 protein in only select nonneuronal cell populations. In particular, the ependymal cells lining ventricles throughout brain, Bergmann glia of cerebellum, and leptomeninges lining brain and glia throughout brain responded to treatment by increasing the level of HO‐1‐like protein. We suggest that when GSH is depleted, an increase in HO‐1 protein content, resulting in increased capacity to form bile pigments, may be of significance to cells with compromised antioxidant capability. Bile pigments are potent antioxidants in biological systems.

To examine the role of carbon monoxide (CO) as a putative neuronal messenger and regulator of cGMP level in vivo, we exploited an animal model to increase brain capability to generate CO. The sole source of CO in mammalian systems is the α-meso carbon bridge of the heme molecule cleaved by heme oxygenase isozymes, HO-1 and HO-2. In adult animals, the noninducible isozyme HO-2 is the predominant form in the brain. We chose to increase, rather than inhibit, brain heme oxygenase activity because synthetic metalloporphyrins, such as Zn-protoporphyrin, which are the only known effective inhibitors of the isozymes, are also potent inhibitors of soluble guanylate cyclase, the enzyme that generates cGMP. In newborn rats both heme oxygenase isozymes were found expressed at low levels, and in the cerebellum heme oxygenase activity could be induced by treatment of 2-day-old animals with a selective depletor of glutathione, buthionine-SR-sulfoximine. The increase in activity was accompanied by marked increases in HO-1 protein and the 1.8 kb HO-1 mRNA in the cerebellum. Despite a pronounced decrease in activity of the hemoprotein nitric oxide synthase, no change in cGMP level was observed. The decrease in the synthase could not be explained by an inhibited heme biosynthesis activity. This unchanged level of cGMP suggests that NO is not the only gaseous heme ligand that can activate guanylate cyclase resulting in the generation of cGMP, but rather that CO may also function in this capacity. Increased capability of select cerebellar cell populations to generate CO, as indicated by an increase in their HO-1 protein content, points to the active role of this isozyme in maintenance of cGMP level under stress conditions, when nitric oxide production is compromised. The cell populations expressing HO-1 protein included those in pia matter and glia, such as astrocytes.