α-NADPH appears to be primarily oxidized by the NADPH-diaphorase activity of nitric oxide synthase (NOS)

Acta histochem. (Jena) 97,313-320 (1995) Gustav Fischer Verlag Jena . Stuttgart· New York

a-NADPH appears to be primarily oxidized by the NADPHdiaphorase activity of nitric oxide synthase (NOS) Zarko Grozdanovic and Reinhart Gossrau Department of Anatomy, Free University of Berlin, Konigin-Luise-Strasse 15, D-141t)5 Berlin, Germany Accepted 7 April 1995

Summary Biochemical studies have shown that the NADPH-diaphorase (NADPH-d) activity of nitric oxide synthase (NOS) represents only a part of the total cellular diaphorase pool. Histochemically, NADPH-d activity can be demonstrated in cells expressing no constitutive NOS. Therefore, attempts aimed to improve the specificity of the NADPH-d reaction are currently being undertaken. In this study, the effect of replacing the natural and common diaphorase substrate P-NADPH with the artificial stereoisomer a-NADPH on the extent of NADPH-d staining was examined. When P-NADPH served as the substrate, discrete populations of central and peripheral neurons as well as numerous non-neural cells in many organs of common laboratory rodents (mouse, rat, gerbil, hamster, guinea pig) and marmosets were found to generate formazan. Substitution of a-NADPH for P-NADPH resulted in reduced staining intensity of nerve cells and muscle fibers. Furthermore, a-NADPH-d staining of macula densa cells, enterocytes and granulocytes varied according to the species examined. No reaction was observed in most other cells which stained positively for P-NADPH-d activity. Examination of adjacent sections, incubated for the demonstration of NOS-immunoreactivity, revealed that aNADPH-d activity and NOS immunostaining are strictly colocalized in neurons, striated muscle fibers and, species-dependently, in macula densa cells. It can thus be concluded that, with the exception of gut granulocytes, a-NADPH is primarily metabolized by the reductase activity of NOS.

Key words: nitric oxide synthase - NOS - NADPH-diaphorase - a-NADPH

Introduction Ongoing studies in our laboratory are designed to improve the specificity of the NADPH/tetrazolium salt system for the selective histochemical visualization of the socalled NADPH-diaphorase (NADPH-d) activity of nitric oxide synthase (NOS; EC 1.14.23.39). NOS is a recently purified and characterized NADPH- and 02-consuming oxidoreductase which catalyzes the transformation of L-arginine to the radical and Correspondence to: Z. Grozdanovic

314

Z. Grozdanovic and R. Gossrau

signalling molecule nitric oxide (NO) and L-citrulline (Bredt and Synder, 1994; Knowles and Moncada, 1994; Masters, 1994; Sessa, 1994). The cloned and expressed NOS has been found to consist of an oxygenase domain, containing Fe2 + /protoporphyrin IX as a prosthetic group, and a reductase domain, exhibiting consensus sequences for NADPH, FAD, and FMN binding. By analogy to cytochrome P-450 reductase (CPR), which shows a close homology to the NOS reductase domain, NADPH-derived electrons are supplied by the flavin-mediated redox-cycle to the heme which directly participates in the activation of molecular oxygen. Moreover, electrons can be passed on to artificial acceptor molecules such as tetrazolium salts, e. g. nitroblue tetrazolium (NBT), resulting in light microscopically visible formazan formation. However, accumulating biochemical and morphological data converge to indicate that this NADPH-d activity is present in cells expressing no NOS under physiological conditions (Matsumoto et aI., 1993; Tracey et aI., 1993; Nakos and Gossrau, 1994; Grozdanovic and Gossrau, 1995). Several histochemical studies have attempted to increase the specificity of the NADPH-d reaction for NOS visualization by adding diverse chemical agents, alone or in combination, into the incubation medium. Thus, it has been claimed that NADP inhibits NADPH-ferrihemoprotein, cytochrome C CPR, or dicoumarol quinon-dependent NADPH dehydrogenases and can, therefore, be used to discriminate between NOSrelated NADPH-d and other diaphorases (for references see Stoward et aI., 1991; Nakos and Gossrau, 1994; Spessart et aI., 1994). At present, the closest correlation between NADPH-d activity and immunoreactivity to neuronal-type NOS, which is the most widely distributed constitutive NOS isoform, is seen when the histochemical reaction is performed in the presence of certain (not all) aldehydes, such as formaldehyde and glutaraldehyde, or the oxidizing compounds hydrogen peroxide, diamide or permanganate; the reason for this differentiation possibility is that the NADPH-d of NOS is more resistent to fixation aldehydes and oxidants than the other diaphorases (NADPH dehydrogenases; Nakos and Gossrau, 1994; Grozdanovic et aI., 1995b; Nakos, Grozdanovic and Gossrau, unpublished observations). Studies dealing with the effects of NADPH analogues, especially the stereoisomer 0/NADPH, on the extent of NADPH-d staining have yielded conflicting results. In a systematic investigation of the histochemical properties of NADPH-d activity in the striatum of rats, Hope and Vincent (1989) found that substitution of O/-NADPH for pNADPH resulted in a slight decrease in formazan production. Persson et aI. (1993) examined the influence of O/-NADPH replacement on NADPH-d reaction in nerve fibers supplying the pig lower urinary tract and found that staining of NADPH-d activity was strongly reduced or even abolished in the presence of O/-NADPH. In a recent analysis of NADPH-d activity in the rat olfactory bulb, Spessert et aI. (1994) observed that when pNADPH was substituted with O/-NADPH, the NADPH-d reaction in neuronal perikarya was absent, whereas it remained unaffected in glomeruli. No data exist for the impact of this artificial redox carrier on NADPH-d staining in extra-neural tissues. The aim of the present study was, therefore, to systematically examine the effects of O/-NADPH on NADPH-d activity in many cells and tissues of several rodent species and marmosets. From our data, it has become clear that O/-NADPH offers a further possibility to differentiate between NOS-related and non-NOS NADPH-d activity.

Materials and Methods Animals, tissue pretreatment. Adult Wistar rats, Han: NMRI mice, hamsters, gerbils, guinea-pigs, and marmosets of either sex were bred on the premises and kept under standardized laboratory conditions (artificiallight/dark cycle, light from 7 a. m. to 7 p. m.; temperature 22 ± 1 0c) with free access to Altromin diet (Altromin, Lippe, FRO) and tap water. The animals were sacrificed under Nembutal anesthesia, and

a-NADPH oxidation by NOS

315

the following organs and tissues were then quickly excised: cerebrum, cerebellum, spinal cord, skin, submandibular and extraorbital gland, tongue, thymus, lymph nodes, trachea, lung, heart, esophagus, stomach, duodenum, jejunum, ileum, ascending, transversal and descending colon, rectum, liver, spleen, pancreas, adrenal gland, kidney, ureter, urinary bladder, urethra, prostate, seminal vesicle, testis, epididymis, ductus deferens, ovary, Fallopian tube, uterus, vagina as well as fore- and hindleg skeletal muscles. Samples of these tissues were placed together onto cork plates, covered with wet filter paper, wrapped with plastic foil, and frozen in liquid N2• Sections 10 Ilm in thickness were cut on a cryostat (Frigocut 2800; Reichert-lung, NuJ31och, FRG) and thawed onto chrome alum/gelatin-coated glass slides. Sections were fixed with 4% formaldehyde for 10- 30 min at 4°C, rinsed in tap water for 10 min and then in distilled water for 5 min. Enzyme histochemistry. The NADPH-d reaction was carried out according to Scherer-Singler et al. (1983) by incubating tissue sections in 0.1 M phosphate buffer (pH 7.6) containing 1 mg fJ-NADPH (Biomol, Hamburg, Germany)/ml, 0.25 mg NBT (Serva, Heidelberg, Germany)/ml and 0.3% Triton X-l00 (v/v) for 30-60 min at 37°C. Sections incubated in the absence of fJ-NADPH served as negative controls. a-NADPH (Sigma, Munich, Germany) was substituted at equimolar concentration for fJ-NADPH in the NADPH-d procedure. In addition, specimens were incubated in the presence of both fJ-NADPH and a-NADPH, at 1 mg/ml each. After incubation, the media were poured off, and the sections were rinsed in tap water for 10 min, then in distilled water for 5 min and mounted in glycerol jelly. , Immunofluorescence. Formaldehyde-fixed cryosections were soaked in phosphate-buffered saline (PBS; 0.1 M, pH 7.4) containing 0.3"70 (v/v) Triton x-tOO (15 min), exposed to a 2"70 solution of normal horse serum (30 min), rinsed in Triton-PBS (15 min) and incubated in a rabbit antiserum to pig brain NOS (Mayer et al. 1990), diluted 1 : 1000, in a humid chamber for 18 - 24 h at room temperature. The bound anti-NOS antibody was detected using a Cy3-conjugated goat antibody to rabbit-IgG (Jackson, West Grove, PA, USA), diluted 1: 100, for 1 h at room temperature. After several PBS washes, the sections were air-dried and coverslipped with glycerol/PBS (1: 1). Controls were performed by substituting non-immune rabbit serum or PBS for the primary antiserum.

Results NADPH-d histochemistry. When ,B-NADPH was used for NADPH-d mapping, NBTformazan production was observed in very particular populations of central and peripheral neurons a well as in many non-neural cells in all animal species examined. Among the latter, epithelial cells in various locations were found to stain positively for NADPH-d activity: acinar and secretory duct cells in the submandibular gland, acinar cells in the lacrimal gland, Clara cells and type II pneumocytes of the lung, surface epithelial cells of the tongue, pharynx and esophagus, parietal and surface epithelial cells of the stomach, hepatocytes, epithelial cells of the gall bladder, acinar cells and endocrine cells of the pancreas. Other NBT-formazan producing epithelial cells were those in the macula densa, proximal and distal tubules as well as in the collecting ducts of the kidney, epithelium of the ureter and urinary bladder, Fallopian tube, uterus, vagina as well as in the epididymis, ductus deferens and penis. Furthermore, diaphorase activity was found in vascular and non-vascular smooth muscle cells, capillary and arterial endothelial cells, Bowman cells of the kidney, renal glomerular cells, endocardial cells, plasma membrane (sarcolemma) of skeletal muscle fibers in all kinds of somatic and visceral muscles, thymic medullary cells and endothelial cells of the spleen. Replacing ,B-NADPH with the a-stereoisomer resulted in a completely different staining pattern. Staining by NADPH-d activity was largely restricted to nerve cell bodies and fibers in the central (Fig. 1) and peripheral nervous system. In comparison with the reaction within central neurons, the staining of myenteric plexus ganglion cells (Figs. 2 - 6) was stronger. Additionally and depending on the species investigated, staining by NADPH-d was also found in small intestinal enterocytes (Fig. 7) and macula densa

3

4

Fig. 1. Using a-NADPH as the substrate formazan is present in perikarya (arrows) and nerve fibers (arrowheads) of the marmoset cerebrum. Figs. 2-4. In the myenteric plexus formazan is found with a-NADPH as the substrate in perikarya (arrows) and nerve fibers (arrowheads). C circular layer, L longitudinal layer of the muscularis; S submucous layer, Cr crypts, thin arrows granulocytes oxidizing a-NADPH. Fig. 2. Marmoset stomach. Fig. 3. Rat, colon. Fig. 4. Hamster, ileum. Magnification of all Figs. x 381.8.

Figs. 5, 6. Formazan is produced with a-NADPH as the substrate in perikarya (arrows) and nerve fibers (arrowheads) of the myenteric plexus. Fig. 5. Guinea-pig, jejenum. Fig. 6. Gerbil, colon. Fig. 7. In gerbils jejunal enterocytes (arrows) produce great amounts of formazan. Fig. 8. In the jejenum of rats granulocytes (arrows) in the lamina propria are heavily stained by formazan with a-NADPH as the substrate. Arrowheads weakly active enterocytes. Magnification of all Figs. x 377.2.

318

Z. Grozdanovic and R. Gossrau

cells, in granulocytes of the gut (Fig. 8) and spleen as well as in the sarcolemma of skeletal muscle fibers, e. g. in the tongue and esophagus. The granulocytes appeared to produce more formazan with a-NADPH as the substrate than with P-NADPH. The other cells and tissues, which stained positively when P-NADPH was employed as the substrate, showed no clear activity. Co-incubation with a-NADPH and P-NADPH yielded results as with P-NADPH alone. NOS-immunoreactivity. With the exception of enterocytes and granulocytes, NOSimmunoreactivity was detected at the same sites which also reacted for a-NADPH-d activity, i. e., in particular populations of central and peripheral neurons, macula densa cells and in the plasma membrane region of somatic and visceral striated muscle fibers. Furthermore, some additional cells were immunolabeled such as Bowman cells, endothelial cells and thymic medullary cells (see Grozdanovic et al., 1995 b).

Discussion This study shows that substitution of a-NADPH for P-NADPH consistently results in a reduced NADPH-d activity in neuronal perikarya and processes as well as in the area of sarcolemma of striated muscle fibers, whereas staining of macula densa cells, enterocytes and gut granulocytes depends on the species investigated. In most of the other cells and tissues, replacement with the a-stereoisomer resulted in lack of NADPHd activity. Our findings appear to corroborate the data of Hope and Vincent (1989), but contrast with the results of Persson et al. (1993) and Spessert et al. (1994), who found no staining of neuronal elements in the presence of a-NADPH. By contrast to the aforementioned studies, which have concentrated on the effect of the a-stereoanalogue on NADPH-d in the nervous tissue of one species, we have extended our investigations to include numerous neural and non-neural cells and tissues in a variety of species. In all species examined, the distribution of a-NADPH-d activity closely resembles that of NOS mapped by immunohistochemistry. Moreover, the pattern of a-NADPH-d staining parallels P-NADPH-d results obtained with the modified - formaldehyde, permanganate - reaction procedures (Nakos and Gossrau, 1994; Grozdanovic et al., 1995 a). Thus, although P-NADPH is the preferable substrate for the NADPH-d activity of NOS, it is the a-isomer which is selectively metabolized by this diaphorase. The observation that in the presence of both a-NADPH and P-NADPH, NADPH-d activity is unaffected indicates that a-NADPH is not an inhibitor of the histochemical reaction, but is rather preferentially accepted by NOS-associated NADPH-d. The absence of a-NADPH-d staining in macula densa cells, Bowman cells, and arterial endothelial cells of certain species, in which P-NADPH-activity and NOS-immunoreactivity have been shown to co-exist, can tentatively be explained by the relatively low oxidation rates of a-NADPH. Thus, cells with too low NADPH-d activity of NOS cannot be visualized by a-NADPH. There is evidence to assume that a-pyridine nucleotides such as a-NADPH, in which the ribosylnicotinamide bond is in the a-configuration, do not occur naturally (Jacobson et al., 1973). Possibly, the NADPH-binding domain of the NOS-associated NADPH-d is less stereospecific than that of other diaphorases, so that both the a- and the p-isomer can be oxidized. Cloning of a cDNA for the neuronal NOS has revealed that the NADPH-ribose and NADPH-adenine recognition sites of NOS are highly homologous to similar sequences in other oxidoreductases, such as NADPH cytochrome P450 reductase, sulphite reductase and ferredoxin NADP+ reductase (Bredt et al., 1991). It may thus be speculated that the ability of NOS to oxidize a-NADPH is dependent on its secondary and/or tertiary structure.

o:-NADPH oxidation by NOS

319

Biochemical studies indicate that NOS is constitutively expressed in epithelial cells of the rat small intestine (Tepperman et aI., 1993) and of the pig small and large intestine (M'Rabet Touil et aI., 1993). NOS purified from the pig enterocytes and colonocytes is found to be Ca 2 + - and flavin-independent (M'Rabet-Touil et aI., 1993). However, there is a lack of correlation between NADPH-d activity and NOS immunochemical staining in intestinal epithelial cells. Enterocytes exhibit a strong histochemical reaction, which may either be caused by a co-reacting NADH-diaphorase activity (Grozdanovic and Gossrau, 1995) or by the NADPH-d activity of NOS not detectable immunohistochemically (Costa et aI., 1992; Springall et al., 1992; Grozdanovic et aI., 1995 b). It may thus be assumed that gut NOS possesses an amino-acid sequence which cannot be recognized by antobodies raised against the neuronal isoenzyme. The high oxidation rates of o:-NADPH, which can exceed those of fJ-NADPH, in granulocytes are presumably not caused by the NADPH-d activity of constitutive NOS. Available information indicates that granulocytes express an inducible isoform of NOS under conditions of increased endotoxin release or/and cytokine production (McCall et ai., 1991 a, b). Therefore, either this staining may represent a low level stimulation of the inducible isoenzyme or another NADPH-oxidizing enzyme appears to be responsible for NBT-formazan production in granulocytes. Summing up, the data referred to above strengthen the idea that o:-NADPH can be used to differentiate between the NADPH-d activity of constitutive NOS and other NADPH oxidizing enzymes. Moreover, this compound may provide further information on the molecular behavior of the reductase activity of NOS making it such a special enzyme molecule.

Acknowledgements We thank Dr. B. Mayer for the donation of the NOS antibody, Ms R. Richter for her technical assistence, Ms U. Sauerbier for photographic work and Ms U. Saykam for her help with the preparation of the manuscript.

References Bredt DS, Hwang PM, Glatt CE, Lowenstein C, Reed RR, and Snyder SH (1991) Cloned and expressed nitric oxide synthase resembles cytochrome P-450 reductase. Nature 351: 714 -718 Bredt DS, and Snyder SH (1994) Nitric oxide: a physiologic messenger molecule. Annu Rev Biochem 63: 175-195 Costa M, Furness JB, Pompolo S, Brookes SJH, Bornstein JC, Bredt DS, and Snyder SH (1992) Projections and chemical coding of neurons with immunoreactivity for nitric oxide synthase in the guineapig small intestine. Neurosci Lett 148: 121 -125 Grozdanovic Z, and Gossrau R (1995) Non-specific alkaline phosphatase activity can be responsible for staining of NADPH-diaphorase activity in certain non-neural cells. Folia Histochem Cytobiol 33: 3-10 Grozdanovic Z, Nakos G, Christova T, Nikolova Z, Mayer B, and Gossrau R (1995a) Demonstration of nitric oxide synthase (NOS) in marmosets by NADPH diaphorase (NADPH-d) histochemistry and NOS immunohistochemistry. Acta histochem 97: 321-332 Grozdanovic Z, Nakos G, Mayer B, and Gossrau R (1995 b) A modified method allows for correlation between NADPH-diaphorase histochemistry and immunohistochemistry for the demonstration of neuronal nitric oxide synthase (nNOS). Folia Histochem Cytobiol 33: 11 -18 Hope BT, and Vincent SR (1989) Histochemical characterization of neuronal NADPH-diaphorase. J Histochem Cytochem 37: 653 -661 Jacobson EL, Jacobson MK, and Bernofsky C (1973) Evidence against the natural occurrence of C
320

Z. Grozdanovic and R. Gossrau

Masters BSS (1994) Nitric oxide synthases: why so complex? Annu Rev Nutr 14: 131-145 Matsumoto T, Nakane M, Pollock JS, Kuk JE, and Fllrstermann U (1993) A correlation between soluble brain nitric oxide synthase and NADPH-diaphorase activity is only seen after exposure of the tissue to fixative. Neurosci Lett 155: 61-64 McCall TB, Feelisch M, Palmer RMJ, and Moncada S (1991 a) Identification of N-iminoethyl-L-ornithine as an irreversible inhibitor of nitric oxide synthase in phagocytic cells. Br J Pharmacol 102: 234 - 238 McCall TB, Palmer RMJ, and Moncada S (1991 b) Induction of nitric oxide synthase in rat peritoneal neutrophils and its inhibition by dexamethasone. Eur J Immunol 21: 2523 - 2527 M'Rabet-Touil H, Blachier F, Morel M-T, Darcy-Vrillon B, and Duee P-H (1993) Characterization and ontogenesis of nitric oxide synthase activity in pig enterocytes. FEBS Lett 331: 243 - 247 Nakos G, and Gossrau R (1994) When NADPH diaphorase (NADPHd) works in the presence of formaldehyde, the enzyme appears to visualize selectively cells with constitutive nitric oxide synthase (NOS). Acta histochem 96: 335 - 343 Persson K, AIm P, Johansson K, Larsson B, and Andersson K-E (1993) Nitric oxide synthase in pig lower urinary tract: immunohistochemistry, NADPH diaphorase histochemistry and functional effects. Br J Pharmacol 110: 521-530 Sessa WC (1994) The nitric oxide synthase family of proteins. J Vasc Res 31: 131-143 Spessert R, Wohlgemuth C, Reuss S, and Layes E (1994) NADPH-diaphorase activity of nitric oxide synthase in the olfactory bulb: co-factor specifcity and characterization regarding the interrelation to NO formation. J Histochem Cytochem 42: 569 - 573 Springall DR, Riveros-Moreno V, Buttery L, Suburo A, Bishop AE, Merrett M, Moncada S, and Polak JM (1992) Immunological detection of nitric oxide synthase(s) in human tissues using heterologous antibodies suggesting different isoforms. Histochemistry 98: 159 - 266 Stoward PJ, Meijer AEFH, Seidler E, and Wohlrab F (1991) Chapter 22. Dehydrogenases. In: Stoward P, Pearse AGE (eds). Histochemistry. Theoretical and applied. 4th edn, vol 3. Churchill Livingstone, Edinburgh, London, pp 27 -71 Tepperman BL, Brown JF, and Whittle BJR (1993) Nitric oxide synthase induction and intestinal epithelial cell viability in rats. Am J Physiol 265: G214-G218 Tracey WR, Nakane M, Pollock JS, and Fllrstermann U (1993) Nitric oxide synthases in neuronal cells, macrophages and endothelium are NADPH diaphorases, but represent only a fraction of total cellular NADPH diaphorase activity. Biochem Biophys Res Commun 195: 1035 -1040