Distribution and Cellular Localization of NOS in the Central Nervous System

Since NO is a highly reactive unstable species our knowledge on the distribution of the cells that use this signaling molecule throughout the nervous system relies on the possibility of localizing NOS, the key enzyme in NO synthesis, in tissue sections. The purification and cloning of the various NOSs allowed for the development of antibodies and nucleic acid probes that have, in turn, permitted the immunocytochemical mapping of NOS and the localization in situ of the NOS mRNA. Moreover, the redox activity of NOS accounted for the histochemical NADPH-d reaction in aldehyde-fixed tissues (30,31). Accordingly, in several areas of the brain a one-to-one correspondence between neurons expressing NOS mRNA or NOS immunoreactivity and those showing NADPH-d positivity was found (30,32). Moreover, the recent demonstration that after knocking out the NOS gene there is a complete loss of NADPH-d activity in the nervous system provided definitive evidence for the correspondence of NADPH-d and NOS activities (33). A comparative evaluation of the different methods so far employed for localizing NO-producing neurons is beyond the purpose of this work. The issue has recently been reviewed in details and readers are referred to the literature on this matter (34).

In the nervous system NOS is mainly localized in neurons. NOS is present in several regions of the mammalian brain including the telencephalon, dienceph-alon, midbrain, pons, medulla, cerebellum, and spinal cord (for review, see Ref. 34). Different neuronal types were found to contain NOS, sometimes associated with other signaling molecules in these locations. Being relevant to our work, we will briefly review the data on spinal cord localization (Figs. 2 and 3). Several investigators have described the distribution of NOS-containing neurons in the rat, mouse, cat, and primate. Positive neurons are concentrated in the dorsal horn (laminae I-IV) and the central gray matter (lamina X) at all segmental levels (35,36). Moreover, in all species studied so far, the cholinergic preganglionic autonomic neurons in the intermediolateral cell columns of the thoracic spinal cord are highly positive after either NOS immunocytochemistry or NADPH-d staining, as well as most sacral parasympathetic preganglionic neurons (35). In the human spinal cord some ventral horn motor neurons are also NOS-immunore-active (37). Many of the NADPH-d positive neurons in the dorsal horn are also GABA- and glycine-immunoreactive (38). The outer lamina II (lamina IIo) is particularly rich of NOS-containing neuronal processes. The major contribution to neuropil staining in lamina II of rats appears to be related to projection fibers

Figure 2 Distribution of NADPH-d-positive neurons in the spinal cord. Positive neurons and processes are evident in the superficial dorsal horn (A), the gray matter surrounding the central canal (B), and deep laminae of the dorsal horn (C). wm, white matter. Roman numerals indicate the laminae of the dorsal horn. Scale bars: A = 50 |J,m; B = 100 |J,m; C = 10 |lm.

Figure 2 Distribution of NADPH-d-positive neurons in the spinal cord. Positive neurons and processes are evident in the superficial dorsal horn (A), the gray matter surrounding the central canal (B), and deep laminae of the dorsal horn (C). wm, white matter. Roman numerals indicate the laminae of the dorsal horn. Scale bars: A = 50 |J,m; B = 100 |J,m; C = 10 |lm.

as well as processes from local circuit neurons (35). However, some staining may arise from primary afferents at least at the thoracic and lumbar levels (39). In other species, such as cats and humans, the contribution from primary afferents to neuropil staining in the dorsal horn might be more substantial (36,40). As mentioned in the previous section, after biochemical analysis both a soluble and a membrane-bound insoluble fraction of NOS were found in the brain. We have recently demonstrated that the use of different tetrazolium salts for the histochem-ical localization of NADPH-d yields different subcellular localization of positive reaction sites (41) (Fig. 4). When observed in the electron microscope, nitroblue tetrazolium (NBT) formazan deposits were scattered throughout the neuronal cell body and processes without any preferential association with cell organelles. In keeping, those who have used NOS immunocytochemistry have not been able to observe a specific association with any subcellular organelle or membrane

Figure 3 Distribution of nNOS-immunoreactive neurons and nNOS mRNA in the spinal cord. nNOS-immunoreactive neurons in lamina IV (A) and X (B) after immunolabeling with a specific monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA). Neurons expressing the nNOS mRNA in lamina II (C) and III (D). Arrows in C indicate the sites of mRNA accumulation. In situ hybridization of nNOS mRNA was performed using a pool of three 45-mer antisense oligonucleotides complementary to nucleotides 211266, 795-840, and 4712-4757 of the cloned cDNA, end-labeled with digoxigeninated-11-dUTP. Scale bars: A, C, D = 25 |lm; B = 50 |lm.

(42). On the other hand, when we have used 2-(2'-benzothiazolyl)-5-styryl-3-(4'-phthalhydrazidyl)tetrazolium chloride (BSPT) as a chromagen, enzyme activity appeared to be localized to the rough endoplasmic reticulum, mitochondria, and nuclear membranes, as previously observed in some other areas of the brain (4345). Therefore, it seems reasonable to assume that both a membrane-bound and a cytosolic fraction of the enzyme exist in neurons in keeping with the already discussed results obtained after biochemical characterization of tissue homoge-nates.

Although nNOS was originally regarded as being constitutively expressed solely in neurons, there is increasing evidence that astrocytes also contain a con-

Figure 4 Ultrastructural visualization of NADPH-d-positive neurons. The diaphorase reaction can be revealed by using different chromagens such as NBT, which gives a floc-cular reaction product uniformly distributed with the cytoplasm of positive cells (A). When BSPT is used, the formazan deposits are clearly localized to the membrane of several cell organelles (B), including the endoplasmic reticulum (insert). Scale bars: A, B = 1 |J,m; insert = 0.5 |J,m.

stitutive, calcium-activated NOS (for review, see Ref. 46). Astrocytes are intimately associated with blood vessels and neurons, and it is now widely recognized that they have a role not only in development and response to injury, but also in normal cell-to-cell signaling (47). Although most of the studies on the localization of NOS in the central nervous system suggested that its distribution was entirely neuronal (see above), the immunocytochemical localization of l-arginine revealed a predominantly astrocytic distribution (48). Moreover, immu-nocytochemical labeling in the cerebellum besides the neurons revealed a positive reaction over several nonneuronal types including the Bergmann glia and astrocytes (49). Furthermore, NADPH-d was found to colocalize with NOS in both neuronal and glial cells, and a nonneuronal expression of NADPH-d was also reported in the perivascular glia (45). At the electronic level it appeared that NADPH-d activity is highly concentrated in neurons and endothelial cells (41,45). We have been unable to detect any glial staining after ultrastructural

NADPH-d staining (41), but it was previously claimed that enzyme activity was ubiquitously distributed in all cells of the central nervous system, albeit at lower concentration than in neurons and endothelial cells (45).

As to the localization of iNOS, it was mainly described in microglial cells (for review, see Ref. 46). Once induced, this NOS isoform is continually active over a period of hours. Generally NO production by the microglia requires transcription and translation, and needs l-arginine and NADPH, but not Ca2+ and calmodulin. Evidence has also been provided for the existence of iNOS in astrocytes in vitro (20,50,51). The astrocyte iNOS cDNA has also been cloned and appeared to share 91% sequence homology with mouse macrophage enzyme (20). In vitro studies suggested that a single astrocyte is likely to produce less NO than a single microglial cell, but since astrocytes are by far more numerous than microglia, their contribution could be substantial (46).

Was this article helpful?

0 0

Post a comment