Disulfiram Administration Affects Substance P-like Immunoreactive and Monoaminergic Neural Systems in Rodent Brain*

The biosynthetic enzyme peptidylglycine cY-amidat-ing monooxygenase catalyzes the formation of a variety of biologically active cY-amidated peptides from respective COOH-terminal glycine-extended peptide precursors. Peptidylglycine a-amidating monooxygenase activity is dependent on copper, ascorbate, and molecular oxygen and is inhibited by the relatively selective copper chelator N,N-diethyldithiocarbamate or its disulfide dimer disulfiram (Antabuse). In the present study, chronic disulfiram treatment (100 mg/ kg/day, for 12-25 days) resulted in significant changes in several neurochemical parameters in the mouse central nervous system, including levels of substance P-like, unamidated substance P-Gly-like, and protease-generated substance P-Gly-Lys-like immunoreactivities (SP-LI, SP-G-LI,

The biosynthetic enzyme peptidylglycine cY-amidating monooxygenase catalyzes the formation of a variety of biologically active cY-amidated peptides from respective COOH-terminal glycine-extended peptide precursors.
Peptidylglycine a-amidating monooxygenase activity is dependent on copper, ascorbate, and molecular oxygen and is inhibited by the relatively selective copper chelator N,N-diethyldithiocarbamate or its disulfide dimer disulfiram (Antabuse).
In the present study, chronic disulfiram treatment (100 mg/ kg/day, for 12-25 days) resulted in significant changes in several neurochemical parameters in the mouse central nervous system, including levels of substance Plike, unamidated substance P-Gly-like, and proteasegenerated substance P-Gly-Lys-like immunoreactivities (SP-LI, SP-G-LI, and SP-G-K-LI, respectively). Combined high performance liquid chromatography/ radioimmunoassay analyses of the extracted SP-LI, SP-G-LI, and SP-G-K-L1 species indicated very similar chromatographic and immunochemical behavior as demonstrated for chemically authentic peptide standards. Additionally, changes in levels of monoamines and their metabolites were observed after drug administration.
Complementary immunohistochemical analyses using affinity-purified anti-SP-G sera localized these drug-induced changes in levels of immunoreactive unamidated precursor to neural elements that normally express SP. As  Substance P (SP),' a neuropeptide with widespread distribution throughout the brain, spinal cord, and autonomic nervous system of many mammalian species, is involved in transmission and modulation of sensory information, regulation of neuroendocrine processes, and integration of motor function (1). SP is typical of a variety of gut-brain peptides that contain an amidated COOH terminus, a posttranslational modification that is required for full biological and pharmacological activity (l-3). It has been demonstrated that precursor forms of amidated peptides are found as COOH-terminal glycine-extended species (4-7), with glycine serving as the nitrogen donor in the amidation reaction (8,9). Conversion of glycine-extended peptide precursors to biologically mature a-mediated peptides is catalyzed by peptidylglycine cY-amidating monooxygenase with cofactor requirements of copper, ascorbate, and molecular oxygen (10). Previous studies have reported a time-dependent loss of peptide amidation in several primary cell culture systems, probably due to cofactor depletion (10, ll), suggesting that this terminal maturation reaction may be rate-limiting in peptide biosynthesis and susceptible to modulation by specific drugs. Additionally, it has been recently demonstrated that administration of the relatively selective copper chelator iV,N-diethyldithiocarbamate or its disulfide dimer, disulfiram (Antabuse), produces a dose-dependent increase in glycine-extended immature forms of Lu-melanocyte-stimulating hormone and of joining peptide in the intermediate pituitary and in cultured mouse corticotrophic tumor cells, via inhibition of peptidylglycine cy-amidating monooxygenase activity (12). However, the effects of inhibitors of amidation on neuropeptide expression within the CNS have not been previously reported. The analysis of COOH-terminal extended immature forms of SP may serve as a sensitive indicator of biosynthetic and posttranslational processing events in SP-containing neural systems. Previous work from this laboratory has provided an immunochemical and chromatographic characterization of the unamidated immediate precursor form to SP, i.e. SP-G-LI, by combined RIA-HPLC analyses of rodent nervous tissues, utilizing highly specific and sensitive anti-SP-G sera (5, 13). In the present study, we have examined the effects of disulfiram administration on several neurochemical parameters in mouse CNS, including levels of SP-LI, SP-G-LI, '  Therefore, in order to monitor any potential contributions of these related COOH-terminal extended tachykinin precursor forms in the RIAs by direct means, NKA-G-K and NKB-G-K were synthesized and purified to over 95% homogeneity by HPLC.
The glycine-extended forms NKA-G and NKB-G were generated from the respective G-K-extended forms by carboxypeptidase B treatment followed by HPLC purification.
The specificities of anti-SP-G and anti-SP-G-K could then be assessed towards COOHterminal extended forms of related tachykinin species.
In addition, we assessed the molar cross-reactivities of SP-free acid and SP-Gly-Lys-Arg in our assays. Finally, molar cross-reactivities of COOHterminal extended forms of the unrelated peptides oxytocin (OT) and arginine vasopressin (AVP) were measured. The AVP analogs AVP-Gly" (AVP-G), AVP-Gly"-Lys" (AVP-G-K), AVP-Gly"'-Lys"-Arg'* (AVP-G-K-R) and the homologous OT analogs OT-G, OT-G-K, and OT-G-K-R were generously provided by Drs. M. Altstein and H. Gainer of the NINCDS. As reported below (Table  I), the G-and G-K-extended forms of NKA and NKB displayed only minimal molar cross-reactivities in our analyses.
The most prominent reactive species were NKA-G and NKB-G-K, which displayed 16 and 20 parts per 10,000 molar crossreactivities with anti-SP-G and anti-SP-G-K, respectively. In addition, SP-G-K-R was found to display approximately 1% molar crossreactivity with anti-SP-G-K. Finally, the extended forms of OT and AVP displayed less than 1 part in 100,000 molar cross-reactivity with the two antisera.
These values indicate that enzymatically generated G-and G-K-extended forms of unrelated neuropeptides or of heterologous global peptide fragments display negligible interferences in the RIAs. Overall, these cross-reactivity data indicate that the RIAs described here possess a very high degree of specificity for the SP precursor determinants SP-G and SP-G-K. Radioimmunoassqysof SP-LI, SP-G-LI, and SP-G-K-LI-The RIA procedures were peifoimed essentially as described (5). In these analvses.
anti-SP, anti-SP-G, and anti-SP-G-K sera were used at a finai dilutions of ' pH 3.15. Amperometric responses were measured after oxidation of analytes at 0.7 V. These included responses to dopamine and its major metabolite 3,5-dihydroxyphenyl- Molar percent cross-reactiuities of peptide analogs with anti-SP-G and anti-SP-G-K sera PeDtide Anti-SP-G Anti-SP-G-K SP-Gly-Lys-Arg SP-free acid (RPKPQQFFGLM) NKA-Gly NKA-Gly-Lys NKB-Gly NKB-Gly-Lys OT-Gly OT-Gly-Lys OT-Gly-Lys-Arg AVP-Gly AVP-Gly-Lys Zmmunoh.istochemistry-Mice were anesthetized with pentobarbital (100 mg/kg) and perfused intracardially with 10 ml of phosphatebuffered saline (PBS), pH 7.2, followed by 200 ml of 2% paraformaldehyde in 0.1 M phosphate buffer, pH 7.2, at 4 "C. After dissection of brain and spinal cord, tissues were post-fixed for 4 h in 15% sucrose in 2% paraformaldehyde, 0.1 M phosphate buffer, pH 7.2. Tissue blocks were then frozen in dry ice and sections were cut on a freezing microtome.
Sections were incubated in primary antisera (either anti-SP or anti-SP-G) diluted in PBS, 0.4% Triton X-100, 5% normal goat serum, for 40 h at 4 "C. Anti-SP sera were used at a final dilution of 1:5000. Anti-SP-G sera were affinity-purified using a Bio-Rad Affi-Gel-10 agarose column containing 1 mg of covalently bound SP-G and repurified and concentrated using an immobilized protein A column (Repligen

RESULTS
Tissue Levels of SP-LI, SP-G-LI, and SP-G-K-LI in Mouse CNS after Drug Treatment-In the present study, we have examined the effects of an inhibitor of peptidylglycine (Yamidating monooxygenase activity, i.e. disulfiram, on neurochemical and physiological parameters in the mouse, including peptide and monoamine levels and pain thresholds. In addition, two of the disulfiram-treated groups were also injected with capsaicin in order to promote increased peptide turnover in peripheral afferents as a consequence of SP depletion from terminals (14). We observed that levels of SP-LI were generally reduced in CNS areas of experimental animals as compared to controls (Fig. LA). Most notably, levels of SP-LI were significantly reduced by approximately 30% in the spinal cords of drug-treated animals as compared to controls. In contrast, levels of SP-G-L1 were dramatically increased at a level of statistical significance of p < 0.01 in four brain and spinal areas of all experimental groups, as compared to vehicle-treated controls (Fig. 1B). Concentrations of SP-G-L1 in the main olfactory bulb and pons of control and experimental groups were found to vary at the limits of detection of the SP-G RIA and could not be accurately quantified. Thus, these values are not included in the presented data. The elevations in levels of SP-G-L1 ranged from approximately 5-to 17-fold in experimentals versus controls with an average increase of 11.7-fold f 1.4 (SE.). To monitor changes in unamidated precursor relative to mature peptide, SP-G-LI:SP-LI ratios were plotted. With coordinate reductions in levels of SP-LI and increases in levels of SP-G-LI, the SP-G-LI:SP-LI ratios were markedly increased in treated versus control animals. In control animals, levels of SP-G-L1 expressed on a normalized basis ranged from 0.38 to 2.3% of the levels of SP-LI across CNS regions. After disulfiram treatment, normalized levels of SP-G-L1 ranged from 4.2 to 21% of the levels of SP-LI across CNS regions, with an average value of 14.3% + 1.8 (S.E.) ( Fig. 2A). On a regional basis, the smallest increases in SP-G-L1 were observed in the medial basal hypothalamus of drug-treated animals versus controls, whereas the largest increases in SP-G-L1 were noted in the medulla and preoptic area of these same animals. In addition, there were no statistically significant differences in the magnitude of increase of the SP-G-LI:SP-LI ratio among the three treatment groups DIS, DIS/CAPl, and DIS/CAPB across CNS areas. Interestingly, levels of heterogeneous SP precursors, normalized as SP-G-K determinants and monitored by RIA of trypsin-treated extracts, were increased in several brain areas of animals concomitantly treated with both drugs, i.e. the DIS/CAPl group (Fig. 1C). These changes were effectively realized as SP-G-K-LI:SP-LI ratios and were

Effects of Disulfiram on SP-LI and Monoaminergic
Systems shown to reach levels of statistical significance in the pons, preoptic area, and spinal cord of animals in the DIS/CAPl group, areas known to contain numerous SP-positive somata (Fig. 2B). Unlike the dramatic IO-fold increases in SP-G-LISP-L1 ratios observed in CNS areas of drug-treated animals, SP-G-K-LI:SP-LI ratios were modestly increased by approximately 2-fold in the CNS areas mentioned above. Finally, there was an observable trend toward increased extended precursor uersus mature SP, normalized as SP-G-K-LI:SP-LI ratios, in the brain stems and spinal cords of animals treated with disulfiram alone (DIS group).
Immunochemical Characterization of SP-LI, SP-G-LI, and SP-G-K-LI Extracted from Mouse Brain Samples-Extensive HPLC/RIA analyses of extracted brain tissues from both control and drug-treated animals yielded very similar chromatographic and immunochemical behavior for SP-LI, SP-G-LI, and SP-G-K-LI, as demonstrated for the chemically authentic peptide standards. After fractionation by gradient reverse-phase HPLC, the major immunoreactive peaks of SP-LI, SP-G-LI, and SP-G-K-L1 were coeluted with authentic standards (Fig. 3). In brain extracts of drug-treated animals only, small peaks of SP-LI and SP-G-L1 were coeluted with synthetic sulfoxide derivatives of SP and SP-G, respectively (Fig. 3, A-D). Normalized SP-G-LI:SP-LI molar ratios as monitored by RIA of HPLC-fractionated brain extracts from control and drug-treated animals were calculated to be 2.9 and 25.7%, respectively. These values are consistent with average IO-fold changes monitored by RIA of individual brain and spinal areas (Fig. 1). In addition, our analyses demonstrated the presence of very low steady state levels of SP-G-K-LI, normalized as 0.21 and 0.44% of the steady state levels of SP-LI in control and drug-treated animals, respectively (Fig. 3, E and F). Here we observed that approximately 70% of the total SP-G-K-L1 found in drug-treated animals was coeluted with authentic SP-G-K and its sulfoxide derivative, whereas 30% of the total immunoreactivity was eluted at earlier retention times (Fig. 3F). Consistent with our previous work (5, 13), SP-G-KLI in protease-treated extracts was increased approximately 75-fold, compared with untreated samples (Fig. 3, G and H). In the analyses of trypsin-treated samples from both control and drug-treated animals, the major peaks of SP-G-K-L1 were coeluted with authentic SP-GK, with small but significant peaks of SP-G-K-L1 (5-10% total immunoreactivity) eluted at retention times intermediate of SP-GK and its sulfoxide derivative. The minor peak of SP-G-K-L1 may represent an amino-terminal-extended SP-G-K determinant resulting from incomplete trypsin digestion, or a previously uncharacterized molecular variant containing the COOH-terminal epitope of SP-G-K. Normalized SP-G-K-LI:SP-LI molar ratios were 15.4 and 24.7% in extracts from control and drug-treated animals, respectively. Thus, disulfiram treatment resulted in an apparent 70% increase in immunoreactivity corresponding to heterogeneous forms of SP precursors monitored by RIA of SP-G-K determinants after trypsin treatment. Overall, the accuracy and precision (92-96% total recovery of chromatographed immunoreactivity) of the HPLC-RIA analyses have demonstrated the feasibility of our method as a general approach for quantification of changes in levels of immunoreactivity corresponding to mature and immature molecular forms of SP after pharmacological or physiological challenge.

HPLC-ECD
Analysis of Biogenic Amine Levels-It has been previously demonstrated that the enzyme responsible for norepinephrine biosynthesis, i.e. dopamine /3-hydroxylase, is inhibited by disulfiram administration via chelation of copper cofactor (21). Thus, the present study included parallel anal-yses of monoamines, in part to monitor efficacy of drug administration via changes in CNS levels of norepinephrine. Here norepinephrine levels were modestly reduced in a variable manner in CNS areas of drug-treated animals as compared to controls (Fig. 4A). Average drug-induced reductions in norepinephrine levels ranged from 42 to 11% of control values in main olfactory bulb and pons, respectively. In addition, norepinephrine levels were reduced in a statistically significant manner in main olfactory bulb, preoptic area, medial basal hypothalamus, and spinal cord of all drug-treated groups, as compared to controls. In medulla and pons, norepinephrine levels were significantly reduced only for the DIS/CAPl group, as compared to controls. Statistically significant increases in 5-HT turnover, as monitored by the 5-HIAA:5-HT ratio, were observed in CNS regions of drug-treated animals, as compared to controls (Fig.  40). These increases in 5-HT turnover were generally attributable to elevated levels of 5-HIAA in most CNS areas of drug-treated animals (Fig. 4C). In groups DIS/CAPl and DIS/CAP2, observed decreases in 5-HT levels with concomitant elevations in 5-HIAA levels yielded dramatic increases in 5-HT turnover of 50-100% versus controls in medulla and medial basal hypothalamus, respectively.
Finally, dopamine levels in main olfactory bulb, preoptic area, and medial basal hypothalamus exhibited an inconsistent pattern of change in all the drug-treated groups, as compared to controls (data not shown). A statistically significant increase in dopamine levels was observed only in the medial basal hypothalamus of animals treated with disulfiram alone (DIS group). However, dopamine turnover as monitored by the 3,5dihydroxyphenylacetic aciddopamine ratio was reduced in a statistically significant manner in the preoptic area of all animals in drug-treated groups (data not shown).

Anatomical and Behavioral
Studies-Complementary immunohistochemical analyses using affinity-purified anti-SP-G sera revealed markedly increased staining at the cellular level in CNS sections from DIS-treated animals, as compared to controls (Fig. 5). In drug-treated cases, immunohistochemical staining of SP-G-LI-containing neural elements was prominent in those areas of the brain exhibiting dense labeling in analyses using anti-SP sera, i.e. ventral pallidurn, substantia n&a, and dorsolateral medulla and the substantia gelatinosa of the spinal cord. Areas of the brain that exhibited moderate to light labeling in analyses using anti-SP sera, i.e. preoptic area, medial basal hypothalamus, periaqueductal gray, nucleus solitarius, and ventral spinal cord, exhibited very light labeling in analyses using anti-SP-G sera. Pronounced staining of SP-G-LI-containing terminal fields was observed in the substantia gelatinosa of the spinal cord (Fig.  5A) and in the trigeminal nucleus caudalis of the medulla (Fig. 5C) of a DIS-treated animal. The general pattern of staining was similar to that seen in spinal cord and medulla in analyses using anti-SP sera (Fig. 5, B and E, respectively). However, some differences in the distribution of reaction product were observed in comparative analyses using the two antisera. With anti-SP sera, dense labeling was observed in the dorsal longitudinal fasciculus, a fiber tract located ventrolateral to the substantia gelatinosa. In sections treated with anti-SP-G sera, little or no reaction product was found in the dorsal longitudinal fasciculus. This difference was also observed in a medullary area, along the ventromedial border of the nucleus caudalis, which was stained in analyses using anti-SP sera but not in those employing anti-SP-G sera. Finally, analyses using anti-SP-G sera performed on CNS sections of vehicle-treated controls or animals administered only CAP yielded no observable immunohistochemical stain-  ing of cellular elements (Fig. 50). Strong presumptive evidence supporting the specificity of the immunohistochemical staining was provided by preasborption of the anti-SP-G sera with 1 pg/ml SP-G, which eliminated all of the cellular staining observed in CNS sections of DIS-treated animals. In contrast, preabsorption of anti-SP-G sera with 1 pg/ml SP, a concentration previously demonstrated to completely eliminate cellular staining in analyses performed with anti-SP sera, produced minimal reduction in staining in parallel assays of CNS sections from drug-treated animals (data not shown  Finally, we observed that disulfiram pretreatment produced a dramatic protective effect on capsaicin-induced mortality in experimental animals. Rates of survival were shown to increase by approximately 300% in the disulfiram/capsaicin groups, as compared to the animals treated with capsaicin alone (data not shown). The biochemical mechanism of this protective effect is not immediately evident, but the recent availability of capsaicin for use in pain therapy (25) warrants its further investigation.

DISCUSSION
In the present study, we have demonstrated significant drug-induced changes in several neurochemical parameters in mouse CNS, including levels of immunoreactivity corresponding to mature and immature forms of SP, monoamines, and their metabolites. Regulatory events in the maturation processes of neuropeptides are not well understood (4). Our data suggest that the observed changes in steady state levels of immunoreactivities corresponding to mature SP and to the immature COOH-terminal extended forms of SP, i.e. SP-G and SP-G-K, are reflective of genuine compensatory biosynthetic and posttranslational processing events in SP-containing neural systems after pharmacological challenge. In addition, these data reflect established functional interrelationships between SP-containing and monoaminergic neural systems in sensory and integrative areas throughout the neuroaxis (1,26,27).
Although amino acid sequence data cannot be conciusively confirmed by combined chromatographic and immunochemical data, we are confident that the accuracy and precision of our analytical methodology have yielded strong presumptive evidence supporting the existence of mature SP and of unamidated SP precursors in nervous tissues. To a best approximation, the dramatic increases in steady state levels of SP-G-L1 accurately reflect modulation of expression of a COOHterminal glycine-extended precursor form of a CNS peptide after administration of an established inhibitor of the biosynthetic enzyme peptidylglycine cY-amidating monooxygenase. In complementary immunohistochemical analyses, these neurochemical changes were localized at the cellular level to neural elements that normally express SP. Disulfiram (Antabuse) is the disulfide dimer of NJ-diethyldithiocarbamate, to which it is rapidly converted via a reductive mechanism (22, 23). The presumptive biologically active form of disulfiram, i.e. N,N-diethyldithiocarbamate, is a potent and relatively specific copper chelator that has been shown to reversibly inhibit pituitary peptidylglycine cu-amidating monooxygenase activity in vitro (10) and in uiuo (12). The ability of N,N-diethyldithiocarbamate to chelate copper is also thought to be responsible for its inhibition of the cuproprotein, dopamine fi-hydroxylase (21), resulting in decreased levels of norepinephrine and epinephrine.
Peptidylglycine a-amidating monooxygenase has been shown to catalyze the formation of a terminal carboxyamide moiety and glyoxylate from a variety of peptidylglycine substrates (8-10, 24). SP-G is probably the major immediate precursor to SP, although we cannot rule out an alternative processing pathway via amidation of large molecular mass glycine-extended forms, followed by endoproteolytic cleavage at paired dibasic residues and release of mature SP (28). We have detected low steady state levels of SP-G-L1 in all CNS areas examined from both mice and rats.' These areas included olfactory bulbs, preoptic area, medial basal hypothalamus, lateral septum, substantia nigra, interpeduncular nu- Photomicrographs of sections of spinal cord and medulla from disulfiram-treated animals reacted with antisera to SP-G (A and C) and with antisera to SP (B and E). Reaction product demonstrating immobilized antigen is localized to substantia gelatinosa (A and B) and nucleus caudalis trigeminalis (C and E). Photomicrograph of medulla section from an animal with no disulfiram treatment reacted with antisera to SP-G demonstrating no reaction product in nucleus caudalis trigeminalis (D). cleus, pons, medulla, dorsal and ventral spinal cord, cortex, and dorsal root ganglia. Normalized SP-G-LISP-L1 ratios displayed approximately lo-fold variation across CNS areas, with steady state levels of SP-G-L1 averaging 2-3% of the levels of SP-LI on a whole brain basis. In the present study, disulfiram treatment produced an average lo-fold increase in steady state levels of SP-G-L1 in CNS areas of all drugtreated groups, as compared to controls. Combined HPLC/ RIA analyses of extracted brain tissue indicated that SP-G-LI and SP-G-K-L1 in control and in drug-treated animals displayed similar chromatographic and immunochemical behavior as demonstrated for chemically authentic standards. In addition, normalized SP-G-LISP-L1 ratios as monitored by RIA of HPLC-fractionated brain extracts from control and drug-treated animals were calculated to be 2.9 and 25.7%, respectively.
These data are consistent with average lo-fold changes monitored by RIA of individual brain and spinal areas. Fmthermore, the magnitude of drug-induced changes in levels of immunoreactivities corresponding to unamidated SP precursor uersus mature SP is very similar to that reported by Mains and co-workers (12) for the unamidated uersus amidated forms of the pituitary peptides cu-melanocyte-stimulating hormone and joining peptide after chronic disulfiram administration (approximately 33%). The presence of relatively modest steady state levels of SP-G-L1 in SP-containing neural systems that are dramatically altered by drug challenge suggests to us that terminal peptide amidation may represent a rate-limiting step in SP maturation and expression. By comparison, steady state levels of SP-G-K-L1 quantified without prior trypsin digestion from extracted brain of control animals were found to be 14 times lower than those of SP-G-L1 and were increased only approximately 2-fold in brain from drug-treated animals (Fig. 3, E and F, respectively).
Free SP-G-K-L1 probably represents a short-lived processing intermediate in SP maturation found in very low steady state concentrations in CNS; our data strongly suggest that the neuronal carboxypeptidase responsible for converting SP-G-K-L1 into SP-G-L1 is not a likely