- Email: [email protected]
Chemical characterization of substance P-like immunoreactivity in primary afferent neurones
Brain Research, 220 (1981) 203-207 Elsevier/North-Holland Biomedical Press
203
Chemical characterization of substance P-like immunoreactivity in primary afferent neurones
A. HARMAR* and P. KEEN Department of Pharmacology, Medical School, University of Bristol, Bristol BS8 1TD (U.K.)
(Accepted April 30th, 1981) Key words: substance P - - dorsal root ganglion - - spinal cord
Substance P-like immunoreactivity (SPLI) in rat dorsal root ganglia and dorsal spinal cord was characterized by high-performance liquid chromatography followed by radioimmunoassay. In spinal cord and ganglia, respectively, 87 ~o and 64 9o/ of SPLI eluted with authentic SP. The remainder of the SPLI in ganglia eluted as a single peak which did not represent the sulphoxide of SP or any of its Cterminal fragments. It has been suggested that substance P (SP) may function as a neurotransmitter in certain primary afferent neurones, particularly those subserving nociception. This suggestion is largely based upon radioimmunoassay and immunohistochemical studies using antisera which react not only with SP but also with C-terminal fragments of the peptide 10. There is some evidence that SP-like immunoreactivity (SPL1) in nervous tissue may be heterogeneous. (1) Dorsal root ganglia ( D R G ) 13 and superior cervical ganglia 14 have been reported to contain a high molecular weight form of SPL1 which has been suggested to be a precursor form of the peptide. (2) In a study of SPLI in a number of regions of the brain, Ben-Ari et al. 2 found evidence for the presence of the Cterminal (2-11) and (3-11) fragments of SP in addition to the intact undecapeptide. (3) SP (5-11) has been proposed as a major, biologically active metabolite of SP in nervous tissue, its formation being due to the action of postproline cleaving enzyme a. (4) Rat D R G incorporate [asS]methionine into two peptides with SPLI. One of these species corresponds to authentic SP; the nature of the second immunoreactive species has yet to be established 6. In view of this evidence for heterogeneity of SPLI in nervous tissue, we have employed H P L C followed by radioimmunoassay to characterize SPLI in two parts of the sensory nervous system. The biosynthesis of neuropeptides is probably confined to neuronal cell bodies: axons and nerve terminals have a very limited capacity for protein synthesis 4. In the case of spinal sensory neurones, SP synthesis takes place in cell bodies in D R G 6: the initial product of translation is probably a high molecular weight precursor protein. SP is then axonally transported to nerve terminals in dorsal spinal cord and in the periphery8, ~5, whence it is released and, presumably, degraded. We have therefore examined SPLI in the D R G , a region in which any SP precursor should be * Present address : MRC Brain Metabolism Unit, University Department of Pharmacology, 1 George Square, Edinburgh EH8 9J2, U.K. 0006-8993/81/0000-0000/$ 02.50 © Elsevier/North-Holland Biomedical Press
204 concentrated, and in dorsal spinal cord, a region likely to be enriched in SP metabolites. Rats (Wistar, 350 g) were anaesthetized with pentobarbitone and killed by transcardiac perfusion with ice-cold Krebs-bicarbonate. The animals were kept on ice while D R G L3-L6 and the dorsal part of the spinal cord (cut through the central canal) at the level of the lumbosacral enlargement were dissected out. Tissues were placed in tubes containing 2 M acetic acid, placed in a boiling water bath for 10 rain, homogenized, centrifuged at 9000 g for 5 min and the supernatants lyophilized. Material derived from 50 mg spinal cord or from the D R G of 6 rats was taken up in I ml of 0.15 M NaC1, pH 2.1, centrifuged at 9000 g for 5 min and the supernatant subjected to HPLC. Samples were loaded onto a 4.6 ~ 250 mm column of Ultrasphere ODS (Altex) using a 2 ml injection loop, and separated with a gradient of acetonitrile in primary solvent of0.15 M NaC1 (pH 2.1) as described by O'Hare and Nice 11. Under these conditions SP was resolved from all of its C-terminal fragments, which eluted in the order shown in Fig. 1. SP was also resolved from its sulphoxide derivative ~ which eluted earlier than the parent peptide. Fractions of 300 #1 were collected and ~u SP [9 CIO C8 C7 C6 C5 C7
C2
C3
r15 [email protected]
:.= 4 0
Dorsal spinal cord
U
L-
o
5
C
E E
'~ I
"3
u
2
13._ E 121 ,,i,,,-, ffl
Dorsal root ganglia
cf)
1
Fraction no.
200
Fig. 1. Elution of SELl from a reverse-phase HPLC column. Peptides from dorsal spinal cord (upper) and dorsal root ganglia (lower) were separated by HPLC and individual samples assayed for SPLI by radioimmunoassay (see text for details). Arrows indicate the positions at which substance P (SP), its C-terminal fragments (Cn where n = number of residues) and their pyroglutamyl (pGlu) derivatives eluted.
205 evaporated to dryness under reduced pressure over N a O H ; the residue was then taken up in 250/zl barbital buffer (pH 8.6) containing 1 mg/ml bovine serum albumin. Aliquots of 2-50 #1 were then assayed for SPLI by radioimmunoassay. The radioimmunoassay system used an antibody directed towards the Cterminal region of SP, and [125I]TyrS-Sp was used as tracer. The affinities of SP (3-11), (4-11) and (5-1 I) for the antibody were comparable to that of the parent peptide, whereas both SP (8-11) and SP free acid had greater than 2000-fold lower affinities. The assay had a sensitivity of 5 pg SP per tube. A solution of synthetic SP (UCB) was used as standard; its SP content was determined by amino acid analysis. Three peaks of SPLI were observed in nervous tissue extracts - - for convenience these will be referred to, in order of elution time on HPLC, as P1, P2 and P3. P2 corresponded in elution time to authentic SP. In spinal cord, 87.3 ~ of the SPLI was contained in the SP peak (P2). Small amounts of SPLI (7.8 ~ and 4.9 ~ of total SPLI) were contained in PI and P3 respectively. In D R G 64.4 ~ of total SPLI was present as authentic SP (P2), the remainder being present in P1. The elution time of P3 on H P L C was different from that of any of the SP fragments and derivatives tested. Although the elution time of P1 was similar to that of SP(8-11), the antiserum used for radioimmunoassay had essentially no affinity for this fragment. P1 and P3 were not due to oxidation of SP to its sulphoxide or to other artefacts of the tissue extraction procedure: when [zH]SP (experimental material, Radiochemical Centre, Amersham) was passed through the extraction procedure, [zH]SP was the only immunoreactive species recovered after HPLC, and neither PI nor P3 corresponded in retention time to SP sulphoxide or SP sulphone. We have reported elsewhere 6 that isolated D R G incorporate [35S]methionine into SP and into a second SPLI species with similar chromatographic properties to P1. In subsequent studies, it was possible to clearly distinguish P1 from all of the Cterminal fragments of SP and to establish that its immunoprecipitation by SP antisera was specific7. The patterns of labelling of SP and P1 with [85S]methionine and [3H]proline, and the failure of P1 to react with antisera directed towards the Nterminal region of SP, suggsted that P1 could represent a peptide related to SP in Cterminal amino acid sequence (and hence in immunological properties) but different in its N-terminal region. In view of recent reports of the existence in nervous tissue of tachykinins other than SP 9 the elucidation of the structure of PI would be of considerable interest. The nature of P3, which was present in spinal cord but not in dorsal root ganglia, is not clear. Since D R G in vitro do not incorporate significant amounts of radiolabelled amino acids into this peak, it is not possible to rigorously establish the specificity of its binding to SP antisera, as has been possible for P1. If binding of P3 is specific, it could represent a peptide related to SP which is synthesised by neurones having their cell bodies outside the DRG. Although the present study is the first in which the chemical nature of SPLI in cell bodies and terminals of primary afferent neurones has been examined, two related studies require comparison with the present results: in a study of SPLI released from the superfused spinal cord, Agaki et al.1 reported that only one species of SPLI was detected by HPLC followed by radioimmunoassay. This report is consistent with our finding
206 that m o r e t h a n 87 % o f the S P L I in spinal c o r d is a p p a r e n t l y identical to synthetic SP. A l t h o u g h we here p r o v i d e evidence for the presence o f two further m i n o r c o m p o n e n t s o f S P L I in spinal cord, they w o u l d be difficult to detect in superfusates o f spinal cord which c o n t a i n only small a m o u n t s o f substance P. P e t t i b o n e et al. 12 analyzed S P L I in rat striatum, which contains the cell bodies o f the S P - p r o d u c i n g neurones o f the striatonigral tract. A l t h o u g h these a u t h o r s r e p o r t e d t h a t SP was the only m a j o r i m m u n o r e a c t i v e p e p t i d e found, there was also evidence for a low c o n c e n t r a t i o n o f a peptide eluting before SP which they identified as SP sulphoxide. Alternatively, it is possible t h a t this p e p t i d e m a y c o r r e s p o n d to the P1 r e p o r t e d here. Since in the present w o r k no evidence was f o u n d for the presence o f SP (3-11) or SP (5-11) in spinal cord, it seems unlikely t h a t the suggestion t h a t these SP fragments m a y p l a y a role in n e u r o t r a n s m i s s i o n is correct, at least in spinal cord. Similarly, the s u b s t a n t i a nigra, which is rich in SP-containing terminals, does n o t c o n t a i n any large a m o u n t o f fragmentsL However, as SP fragments have been r e p o r t e d in other areas o f the C N S 2 their role in other SP-containing tracts in the nervous system c a n n o t be discounted. This w o r k was s u p p o r t e d by a project g r a n t f r o m the M e d i c a l Research Council.
1 Agaki, H., Otsuka, M. and Yanagisawa, M., Identification by high-performance liquid chromatography of immunoreactive substance P released from isolated rat spinal cord, Neurosci. Lett., 20 (1980) 259-263. 2 Ben-Ari, Y., Pradelles, P., Gros, C. and Dray, F., Identification of authentic substance P in striatonigral and amygdaloid nuclei using combined high-performance liquid chromatography and radioimmunoassay, Brain Research, 173 (1979) 360-363. 3 Blumberg, S., Teichberg, Y. 1., Charli, J. L., Hersh, L. B. and McKelvy, J. F., Cleavage of substance P to an N-terminal tetrapeptide and a C-terminal heptapeptide by a post-proline cleaving enzyme from bovine brain, Brain Research, 192 (1980) 477486. 4 Droz, B., Renewal of synaptic proteins, Brain Research, 62 (1973) 383-394. 5 Floor, E. and Leeman, S. E., Substance P sulfoxide: separation from substance P by high-pressure liquid chromatography, biological and immunological activities, and chemical reduction, Analyt. Biochem., 101 (1980)498-503. 6 Harmar, A., Schofield, J. G. and Keen, P., Cycloheximide-sensitivesynthesis of substance P by isolated dorsal root ganglia, Nature (Lond.), 284 (1980) 267-269. 7 Harmar, A., Schofield, J. G. and Keen, P., Substance P biosynthesis in dorsal root ganglia: an immunochemical study of [3~S]methionine and [3H]proline incorporation in vitro, Neuroscience, in press. 8 Keen, P. and Harmar, A., Turnover and central and peripheral axonal transport of substance P in primary afferent neurones in vitro, Neurosci. Lett., Suppl. 5 (1980) $208. 9 Lazarus, L. H., Linnoila, R. I., Hernandez, O. and DiAugustine, R. P., A neuropeptide in mammalian tissues with physalaemin-like immunoreactivity, Nature (Lond.), 287 (1980) 555-558. 10 Mroz, E. A. and Leeman, S. E., Substance P. In B. M. Jaffe and H. R. Behrman (Eds.), Methods of Hormone Radioimmunoassay, 2nd edn., Academic Press, New York, 1979, pp. 121-137. 11 O'Hare, M. J. and Nice, E. C., Hydrophobic high-performance liquid chromatography of hormonal polypeptides and proteins on alkysilane-bonded silica, J. Chromatog., 171 (1979) 209-226. 12 Pettibone, D. J., Wurtman, R. J. and Leeman, S. E., Striatal substance P-like immunoreactivity : characterisation by high-performance liquid chromatography and aspects of subcellular distribution and in vitro release by potassium, Life Sci., 27 (1980) 1593-1602. 13 Robinson, S. E., Schwartz, J. P. and Costa, E., Substance P in the superior cervical ganglion and the submaxillary gland of the rat, Brain Research, 182 (1980) 11-17.
207 14 Scl~wartz, J. P. and Costa, E., Nerve growth factor-mediated increase of the substance P content of chick embryo dorsal root ganglia, Brain Research, 170 (1979) 198-202. 15 Takahashi, T. and Otsuka, M., Regional distribution of substance P in the spinal cord and nerve roots of the cat and the effect of dorsal root section, Brain Research, 87 (1975) 1-11.
203
Chemical characterization of substance P-like immunoreactivity in primary afferent neurones
A. HARMAR* and P. KEEN Department of Pharmacology, Medical School, University of Bristol, Bristol BS8 1TD (U.K.)
(Accepted April 30th, 1981) Key words: substance P - - dorsal root ganglion - - spinal cord
Substance P-like immunoreactivity (SPLI) in rat dorsal root ganglia and dorsal spinal cord was characterized by high-performance liquid chromatography followed by radioimmunoassay. In spinal cord and ganglia, respectively, 87 ~o and 64 9o/ of SPLI eluted with authentic SP. The remainder of the SPLI in ganglia eluted as a single peak which did not represent the sulphoxide of SP or any of its Cterminal fragments. It has been suggested that substance P (SP) may function as a neurotransmitter in certain primary afferent neurones, particularly those subserving nociception. This suggestion is largely based upon radioimmunoassay and immunohistochemical studies using antisera which react not only with SP but also with C-terminal fragments of the peptide 10. There is some evidence that SP-like immunoreactivity (SPL1) in nervous tissue may be heterogeneous. (1) Dorsal root ganglia ( D R G ) 13 and superior cervical ganglia 14 have been reported to contain a high molecular weight form of SPL1 which has been suggested to be a precursor form of the peptide. (2) In a study of SPLI in a number of regions of the brain, Ben-Ari et al. 2 found evidence for the presence of the Cterminal (2-11) and (3-11) fragments of SP in addition to the intact undecapeptide. (3) SP (5-11) has been proposed as a major, biologically active metabolite of SP in nervous tissue, its formation being due to the action of postproline cleaving enzyme a. (4) Rat D R G incorporate [asS]methionine into two peptides with SPLI. One of these species corresponds to authentic SP; the nature of the second immunoreactive species has yet to be established 6. In view of this evidence for heterogeneity of SPLI in nervous tissue, we have employed H P L C followed by radioimmunoassay to characterize SPLI in two parts of the sensory nervous system. The biosynthesis of neuropeptides is probably confined to neuronal cell bodies: axons and nerve terminals have a very limited capacity for protein synthesis 4. In the case of spinal sensory neurones, SP synthesis takes place in cell bodies in D R G 6: the initial product of translation is probably a high molecular weight precursor protein. SP is then axonally transported to nerve terminals in dorsal spinal cord and in the periphery8, ~5, whence it is released and, presumably, degraded. We have therefore examined SPLI in the D R G , a region in which any SP precursor should be * Present address : MRC Brain Metabolism Unit, University Department of Pharmacology, 1 George Square, Edinburgh EH8 9J2, U.K. 0006-8993/81/0000-0000/$ 02.50 © Elsevier/North-Holland Biomedical Press
204 concentrated, and in dorsal spinal cord, a region likely to be enriched in SP metabolites. Rats (Wistar, 350 g) were anaesthetized with pentobarbitone and killed by transcardiac perfusion with ice-cold Krebs-bicarbonate. The animals were kept on ice while D R G L3-L6 and the dorsal part of the spinal cord (cut through the central canal) at the level of the lumbosacral enlargement were dissected out. Tissues were placed in tubes containing 2 M acetic acid, placed in a boiling water bath for 10 rain, homogenized, centrifuged at 9000 g for 5 min and the supernatants lyophilized. Material derived from 50 mg spinal cord or from the D R G of 6 rats was taken up in I ml of 0.15 M NaC1, pH 2.1, centrifuged at 9000 g for 5 min and the supernatant subjected to HPLC. Samples were loaded onto a 4.6 ~ 250 mm column of Ultrasphere ODS (Altex) using a 2 ml injection loop, and separated with a gradient of acetonitrile in primary solvent of0.15 M NaC1 (pH 2.1) as described by O'Hare and Nice 11. Under these conditions SP was resolved from all of its C-terminal fragments, which eluted in the order shown in Fig. 1. SP was also resolved from its sulphoxide derivative ~ which eluted earlier than the parent peptide. Fractions of 300 #1 were collected and ~u SP [9 CIO C8 C7 C6 C5 C7
C2
C3
r15 [email protected]
:.= 4 0
Dorsal spinal cord
U
L-
o
5
C
E E
'~ I
"3
u
2
13._ E 121 ,,i,,,-, ffl
Dorsal root ganglia
cf)
1
Fraction no.
200
Fig. 1. Elution of SELl from a reverse-phase HPLC column. Peptides from dorsal spinal cord (upper) and dorsal root ganglia (lower) were separated by HPLC and individual samples assayed for SPLI by radioimmunoassay (see text for details). Arrows indicate the positions at which substance P (SP), its C-terminal fragments (Cn where n = number of residues) and their pyroglutamyl (pGlu) derivatives eluted.
205 evaporated to dryness under reduced pressure over N a O H ; the residue was then taken up in 250/zl barbital buffer (pH 8.6) containing 1 mg/ml bovine serum albumin. Aliquots of 2-50 #1 were then assayed for SPLI by radioimmunoassay. The radioimmunoassay system used an antibody directed towards the Cterminal region of SP, and [125I]TyrS-Sp was used as tracer. The affinities of SP (3-11), (4-11) and (5-1 I) for the antibody were comparable to that of the parent peptide, whereas both SP (8-11) and SP free acid had greater than 2000-fold lower affinities. The assay had a sensitivity of 5 pg SP per tube. A solution of synthetic SP (UCB) was used as standard; its SP content was determined by amino acid analysis. Three peaks of SPLI were observed in nervous tissue extracts - - for convenience these will be referred to, in order of elution time on HPLC, as P1, P2 and P3. P2 corresponded in elution time to authentic SP. In spinal cord, 87.3 ~ of the SPLI was contained in the SP peak (P2). Small amounts of SPLI (7.8 ~ and 4.9 ~ of total SPLI) were contained in PI and P3 respectively. In D R G 64.4 ~ of total SPLI was present as authentic SP (P2), the remainder being present in P1. The elution time of P3 on H P L C was different from that of any of the SP fragments and derivatives tested. Although the elution time of P1 was similar to that of SP(8-11), the antiserum used for radioimmunoassay had essentially no affinity for this fragment. P1 and P3 were not due to oxidation of SP to its sulphoxide or to other artefacts of the tissue extraction procedure: when [zH]SP (experimental material, Radiochemical Centre, Amersham) was passed through the extraction procedure, [zH]SP was the only immunoreactive species recovered after HPLC, and neither PI nor P3 corresponded in retention time to SP sulphoxide or SP sulphone. We have reported elsewhere 6 that isolated D R G incorporate [35S]methionine into SP and into a second SPLI species with similar chromatographic properties to P1. In subsequent studies, it was possible to clearly distinguish P1 from all of the Cterminal fragments of SP and to establish that its immunoprecipitation by SP antisera was specific7. The patterns of labelling of SP and P1 with [85S]methionine and [3H]proline, and the failure of P1 to react with antisera directed towards the Nterminal region of SP, suggsted that P1 could represent a peptide related to SP in Cterminal amino acid sequence (and hence in immunological properties) but different in its N-terminal region. In view of recent reports of the existence in nervous tissue of tachykinins other than SP 9 the elucidation of the structure of PI would be of considerable interest. The nature of P3, which was present in spinal cord but not in dorsal root ganglia, is not clear. Since D R G in vitro do not incorporate significant amounts of radiolabelled amino acids into this peak, it is not possible to rigorously establish the specificity of its binding to SP antisera, as has been possible for P1. If binding of P3 is specific, it could represent a peptide related to SP which is synthesised by neurones having their cell bodies outside the DRG. Although the present study is the first in which the chemical nature of SPLI in cell bodies and terminals of primary afferent neurones has been examined, two related studies require comparison with the present results: in a study of SPLI released from the superfused spinal cord, Agaki et al.1 reported that only one species of SPLI was detected by HPLC followed by radioimmunoassay. This report is consistent with our finding
206 that m o r e t h a n 87 % o f the S P L I in spinal c o r d is a p p a r e n t l y identical to synthetic SP. A l t h o u g h we here p r o v i d e evidence for the presence o f two further m i n o r c o m p o n e n t s o f S P L I in spinal cord, they w o u l d be difficult to detect in superfusates o f spinal cord which c o n t a i n only small a m o u n t s o f substance P. P e t t i b o n e et al. 12 analyzed S P L I in rat striatum, which contains the cell bodies o f the S P - p r o d u c i n g neurones o f the striatonigral tract. A l t h o u g h these a u t h o r s r e p o r t e d t h a t SP was the only m a j o r i m m u n o r e a c t i v e p e p t i d e found, there was also evidence for a low c o n c e n t r a t i o n o f a peptide eluting before SP which they identified as SP sulphoxide. Alternatively, it is possible t h a t this p e p t i d e m a y c o r r e s p o n d to the P1 r e p o r t e d here. Since in the present w o r k no evidence was f o u n d for the presence o f SP (3-11) or SP (5-11) in spinal cord, it seems unlikely t h a t the suggestion t h a t these SP fragments m a y p l a y a role in n e u r o t r a n s m i s s i o n is correct, at least in spinal cord. Similarly, the s u b s t a n t i a nigra, which is rich in SP-containing terminals, does n o t c o n t a i n any large a m o u n t o f fragmentsL However, as SP fragments have been r e p o r t e d in other areas o f the C N S 2 their role in other SP-containing tracts in the nervous system c a n n o t be discounted. This w o r k was s u p p o r t e d by a project g r a n t f r o m the M e d i c a l Research Council.
1 Agaki, H., Otsuka, M. and Yanagisawa, M., Identification by high-performance liquid chromatography of immunoreactive substance P released from isolated rat spinal cord, Neurosci. Lett., 20 (1980) 259-263. 2 Ben-Ari, Y., Pradelles, P., Gros, C. and Dray, F., Identification of authentic substance P in striatonigral and amygdaloid nuclei using combined high-performance liquid chromatography and radioimmunoassay, Brain Research, 173 (1979) 360-363. 3 Blumberg, S., Teichberg, Y. 1., Charli, J. L., Hersh, L. B. and McKelvy, J. F., Cleavage of substance P to an N-terminal tetrapeptide and a C-terminal heptapeptide by a post-proline cleaving enzyme from bovine brain, Brain Research, 192 (1980) 477486. 4 Droz, B., Renewal of synaptic proteins, Brain Research, 62 (1973) 383-394. 5 Floor, E. and Leeman, S. E., Substance P sulfoxide: separation from substance P by high-pressure liquid chromatography, biological and immunological activities, and chemical reduction, Analyt. Biochem., 101 (1980)498-503. 6 Harmar, A., Schofield, J. G. and Keen, P., Cycloheximide-sensitivesynthesis of substance P by isolated dorsal root ganglia, Nature (Lond.), 284 (1980) 267-269. 7 Harmar, A., Schofield, J. G. and Keen, P., Substance P biosynthesis in dorsal root ganglia: an immunochemical study of [3~S]methionine and [3H]proline incorporation in vitro, Neuroscience, in press. 8 Keen, P. and Harmar, A., Turnover and central and peripheral axonal transport of substance P in primary afferent neurones in vitro, Neurosci. Lett., Suppl. 5 (1980) $208. 9 Lazarus, L. H., Linnoila, R. I., Hernandez, O. and DiAugustine, R. P., A neuropeptide in mammalian tissues with physalaemin-like immunoreactivity, Nature (Lond.), 287 (1980) 555-558. 10 Mroz, E. A. and Leeman, S. E., Substance P. In B. M. Jaffe and H. R. Behrman (Eds.), Methods of Hormone Radioimmunoassay, 2nd edn., Academic Press, New York, 1979, pp. 121-137. 11 O'Hare, M. J. and Nice, E. C., Hydrophobic high-performance liquid chromatography of hormonal polypeptides and proteins on alkysilane-bonded silica, J. Chromatog., 171 (1979) 209-226. 12 Pettibone, D. J., Wurtman, R. J. and Leeman, S. E., Striatal substance P-like immunoreactivity : characterisation by high-performance liquid chromatography and aspects of subcellular distribution and in vitro release by potassium, Life Sci., 27 (1980) 1593-1602. 13 Robinson, S. E., Schwartz, J. P. and Costa, E., Substance P in the superior cervical ganglion and the submaxillary gland of the rat, Brain Research, 182 (1980) 11-17.
207 14 Scl~wartz, J. P. and Costa, E., Nerve growth factor-mediated increase of the substance P content of chick embryo dorsal root ganglia, Brain Research, 170 (1979) 198-202. 15 Takahashi, T. and Otsuka, M., Regional distribution of substance P in the spinal cord and nerve roots of the cat and the effect of dorsal root section, Brain Research, 87 (1975) 1-11.