Neuroscience
Vol. 49, No. 3, pp. 621-633,
0306-4522/92
1992
$5.00 + 0.00
Pergamon Press Ltd 0 1992IBRO
Printed in Great Britain
NEUROTENSIN INCREASES TYROSINE HYDROXYLASE MESSENGER RNA-POSITIVE NEURONS IN SUBSTANTIA NIGRA AFTER RETROGRADE AXONAL TRANSPORT M.-C. BURGEVIN, M.-N. CASTEL, D. QUARTERONET,T. CHEVJZTand P. M. LADURON* Research Center, RhBne-Poulenc Rorer, 13 Quai Jules Guesde, 94403 Vitry sur Seine Cedex, France Abstract-In previous studies we have shown that labelled neurotensin injected into the rat striatum was found to be transported retrogradely in dopaminergic neurons through a process which was receptor and microtubule dependent. Now, we show, by in situ hybridization, the consequences of the striatal injection of neurotensin on the gene expression of tyrosine hydroxylase in the substantia nigra. Rats were injected with neurotensin or its fragments in the striatum of one side and with saline or the inactive fragment on the other. The number of nigral cells expressing tyrosine hydroxylase. mRNA was found to increase by 40% after injection of neurotensin or its active fragment (neurotensin 8-13). In the same experimental conditions, the inactive fragment (neurotensin l-8) was without effect. Time-course experiments revealed that the tyrosine hydroxylase mRNA was increased 4 h after neurotensin injection but not at 1 or 16 h. The fact that the increase of mRNA parallels the appearance of labelled neurotensin in the substantia nigra indicates that the changes in the gene expression of tyrosine hydroxylase might lx the consequence of the retrograde axonal transport of neurotensin. These results represent the first evidence for the existence of a long-distance retrograde signalling process in which the neuropeptide and presumably its receptor may serve as information molecule between synapses and the cell body.
It has been demonstrated that regulatory influences exist between neurotensin and dopaminergic pathways in several cerebral structures, especially the basal ganglia and the limbic structures.36~37 For instance, neurotensin potentiates dopamine release induced or not by K + from striatal tissues in vitro and in vivo.16~‘*~37 Moreover, lesion studies and subcellular fractionation analysis have shown that neurotensin receptors are located on dopaminergic nerve terminals in the striatum.z0*26*38*42 Fast anterograde axonal transport is required to supply nerve terminals with enzymes and other neuronal constituents.21 Ligature experiments in peripheral nerves have demonstrated fast and bidirectional axonal transport of muscarinic receptors32 and of various presynaptic receptors.” Moreover, in vivo experiments have shown that nerve growth factor (NGF)2’ and opiate3’ can move retrogradely from nerve terminals to the cell body through a process involving the presence of receptors.“,33@ Thus, a possible role for the retrograde transport of receptors is to convey signal molecules from the synapse to the cell body. Recently, we demonstrated a retrograde axonal transport of neurotensin in the central nervous system.6 Labelled neurotensin injected into the rat stria-
*To whom correspondence should be addressed. Abbreuiations: EDTA, ethylenediaminetetra-acetate; NGF, nerve growth factor; PBS, phosphate-buffered saline; TH, tyrosine hydroxylase; SSC, sodium chloride-sodium citrate buffer.
turn was recovered as such in the substantia nigra after retrograde transport in dopaminergic neurons.6*7 This process was microtubule dependent and appeared to involve neurotensin receptors. The retrograde transport of neurotensin was thought to occur after binding of labelled neurotensin to presynaptic receptors in the striatum and its subsequent internalization presumably as ligandreceptor complex. The dopaminergic nigrostriatal pathway is an ideal model for examining the consequences of the retrograde transport process on gene expression owing to the relatively long distance between the striatal nerve terminals and the nigral cell bodies which allow one to expect a sufficiently long delay between neurotensin injection into the striatum and the appearance of biological responses in the substantia nigra. However, the physiological role of the transported material in the cell body remains to be elucidated. EXPERIMENTAL.PROCEDURES Animals Adult male Sprague+Dawley (Charles River, France) rats under pentobarbital anaesthesia (6Omg/kg, i.p.) were injected bilaterally with 2 ~1 of a peptidase inhibitor, kelatorphan (30~~) in both striata at a rate of 0.2 pl/min (coordinates: AI’-0.3mm; L -3.6mm; V- 6mm with respect to bregma). Ten minutes later, 15 pmol of neurotensin or its different fragments were infused into one striatum, whereas saline or neurotensin fragments were injected into the other striatum. The peptides (neurotensin 1-13, neurotensin 8-13 and neurotensin l-8 from Sigma) were delivered as previously described.6 621
628
M.-C‘. BURGEVIN
Tissue preparation
After decapitation at 1, 4 or 16 h after the end 01 neurotensin injections, brains were rapidly removed, frozen immediately in isopentane at - 25°C and stored at - 70°C. The brains were cut coronally in a cryostat into lo-pm sections which were thaw-mounted on RNAse-free gelatinsubbed glass microscope slides and kept at -70°C until further use. Adjacent sections of 40 pm were processed for histological staining. Frozen sections were warmed from - 70°C to room temperature, fixed with 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) without Ca’+ and Mg*+ (PH 7.4) for 10min at room temperature and rinsed four times with 0.1 M PBS. Sections were then pretreated with 25 pg/ml pronase E in 50 mM Tris-5 mM EDTA (pH 7.5) for 10min at room temperature. The reaction was stopped by 2 mg/ml glycine for 1min. After this treatment, the sections were refixed with 4% paraformaldehyde for 10 min and washed three times with 0.1 M PBS for 1 min in each bath, then dehydrated through 70 and 95% ethanol (three baths of I min each) and dried. Preparation of cDNA probes
The construction and characterization of the recombinant pEM-TH19 clone in plasmid pBR322 have been described.” The Barn HI, Hind III insert of this clone which contains both the coding and the entire 3’-non coding regions of tyrosine hydroxylase (TH) mRNAZ3 was purified by electroelution after agarose gel electrophoresis and labelled by nick-translation4’ to specific activities greater than lo* d.p.m./pg by the addition of 35S-substituted adenosine and cytidine 5’-[cc-thio]-triphosphates (Amersham). A control pBR322 probe was labelled in the same way. Hybridization procedure
In this study hybridization was performed using a slight modification of the technique described by Berod et al.’ Briefly, dried sections were prehybridized for 1h at room temperature in a solution containing 4 x SSC (1 x SSC = 0.15 M sodium chloride-O.0015 M sodium citrate, pH 7.0) and 1 x Denhardt’s solution (0.02 Ficoll/O.OZ% polyvinylpyrrolidone/O.O2% bovine serum albumin) and air dried. About 1 ng of boiled probe in 15 ~1
of hybridization solution (50% formamide, 4 x SSC, 1 x Denhardt, 1% sarcosyl, 10mM dithiothreitol, 1OOng pBR322, 1OOng yeast tRNA and 1OOng herring sperm DNA) was applied per section. After hybridization for 18 h at 42”C, sections were dipped in 4 x SSC to remove hybridization buffer, washed stringently five times in 4 x SSC at 42°C for 20 min, 2 x SSC for 20 min, 1 x SSC for 30 min, 0.5 x SSC for 20 min and 0.1 x SSC for 30 min. They were then dehydrated with 70 and 95% ethanol prior to air drying. Autoradiograms were generated by apposition of “Slabelled sections to /?-max film (Amersham) for three days. The cellular radiolabelling was detected by dipping in NTB2 emulsion (Kodak) and developing after 15 days. Tissue sections were then stained with Toluidine Blue. For negative controls, adjacent sections were hybridized with the pBR322 probe and, in this case, no signal was detected on the cells (Fig. lC, D). Image analysis The content of TH mRNA in the substantia nigra neurons was studied on emulsion-coated sections stained with Toltidine Blue. Grain counting was performed under polarized light and assessed using the computer-based image analysis system (HISTO ZOO)from BIOCOM (France).’ The system allows a rapid estimation of the number of grains over a defined region, in the present case a neuronal cell. A standard curve of optical density as a function of a defined number of grains was established for each experiment. From the standard curve, the computer converted the optical density into a number of grains. Histological staining
et (I/
defined the boundaries of the cells. However, grain clusters usually overlapped the histological definition of the cell boundaries (see Fig. lA, B). Therefore, optical density per cell was measured over the area corresponding to the grain clusters. For this, a circle of 20pm diameter, which is generally sufficient to enclose all the clusters of grains over the cell, was chosen. Background grain density wac estimated over neighbouring cells without labelling. Moreover, the computer permits the definition of an identical area, a rectangle of 1mm* positioned on the same area of the right and left substantia nigra on the same slice. Cells were selected throughout this entire rectangle and the numbers of cells labelled on the left and right sides were compared using pooled data obtained from different determinations and subjected to a Student’s paired f-test.
RESULTS First, the specificity of the probe for detecting TH mRNA was examined. The cellular localization of TH mRNA in the substantia nigra pars compacta, with the dense clusters of autoradiographic grains, which are superimposed over the Toluidine Blue colouration of cell bodies is shown in Fig. I A and B. In the same area of adjacent section, the control probe had no specific labelling (Fig. lC, D). It is of note that in the whole coronal section, only substantia nigra pars compacta and ventral tegmental area were labelled with TH cDNA probe. Then, autoradiograms of different experiments were examined. A more intense labelling of TH mRNA on the side injected with neurotensin 1-13 (indicated by an arrow) than on the control side injected with saline is shown in Fig. 2A. In the same manner, there is more labelling on the side injected with neurotensin 8-13, an active fragment (ibdicated by the arrow in Fig. 2B) than on the side injected with the inactive fragment of neurotensin (neurotensin l-8). In order to quantify more precisely the differences between the right and left substantia nigra, grain counting analysis was carried out. Four hours after injection into the striatum (Table l), there was a highly significant increase (+39%; P < 0.001) in the number of cells expressing a detectable level of TH mRNA in the substantia nigra ipsilateral to the neurotensin l-13-injected side. Similarly, significant differences were observed between sides when rats were bilaterally injected with the active neurotensin fragment (neurotensin 8-13) (+39%; P < 0.05) in one side and saline in the control hemisphere or with neurotensin l-l 3 (+43%; P < 0.05) as compared to the inactive fragment l-8. When animals were injected with the inactive fragment of neurotensin in one side and saline in the other, no significant modification in the labelled cell number was observed. These results strongly suggest that neurotensin receptors are involved in this increase of TH mRNA levels. In parallel, the total number of Toluidine Blue-stained cells was determine@ in the counting area in both sides; we never detected significant difference between both sides in the same
mRNA-positive neurons in substantia nigra
629
Fig. 1.Controlexperimentsto assessthespeciticityof theTH cDNA probe. Bright-field (A)and dark-field (B)photomicrogaphsallowthe visualizationof TH mRNA in cellsin substantia nigra pars compacta after hybridization with TH probe. C and D are photomicrographs of an adjacent section which has been hybridized with a control pBR322DNA probe. Radiolabellingwas detected by dipping in NTB, emulsion.
section. Since the majority of eefls in the substantia nigra are dopaminergic4’it is likely that neurotensin increased the level of TH mRNA in all the cells even in those where no labelling was apparently found in control experiments owing to the limitations of the method used. In the present study, a very large number of slices were examined so that more than 3000 cells were analysed in each substantia nigra.
When neurotensin accumulation in the substantia nigra was examined at 1, 4 and 16h post-injection into the striatum then maximum accumulation was seen after 4 h, which paralleled the increase (t 49%; P c 0.001) in the number of oells expressing TH mRNA (Fig. 3R). No sign&ant change in TH mRNA level in the substantia nigra was seen at 1 and 16h post-injection which was found to parallel the absence of labelled neurotensin (Fig. 3A).
630
Fig. 2. Autoradiograms of rat ventral mesencephalon obtained 4 h after the injection of I5 pmol of neurotensin 1-13 (A) and neurotensin 8-13 (B) into the right striatum; saline (A) and neurotensin l-8 (B) into the left striatum. Autoradiograms were generated by opposition of labelled sections to hyper /I-max film. Arrows indicate the right substantia nigra.
DISCUSSION
other, in such a way that each animal acted as its own control. Therefore, one may exclude the possibility that the observed modifications of TH mRNA levels resulted from the influence of external factors (anaesthesia, stress and effects of kelatorphan or of injection). In this regard, it is noteworthy that the physiological stress of cold exposure has recently been reported to increase TH mRNA in the rat locus coeruleus but not in the substantia nigra.40
The foregoing results provide evidence that the injection of neurotensin or of its active fragment into the striatum led to an increase of TH mRNA in dopaminergic cell bodies in the substantia nigra. All the in situ hybridization experiments were performed in parallel in both hemispheres; neurotensin and its analogues were injected into one side and saline in the Table
1. Tyrosine Bilateral
Side 1 Neurotensin Neurotensin Neurotensin Neurotensin
hydroxylase
messenger
injection Side 2
I- 13 I- 13 8-13 l-8
RNA-expressing Cell number/mm2
NaCl Neurotensin NaCl NaCl
Side 1-8
39 33 43 38
I
& 3*** + 4* & s* f 4 n.s.
cells in the substantta
nigra
+ S.E.M. Side 2 28 k 2 23 k 3 31&4 43 * 3
Slice number 37 IO 18 22
Effects of bilateral injections of neurotensin and two neurotensin fragments (1-8 and 8- 13) or saline on the number of cells expressing TH mRNA in the substantia nigra. The results are expressed in terms of positive cell number/mm2 * S.E.M. Data were analysed using the Student’s r-test for paired values (*P i 0.05, ***P < 0.001). The sample number (n) is the total number of slices analysed coming from five different experiments for neurotensin I-13/NaCl, four experiments for neurotensin l13/neurotensin l-8 and five experiments for neurotensin 8- 13/NaCl and neurotensin I-8/NaCl using a total number of 12 rats. Note that for each separate experiment the difference between the neurotensin-treated hemisphere and the control hemisphere was also significant.
mRNA-positive neurons in substantia nigra
1 (A)
-D-NT
I-13
4
I
16
Time (hours.1
i3X NT 1-13
(9)
0 60
NoCl
i
Tim
(hours)
Fig. 3. (A) Timecourse of labelling in both substantia nigra after bilateral injections of 0.16 pmol monoiodo-Tyr3neurotension 1-13 (solid line) and saline (dashed line) into the striatmn. (B) Time-course of neurotensin l- 13 (15 nmol) effect on tbe TH mRNA. The results are expressed in terms of number of labelled cells in the substantia nigra at different times atIer neurotensin l-1 3 injection (batched bar) or saline (white bar) into the striatum. Data analysis was performed with Student’s t-test for paired values (***P < 0.001). NT, neurotensin.
The main question is to know whether the increase
of TH mRNA is the result of the retrograde axonal transport of neurotensin and of its subsequent ap pearance in the cell body. First, a significant increase (+49%; P > 0.001) in the number of cells expressing TH mRNA was observed at 4 h post-injection, whereas no modification was detected at 1 and 15 h post-injection (Fig. 2B). This corresponds with the time-course of the accumulation of labelled neurotensin in the substantia nigra since the maximal amount of labelling was observed 4 h after neurotensin injection whereas no radioactivity was detected at 1 and 16 h (Fig. 2A). The time of 1 h is crucial because it makes unlikely the possibility that TH mRNA changes occur through firing along dopaminergic or other axons. Indeed, as a rule, gene expression changes induced by sensory stimuli, kindling, electroshock, stress or drugs are much more rapid, reaching often maximal levels after 1S-60 min.1L~17*43 Moreover, a rapid stimulation (less than 1 h) of TH transcription by protein kinase C has recently been reported” so that TH gene can be
631
regarded as an immediate-early gene. In contrast here, there was no increase in TH mRNA at 1 h after injection of neurotensin into the striatum. A second point is that all the neurotensin receptors in the rat caudate nucleus are located on dopaminergic terminals.26*42As neurotensin is known to increase dopamine release presumably in stimulating inositol triphosphate-mediated calcium mobilization’ one might expect that neurotensin-induced dopamine release can increase TH mRNA in the substantia through retrograde firing along nondopaminergic axons. Since haloperidol, a dopamine antagonist, did not alter the level of TH mRNA in midbrain dopaminergic neurons in the rat,” one may exclude a retrograde firing along non-dopaminergic axons to explain our results. Consequently, the present data strongly suggest that the arrival of neurotensin in the substantia nigra is the signal to increase the TH mRNA levels. Interestingly, the amount of neurotensin retrogradely transported in dopaminergic neurons is reduced in senescent rats* and totally abolished after lesion with 6-hydroxydopamine.6 The TH mRNA in the remaining dopaminergic cells of the substantia nigra is not increased in Parkinson’s disease; in fact, it fall8’ suggesting a possible reduction of the signalling system, perhaps due to a decrease in neurotensin retrograde transport. Now the question arises-what is the sequence of events from the injection of neurotensin into the striatum to the subsequent effect on the TH mRNA? First, neurotensin has to bind on presynaptic receptors in the striatum before inducing internalization of the ligand-receptor complex. Such a process was recently found for neurotensin receptors in neuronal cells in culture.46 The second step corresponds to the retrograde transport of neurotensin in dopaminergic neurons6g7 a process which has previously been found with NGF.25*MIn the present experiments, colchicine, which has been found to block the retrograde transport of neurotensin,6 was not used owing to is great diversity of effects on mRNAs.‘“s3LWhat is the signal for this retrograde transport? We speculate that phosphorylation of neurotensin receptors, a process common for numerous receptors” should serve as polarity signal for the retrograde transport just like glycosylation seems to be essential for correct trathcking of fl-adrenergic receptors to the cell membrane.39 The third step concerns the fate of neurotensin and its receptor in the cell body with the subsequent consequence on TH mRNA. As already discussed,” three pathways may be postulated: one direct to the nucleus and two others passing through the lysosoma1 system and/or the Golgi apparatus. Although one cannot exclude post-transcriptional mechanisms, in particular the mRNA stabilization which has been reported for tubulin mRNAs with insulin,i9 the most likely hypothesis is that the neurotensin receptor complex enters the nucleus and should serve as
632
M.-C.
hKGEVlN
et
al
nuclear protein or transcription factor in the reguneeded in order to elucidate the mechanisms whereby lation of TH gene expression. Supporting such an neurotensin can induce TH mRNA in dopaminergic hypothesis are the recent electron-microscopic studies neurons. However, the most likely explanation reshowing that about 15% of labelling recovered in mains a nuclear translocation of neurotensin-bound receptor. perikarya in substantia nigra was associated with nuclei 4 h after intrastriatal injection of labelled neurotensin.’ CONCLUSION In addition to the well-known nuclear receptors belonging to the superfamily of ligand-activated The foregoing results are compatible with the enhancer binding factors (steroid and thyroid hypothesis that the retrograde axonal transport of hormones, retinoic acid and vitamin D3), there is a receptor-mediated neuropeptide can lead to changes second group of nuclear receptors considered as in gene expression. Since second messengers do not putative because there is only indirect evidence supundergo fast retrograde axonal transport, these data represent the first evidence for a long-distance retroporting their presence in the nucleus (see ref. 34). In grade signalling process involving a neuropeptide and this regard, nuclear location has been reported for its receptor. Here, the ligand-receptor complex thyrotropin-releasing hormonal35 prolactin,12 vasoshould serve as “third messenger” or information active intestinal polypeptide4 and fibroblast growth molecule between the synapse and the cell body. Such factor’ but the most convincing evidence for nuclear a long-distance retrograde signalling system which is translocation from cell surface receptors was recently provided from experiments using interleukin- 1.‘5,22.24 particularly well adapted to neurons may be of importance, in neurogenesis, in long-term memory In absence of ligand, the majority of interleukin-1 and possibly in pathologies such as Parkinson’s receptors were located on the cell surface but after disease and schizophrenia. ligand-induced internalization, they were found in the nuclei. Mutant receptors lacking most of the Acknowledgements-We thank J. Mallet for providing the cytoplasmic domain were expressed at the cell surface cDNA probe, S. Dumas for her help with the in situ and cquld undergo nuclear translocation but they hybridization technique, G. Corkidi and J. P. Hermelin did lead to biological responses, the interleukin-l(BIOCOM, France) for helpful discussions on computermediated induction of interleukin-2 and SV40 assisted analysis and J. Pratt for his help in preparing the manuscript. promoters. For neurotensin. further work will be
REFERENCES
1. Baldin V., Roman A. M., Bose-Bieme I., Almaric F. and BouchC G. (1990) Translocation of b FGF to the nucleus is Gl phase cell cycle specific in bovine aortic endothelial cells. Eur. mulec. Biol. Org. J. 9, 1511-1517. 2. Berod A., Faucon-Biguet N., Dumas S., Block B. and Mallet J. (1987) Modulation of tyrosine hydroxylase gene expression in the central nervous system visualized by in situ hybridization. Proc. n&n. Acad. Sci. U.S.A. 84, 1699-1703. 3. Bisconte J. C. (1979) Kinetic analysis of cellular populations by means of the quantitative radioautography. ht. Reo. Cytol. 57, 75-126. 4. Bishr Omary M. and Kagnoff M. F. (1987) Identification of nuclear receptors for VIP on a human colonic adenocarcinoma cell line. Science 238, 1578-I 58 1. 5. Bozou J. C., Rochet N., Magnaldo I., Vincent J. P. and Kitabgi P. (1989) Neurotensin stimulates inositol triphosphate-mediated calcium mobilization but not protein kinase C activation in HT29 cells. Biochem. J. 264, 871-878. 6. Caste1 M. N., Malgouris C., Blanchard J. C. and Laduron P. M. (1990) Retrograde axonal transport of newotensin in the dopaminergic nigrostriatal pathway in the rat. Neuroscience 36, 425-430. 7. Caste1 M. N., Faucher D., Cuint F., Dutidat P., Boireau A. and Laduron P. M. (1991) Identification of intact neurotensin in the substantia nigra after its retrograde axonal transport in dopaminergic neurons. J. Neurochem. 56, 1816-1818. 8. Caste1 M. N., Lechardeur D., Blanchard J. C. and Laduron P. M. (1991) Reduction of the amount of neurotensin retrogradely transported in dopaminergic neurons of senescent rat brain. Neurosci. Letr. 124, 11l-l 14. 9. Caste1 M. N., Woulfe J., Wang X., Laduron P. M. and Beaudet A. (1992) Light and electron microscopic localization of retrogradely-transported neurotensin in dopaminergic nigrostriatal neurons of the rat. Neuroscience (in press). 10. Ceccatelli S., Cortes R. and Hiikfelt T. (1991) Effect of reserpine and colchicine on neuropeptide mRNA levels in the rat hypothalamic paraventricular nucleus. Molec. Brain Res. 9, 57-69. 11. Chang S. L., Squint0 S. P. and Harian R. E. (1988) Morphine activation of c-Fos expression in rat brain. Biochem. Biophys. Rex Commun. 157, 698-704. 12. Clevenger C. V., Altmann S. W. and Prystowski M. B. (1991) Requirement of nuclear prolactin for interleukin 2-stimulated proliferation of T lymphocytes. Science 253 77-79. 13. Cole A. J., Abu-Shakra S., SalTen D. W., Baraban J. M. and Worley P. F. (1990) Rapid rise in transcription factor mRNAs in rat brain after electroshock-induced seizures. J. Neurochem 55, 1920-1927. 14. Cottingham S. L., Pickar D., Shimotake T. K., Montpied P., Paul S. M. and Crawley J. N. (1990) Tyrosine hydroxylase
and cholecystokinin mRNA levels in the substantia nigra, ventral tegmental area, and locus coeruleus are unaffected by acute and chronic haloperidol administration. Cell. molec. Neurobiol. 10, 41-50. 15. Curtis B. M., Widmer M. B., de Roos P. and Qwamstrom E. E. (1990) IL-l and its receptor are translocated to the nucleus J. Immunol. 144, 1295-1303.
mRNA-positive
neurons in substantia nigra
633
16. De Quidt M. E. and Emson P. C. (1983) Neurotensin facilitates dopamine release in vitro from rat slices. Brain Res. 274, 376-380. 17. Draisci G. and Iadarola M. J. (1989) Temporal analysis of increase in c-fos, predynorphin and preproenkephalin mRNAs in rat spinal cord. Motet. Brain Res. 6, 31-37. 18. Faggin 8. M. and Cubbedu L. X. (1990) Rapid desensitization of dopamine release induced by ne~otensin and neurotensin fragments, J. Pharmoc. exp. Ther. 253, 812-818. 19. Femyhough P., Mill J. F., Roberts J. L. and Ishil D. N. (1989) Stabilization of tubulin RNAs by insulin and insulin-like growth factor I during neurite formation. Molec. Brain Res. 6, 109-120. 20. Goedert M., Pittaway K. and Emson P. C. (1984) Neurotensin receptors in the rat striatum: lesion studies. Brain Res. 299, 164-168. 21. Grafstein B. and Forman D. S. (1980) Intracellular transport in neurons. Physiol. Reo. 60, 1167-1283. 22. Grenfell S., Smithers N., Miller K. and Solari R. (1989) Receptor-mediated endocytosis and nuclear transport of human interleukin-1. Rio&em. J. 244, 813822. 23. Grima B., Lamouroux A., Blanot F., Faucon-Biguet N. and Mallet J. (1985) Complete coding sequence of rat tyrosine hydroxylase mRNA. Proc. natn. Acad. Sci. U.S.A. 2, 617-621. 24. Heguy A., Baldari C., Bush K., Nagele R., Newton R. C., Robb R. J., Horuk R., Telford J. L. and Melli M. (1991) Internalization and nuclear localization of interleukin-1 are not sulIlcient for function. Cell Growth Difir 2, 31 l-31 5. 25. Hendry 1. A., StaCkel K., Thoenen H. and Iversen L. L. (1974) The retrograde axonal transport of nerve growth factor. Brain Res. 68, 103-121. 26. Herve D., Tassin J. P., Studler J. M., Dana D., Kitabgi P., Vincent J. P., Glowinski J. and Rostkne W. (1986) Dopaminergic control of iz51labeled neurotensin binding site density in corticolimbic structures in the rat brain. Proc. nom. Acad. Sci. U.S.A. 83, 6203-6207. 27. Horellou P., Guibert B., Leviel V. and Mallet J. (1986) A single RNA species injected in Xenopus oocyte directs the synthesis of active tyrosine hydroxylase. Fedn Eur. biochem. Sots Left. UM, 6-10. 28. Hugganir R. L. and Greengard P. (1990) Regulation of neurotransmitter receptor desensitization by protein phosphorylation. Natron 5, 555-567. 29. Javoy-Agid F., Hirsh E. C., Dumas S., Duyckaerts C., Mallet J. and Agid Y. (1990) Decreased tyrosine hydroxylase messenger RNA in the surviving dopamine neurons of the subs~ntia nigra in Parkinson’s disease: an in situ hybridization study. Neurosctince 38, 245-253. 30. Johnson E. M., Taniuchi M., Clark H. B., Springer J. E., Koh S., Tayrien M. W. and Loy R. (1987) Demonstration of the retrograde receptor of nerve growth factor in peripheral and central nervous system. J. Neurosci. 7, 923-929. 31. Kiyama H. and Emson P. (1991) Colchicine-induced expression of proneurotensin mRNA in rat striatum and hypothalamus. Brain Res. 9, 353-358. 32. Laduron P. M. (1980) Axoplasmic transport of muscarinic receptors. Nufure 286, 287-288. 33. Laduron P. M. (1987) Axonal transport of receptors: possible involvement in long-term memory. Neuroscience 22, 767-779. 34. Laduron P. M. (1991) Towards genomic pharmacology: from membranal to nuclear receptors. In Adounces in Drug Research (Testa -B. ed.), Vol. 22;~~. 107Ll48. Academic Press, London. 35. Laverriere J. N.. Gourdii D.. Picart R. and Tixier-Vidal A. (1981) Thvroliberin is ranidlv transferred to the nucleus of GH3 pituitary cells at both 4°C and 37°C. Biochem. Biophys. Res.-Commun. 103,-833-840. 36. Nemeroff C. B. and Cain S. T. (1985) Neurotensin-dopamine interactions in the CNS. Trena!sPharmac. Sci. 6,201-205. 37. Quirion R. (1983) Interactions between neurotensin and dopamine in the brain: an overview. Peptides 4, 609-615. 38. Guirion R., Chiueh C. C., Everist H. D. and Pert A. (1985) Comparative l~li~tion of neurotensin receptors on nigrostriatal and mesolimbic dopaminergic terminals. Brain Res. 327, 385-389. 39. Rands E.. Candelore M. R.. Cheuna A. H.. Hill W. S.. Strader C. D. and Dixon R. A. F. (1990) . , Mutational analysis of /I-adrenergic receptor glycosylat&n. J. biol. Chem. 266, 10,759-10,764. 40. Richard F., Faucon-Biguet N., Labatut R., Rollet D., Mallet J. and Buda M. (1988) Modulation of tyrosine hydroxylase gene expression in rat brain and adrenals by exposure to cold. J. Neurosci. Res. 28, 32-37. 41. Rigby P. W. J., Dieckmann M., Rhodes C, and Berg P. (1977) Labeling deoxyribonucleic acid to high specific activity in &fro by nick-translation with DNA polymerase I. J. molec. Biol. 113, 237-251. 42. Schotte A., Rost&ne W. and Laduron P. M. (1988) Different su~ll~ar localization of neurotensin receptor and neurotensin acceptor-sites in the rat brain dopaminergic system. J. Neurochem. 50, 1026-1031. 43. Smith M. A., Brady L. S., Glowa J., Gold P. Q. and Herkenham (1991) ERects of stress and adrenalectomy on tyrosine hydroxylase mRNA levels in the locus coeruleus by in situ hybridization. Brain Res. 544, 26-32. 44. Sonnenberg J. L., Rauscher F. J. III, Morgan J. I. and Curran T. (1989) Regulation of proenkephalin by Fos and Jun. Science 246, 1622-1625. 45. Swanson L. W. (1982) The projections of the ventral te8mental area and adjacent regions: a combined fluorescent retrograde tracer and immunofluor~n~ study in the rat. Brain Res. Bull. 9, 321-353. 46. Vanisberg M. A., Maloteaux J. M., Octave J, N. and Laduron P. M. (1991) Rapid agonist-indu~ interaction of neurotensin receptors in rat cultured neurons. B&hem. Pharmac. 42, 2265-2274. 47. Vyas S., Faucon Biguet N. and Mallet J. (1990) Transcriptional and post-transcriptional regulation of tyrosine hydroxylase gene by-protein kinase C. Eur. molec: biol. Org: J. 9, 3707-3712. 48. Young W. S., Wamsley J. K., Zarbin M. A. and Kuhar M. J. (1980) Opioid receptors undergo axonal flow. Science 210, 76-78. (Accepted 19 February 1992)
Vol. 49, No. 3, pp. 621-633,
0306-4522/92
1992
$5.00 + 0.00
Pergamon Press Ltd 0 1992IBRO
Printed in Great Britain
NEUROTENSIN INCREASES TYROSINE HYDROXYLASE MESSENGER RNA-POSITIVE NEURONS IN SUBSTANTIA NIGRA AFTER RETROGRADE AXONAL TRANSPORT M.-C. BURGEVIN, M.-N. CASTEL, D. QUARTERONET,T. CHEVJZTand P. M. LADURON* Research Center, RhBne-Poulenc Rorer, 13 Quai Jules Guesde, 94403 Vitry sur Seine Cedex, France Abstract-In previous studies we have shown that labelled neurotensin injected into the rat striatum was found to be transported retrogradely in dopaminergic neurons through a process which was receptor and microtubule dependent. Now, we show, by in situ hybridization, the consequences of the striatal injection of neurotensin on the gene expression of tyrosine hydroxylase in the substantia nigra. Rats were injected with neurotensin or its fragments in the striatum of one side and with saline or the inactive fragment on the other. The number of nigral cells expressing tyrosine hydroxylase. mRNA was found to increase by 40% after injection of neurotensin or its active fragment (neurotensin 8-13). In the same experimental conditions, the inactive fragment (neurotensin l-8) was without effect. Time-course experiments revealed that the tyrosine hydroxylase mRNA was increased 4 h after neurotensin injection but not at 1 or 16 h. The fact that the increase of mRNA parallels the appearance of labelled neurotensin in the substantia nigra indicates that the changes in the gene expression of tyrosine hydroxylase might lx the consequence of the retrograde axonal transport of neurotensin. These results represent the first evidence for the existence of a long-distance retrograde signalling process in which the neuropeptide and presumably its receptor may serve as information molecule between synapses and the cell body.
It has been demonstrated that regulatory influences exist between neurotensin and dopaminergic pathways in several cerebral structures, especially the basal ganglia and the limbic structures.36~37 For instance, neurotensin potentiates dopamine release induced or not by K + from striatal tissues in vitro and in vivo.16~‘*~37 Moreover, lesion studies and subcellular fractionation analysis have shown that neurotensin receptors are located on dopaminergic nerve terminals in the striatum.z0*26*38*42 Fast anterograde axonal transport is required to supply nerve terminals with enzymes and other neuronal constituents.21 Ligature experiments in peripheral nerves have demonstrated fast and bidirectional axonal transport of muscarinic receptors32 and of various presynaptic receptors.” Moreover, in vivo experiments have shown that nerve growth factor (NGF)2’ and opiate3’ can move retrogradely from nerve terminals to the cell body through a process involving the presence of receptors.“,33@ Thus, a possible role for the retrograde transport of receptors is to convey signal molecules from the synapse to the cell body. Recently, we demonstrated a retrograde axonal transport of neurotensin in the central nervous system.6 Labelled neurotensin injected into the rat stria-
*To whom correspondence should be addressed. Abbreuiations: EDTA, ethylenediaminetetra-acetate; NGF, nerve growth factor; PBS, phosphate-buffered saline; TH, tyrosine hydroxylase; SSC, sodium chloride-sodium citrate buffer.
turn was recovered as such in the substantia nigra after retrograde transport in dopaminergic neurons.6*7 This process was microtubule dependent and appeared to involve neurotensin receptors. The retrograde transport of neurotensin was thought to occur after binding of labelled neurotensin to presynaptic receptors in the striatum and its subsequent internalization presumably as ligandreceptor complex. The dopaminergic nigrostriatal pathway is an ideal model for examining the consequences of the retrograde transport process on gene expression owing to the relatively long distance between the striatal nerve terminals and the nigral cell bodies which allow one to expect a sufficiently long delay between neurotensin injection into the striatum and the appearance of biological responses in the substantia nigra. However, the physiological role of the transported material in the cell body remains to be elucidated. EXPERIMENTAL.PROCEDURES Animals Adult male Sprague+Dawley (Charles River, France) rats under pentobarbital anaesthesia (6Omg/kg, i.p.) were injected bilaterally with 2 ~1 of a peptidase inhibitor, kelatorphan (30~~) in both striata at a rate of 0.2 pl/min (coordinates: AI’-0.3mm; L -3.6mm; V- 6mm with respect to bregma). Ten minutes later, 15 pmol of neurotensin or its different fragments were infused into one striatum, whereas saline or neurotensin fragments were injected into the other striatum. The peptides (neurotensin 1-13, neurotensin 8-13 and neurotensin l-8 from Sigma) were delivered as previously described.6 621
628
M.-C‘. BURGEVIN
Tissue preparation
After decapitation at 1, 4 or 16 h after the end 01 neurotensin injections, brains were rapidly removed, frozen immediately in isopentane at - 25°C and stored at - 70°C. The brains were cut coronally in a cryostat into lo-pm sections which were thaw-mounted on RNAse-free gelatinsubbed glass microscope slides and kept at -70°C until further use. Adjacent sections of 40 pm were processed for histological staining. Frozen sections were warmed from - 70°C to room temperature, fixed with 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) without Ca’+ and Mg*+ (PH 7.4) for 10min at room temperature and rinsed four times with 0.1 M PBS. Sections were then pretreated with 25 pg/ml pronase E in 50 mM Tris-5 mM EDTA (pH 7.5) for 10min at room temperature. The reaction was stopped by 2 mg/ml glycine for 1min. After this treatment, the sections were refixed with 4% paraformaldehyde for 10 min and washed three times with 0.1 M PBS for 1 min in each bath, then dehydrated through 70 and 95% ethanol (three baths of I min each) and dried. Preparation of cDNA probes
The construction and characterization of the recombinant pEM-TH19 clone in plasmid pBR322 have been described.” The Barn HI, Hind III insert of this clone which contains both the coding and the entire 3’-non coding regions of tyrosine hydroxylase (TH) mRNAZ3 was purified by electroelution after agarose gel electrophoresis and labelled by nick-translation4’ to specific activities greater than lo* d.p.m./pg by the addition of 35S-substituted adenosine and cytidine 5’-[cc-thio]-triphosphates (Amersham). A control pBR322 probe was labelled in the same way. Hybridization procedure
In this study hybridization was performed using a slight modification of the technique described by Berod et al.’ Briefly, dried sections were prehybridized for 1h at room temperature in a solution containing 4 x SSC (1 x SSC = 0.15 M sodium chloride-O.0015 M sodium citrate, pH 7.0) and 1 x Denhardt’s solution (0.02 Ficoll/O.OZ% polyvinylpyrrolidone/O.O2% bovine serum albumin) and air dried. About 1 ng of boiled probe in 15 ~1
of hybridization solution (50% formamide, 4 x SSC, 1 x Denhardt, 1% sarcosyl, 10mM dithiothreitol, 1OOng pBR322, 1OOng yeast tRNA and 1OOng herring sperm DNA) was applied per section. After hybridization for 18 h at 42”C, sections were dipped in 4 x SSC to remove hybridization buffer, washed stringently five times in 4 x SSC at 42°C for 20 min, 2 x SSC for 20 min, 1 x SSC for 30 min, 0.5 x SSC for 20 min and 0.1 x SSC for 30 min. They were then dehydrated with 70 and 95% ethanol prior to air drying. Autoradiograms were generated by apposition of “Slabelled sections to /?-max film (Amersham) for three days. The cellular radiolabelling was detected by dipping in NTB2 emulsion (Kodak) and developing after 15 days. Tissue sections were then stained with Toluidine Blue. For negative controls, adjacent sections were hybridized with the pBR322 probe and, in this case, no signal was detected on the cells (Fig. lC, D). Image analysis The content of TH mRNA in the substantia nigra neurons was studied on emulsion-coated sections stained with Toltidine Blue. Grain counting was performed under polarized light and assessed using the computer-based image analysis system (HISTO ZOO)from BIOCOM (France).’ The system allows a rapid estimation of the number of grains over a defined region, in the present case a neuronal cell. A standard curve of optical density as a function of a defined number of grains was established for each experiment. From the standard curve, the computer converted the optical density into a number of grains. Histological staining
et (I/
defined the boundaries of the cells. However, grain clusters usually overlapped the histological definition of the cell boundaries (see Fig. lA, B). Therefore, optical density per cell was measured over the area corresponding to the grain clusters. For this, a circle of 20pm diameter, which is generally sufficient to enclose all the clusters of grains over the cell, was chosen. Background grain density wac estimated over neighbouring cells without labelling. Moreover, the computer permits the definition of an identical area, a rectangle of 1mm* positioned on the same area of the right and left substantia nigra on the same slice. Cells were selected throughout this entire rectangle and the numbers of cells labelled on the left and right sides were compared using pooled data obtained from different determinations and subjected to a Student’s paired f-test.
RESULTS First, the specificity of the probe for detecting TH mRNA was examined. The cellular localization of TH mRNA in the substantia nigra pars compacta, with the dense clusters of autoradiographic grains, which are superimposed over the Toluidine Blue colouration of cell bodies is shown in Fig. I A and B. In the same area of adjacent section, the control probe had no specific labelling (Fig. lC, D). It is of note that in the whole coronal section, only substantia nigra pars compacta and ventral tegmental area were labelled with TH cDNA probe. Then, autoradiograms of different experiments were examined. A more intense labelling of TH mRNA on the side injected with neurotensin 1-13 (indicated by an arrow) than on the control side injected with saline is shown in Fig. 2A. In the same manner, there is more labelling on the side injected with neurotensin 8-13, an active fragment (ibdicated by the arrow in Fig. 2B) than on the side injected with the inactive fragment of neurotensin (neurotensin l-8). In order to quantify more precisely the differences between the right and left substantia nigra, grain counting analysis was carried out. Four hours after injection into the striatum (Table l), there was a highly significant increase (+39%; P < 0.001) in the number of cells expressing a detectable level of TH mRNA in the substantia nigra ipsilateral to the neurotensin l-13-injected side. Similarly, significant differences were observed between sides when rats were bilaterally injected with the active neurotensin fragment (neurotensin 8-13) (+39%; P < 0.05) in one side and saline in the control hemisphere or with neurotensin l-l 3 (+43%; P < 0.05) as compared to the inactive fragment l-8. When animals were injected with the inactive fragment of neurotensin in one side and saline in the other, no significant modification in the labelled cell number was observed. These results strongly suggest that neurotensin receptors are involved in this increase of TH mRNA levels. In parallel, the total number of Toluidine Blue-stained cells was determine@ in the counting area in both sides; we never detected significant difference between both sides in the same
mRNA-positive neurons in substantia nigra
629
Fig. 1.Controlexperimentsto assessthespeciticityof theTH cDNA probe. Bright-field (A)and dark-field (B)photomicrogaphsallowthe visualizationof TH mRNA in cellsin substantia nigra pars compacta after hybridization with TH probe. C and D are photomicrographs of an adjacent section which has been hybridized with a control pBR322DNA probe. Radiolabellingwas detected by dipping in NTB, emulsion.
section. Since the majority of eefls in the substantia nigra are dopaminergic4’it is likely that neurotensin increased the level of TH mRNA in all the cells even in those where no labelling was apparently found in control experiments owing to the limitations of the method used. In the present study, a very large number of slices were examined so that more than 3000 cells were analysed in each substantia nigra.
When neurotensin accumulation in the substantia nigra was examined at 1, 4 and 16h post-injection into the striatum then maximum accumulation was seen after 4 h, which paralleled the increase (t 49%; P c 0.001) in the number of oells expressing TH mRNA (Fig. 3R). No sign&ant change in TH mRNA level in the substantia nigra was seen at 1 and 16h post-injection which was found to parallel the absence of labelled neurotensin (Fig. 3A).
630
Fig. 2. Autoradiograms of rat ventral mesencephalon obtained 4 h after the injection of I5 pmol of neurotensin 1-13 (A) and neurotensin 8-13 (B) into the right striatum; saline (A) and neurotensin l-8 (B) into the left striatum. Autoradiograms were generated by opposition of labelled sections to hyper /I-max film. Arrows indicate the right substantia nigra.
DISCUSSION
other, in such a way that each animal acted as its own control. Therefore, one may exclude the possibility that the observed modifications of TH mRNA levels resulted from the influence of external factors (anaesthesia, stress and effects of kelatorphan or of injection). In this regard, it is noteworthy that the physiological stress of cold exposure has recently been reported to increase TH mRNA in the rat locus coeruleus but not in the substantia nigra.40
The foregoing results provide evidence that the injection of neurotensin or of its active fragment into the striatum led to an increase of TH mRNA in dopaminergic cell bodies in the substantia nigra. All the in situ hybridization experiments were performed in parallel in both hemispheres; neurotensin and its analogues were injected into one side and saline in the Table
1. Tyrosine Bilateral
Side 1 Neurotensin Neurotensin Neurotensin Neurotensin
hydroxylase
messenger
injection Side 2
I- 13 I- 13 8-13 l-8
RNA-expressing Cell number/mm2
NaCl Neurotensin NaCl NaCl
Side 1-8
39 33 43 38
I
& 3*** + 4* & s* f 4 n.s.
cells in the substantta
nigra
+ S.E.M. Side 2 28 k 2 23 k 3 31&4 43 * 3
Slice number 37 IO 18 22
Effects of bilateral injections of neurotensin and two neurotensin fragments (1-8 and 8- 13) or saline on the number of cells expressing TH mRNA in the substantia nigra. The results are expressed in terms of positive cell number/mm2 * S.E.M. Data were analysed using the Student’s r-test for paired values (*P i 0.05, ***P < 0.001). The sample number (n) is the total number of slices analysed coming from five different experiments for neurotensin I-13/NaCl, four experiments for neurotensin l13/neurotensin l-8 and five experiments for neurotensin 8- 13/NaCl and neurotensin I-8/NaCl using a total number of 12 rats. Note that for each separate experiment the difference between the neurotensin-treated hemisphere and the control hemisphere was also significant.
mRNA-positive neurons in substantia nigra
1 (A)
-D-NT
I-13
4
I
16
Time (hours.1
i3X NT 1-13
(9)
0 60
NoCl
i
Tim
(hours)
Fig. 3. (A) Timecourse of labelling in both substantia nigra after bilateral injections of 0.16 pmol monoiodo-Tyr3neurotension 1-13 (solid line) and saline (dashed line) into the striatmn. (B) Time-course of neurotensin l- 13 (15 nmol) effect on tbe TH mRNA. The results are expressed in terms of number of labelled cells in the substantia nigra at different times atIer neurotensin l-1 3 injection (batched bar) or saline (white bar) into the striatum. Data analysis was performed with Student’s t-test for paired values (***P < 0.001). NT, neurotensin.
The main question is to know whether the increase
of TH mRNA is the result of the retrograde axonal transport of neurotensin and of its subsequent ap pearance in the cell body. First, a significant increase (+49%; P > 0.001) in the number of cells expressing TH mRNA was observed at 4 h post-injection, whereas no modification was detected at 1 and 15 h post-injection (Fig. 2B). This corresponds with the time-course of the accumulation of labelled neurotensin in the substantia nigra since the maximal amount of labelling was observed 4 h after neurotensin injection whereas no radioactivity was detected at 1 and 16 h (Fig. 2A). The time of 1 h is crucial because it makes unlikely the possibility that TH mRNA changes occur through firing along dopaminergic or other axons. Indeed, as a rule, gene expression changes induced by sensory stimuli, kindling, electroshock, stress or drugs are much more rapid, reaching often maximal levels after 1S-60 min.1L~17*43 Moreover, a rapid stimulation (less than 1 h) of TH transcription by protein kinase C has recently been reported” so that TH gene can be
631
regarded as an immediate-early gene. In contrast here, there was no increase in TH mRNA at 1 h after injection of neurotensin into the striatum. A second point is that all the neurotensin receptors in the rat caudate nucleus are located on dopaminergic terminals.26*42As neurotensin is known to increase dopamine release presumably in stimulating inositol triphosphate-mediated calcium mobilization’ one might expect that neurotensin-induced dopamine release can increase TH mRNA in the substantia through retrograde firing along nondopaminergic axons. Since haloperidol, a dopamine antagonist, did not alter the level of TH mRNA in midbrain dopaminergic neurons in the rat,” one may exclude a retrograde firing along non-dopaminergic axons to explain our results. Consequently, the present data strongly suggest that the arrival of neurotensin in the substantia nigra is the signal to increase the TH mRNA levels. Interestingly, the amount of neurotensin retrogradely transported in dopaminergic neurons is reduced in senescent rats* and totally abolished after lesion with 6-hydroxydopamine.6 The TH mRNA in the remaining dopaminergic cells of the substantia nigra is not increased in Parkinson’s disease; in fact, it fall8’ suggesting a possible reduction of the signalling system, perhaps due to a decrease in neurotensin retrograde transport. Now the question arises-what is the sequence of events from the injection of neurotensin into the striatum to the subsequent effect on the TH mRNA? First, neurotensin has to bind on presynaptic receptors in the striatum before inducing internalization of the ligand-receptor complex. Such a process was recently found for neurotensin receptors in neuronal cells in culture.46 The second step corresponds to the retrograde transport of neurotensin in dopaminergic neurons6g7 a process which has previously been found with NGF.25*MIn the present experiments, colchicine, which has been found to block the retrograde transport of neurotensin,6 was not used owing to is great diversity of effects on mRNAs.‘“s3LWhat is the signal for this retrograde transport? We speculate that phosphorylation of neurotensin receptors, a process common for numerous receptors” should serve as polarity signal for the retrograde transport just like glycosylation seems to be essential for correct trathcking of fl-adrenergic receptors to the cell membrane.39 The third step concerns the fate of neurotensin and its receptor in the cell body with the subsequent consequence on TH mRNA. As already discussed,” three pathways may be postulated: one direct to the nucleus and two others passing through the lysosoma1 system and/or the Golgi apparatus. Although one cannot exclude post-transcriptional mechanisms, in particular the mRNA stabilization which has been reported for tubulin mRNAs with insulin,i9 the most likely hypothesis is that the neurotensin receptor complex enters the nucleus and should serve as
632
M.-C.
hKGEVlN
et
al
nuclear protein or transcription factor in the reguneeded in order to elucidate the mechanisms whereby lation of TH gene expression. Supporting such an neurotensin can induce TH mRNA in dopaminergic hypothesis are the recent electron-microscopic studies neurons. However, the most likely explanation reshowing that about 15% of labelling recovered in mains a nuclear translocation of neurotensin-bound receptor. perikarya in substantia nigra was associated with nuclei 4 h after intrastriatal injection of labelled neurotensin.’ CONCLUSION In addition to the well-known nuclear receptors belonging to the superfamily of ligand-activated The foregoing results are compatible with the enhancer binding factors (steroid and thyroid hypothesis that the retrograde axonal transport of hormones, retinoic acid and vitamin D3), there is a receptor-mediated neuropeptide can lead to changes second group of nuclear receptors considered as in gene expression. Since second messengers do not putative because there is only indirect evidence supundergo fast retrograde axonal transport, these data represent the first evidence for a long-distance retroporting their presence in the nucleus (see ref. 34). In grade signalling process involving a neuropeptide and this regard, nuclear location has been reported for its receptor. Here, the ligand-receptor complex thyrotropin-releasing hormonal35 prolactin,12 vasoshould serve as “third messenger” or information active intestinal polypeptide4 and fibroblast growth molecule between the synapse and the cell body. Such factor’ but the most convincing evidence for nuclear a long-distance retrograde signalling system which is translocation from cell surface receptors was recently provided from experiments using interleukin- 1.‘5,22.24 particularly well adapted to neurons may be of importance, in neurogenesis, in long-term memory In absence of ligand, the majority of interleukin-1 and possibly in pathologies such as Parkinson’s receptors were located on the cell surface but after disease and schizophrenia. ligand-induced internalization, they were found in the nuclei. Mutant receptors lacking most of the Acknowledgements-We thank J. Mallet for providing the cytoplasmic domain were expressed at the cell surface cDNA probe, S. Dumas for her help with the in situ and cquld undergo nuclear translocation but they hybridization technique, G. Corkidi and J. P. Hermelin did lead to biological responses, the interleukin-l(BIOCOM, France) for helpful discussions on computermediated induction of interleukin-2 and SV40 assisted analysis and J. Pratt for his help in preparing the manuscript. promoters. For neurotensin. further work will be
REFERENCES
1. Baldin V., Roman A. M., Bose-Bieme I., Almaric F. and BouchC G. (1990) Translocation of b FGF to the nucleus is Gl phase cell cycle specific in bovine aortic endothelial cells. Eur. mulec. Biol. Org. J. 9, 1511-1517. 2. Berod A., Faucon-Biguet N., Dumas S., Block B. and Mallet J. (1987) Modulation of tyrosine hydroxylase gene expression in the central nervous system visualized by in situ hybridization. Proc. n&n. Acad. Sci. U.S.A. 84, 1699-1703. 3. Bisconte J. C. (1979) Kinetic analysis of cellular populations by means of the quantitative radioautography. ht. Reo. Cytol. 57, 75-126. 4. Bishr Omary M. and Kagnoff M. F. (1987) Identification of nuclear receptors for VIP on a human colonic adenocarcinoma cell line. Science 238, 1578-I 58 1. 5. Bozou J. C., Rochet N., Magnaldo I., Vincent J. P. and Kitabgi P. (1989) Neurotensin stimulates inositol triphosphate-mediated calcium mobilization but not protein kinase C activation in HT29 cells. Biochem. J. 264, 871-878. 6. Caste1 M. N., Malgouris C., Blanchard J. C. and Laduron P. M. (1990) Retrograde axonal transport of newotensin in the dopaminergic nigrostriatal pathway in the rat. Neuroscience 36, 425-430. 7. Caste1 M. N., Faucher D., Cuint F., Dutidat P., Boireau A. and Laduron P. M. (1991) Identification of intact neurotensin in the substantia nigra after its retrograde axonal transport in dopaminergic neurons. J. Neurochem. 56, 1816-1818. 8. Caste1 M. N., Lechardeur D., Blanchard J. C. and Laduron P. M. (1991) Reduction of the amount of neurotensin retrogradely transported in dopaminergic neurons of senescent rat brain. Neurosci. Letr. 124, 11l-l 14. 9. Caste1 M. N., Woulfe J., Wang X., Laduron P. M. and Beaudet A. (1992) Light and electron microscopic localization of retrogradely-transported neurotensin in dopaminergic nigrostriatal neurons of the rat. Neuroscience (in press). 10. Ceccatelli S., Cortes R. and Hiikfelt T. (1991) Effect of reserpine and colchicine on neuropeptide mRNA levels in the rat hypothalamic paraventricular nucleus. Molec. Brain Res. 9, 57-69. 11. Chang S. L., Squint0 S. P. and Harian R. E. (1988) Morphine activation of c-Fos expression in rat brain. Biochem. Biophys. Rex Commun. 157, 698-704. 12. Clevenger C. V., Altmann S. W. and Prystowski M. B. (1991) Requirement of nuclear prolactin for interleukin 2-stimulated proliferation of T lymphocytes. Science 253 77-79. 13. Cole A. J., Abu-Shakra S., SalTen D. W., Baraban J. M. and Worley P. F. (1990) Rapid rise in transcription factor mRNAs in rat brain after electroshock-induced seizures. J. Neurochem 55, 1920-1927. 14. Cottingham S. L., Pickar D., Shimotake T. K., Montpied P., Paul S. M. and Crawley J. N. (1990) Tyrosine hydroxylase
and cholecystokinin mRNA levels in the substantia nigra, ventral tegmental area, and locus coeruleus are unaffected by acute and chronic haloperidol administration. Cell. molec. Neurobiol. 10, 41-50. 15. Curtis B. M., Widmer M. B., de Roos P. and Qwamstrom E. E. (1990) IL-l and its receptor are translocated to the nucleus J. Immunol. 144, 1295-1303.
mRNA-positive
neurons in substantia nigra
633
16. De Quidt M. E. and Emson P. C. (1983) Neurotensin facilitates dopamine release in vitro from rat slices. Brain Res. 274, 376-380. 17. Draisci G. and Iadarola M. J. (1989) Temporal analysis of increase in c-fos, predynorphin and preproenkephalin mRNAs in rat spinal cord. Motet. Brain Res. 6, 31-37. 18. Faggin 8. M. and Cubbedu L. X. (1990) Rapid desensitization of dopamine release induced by ne~otensin and neurotensin fragments, J. Pharmoc. exp. Ther. 253, 812-818. 19. Femyhough P., Mill J. F., Roberts J. L. and Ishil D. N. (1989) Stabilization of tubulin RNAs by insulin and insulin-like growth factor I during neurite formation. Molec. Brain Res. 6, 109-120. 20. Goedert M., Pittaway K. and Emson P. C. (1984) Neurotensin receptors in the rat striatum: lesion studies. Brain Res. 299, 164-168. 21. Grafstein B. and Forman D. S. (1980) Intracellular transport in neurons. Physiol. Reo. 60, 1167-1283. 22. Grenfell S., Smithers N., Miller K. and Solari R. (1989) Receptor-mediated endocytosis and nuclear transport of human interleukin-1. Rio&em. J. 244, 813822. 23. Grima B., Lamouroux A., Blanot F., Faucon-Biguet N. and Mallet J. (1985) Complete coding sequence of rat tyrosine hydroxylase mRNA. Proc. natn. Acad. Sci. U.S.A. 2, 617-621. 24. Heguy A., Baldari C., Bush K., Nagele R., Newton R. C., Robb R. J., Horuk R., Telford J. L. and Melli M. (1991) Internalization and nuclear localization of interleukin-1 are not sulIlcient for function. Cell Growth Difir 2, 31 l-31 5. 25. Hendry 1. A., StaCkel K., Thoenen H. and Iversen L. L. (1974) The retrograde axonal transport of nerve growth factor. Brain Res. 68, 103-121. 26. Herve D., Tassin J. P., Studler J. M., Dana D., Kitabgi P., Vincent J. P., Glowinski J. and Rostkne W. (1986) Dopaminergic control of iz51labeled neurotensin binding site density in corticolimbic structures in the rat brain. Proc. nom. Acad. Sci. U.S.A. 83, 6203-6207. 27. Horellou P., Guibert B., Leviel V. and Mallet J. (1986) A single RNA species injected in Xenopus oocyte directs the synthesis of active tyrosine hydroxylase. Fedn Eur. biochem. Sots Left. UM, 6-10. 28. Hugganir R. L. and Greengard P. (1990) Regulation of neurotransmitter receptor desensitization by protein phosphorylation. Natron 5, 555-567. 29. Javoy-Agid F., Hirsh E. C., Dumas S., Duyckaerts C., Mallet J. and Agid Y. (1990) Decreased tyrosine hydroxylase messenger RNA in the surviving dopamine neurons of the subs~ntia nigra in Parkinson’s disease: an in situ hybridization study. Neurosctince 38, 245-253. 30. Johnson E. M., Taniuchi M., Clark H. B., Springer J. E., Koh S., Tayrien M. W. and Loy R. (1987) Demonstration of the retrograde receptor of nerve growth factor in peripheral and central nervous system. J. Neurosci. 7, 923-929. 31. Kiyama H. and Emson P. (1991) Colchicine-induced expression of proneurotensin mRNA in rat striatum and hypothalamus. Brain Res. 9, 353-358. 32. Laduron P. M. (1980) Axoplasmic transport of muscarinic receptors. Nufure 286, 287-288. 33. Laduron P. M. (1987) Axonal transport of receptors: possible involvement in long-term memory. Neuroscience 22, 767-779. 34. Laduron P. M. (1991) Towards genomic pharmacology: from membranal to nuclear receptors. In Adounces in Drug Research (Testa -B. ed.), Vol. 22;~~. 107Ll48. Academic Press, London. 35. Laverriere J. N.. Gourdii D.. Picart R. and Tixier-Vidal A. (1981) Thvroliberin is ranidlv transferred to the nucleus of GH3 pituitary cells at both 4°C and 37°C. Biochem. Biophys. Res.-Commun. 103,-833-840. 36. Nemeroff C. B. and Cain S. T. (1985) Neurotensin-dopamine interactions in the CNS. Trena!sPharmac. Sci. 6,201-205. 37. Quirion R. (1983) Interactions between neurotensin and dopamine in the brain: an overview. Peptides 4, 609-615. 38. Guirion R., Chiueh C. C., Everist H. D. and Pert A. (1985) Comparative l~li~tion of neurotensin receptors on nigrostriatal and mesolimbic dopaminergic terminals. Brain Res. 327, 385-389. 39. Rands E.. Candelore M. R.. Cheuna A. H.. Hill W. S.. Strader C. D. and Dixon R. A. F. (1990) . , Mutational analysis of /I-adrenergic receptor glycosylat&n. J. biol. Chem. 266, 10,759-10,764. 40. Richard F., Faucon-Biguet N., Labatut R., Rollet D., Mallet J. and Buda M. (1988) Modulation of tyrosine hydroxylase gene expression in rat brain and adrenals by exposure to cold. J. Neurosci. Res. 28, 32-37. 41. Rigby P. W. J., Dieckmann M., Rhodes C, and Berg P. (1977) Labeling deoxyribonucleic acid to high specific activity in &fro by nick-translation with DNA polymerase I. J. molec. Biol. 113, 237-251. 42. Schotte A., Rost&ne W. and Laduron P. M. (1988) Different su~ll~ar localization of neurotensin receptor and neurotensin acceptor-sites in the rat brain dopaminergic system. J. Neurochem. 50, 1026-1031. 43. Smith M. A., Brady L. S., Glowa J., Gold P. Q. and Herkenham (1991) ERects of stress and adrenalectomy on tyrosine hydroxylase mRNA levels in the locus coeruleus by in situ hybridization. Brain Res. 544, 26-32. 44. Sonnenberg J. L., Rauscher F. J. III, Morgan J. I. and Curran T. (1989) Regulation of proenkephalin by Fos and Jun. Science 246, 1622-1625. 45. Swanson L. W. (1982) The projections of the ventral te8mental area and adjacent regions: a combined fluorescent retrograde tracer and immunofluor~n~ study in the rat. Brain Res. Bull. 9, 321-353. 46. Vanisberg M. A., Maloteaux J. M., Octave J, N. and Laduron P. M. (1991) Rapid agonist-indu~ interaction of neurotensin receptors in rat cultured neurons. B&hem. Pharmac. 42, 2265-2274. 47. Vyas S., Faucon Biguet N. and Mallet J. (1990) Transcriptional and post-transcriptional regulation of tyrosine hydroxylase gene by-protein kinase C. Eur. molec: biol. Org: J. 9, 3707-3712. 48. Young W. S., Wamsley J. K., Zarbin M. A. and Kuhar M. J. (1980) Opioid receptors undergo axonal flow. Science 210, 76-78. (Accepted 19 February 1992)
