Brain Research, 556 (1991) 108-116 Elsevier Science Publishers B.V. A DONIS 000689939116885G
108
BRES 16885
Serotonergic sprouting is induced by dopamine-lesion in substantia nigra of adult rat brain Feng C. Zhou 1, Sharon
Bledsoe I
and James Murphy 2
1Department of Anatomy, Indiana University School of Medicine and 2Department of Psychology, Purdue School of Science, Indianapolis, IN (U.S.A.) (Accepted 26 March 1991) Key words: 5-HT hyperinnervation; Striatum; Heterotypic sprouting; Neuronal regeneration; CNS plasticity; Image analysis
We have previously extracted a serotonin (5-HT) neurotrophic supernatant from the 5,7-DHT lesioned hippocampus. The current study shows that a new 5-HT neurotrophic signal was monitored in the striatum and nigra after DA-denervation. Such a signal may be involved in the heterotypic sprouting. Dopaminergie neurotoxin, 6-hydroxydopamine (6-OHDA), was injected directly into the substantia nigra of adult rats. Two months after surgery, immtmocytoehemical staining showed that tyrosine hydroxylase (TH)-positive cell bodies had mostly disappeared in the substantia nigra, and TH-positive terminals in the striatum were almost completely depleted. Meanwhile, the 5-HT fibers, which exist in the same areas with low density, sprouted in the nigra as well as in the striatum and became dense. Normally 5-HT fibers innervate the striatum sparsely and the globus pallidus densely with sharp delineation (in the control side), and become dense across both areas with no appreciable delineation (in the lesion side). The increase of 5-HT fibers was more prominent in the posterior than in the anterior striatum. A significant increase in 5-HT and 5-HIAA levels was also evident in the posterior striatum when the decrease in DA level exceeded 90% in the nigra and striatum. In addition, we found that induction of 5-HT sprouting requires a >90% decrease of DA level. Current data support that 6-OHDA injection in the substantia nigra of adult rats triggered atrophic signal or removed an inhibition for the growth of 5-HT neurons which responded with sprouting in the nigra as well as in the striatum.
INTRODUCTION Nerve fibers grow vigorously during development, but growth ceases or equilibrium is reached when the brain matures. By the end of the developmental process, each neuronal system has constructed a precise and consistent innervation pattern and terminal density in every brain region. For example, in the hippocampus, the serotonin (5-HT), norepinephrine, cholinergic, and entorhinal inputs form an orderly laminated territory with distinct individual density (for review see Cotman et al.8). In the striatum and in the globus pallidus, the number of dopamine ( D A ) and 5-HT fibers is committed in the normal brain. The dopamine fibers are extremely dense and the 5-HT fibers extremely sparse in the striatum, while the opposite is true in the globus paUidus. One hypothesis which accounts for these observations is that in a given brain region, the density of each transmitter fiber system is a function of the amount of target factors present (such as specific receptors, extracellular matrix, and/or trophic factors etc.). A change in the dynamic equilibrium, such as a decrease in innervation density, often evokes either an upregulation of target factors (ex.
of amount receptors or its sensitivity) in the initial stage or resumption of innervation in the long term, if the neurons for reinnervation are available and/or a certain number of fibers is spared in the target field. This hypothesis clearly explains our and others' early observations on homotypic sprouting e.g. that adult intact fibers reinitiate growth in response to partial denervation of the same transmitter-type nerve fibers. Removing 5-HT fibers from the afferent path, the cingulum bundle, induced a collateral sprouting of the same type fibers from an alternative pathway, the fimbria-fornix into the dorsal hippocampus 1'2'2°'21. Similar observations were reported on noradrenergic and cholinergic systems in the hippocampus 4,9. However, the hypothesis of dynamic equilibrium does not extrapolate to the situation in which all or a majority of the viable neurons are degenerated. Here homotypic sprouting will not occur and reinnervation will not be achieved, such as in some experimental conditions and diseased states (Parkinsonism and Huntington's Disease). The dynamic equilibrium within a neuronal system is thus disrupted. We propose that in the striatum, a new dynamic equilibrium of fiber innervations within a given
Correspondence: EC. Zhou, Indiana University School of Medicine, Department of Anatomy, MS 258, Indianapolis, IN 46202, U.S.A.
109 region can be achieved between 2 different neuronal types when one type of neuron is not available. This hypothesis prevails in the developmental stage 6'12A6'17. We show that it can be extended into the mature brain, which may provide a rationale for heterotypic sprouting in the adult, a sprouting of intact fibers in response to damage of a different type of fibers (for review see C o t m a n et al.S). Two adjacent brain regions, the globus pallidus (GP) and the striatum, were chosen as the testing ground. The D A and 5-HT systems having clear neuronal pathways, discrete innervation patterns, and contrasting innervation densities were used to test dynamic changes. In the normal brain, D A terminal density is extremely high in the striatum and low in the GP, while 5-HT terminal density is high in the GP and low in the striatum. A specific neuronal toxin, 6-hydroxydopamine ( 6 - O H D A ) , was used to lesion the D A fibers in the striatum. Thus,
a specific manipulation of a single type fiber system can be performed, and the response to such unique changes will be specific. The dynamic change of D A and 5-HT systems in the striatum were monitored qualitatively and quantitatively with high performance liquid chromatography (HPLC) and imaging computer-assisted immunocytochemistry. MATERIALS AND METHODS 1. Animals and lesions
Young adult Sprague-Dawley rats (200-250 g) were used in this study. Animals were pretreated with desipramine (10 mg/kg, i.p.) 40 rain before neurotoxic lesion. The neurotoxin 6-hydroxydopamine (6-OHDA) was injected into the substantia nigra (SN) on the fight side of the brain (50/~g/10 ~ul ascorbic safine each, two ipsilateral injections) to remove DA innervation in the striatum. Our injections assured degeneration of the vast majority of DA neurons in the nigral region and DA fibers in the striatum as examined by our immunocytochemistry. A total of 24 rats were used, 4 for immunocytochemical examination of DA and 5-HT fiber density, and 16
Fig. 1. Degeneration of DA-nigro-striatal system. The DA neurons in the nigra were positive-stained with anti-TH antiserum in the substantia nigra (b,c; arrowheads). Two months after injection of neurotoxin 6-OHDA in the substantia nigra, most of the DA neurons in the injected nigra were degenerated into debris (a), while DA neurons in the contralateral side were mostly intact (a,b); meanwhile, the TH positive terminals had mostly disappeared in the striatum which is a major target region of the DA-nigral neurons. SN, substantia nigra; St, striatum; GP, globus pallidus. Asterisk, lesion side; Star, contralateral side to lesion. Bar = 300 ~tm in (a) and (b).
110
for HPLC measurement of DA and 5-HT levels. Intrinsic controls were used. In each animal, the contralateral side of the brain was used as control. To ensure that the plastic change of 5-HT fibers in the striatum was a result of 6-OHDA damage on DA fibers and not a non-specific lesion of 6-OHDA on 5-HT fibers, a control group, pretreated with desipramine (1 mg/100 g body weight) received 5-HT neurotoxin 5,7-dihydroxytryptamine (5,7-DHT) injection2°.21 into the nigral regions with the same coordinates (n = 4). The 5-HT fibers of this group were examined with immunocytochemistry 2 months after lesion.
Four animals were used for data analysis in a given region. Control. Precise consistency in section-thickness and Triton X-100 penetration is important for density and intensity measurements, and was strictly controlled among groups. The immunopenetration in this laboratory, estimated by Epon-embedded cross-section, is approximately 10 pm. In each experiment, the control and experimental groups were processed simultaneously under identical conditions. The variability between experiments was controlled by comparing the referenced brain regions in 2 experimental samples. The above method is routine in our laboratory 23,24.
2. lmmunocytochemistry of 5-HT and TH
4. High performance liquid chromatography (HPLC) analysis of 5-HT, DA and their metabolites
Two months after 6-OHDA or 5,7-DHT lesion, all animals were pretreated with pargyline, (Sigma, St. Louis, MO, 200 mg/kg, an inhibitor of the major 5-HT catabolic enzyme) 80 min prior, and L-tryptophan (Sigma, 200 mg/kg, a precursor of 5-HT) 60 min prior to perfusion. These animals were then perfused with formaldehyde (Reagent grade, Fisher Scientific), freshly made from 4% paraformaldehyde and 0.1 M phosphate buffered saline (PBS), intracardially under deep anesthesia. Their brains were then removed, left in the same fixative overnight and sectioned at 40/~m for immunocytochemical staining. 5-HT antiserum (produced against 5-HTLimulus hemocyanin conjugate in rabbit and characterized in our and Dr. E.C. Azmitia's laboratories) was used to stain 5-HT neurons and tyrosine hydroxylase (TH) antiserum (Eugene Tech, AUendale, NJ; produced in rabbit) was used to stain DA and NE neurons. The Sternberger's peroxidase-antiperoxidase (PAP) indirect-enzyme method (1979) was used for staining. The PAP reaction was done with 0.003% H20 2 and 0.05% 3"3-diaminobenzidine. The primary, secondary and marker antibodies were diluted with PBS containing 0.2% Triton X-100 and 1% normal sheep serum. The primary antibodies were incubated overnight, and the second and third for one hour.
3. Image-analysis of 5-HT fiber density The density of 5-HT fiber-ingrowth in the striatum and/or substantia nigra was examined as previously published23'24. In brief, a box-measurement was taken under a 10x objective lens under dark-field microscopy. Boxes were chosen in the same areas of brain-regions of both hemispheres in each section, and.were recorded for at least 12 consecutive sections of the brain-region on similar levels.
Rats were killed by decapitation, and the head was rapidly cooled by dipping into liquid N 2 for 4 s. The brains were removed and dissected on ice. Brain parts, substantia nigra, globus pallidus, anterior (to bregma) and posterior (to bregma) striata of each hemisphere, were collected separately, frozen in liquid Nz, sealed in aluminum foil, and stored at -70 °C until assayed for the contents of DA, and its metabolites, dihydroxyphenylacetic acid (DOPAC), homovanillic acid (HVA), and 5-HT and its metabolite, 5-hydroxyindoleacetic acid (5-HIAA) by HPLC with electrochemical detection according to published Procedures 13,14. Briefly, the brain areas were homogenized in 1 N formic acid/acetone (15:85, v/v) containing dihydroxybenzylamine (DHBA, Sigma Chemical Co., St. Louis, MO) as an internal standard. The samples were centrifuged at 900 g for 15 min. A 750/tl aliquot of the supernatant was washed with heptane/chloroform (8:1; v/v), and the organic phase was aspirated and discarded. The aqueous phase containing the compounds of interest was dried in a vacuum centrifuge and resuspended in 200 ~1 of the citric acid-sodium phosphate HPLC buffer, and 20 pi of each sample was injected into the HPLC apparatus. The mobile phase (pH 3.5) contained 17.5% (v/v) methanol, 29.6 mM citric acid, 14.3 mM disodium phosphate, 0.275 mM octyl sodium sulfate and 0.1 mM EDTA. Flow rate on the HPLC system was maintained at 1.0 ml/min through a 10 cm, 3p C18 reversed-phase column. The glassy carbon working electrode was set at 0.8 V versus the Ag/AgCI reference electrode. Peak areas were determined by integration, and sample compound contents were corrected for recovery with the internal standard and quantified against standard curves of true external standards for each compound (Sigma, St. Louis, MO).
TABLE I
Regional brain differences (intact-lesion) in contents of DA, DOPAC, HVA, 5-HT and 5-HIAA for rats receiving unilateral 6-OHDA lesion in substantia nigra HPLC assay showed that DA and its metabolites, DOPAC and HVA, were substantially decreased in the anterior and posterior striatum after the 6-OHDA injection in the nigra. Animals having greater than 90% DA-depletion in the nigra had greater levels of decrease of DA and metabolites in the striatum. The level of the 5-HT and its metabolite, 5-HIAA, in the striatum were significantly increased when greater than 90% of the DA were depleted in the nigra by 6-OHDA injection.
Mean + S. E. M. DA
DOPA C
HVA
5-HT
5-HIAA
Animals with less than 90% DA depletion in substantia nigra, n = 8 Substantia nigra Globuspallidus Anterior striatum Posterior striatum
0.48 + 13.61 + 44.05 + 47.33 +
1.28 6.78 6.69** 12.30"
-0.03 1.32 3.56 4.14
+ + + +
0.37 0.48 0.60** 1.39"
0.12 -0.29 1.63 2.51
+ + + +
0.16 0.33 0.27** 0.41"*
3.83 -0.86 -0.65 0.82
+ + + +
3.73 1.86 0.24 0.60
1.33 + 1.17 + -0.10 + 0.48 +
0.92 1.80 0.28 0.97
0.07 0.74 3.67 3.47
+ + + +
0.06 0.41 0.24** 0.70**
0.01 0.17 1.00 -2.30
+ + + +
2.08 1.09 0.73 0.75*
-0.73 0.30 0.75 -2.29
0.95 0.75 0.41 0.87*
Animals with greater than 90% DA depletion in substantia nigra, n = 8 Substantia nigra Globus pallidus Anterior striatum Posterior striatum
1.78 5.98 87.15 49.46
+ + + +
0.61" 2.57 2.59** 9.52**
0.21 0.75 5.58 4.60
+ + + +
0.12 0.69 0.51"* 1.02"*
* = P < 0.05 and ** = P < 0.01 by paired t-test for intact compared with lesion; unit = nmol/g wet tissue.
+ + + +
q
Lesionside suiaurm /-J Conualateralside Fig. 2. Dark-field microscopy shows that the 54-U immunoreactive (S-HT-IR) fibers turned from a sparse pattern in the normal striatum mto a dense pattern when the DA neurons in nigra were mostly degenerated and innervation in striatum was mostly depleted. The dense S-HT-JR fibers in the lesion side (b,d,f) form a sharp contrast to their contralateral counterparts (a,c,e). Paired photographs (a,b; c,d; and e,f’) in this panel were taken from same brain sections of each level indicated by the drawing. Notice that 5-HT-IR, normally sparse in striatum and dense in globus pallidus became dense in both areas (c,d). IC, internal capsule bundle; CC, corpus callosum. Bar = 300 ym for all photographs.
112
Fig. 3. The density difference of the 5-HT fibers between the normal and the lesioned side were obviously demonstrated using a 100x oil lens under bright-field microscopy. Normally, 5-HT fibers are sparse in striatum (a) and dense in globus pallidus (c). After 6-OHDA lesion of DA fibers, the 5-HT-immunoreactive fiber- density in the striatum (St) of the lesioned side were increased to a similar level of globus patlidus (GP). In addition, the 5-HT terminals in the lesion site had more prominent varicosities (arrowheads) than those of contralaterai controls (arrows). Bar = 20/tm for all photographs.
RESULTS
Dopaminergic degeneration The 6-hydroxydopamine (6-OHDA, 50 ~tg/10/~1 ascorbic saline, 2 ipsilateral injections) lesion in the nigra specifically removed the dopamine (DA) neurons in the nigra and the majority of the DA-terminal fibers in the striatum (stained with antiserum to tyrosine hydroxylase, Fig. 1). Immunocytochemistry showed that ~ilmost all the D A neurons in the substantia nigra area (A9) were degenerated, while a number of D A neurons were spared in the ventral tegmental area (A10). In the lesion side, D A fiber terminals in the striatum had mostly disappeared with very few spared fibers near the globus paUidus (GP); in the GP, the majority of the TH-positive fibers remained but had a twisted appearance (not shown). This is in contrast with the contralateral side where D A fibers were packed in the striatum, and appeared straight and sparsely distributed in the GP.
HPLC measurement revealed that levels of D A and its metabolites D O P A C and H V A were drastically decreased in substantia nigra, striatum and GP, but by varying degrees (Table I). By examining these changes, we found that only when D A fibers in the substantia nigra were decreased by more than 90%, would the 5-HT level increase significantly in the striatum (Table I).
Serotonergic hyperinnervation Two regions of 5-HT fibers were observed to increase after 6 - O H D A lesion in the SN - - the lesion site in the substantia nigra (Fig. 5), and the major DA-nigral target site, the striatum (Figs. 2 and 3). The 5-HT fibers accumulated densely in the 6 - O H D A injection site where normally they were sparsely distributed. Meanwhile, the 5-HT fibers were clearly increased in the striatum where D A fibers were almost completely depleted. Normally, 5-HT fibers are sparse in the striatum and dense in the GP with sharp contrast and delineation at the border; this
113 5-HT immunostaining Fiber Density is Increased in the Striatum after 6-OHDA Lesion in Nigra
a
80"
Anterior Striata
Anterior Pallidus
60
40"
[]
S~M
[]
MEAN
20' :):
:
0 Lesion
Contralateral
Lesion
Conwalateral
b 8O
60
Posterior Striata
Posterior Pallidus
*
~ 40'
°
i i,i iJ
20,
[] []
SEM Mean
0 Lesion
Contralateral
Lesion
C~ntralateral
Fig. 4. The density of 5-HT fibers in the lesioned and contralateral sides of anterior (a) and posterior striatum (b) and their correspondent globus pallidus were analyzed using an AIC 3000 image analyzer. The density of 5-HT fibers in the striatum of the lesioned side is significantly higher than that in the contralateral side, and is similar to those in both sides of the giobus pallidus (ANOVA, n = 4 in each case).
is shown in the striatum and G P in the side contralateral to the lesion. A dramatic change occurred in the striatum of the lesioned side - - 5-HT fiber density increased to such a level that the clear contrast and sharp borderline between the striatum and G P no longer existed (Fig. 2). D e n s i t o m e t r y image analysis showed the 5-HT fiber density in the ipsilateral (to lesion) striatum is significantly different from the contralateral striatum, and statistically similar to the G P of both hemispheres (Fig. 4). These changes were more dramatic in the posterior striatum than anterior striatum. The normal and altered D A and 5-HT innervation in the striatum and G P in response to the single sided 6 - O H D A - l e s i o n is summarized in Fig. 7. H P L C m e a s u r e m e n t supported this observation (Table I). A significant increase in 5-HT and 5 - H I A A levels was evident in the posterior striatum when the decrease in
Fig. 5. The 5-HT-IR fiber density also increased in the substantia nigra (SN) after 6-OHDA injection in this region. The 5-HT-IR terminals in the boxed area appeared denser in the lesion side as compared to those in the contralateral side to the lesion (a). Bar = 30/~m for (a) and (b).
5-HT Fiber Density in the Striatum after 5,7-DHT Lesion in Nigra l0 o
&
6-
:;-;,,;-;,;,;,:,,;.:,;,,-
4
ii!iii!!iiii!iiiii!ii
2
*
Lesion
[]
SEM
[]
MEAN
""'":':"'"'""'":
Contralatcnd
Fig. 6. The sprouting of 5-HT fibers in the striatum after 6-OHDA injection was due to DA and not the 5-HT neuronal lesion. The 5-HT fiber density is not increased, but decreased (*P < 0.01, Student's t-test) in the striatum two months after 5-HT neurotoxin, 5,7-DHT lesion in the nigra.
114
Normal Striatum-Pallidus
Striatum-Pallidus after 6-OHDA Lesion in Nigra
density~
fiberdensity
DA
~ raphe { ~ nigra
ng l ra
DP
~ substantia
1
~.JLesloneldwith6-OHDA
raphe
mgra
6-OHDA
[ ] Highfiber density [ ] Low fiber density [ ] Extremelylow fiber density
Fig. 7. Schematic drawing shows that the density-pattern of 5-HT and DA fibers (stained with anti-tyrosine hydroxylase) in the striatum (St) and globus pallidus (Gp) are mutually compensated high DA with low 5-HT in St, and low DA with high 5-HT in Gp. Partial lesion of DA neurons in the nigra results in low DA and low 5-HT in the St. However, when DA is decreased to less than 10% of the normal level, the 5-HT rises to high levels in the striatum. The net result is low DA with high 5-HT in both St and Gp.
D A level exceeded 90% in the nigra and 95% in striatum. In addition, we found that when the decreases in D A level in the substantia nigra and striatum were less than 90%, the 5-HT and 5-HIAA levels were not significantly changed in the striatum or globus pallidus. Control. To assure that the increase of 5-HT fiber terminals in the striatum was not due to a non-specific damage (by 6-OHDA, if any) of 5-HT fibers coursing through the substantia nigra, a 5,7-DHT lesion was made into the the same area to check if sprouting of 5-HT fibers in the striatum would be induced. This was not the case. The 5-HT fiber density in the striatum of the lesion side is lower than that of contralateral side over the 2 month period of time (Fig. 6) Thus, the sprouting of 5-HT fibers in the striatum was not due to the damage of 5-HT fibers in the nigral region within the 2 month period of time. DISCUSSION Injection of 6 - O H D A at the substantia nigra resulted in an increase of a number of 5-HT fibers in the nigra and striatum. Specific 6 - O H D A lesioning with desipramine blocking and a minor non-specific neuronal injury in the needle tract (40-80 /~m micropipet) in the substantia nigra result in a degeneration of D A neurons and a subsequent D A fiber-denervation of striatum. This degenerative manipulation involves fairly little damage to other neuronal systems. Thus, the expanding of 5-HT
fibers in the striatum is a sprouting specific to the dopaminergic degeneration in the striatum and nigra, and is a heterotypic collateral sprouting. Our control study with 5,7-DHT injection at the same coordinates as 6-OHDA injection to specifically lesion 5-HT fibers in the same area which resulted in a decrease (but not increase) of 5-HT fibers in the striatum in a 2 month period, indicates that the increase of 5-HT fiber density is not a result of non-specific injury of 5-HT fibers in the substantia nigra. Thus, in the adult brain, the expansion of 5-HT innervation of the striatum is a consequence of DA denervation in this area. This heterotypic sprouting is well demonstrated in neonatal rats, in which D A and 5-HT innervation has not taken shape and the neurons are still developing and are very plastic. Injection of 6 - O H D A in the fourth ventricle induced 5-HT sprouting into the striatum 6"12"16'17. However, no hyperinnervation of 5-HT fibers was reported in the adult brain after 6 - O H D A injection into the lateral ventricle 16"17. Our study with 6-OHDA injection in the substantia nigra showed that vigorous 5-HT hyperinnervation can be achieved in the striatum as well as in the nigra, which is not a result of damage to 5-HT fibers as indicated by our 5,7-DHT control. The plasticity of 5-HT fibers in the adult did not respond entirely differently from those in the new-born. The degree of manipulation may be taken into consideration for the discrepancy in the response. The different approaches of 6-OHDA injection (e.g. in substantia nigra versus lateral ventricles) and different levels of decrease of D A in the mesencephalon (48%) 12 may result in the varying degrees of 5-HT plasticity 16, especially in adult animals in which neurons normally cease to grow. A greater and closer stimulation may be required for triggering growth in the adult brain. A number of different phenomena were also observed between the adult and neonatal plasticity: (i) The 5-HT immunofiber density was increased in the nigra, and (ii) increased throughout the entire striatum, but (iii) the major increase of 5-HT was more in the caudal rather than the rostral. Current data indicate that the adult 5-HT neurons which cease to grow in the intact brain are still capable of reacting to the drastic degeneration of D A neurons in the nigra and denervation of D A fibers in the striatum. Further examination of the extent of lesion also indicated that the significant biochemical increase of 5-HT level and its metabolite 5 - H I A A only occurred when more than 90% of the D A level in the nigra and 95-98% in the striatum were depleted. The significant increase of 5-HT/5-HIAA level was mainly in the caudal striatum. In comparison with immunodensity, we found that the 5-HT fiber-density is significantly increased in both the rostrai and caudal striatum, and the level of
115 elevation of immunofiber density is greater than the elevation of 5-HT content. The increase in 5-HT fiber density may be on a larger scale than that of 5-HT content. This was also observed by Towle et al. TM. Our study indicates that the adult striatum neurons can support extra amounts of 5-HT fiber terminals when D A fiber reinnervation is not available. Not only the 5-HT system, but also Met-enkephalin was observed to have increased in the adult rats with 6 - O H D A lesion ~5. Functionally, whether the 5-HT transmission replaces D A transmission is uncertain. It is reported that the 5-HT hyperinnervation does not participate in locomotive dysfunction after denervation of D A fibers from the striatum 6'7. A series of specific chain reactions occur as the consequences of the D A denervation in the striatum. The D2 receptor m R N A of striato-pallidal neurons was elevated and D1 receptor m R N A of striatonigral neurons was reduced ~°. The information on 5-HT receptors in this paradigm is not yet available. This model of DA-5-HT degeneration-sprouting will be a very intriguing model for the study of the mechanism of heterotypic sprouting, and also interesting in relation to the establishment of new circuitry of different kinds. During growth of a neuronal system into its innervating territory, it is generally observed that 2 phenomena seem to be committed in a particular neuronal system in the normal mature brain - - the distribution of a given fiber terminal and the amount of innervation in a given location. It is speculated that this 'programmed terminal field' of a nervous system is a result of highly specific collaboration between the innervating neurons and the target tissue. This hypothesis is demonstrated in normal development, and is further supported by the transplant of fetal striatum into the adult striatum. In transplant, D A fibers and 5-HT fibers from adult striatum innervate the fetal striatum grafts with distribution and density patterns similar to those in normal development - - DA, extremely dense, and 5-HT, very sparse 11,19,23. The adult D A fibers even form a patchy pattern in the fetal striatum, just as when they were in their young developing stage. The example of specificity between innervating neurons and target tissue is extensive. The specificity between innervating neurons and target tissue is not only formed during development, but also preserved in the adult. In the case that innervation is reduced, efforts to restore it are often attempted. After injury, in addition REFERENCES 1 Azmitia, E.C., Buchan, A.M. and Williams, J.H., Structure and functional restoration by collateral sprouting of hippocampal 5-HT axons, Nature, 274 (1978) 374-376. 2 Azmitia, E.C. and Zhou, EC., Induced homotypic collateral sprouting of hippocampal serotonergic fibers. In G.M. Gilad, A.
to 5 types of non-neuronal reaction (see discussion in Zhou and Azmitia, 199024), there are at least 5 types of neuronal reaction. Four of them are regeneration, homotypic sprouting, reactive synaptogenesis, and receptor supersensitivity to attempt to return the minimal innervating neurons/target tissue equilibrium and to anatomically and/or physiologically 'recuperate' from the denervation. However, when reinnervating neurons are not available (e.g. completely degenerated), an attempt of ectopic balance between the target neurons and new innervating neurons can occur. The current study deafly demonstrates the fifth neuronal reaction to injury, which is that heterotypic neurons can sprout in response to specific degeneration of other types of neurons. This hypothesis is further supported in the similar DA-5-HT model in the cortex 5. These observations indicate that the specific relation and dynamic balance between innervating neurons and target tissue can be disrupted, and territory can be filled by different types of innervating neurons. To date, the mechanism of heterotypic sprouting remains unknown. A number of causes of the 5-HT sprouting in this model are speculated: (i) microenvironment of striatum is altered after denervation of D A fibers, such as transmission balance, calcium or protease concentration and/or electrical field, which trigger the mobility of the 5-HT growth cone; (ii) an inhibition is removed as a consequence of 6 - O H D A lesion and/or DA-denervation; and (iii) a trophic signal(s) to 5-HT neurons is (are) induced by 6 - O H D A which triggers or favors the growth of 5-HT fibers. Although 6 - O H D A causes local increase of 5-HT immunofiber density, it is not itself likely to be the best candidate because the 5-HT manifestation also occurred in striatum which is far from the injection site. This study provides ground work for important future studies on how heterotypic sprouting occurs in the adult brain, the functional consequences of this heterotypic sprouting, and how and if the receptors change to accommodate the new innervating neuronal type in the adult CNS. This information is fundamentally significant for injury and repair in the central nervous system. Acknowledgements. This study is supported by NIH grant NS23027 to EC.Z. The authors thank Ms. Shannon Svaglic for editing the manuscript.
Gorio and G.W. Kreutzberg (Eds.), Processes of Recovery from Neuronal Trauma, Elsevier, Netherlands, Exp. Brain Res. Suppl. 13, 1986, pp 129-141. 3 Berger, T.W., Kaul, S., Stricker, E.M. and Zigmond, M.J., Hyperinnervation of the striatum by dorsal raphe afferents after dopamine-depleting brain lesions in neonatal rats, Brain Research, 336 (1985) 354-358.
116 4 BjOrklund, A. and Stenevi, U., In vivo evidence for hippocampal adrenergic neuronotrophic factor specificallyreleased on septal deafferentation, Brain Research, 229 (1981) 403-428. 5 Blue, M.E. and Molliver, M.E., 6-Hydroxydopamine induces serotonergic axon sprouting in cerebral cortex of newborn rat, Dev. Brain Res., 32 (1987) 255-269. 6 Breeze, G.R., Baumeister, A.A., McCown, T.J., Emerick, S.G., Frye, G.D., Crotty, K. and Mueller, R.A., Behavioral differences between neonatal and adult 6-hydroxydopaminetreated rats to dopamine agonists: relevance to neurological symptoms in clinical syndromes with reduced brain dopamine, J. Pharmacol. Exp. Ther., 231 (1984) 343-354. 7 Bruno, J.P., Jackson, D., Zigmond, M.J. and Stricker, E.M., Effect of dopamine depleting brain lesions in rat pups: role of striatal serotonergic neurons in behavior, Behav. Neurosci., 10 (1987) 806-811. 8 Cotman, C.W., Nieto-Sampedro, M. and Harris, E.W., Synapse replacement in the nervous system of adult vertebrates, Physiol. Rev., 61 (1981) 684-784. 9 Gage, EH., Bjfrklund, A. and Stenevi, U., Reinnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and nor-adrenergic afferents, Brain Research, 268 (1983) 27-37. 10 Gerfen, C.R., Engber, T.M., Mahan, L.C., Susel, Z., Chase, T.N., Monsma Jr., EJ. and Sibley, D.R., D1 and D2 dopamine receptor-regulated gene expression of striatonigral and striatopallidal neurons, Science, 250 (1990) 1429-1432. 11 Graybiel, A.M., Pickel, V.M., Joh, T.H., Reis, D.J. and Ragsdale, C.W., Direct demonstrations of a correspondence between the dopamine islands and acetylcholinesterase patches in the developing striatum, Proc. Natl. Acad. Sci. U.S.A., 78 (1981) 5871-5875. 12 Luthman, J., Bolioli, B., Tsutsumi, T., Verhofstad, A. and Jonsson, G., Sprouting of striatal serotonin nerve terminals following selective lesions of nigro-striatal dopamine neurons in neonatal rats, Brain Res. Bull., 19 (1987) 269-274. 13 Murphy, J.M., McBride, W.J., Lumeng, L. and Li, T.-K., Regional brain levels of monoamines in alcohol-preferring and non-preferring lines of rats, Pharmacol. Biochem. Behav., 16 (1982) 145-149. 14 Murphy, J.M., McBride, W.J., Lumeng, L. and Li, T.-K.,
Monoamine and metabolite levels in CNS regions of the P line of alcohol-preferring rats after acute and chronic ethanol treatment, Pharmacol. Biochem. Behav., 19 (1983) 849-856. 15 Sivam, S.P., Breese, G.R., Krayse, J.E., Napier, T.C., Mueiler, R.A. and Hong, J.-S., Neonatal and adult 6-hydroxydopamineinduced lesions differentially alter tachykinin and enkephalin gene expression, J. Neurochem., 49 (1987) 1623-1632. 16 Snyder, A.M., Zigmond, M.J. and Lund, R.D., Sprouting of serotoninergic afferents into striatum after dopamine-depleting lesions in infant rats; a retrograde transport and immunocytochemical study, J. Comp. Neurol., 245 (1986) 274-281. 17 Stachowiak, M.K., Bruno, J.P., Snyder, A.M., Stricker, E.M. and Zigmond, M.J., Apparent sprouting of striatal serotonergic terminals after dopamine-depleting brain lesions in neonatal rats, Brain Research, 291 (1984) 164-167. 18 Towle, A.C., Crisweli, H.E., Maynard, E.H., Lander, J.M., Joh, T.H., Mueller, R.A. and Breese, G.R., Serotonergic innervation of the rat caudate following a neonatal 6-hydroxydopamine lesion: an anatomical, biochemical and pharmacological study, Pharm. Biochem. Behav., 34 (1989) 167-374. 19 Wictorin, K., Isacson, O., Fischer, W., Nothias, E, Peschanski, M. and BjSrklund, A., Connectivity of striatal grafts implanted into the ibotenic acid-lesioned striatum - - I subcortical afferents, Neuroscience, 27 (1988) 547-562. 20 Zhou, F.C. and Azmitia, E.C., Induced homotypic collateral sprouting of serotonergic fibers in hippocampus, Brain Research, 308 (1984) 53-62. 21 Zhou, EC. and Azmitia, E.C., Induced homotypic sprouting of serotonergic fibers in hippocampus: II. An immunocytochemical study, Brain Research, 373 (1986) 337-348. 22 Zhou, EC., Auerbach, S. and Azmitia, E.C., Injury of serotonergic fibers induces a factor which promotes the maturation of serotonergic neurons but not norepinephrine neurons, J. Neurosci. Res., 17 (1987)235-246. 23 Zhou, EC. and Buchwald, N., Connectivities of striatal grafts in the adult rat brain: a rich afference and scant nigrostriatal efference, Brain Research, 504 (1989) 15-30. 24 Zhou, EC. and Azmitia, E.C., Neurotrophic factor for serotonergic neurons prevents degeneration of grafted raphe neurons in the cerebellum, Brain Research, 507 (1990) 301-308.
108
BRES 16885
Serotonergic sprouting is induced by dopamine-lesion in substantia nigra of adult rat brain Feng C. Zhou 1, Sharon
Bledsoe I
and James Murphy 2
1Department of Anatomy, Indiana University School of Medicine and 2Department of Psychology, Purdue School of Science, Indianapolis, IN (U.S.A.) (Accepted 26 March 1991) Key words: 5-HT hyperinnervation; Striatum; Heterotypic sprouting; Neuronal regeneration; CNS plasticity; Image analysis
We have previously extracted a serotonin (5-HT) neurotrophic supernatant from the 5,7-DHT lesioned hippocampus. The current study shows that a new 5-HT neurotrophic signal was monitored in the striatum and nigra after DA-denervation. Such a signal may be involved in the heterotypic sprouting. Dopaminergie neurotoxin, 6-hydroxydopamine (6-OHDA), was injected directly into the substantia nigra of adult rats. Two months after surgery, immtmocytoehemical staining showed that tyrosine hydroxylase (TH)-positive cell bodies had mostly disappeared in the substantia nigra, and TH-positive terminals in the striatum were almost completely depleted. Meanwhile, the 5-HT fibers, which exist in the same areas with low density, sprouted in the nigra as well as in the striatum and became dense. Normally 5-HT fibers innervate the striatum sparsely and the globus pallidus densely with sharp delineation (in the control side), and become dense across both areas with no appreciable delineation (in the lesion side). The increase of 5-HT fibers was more prominent in the posterior than in the anterior striatum. A significant increase in 5-HT and 5-HIAA levels was also evident in the posterior striatum when the decrease in DA level exceeded 90% in the nigra and striatum. In addition, we found that induction of 5-HT sprouting requires a >90% decrease of DA level. Current data support that 6-OHDA injection in the substantia nigra of adult rats triggered atrophic signal or removed an inhibition for the growth of 5-HT neurons which responded with sprouting in the nigra as well as in the striatum.
INTRODUCTION Nerve fibers grow vigorously during development, but growth ceases or equilibrium is reached when the brain matures. By the end of the developmental process, each neuronal system has constructed a precise and consistent innervation pattern and terminal density in every brain region. For example, in the hippocampus, the serotonin (5-HT), norepinephrine, cholinergic, and entorhinal inputs form an orderly laminated territory with distinct individual density (for review see Cotman et al.8). In the striatum and in the globus pallidus, the number of dopamine ( D A ) and 5-HT fibers is committed in the normal brain. The dopamine fibers are extremely dense and the 5-HT fibers extremely sparse in the striatum, while the opposite is true in the globus paUidus. One hypothesis which accounts for these observations is that in a given brain region, the density of each transmitter fiber system is a function of the amount of target factors present (such as specific receptors, extracellular matrix, and/or trophic factors etc.). A change in the dynamic equilibrium, such as a decrease in innervation density, often evokes either an upregulation of target factors (ex.
of amount receptors or its sensitivity) in the initial stage or resumption of innervation in the long term, if the neurons for reinnervation are available and/or a certain number of fibers is spared in the target field. This hypothesis clearly explains our and others' early observations on homotypic sprouting e.g. that adult intact fibers reinitiate growth in response to partial denervation of the same transmitter-type nerve fibers. Removing 5-HT fibers from the afferent path, the cingulum bundle, induced a collateral sprouting of the same type fibers from an alternative pathway, the fimbria-fornix into the dorsal hippocampus 1'2'2°'21. Similar observations were reported on noradrenergic and cholinergic systems in the hippocampus 4,9. However, the hypothesis of dynamic equilibrium does not extrapolate to the situation in which all or a majority of the viable neurons are degenerated. Here homotypic sprouting will not occur and reinnervation will not be achieved, such as in some experimental conditions and diseased states (Parkinsonism and Huntington's Disease). The dynamic equilibrium within a neuronal system is thus disrupted. We propose that in the striatum, a new dynamic equilibrium of fiber innervations within a given
Correspondence: EC. Zhou, Indiana University School of Medicine, Department of Anatomy, MS 258, Indianapolis, IN 46202, U.S.A.
109 region can be achieved between 2 different neuronal types when one type of neuron is not available. This hypothesis prevails in the developmental stage 6'12A6'17. We show that it can be extended into the mature brain, which may provide a rationale for heterotypic sprouting in the adult, a sprouting of intact fibers in response to damage of a different type of fibers (for review see C o t m a n et al.S). Two adjacent brain regions, the globus pallidus (GP) and the striatum, were chosen as the testing ground. The D A and 5-HT systems having clear neuronal pathways, discrete innervation patterns, and contrasting innervation densities were used to test dynamic changes. In the normal brain, D A terminal density is extremely high in the striatum and low in the GP, while 5-HT terminal density is high in the GP and low in the striatum. A specific neuronal toxin, 6-hydroxydopamine ( 6 - O H D A ) , was used to lesion the D A fibers in the striatum. Thus,
a specific manipulation of a single type fiber system can be performed, and the response to such unique changes will be specific. The dynamic change of D A and 5-HT systems in the striatum were monitored qualitatively and quantitatively with high performance liquid chromatography (HPLC) and imaging computer-assisted immunocytochemistry. MATERIALS AND METHODS 1. Animals and lesions
Young adult Sprague-Dawley rats (200-250 g) were used in this study. Animals were pretreated with desipramine (10 mg/kg, i.p.) 40 rain before neurotoxic lesion. The neurotoxin 6-hydroxydopamine (6-OHDA) was injected into the substantia nigra (SN) on the fight side of the brain (50/~g/10 ~ul ascorbic safine each, two ipsilateral injections) to remove DA innervation in the striatum. Our injections assured degeneration of the vast majority of DA neurons in the nigral region and DA fibers in the striatum as examined by our immunocytochemistry. A total of 24 rats were used, 4 for immunocytochemical examination of DA and 5-HT fiber density, and 16
Fig. 1. Degeneration of DA-nigro-striatal system. The DA neurons in the nigra were positive-stained with anti-TH antiserum in the substantia nigra (b,c; arrowheads). Two months after injection of neurotoxin 6-OHDA in the substantia nigra, most of the DA neurons in the injected nigra were degenerated into debris (a), while DA neurons in the contralateral side were mostly intact (a,b); meanwhile, the TH positive terminals had mostly disappeared in the striatum which is a major target region of the DA-nigral neurons. SN, substantia nigra; St, striatum; GP, globus pallidus. Asterisk, lesion side; Star, contralateral side to lesion. Bar = 300 ~tm in (a) and (b).
110
for HPLC measurement of DA and 5-HT levels. Intrinsic controls were used. In each animal, the contralateral side of the brain was used as control. To ensure that the plastic change of 5-HT fibers in the striatum was a result of 6-OHDA damage on DA fibers and not a non-specific lesion of 6-OHDA on 5-HT fibers, a control group, pretreated with desipramine (1 mg/100 g body weight) received 5-HT neurotoxin 5,7-dihydroxytryptamine (5,7-DHT) injection2°.21 into the nigral regions with the same coordinates (n = 4). The 5-HT fibers of this group were examined with immunocytochemistry 2 months after lesion.
Four animals were used for data analysis in a given region. Control. Precise consistency in section-thickness and Triton X-100 penetration is important for density and intensity measurements, and was strictly controlled among groups. The immunopenetration in this laboratory, estimated by Epon-embedded cross-section, is approximately 10 pm. In each experiment, the control and experimental groups were processed simultaneously under identical conditions. The variability between experiments was controlled by comparing the referenced brain regions in 2 experimental samples. The above method is routine in our laboratory 23,24.
2. lmmunocytochemistry of 5-HT and TH
4. High performance liquid chromatography (HPLC) analysis of 5-HT, DA and their metabolites
Two months after 6-OHDA or 5,7-DHT lesion, all animals were pretreated with pargyline, (Sigma, St. Louis, MO, 200 mg/kg, an inhibitor of the major 5-HT catabolic enzyme) 80 min prior, and L-tryptophan (Sigma, 200 mg/kg, a precursor of 5-HT) 60 min prior to perfusion. These animals were then perfused with formaldehyde (Reagent grade, Fisher Scientific), freshly made from 4% paraformaldehyde and 0.1 M phosphate buffered saline (PBS), intracardially under deep anesthesia. Their brains were then removed, left in the same fixative overnight and sectioned at 40/~m for immunocytochemical staining. 5-HT antiserum (produced against 5-HTLimulus hemocyanin conjugate in rabbit and characterized in our and Dr. E.C. Azmitia's laboratories) was used to stain 5-HT neurons and tyrosine hydroxylase (TH) antiserum (Eugene Tech, AUendale, NJ; produced in rabbit) was used to stain DA and NE neurons. The Sternberger's peroxidase-antiperoxidase (PAP) indirect-enzyme method (1979) was used for staining. The PAP reaction was done with 0.003% H20 2 and 0.05% 3"3-diaminobenzidine. The primary, secondary and marker antibodies were diluted with PBS containing 0.2% Triton X-100 and 1% normal sheep serum. The primary antibodies were incubated overnight, and the second and third for one hour.
3. Image-analysis of 5-HT fiber density The density of 5-HT fiber-ingrowth in the striatum and/or substantia nigra was examined as previously published23'24. In brief, a box-measurement was taken under a 10x objective lens under dark-field microscopy. Boxes were chosen in the same areas of brain-regions of both hemispheres in each section, and.were recorded for at least 12 consecutive sections of the brain-region on similar levels.
Rats were killed by decapitation, and the head was rapidly cooled by dipping into liquid N 2 for 4 s. The brains were removed and dissected on ice. Brain parts, substantia nigra, globus pallidus, anterior (to bregma) and posterior (to bregma) striata of each hemisphere, were collected separately, frozen in liquid Nz, sealed in aluminum foil, and stored at -70 °C until assayed for the contents of DA, and its metabolites, dihydroxyphenylacetic acid (DOPAC), homovanillic acid (HVA), and 5-HT and its metabolite, 5-hydroxyindoleacetic acid (5-HIAA) by HPLC with electrochemical detection according to published Procedures 13,14. Briefly, the brain areas were homogenized in 1 N formic acid/acetone (15:85, v/v) containing dihydroxybenzylamine (DHBA, Sigma Chemical Co., St. Louis, MO) as an internal standard. The samples were centrifuged at 900 g for 15 min. A 750/tl aliquot of the supernatant was washed with heptane/chloroform (8:1; v/v), and the organic phase was aspirated and discarded. The aqueous phase containing the compounds of interest was dried in a vacuum centrifuge and resuspended in 200 ~1 of the citric acid-sodium phosphate HPLC buffer, and 20 pi of each sample was injected into the HPLC apparatus. The mobile phase (pH 3.5) contained 17.5% (v/v) methanol, 29.6 mM citric acid, 14.3 mM disodium phosphate, 0.275 mM octyl sodium sulfate and 0.1 mM EDTA. Flow rate on the HPLC system was maintained at 1.0 ml/min through a 10 cm, 3p C18 reversed-phase column. The glassy carbon working electrode was set at 0.8 V versus the Ag/AgCI reference electrode. Peak areas were determined by integration, and sample compound contents were corrected for recovery with the internal standard and quantified against standard curves of true external standards for each compound (Sigma, St. Louis, MO).
TABLE I
Regional brain differences (intact-lesion) in contents of DA, DOPAC, HVA, 5-HT and 5-HIAA for rats receiving unilateral 6-OHDA lesion in substantia nigra HPLC assay showed that DA and its metabolites, DOPAC and HVA, were substantially decreased in the anterior and posterior striatum after the 6-OHDA injection in the nigra. Animals having greater than 90% DA-depletion in the nigra had greater levels of decrease of DA and metabolites in the striatum. The level of the 5-HT and its metabolite, 5-HIAA, in the striatum were significantly increased when greater than 90% of the DA were depleted in the nigra by 6-OHDA injection.
Mean + S. E. M. DA
DOPA C
HVA
5-HT
5-HIAA
Animals with less than 90% DA depletion in substantia nigra, n = 8 Substantia nigra Globuspallidus Anterior striatum Posterior striatum
0.48 + 13.61 + 44.05 + 47.33 +
1.28 6.78 6.69** 12.30"
-0.03 1.32 3.56 4.14
+ + + +
0.37 0.48 0.60** 1.39"
0.12 -0.29 1.63 2.51
+ + + +
0.16 0.33 0.27** 0.41"*
3.83 -0.86 -0.65 0.82
+ + + +
3.73 1.86 0.24 0.60
1.33 + 1.17 + -0.10 + 0.48 +
0.92 1.80 0.28 0.97
0.07 0.74 3.67 3.47
+ + + +
0.06 0.41 0.24** 0.70**
0.01 0.17 1.00 -2.30
+ + + +
2.08 1.09 0.73 0.75*
-0.73 0.30 0.75 -2.29
0.95 0.75 0.41 0.87*
Animals with greater than 90% DA depletion in substantia nigra, n = 8 Substantia nigra Globus pallidus Anterior striatum Posterior striatum
1.78 5.98 87.15 49.46
+ + + +
0.61" 2.57 2.59** 9.52**
0.21 0.75 5.58 4.60
+ + + +
0.12 0.69 0.51"* 1.02"*
* = P < 0.05 and ** = P < 0.01 by paired t-test for intact compared with lesion; unit = nmol/g wet tissue.
+ + + +
q
Lesionside suiaurm /-J Conualateralside Fig. 2. Dark-field microscopy shows that the 54-U immunoreactive (S-HT-IR) fibers turned from a sparse pattern in the normal striatum mto a dense pattern when the DA neurons in nigra were mostly degenerated and innervation in striatum was mostly depleted. The dense S-HT-JR fibers in the lesion side (b,d,f) form a sharp contrast to their contralateral counterparts (a,c,e). Paired photographs (a,b; c,d; and e,f’) in this panel were taken from same brain sections of each level indicated by the drawing. Notice that 5-HT-IR, normally sparse in striatum and dense in globus pallidus became dense in both areas (c,d). IC, internal capsule bundle; CC, corpus callosum. Bar = 300 ym for all photographs.
112
Fig. 3. The density difference of the 5-HT fibers between the normal and the lesioned side were obviously demonstrated using a 100x oil lens under bright-field microscopy. Normally, 5-HT fibers are sparse in striatum (a) and dense in globus pallidus (c). After 6-OHDA lesion of DA fibers, the 5-HT-immunoreactive fiber- density in the striatum (St) of the lesioned side were increased to a similar level of globus patlidus (GP). In addition, the 5-HT terminals in the lesion site had more prominent varicosities (arrowheads) than those of contralaterai controls (arrows). Bar = 20/tm for all photographs.
RESULTS
Dopaminergic degeneration The 6-hydroxydopamine (6-OHDA, 50 ~tg/10/~1 ascorbic saline, 2 ipsilateral injections) lesion in the nigra specifically removed the dopamine (DA) neurons in the nigra and the majority of the DA-terminal fibers in the striatum (stained with antiserum to tyrosine hydroxylase, Fig. 1). Immunocytochemistry showed that ~ilmost all the D A neurons in the substantia nigra area (A9) were degenerated, while a number of D A neurons were spared in the ventral tegmental area (A10). In the lesion side, D A fiber terminals in the striatum had mostly disappeared with very few spared fibers near the globus paUidus (GP); in the GP, the majority of the TH-positive fibers remained but had a twisted appearance (not shown). This is in contrast with the contralateral side where D A fibers were packed in the striatum, and appeared straight and sparsely distributed in the GP.
HPLC measurement revealed that levels of D A and its metabolites D O P A C and H V A were drastically decreased in substantia nigra, striatum and GP, but by varying degrees (Table I). By examining these changes, we found that only when D A fibers in the substantia nigra were decreased by more than 90%, would the 5-HT level increase significantly in the striatum (Table I).
Serotonergic hyperinnervation Two regions of 5-HT fibers were observed to increase after 6 - O H D A lesion in the SN - - the lesion site in the substantia nigra (Fig. 5), and the major DA-nigral target site, the striatum (Figs. 2 and 3). The 5-HT fibers accumulated densely in the 6 - O H D A injection site where normally they were sparsely distributed. Meanwhile, the 5-HT fibers were clearly increased in the striatum where D A fibers were almost completely depleted. Normally, 5-HT fibers are sparse in the striatum and dense in the GP with sharp contrast and delineation at the border; this
113 5-HT immunostaining Fiber Density is Increased in the Striatum after 6-OHDA Lesion in Nigra
a
80"
Anterior Striata
Anterior Pallidus
60
40"
[]
S~M
[]
MEAN
20' :):
:
0 Lesion
Contralateral
Lesion
Conwalateral
b 8O
60
Posterior Striata
Posterior Pallidus
*
~ 40'
°
i i,i iJ
20,
[] []
SEM Mean
0 Lesion
Contralateral
Lesion
C~ntralateral
Fig. 4. The density of 5-HT fibers in the lesioned and contralateral sides of anterior (a) and posterior striatum (b) and their correspondent globus pallidus were analyzed using an AIC 3000 image analyzer. The density of 5-HT fibers in the striatum of the lesioned side is significantly higher than that in the contralateral side, and is similar to those in both sides of the giobus pallidus (ANOVA, n = 4 in each case).
is shown in the striatum and G P in the side contralateral to the lesion. A dramatic change occurred in the striatum of the lesioned side - - 5-HT fiber density increased to such a level that the clear contrast and sharp borderline between the striatum and G P no longer existed (Fig. 2). D e n s i t o m e t r y image analysis showed the 5-HT fiber density in the ipsilateral (to lesion) striatum is significantly different from the contralateral striatum, and statistically similar to the G P of both hemispheres (Fig. 4). These changes were more dramatic in the posterior striatum than anterior striatum. The normal and altered D A and 5-HT innervation in the striatum and G P in response to the single sided 6 - O H D A - l e s i o n is summarized in Fig. 7. H P L C m e a s u r e m e n t supported this observation (Table I). A significant increase in 5-HT and 5 - H I A A levels was evident in the posterior striatum when the decrease in
Fig. 5. The 5-HT-IR fiber density also increased in the substantia nigra (SN) after 6-OHDA injection in this region. The 5-HT-IR terminals in the boxed area appeared denser in the lesion side as compared to those in the contralateral side to the lesion (a). Bar = 30/~m for (a) and (b).
5-HT Fiber Density in the Striatum after 5,7-DHT Lesion in Nigra l0 o
&
6-
:;-;,,;-;,;,;,:,,;.:,;,,-
4
ii!iii!!iiii!iiiii!ii
2
*
Lesion
[]
SEM
[]
MEAN
""'":':"'"'""'":
Contralatcnd
Fig. 6. The sprouting of 5-HT fibers in the striatum after 6-OHDA injection was due to DA and not the 5-HT neuronal lesion. The 5-HT fiber density is not increased, but decreased (*P < 0.01, Student's t-test) in the striatum two months after 5-HT neurotoxin, 5,7-DHT lesion in the nigra.
114
Normal Striatum-Pallidus
Striatum-Pallidus after 6-OHDA Lesion in Nigra
density~
fiberdensity
DA
~ raphe { ~ nigra
ng l ra
DP
~ substantia
1
~.JLesloneldwith6-OHDA
raphe
mgra
6-OHDA
[ ] Highfiber density [ ] Low fiber density [ ] Extremelylow fiber density
Fig. 7. Schematic drawing shows that the density-pattern of 5-HT and DA fibers (stained with anti-tyrosine hydroxylase) in the striatum (St) and globus pallidus (Gp) are mutually compensated high DA with low 5-HT in St, and low DA with high 5-HT in Gp. Partial lesion of DA neurons in the nigra results in low DA and low 5-HT in the St. However, when DA is decreased to less than 10% of the normal level, the 5-HT rises to high levels in the striatum. The net result is low DA with high 5-HT in both St and Gp.
D A level exceeded 90% in the nigra and 95% in striatum. In addition, we found that when the decreases in D A level in the substantia nigra and striatum were less than 90%, the 5-HT and 5-HIAA levels were not significantly changed in the striatum or globus pallidus. Control. To assure that the increase of 5-HT fiber terminals in the striatum was not due to a non-specific damage (by 6-OHDA, if any) of 5-HT fibers coursing through the substantia nigra, a 5,7-DHT lesion was made into the the same area to check if sprouting of 5-HT fibers in the striatum would be induced. This was not the case. The 5-HT fiber density in the striatum of the lesion side is lower than that of contralateral side over the 2 month period of time (Fig. 6) Thus, the sprouting of 5-HT fibers in the striatum was not due to the damage of 5-HT fibers in the nigral region within the 2 month period of time. DISCUSSION Injection of 6 - O H D A at the substantia nigra resulted in an increase of a number of 5-HT fibers in the nigra and striatum. Specific 6 - O H D A lesioning with desipramine blocking and a minor non-specific neuronal injury in the needle tract (40-80 /~m micropipet) in the substantia nigra result in a degeneration of D A neurons and a subsequent D A fiber-denervation of striatum. This degenerative manipulation involves fairly little damage to other neuronal systems. Thus, the expanding of 5-HT
fibers in the striatum is a sprouting specific to the dopaminergic degeneration in the striatum and nigra, and is a heterotypic collateral sprouting. Our control study with 5,7-DHT injection at the same coordinates as 6-OHDA injection to specifically lesion 5-HT fibers in the same area which resulted in a decrease (but not increase) of 5-HT fibers in the striatum in a 2 month period, indicates that the increase of 5-HT fiber density is not a result of non-specific injury of 5-HT fibers in the substantia nigra. Thus, in the adult brain, the expansion of 5-HT innervation of the striatum is a consequence of DA denervation in this area. This heterotypic sprouting is well demonstrated in neonatal rats, in which D A and 5-HT innervation has not taken shape and the neurons are still developing and are very plastic. Injection of 6 - O H D A in the fourth ventricle induced 5-HT sprouting into the striatum 6"12"16'17. However, no hyperinnervation of 5-HT fibers was reported in the adult brain after 6 - O H D A injection into the lateral ventricle 16"17. Our study with 6-OHDA injection in the substantia nigra showed that vigorous 5-HT hyperinnervation can be achieved in the striatum as well as in the nigra, which is not a result of damage to 5-HT fibers as indicated by our 5,7-DHT control. The plasticity of 5-HT fibers in the adult did not respond entirely differently from those in the new-born. The degree of manipulation may be taken into consideration for the discrepancy in the response. The different approaches of 6-OHDA injection (e.g. in substantia nigra versus lateral ventricles) and different levels of decrease of D A in the mesencephalon (48%) 12 may result in the varying degrees of 5-HT plasticity 16, especially in adult animals in which neurons normally cease to grow. A greater and closer stimulation may be required for triggering growth in the adult brain. A number of different phenomena were also observed between the adult and neonatal plasticity: (i) The 5-HT immunofiber density was increased in the nigra, and (ii) increased throughout the entire striatum, but (iii) the major increase of 5-HT was more in the caudal rather than the rostral. Current data indicate that the adult 5-HT neurons which cease to grow in the intact brain are still capable of reacting to the drastic degeneration of D A neurons in the nigra and denervation of D A fibers in the striatum. Further examination of the extent of lesion also indicated that the significant biochemical increase of 5-HT level and its metabolite 5 - H I A A only occurred when more than 90% of the D A level in the nigra and 95-98% in the striatum were depleted. The significant increase of 5-HT/5-HIAA level was mainly in the caudal striatum. In comparison with immunodensity, we found that the 5-HT fiber-density is significantly increased in both the rostrai and caudal striatum, and the level of
115 elevation of immunofiber density is greater than the elevation of 5-HT content. The increase in 5-HT fiber density may be on a larger scale than that of 5-HT content. This was also observed by Towle et al. TM. Our study indicates that the adult striatum neurons can support extra amounts of 5-HT fiber terminals when D A fiber reinnervation is not available. Not only the 5-HT system, but also Met-enkephalin was observed to have increased in the adult rats with 6 - O H D A lesion ~5. Functionally, whether the 5-HT transmission replaces D A transmission is uncertain. It is reported that the 5-HT hyperinnervation does not participate in locomotive dysfunction after denervation of D A fibers from the striatum 6'7. A series of specific chain reactions occur as the consequences of the D A denervation in the striatum. The D2 receptor m R N A of striato-pallidal neurons was elevated and D1 receptor m R N A of striatonigral neurons was reduced ~°. The information on 5-HT receptors in this paradigm is not yet available. This model of DA-5-HT degeneration-sprouting will be a very intriguing model for the study of the mechanism of heterotypic sprouting, and also interesting in relation to the establishment of new circuitry of different kinds. During growth of a neuronal system into its innervating territory, it is generally observed that 2 phenomena seem to be committed in a particular neuronal system in the normal mature brain - - the distribution of a given fiber terminal and the amount of innervation in a given location. It is speculated that this 'programmed terminal field' of a nervous system is a result of highly specific collaboration between the innervating neurons and the target tissue. This hypothesis is demonstrated in normal development, and is further supported by the transplant of fetal striatum into the adult striatum. In transplant, D A fibers and 5-HT fibers from adult striatum innervate the fetal striatum grafts with distribution and density patterns similar to those in normal development - - DA, extremely dense, and 5-HT, very sparse 11,19,23. The adult D A fibers even form a patchy pattern in the fetal striatum, just as when they were in their young developing stage. The example of specificity between innervating neurons and target tissue is extensive. The specificity between innervating neurons and target tissue is not only formed during development, but also preserved in the adult. In the case that innervation is reduced, efforts to restore it are often attempted. After injury, in addition REFERENCES 1 Azmitia, E.C., Buchan, A.M. and Williams, J.H., Structure and functional restoration by collateral sprouting of hippocampal 5-HT axons, Nature, 274 (1978) 374-376. 2 Azmitia, E.C. and Zhou, EC., Induced homotypic collateral sprouting of hippocampal serotonergic fibers. In G.M. Gilad, A.
to 5 types of non-neuronal reaction (see discussion in Zhou and Azmitia, 199024), there are at least 5 types of neuronal reaction. Four of them are regeneration, homotypic sprouting, reactive synaptogenesis, and receptor supersensitivity to attempt to return the minimal innervating neurons/target tissue equilibrium and to anatomically and/or physiologically 'recuperate' from the denervation. However, when reinnervating neurons are not available (e.g. completely degenerated), an attempt of ectopic balance between the target neurons and new innervating neurons can occur. The current study deafly demonstrates the fifth neuronal reaction to injury, which is that heterotypic neurons can sprout in response to specific degeneration of other types of neurons. This hypothesis is further supported in the similar DA-5-HT model in the cortex 5. These observations indicate that the specific relation and dynamic balance between innervating neurons and target tissue can be disrupted, and territory can be filled by different types of innervating neurons. To date, the mechanism of heterotypic sprouting remains unknown. A number of causes of the 5-HT sprouting in this model are speculated: (i) microenvironment of striatum is altered after denervation of D A fibers, such as transmission balance, calcium or protease concentration and/or electrical field, which trigger the mobility of the 5-HT growth cone; (ii) an inhibition is removed as a consequence of 6 - O H D A lesion and/or DA-denervation; and (iii) a trophic signal(s) to 5-HT neurons is (are) induced by 6 - O H D A which triggers or favors the growth of 5-HT fibers. Although 6 - O H D A causes local increase of 5-HT immunofiber density, it is not itself likely to be the best candidate because the 5-HT manifestation also occurred in striatum which is far from the injection site. This study provides ground work for important future studies on how heterotypic sprouting occurs in the adult brain, the functional consequences of this heterotypic sprouting, and how and if the receptors change to accommodate the new innervating neuronal type in the adult CNS. This information is fundamentally significant for injury and repair in the central nervous system. Acknowledgements. This study is supported by NIH grant NS23027 to EC.Z. The authors thank Ms. Shannon Svaglic for editing the manuscript.
Gorio and G.W. Kreutzberg (Eds.), Processes of Recovery from Neuronal Trauma, Elsevier, Netherlands, Exp. Brain Res. Suppl. 13, 1986, pp 129-141. 3 Berger, T.W., Kaul, S., Stricker, E.M. and Zigmond, M.J., Hyperinnervation of the striatum by dorsal raphe afferents after dopamine-depleting brain lesions in neonatal rats, Brain Research, 336 (1985) 354-358.
116 4 BjOrklund, A. and Stenevi, U., In vivo evidence for hippocampal adrenergic neuronotrophic factor specificallyreleased on septal deafferentation, Brain Research, 229 (1981) 403-428. 5 Blue, M.E. and Molliver, M.E., 6-Hydroxydopamine induces serotonergic axon sprouting in cerebral cortex of newborn rat, Dev. Brain Res., 32 (1987) 255-269. 6 Breeze, G.R., Baumeister, A.A., McCown, T.J., Emerick, S.G., Frye, G.D., Crotty, K. and Mueller, R.A., Behavioral differences between neonatal and adult 6-hydroxydopaminetreated rats to dopamine agonists: relevance to neurological symptoms in clinical syndromes with reduced brain dopamine, J. Pharmacol. Exp. Ther., 231 (1984) 343-354. 7 Bruno, J.P., Jackson, D., Zigmond, M.J. and Stricker, E.M., Effect of dopamine depleting brain lesions in rat pups: role of striatal serotonergic neurons in behavior, Behav. Neurosci., 10 (1987) 806-811. 8 Cotman, C.W., Nieto-Sampedro, M. and Harris, E.W., Synapse replacement in the nervous system of adult vertebrates, Physiol. Rev., 61 (1981) 684-784. 9 Gage, EH., Bjfrklund, A. and Stenevi, U., Reinnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and nor-adrenergic afferents, Brain Research, 268 (1983) 27-37. 10 Gerfen, C.R., Engber, T.M., Mahan, L.C., Susel, Z., Chase, T.N., Monsma Jr., EJ. and Sibley, D.R., D1 and D2 dopamine receptor-regulated gene expression of striatonigral and striatopallidal neurons, Science, 250 (1990) 1429-1432. 11 Graybiel, A.M., Pickel, V.M., Joh, T.H., Reis, D.J. and Ragsdale, C.W., Direct demonstrations of a correspondence between the dopamine islands and acetylcholinesterase patches in the developing striatum, Proc. Natl. Acad. Sci. U.S.A., 78 (1981) 5871-5875. 12 Luthman, J., Bolioli, B., Tsutsumi, T., Verhofstad, A. and Jonsson, G., Sprouting of striatal serotonin nerve terminals following selective lesions of nigro-striatal dopamine neurons in neonatal rats, Brain Res. Bull., 19 (1987) 269-274. 13 Murphy, J.M., McBride, W.J., Lumeng, L. and Li, T.-K., Regional brain levels of monoamines in alcohol-preferring and non-preferring lines of rats, Pharmacol. Biochem. Behav., 16 (1982) 145-149. 14 Murphy, J.M., McBride, W.J., Lumeng, L. and Li, T.-K.,
Monoamine and metabolite levels in CNS regions of the P line of alcohol-preferring rats after acute and chronic ethanol treatment, Pharmacol. Biochem. Behav., 19 (1983) 849-856. 15 Sivam, S.P., Breese, G.R., Krayse, J.E., Napier, T.C., Mueiler, R.A. and Hong, J.-S., Neonatal and adult 6-hydroxydopamineinduced lesions differentially alter tachykinin and enkephalin gene expression, J. Neurochem., 49 (1987) 1623-1632. 16 Snyder, A.M., Zigmond, M.J. and Lund, R.D., Sprouting of serotoninergic afferents into striatum after dopamine-depleting lesions in infant rats; a retrograde transport and immunocytochemical study, J. Comp. Neurol., 245 (1986) 274-281. 17 Stachowiak, M.K., Bruno, J.P., Snyder, A.M., Stricker, E.M. and Zigmond, M.J., Apparent sprouting of striatal serotonergic terminals after dopamine-depleting brain lesions in neonatal rats, Brain Research, 291 (1984) 164-167. 18 Towle, A.C., Crisweli, H.E., Maynard, E.H., Lander, J.M., Joh, T.H., Mueller, R.A. and Breese, G.R., Serotonergic innervation of the rat caudate following a neonatal 6-hydroxydopamine lesion: an anatomical, biochemical and pharmacological study, Pharm. Biochem. Behav., 34 (1989) 167-374. 19 Wictorin, K., Isacson, O., Fischer, W., Nothias, E, Peschanski, M. and BjSrklund, A., Connectivity of striatal grafts implanted into the ibotenic acid-lesioned striatum - - I subcortical afferents, Neuroscience, 27 (1988) 547-562. 20 Zhou, F.C. and Azmitia, E.C., Induced homotypic collateral sprouting of serotonergic fibers in hippocampus, Brain Research, 308 (1984) 53-62. 21 Zhou, EC. and Azmitia, E.C., Induced homotypic sprouting of serotonergic fibers in hippocampus: II. An immunocytochemical study, Brain Research, 373 (1986) 337-348. 22 Zhou, EC., Auerbach, S. and Azmitia, E.C., Injury of serotonergic fibers induces a factor which promotes the maturation of serotonergic neurons but not norepinephrine neurons, J. Neurosci. Res., 17 (1987)235-246. 23 Zhou, EC. and Buchwald, N., Connectivities of striatal grafts in the adult rat brain: a rich afference and scant nigrostriatal efference, Brain Research, 504 (1989) 15-30. 24 Zhou, EC. and Azmitia, E.C., Neurotrophic factor for serotonergic neurons prevents degeneration of grafted raphe neurons in the cerebellum, Brain Research, 507 (1990) 301-308.
