Effect of Prior Dopamine Denervation on Survival and Fiber Outgrowth from Intrastriatal Fetal...

European Journal of Neuroscience, Vol. 2 , pp. 279-290 @European Neuroscience Association 09.53-816~/90 $3.00

Effect of Prior Dopamine Denervation on Survival and Fiber Outgrowth from lntrastriatal Fetal Mesencephalic Grafts

G. Doucet, P. Brundin, L. Descarries’, and A. Bjorklund2

Qepartement de physiologie, Universite de Montreal, Montreal, Canada Department of Medical Cell Research, Section of Neurobiology, University of Lund, Lund, Sweden

Key words: transplantation, dopamine varicosities, dopamine neurons, [3HlDA uptake radioautography, tyrosine hydroxytase irnrnunocytochemistry

Abstract

[3HlDopamine (DA) uptake radioautography and tyrosine hydroxylase (TH) immunocytochemistry were used to assess quantitatively the effects of the presence or absence of host mesostriatal DA afferents on the survival and fiber outgrowth from fetal ventral mesencephalic DA neurons grafted into the neostriatum of adult recipient rats. Rats received bilateral intrastriatal transplants of fetal ventral mesencephalic tissue 1 month after a unilateral injection of 6-hydroxydopamine (6-OHDA) into the right nigrostriatal bundle (denervated side). Five to six months later, some of the grafted rats received a second 6-OHDA injection in the left nigrostriatal bundle (acutely denervated or ‘intact’ side). After a further 7 days, slices of each hemisphere from the latter rats were incubated with t3H]DA and processed for film and high resolution radioautography. The density of the film radioautographs was measured with a computerized image analysis system and calibrated by silver grain cluster (i.e. DA terminal) counting over selected areas of the same sections in light microscope radioautographs. The brains of the remaining grafted rats were processed for TH immunoreactivity 6- 12 months after graft surgery. Neither the size of the grafts, nor the number of surviving TH-positive graft neurons showed any significant difference between the nondenervated and the denervated sides. However, the size of the TH-positive cell bodies was significantly greater in the grafts on the denervated side. In the t3H]DA uptake radioautographs, considerable outgrowth of DA fibers was evident in the neostriatum on the ‘intact’ side in spite of the presence of an intact host DA innervation until 7 days before sacrifice. The overall DA fiber outgrowth was nevertheless almost two-fold greater on the denervated side, and extended deeper into the host neostriatum than on the ‘intact’ side; only 7% of the total neostriatal area, on average, was at background level compared to 30% on the ‘intact’ side, and the overall density of neostriatal DA innervation amounted to 36% of normal as compared to 20% on the ‘intact’ side. The correlation between the overall density of graft-derived DA innervation and the size of the grafts was linear on the ‘intact’ side, but reached a plateau with relatively small grafts on the denervated side. However, the ventral striatum on both sides was very poorly innervated by these grafts. These findings demonstrate that the mature neostriatal tissue can support axonal growth and innervation from grafted fetal DA neurons even in the presence of a normal complement of endogenous DA fibers. Prior removal of the host striatal DA innervation does not influence the overall size of the grafts nor the number of surviving DA neurons, but induces an increase in the cell body size and fiber outgrowth of the grafted DA neurons.

Introduction

Most studies on intrastriatal dopamine (DA) rich transplants of ventral mesencephalic tissue have focused on graft morphology and function in the striatum previously denervated of its endogeneous DA afferents by a complete neurotoxic lesion (see Brundin and Bjorklund 1987 for a recent review). However, in a recent investigation in weaver mice

(Doucet er al., 1989), which are characterized by a partial loss of the nigrostriatal DA neurons, we described a complementary distribution of the graft-derived DA fibers and of the residual host DA afferents within the target, suggesting that spared host DA neurons may influence the growth characteristics of intrastriatal DA neuron grafts. Schmidt

Correspondence to: G . Doucet, UniversitC de Montreal, FacultC de mtdecine, Centre de recherche en sciences neurologiques, C.P. 6128, succursale A, Montreal (Quebec), Canada H3C 337

Received 21 September 1989, revised 25 October 1989, accepted I3 November 1989

280 Fiber outgrowth from dopamine transplants

et al. (1981) and Gage et al. (1983) had already reported that DA fiber outgrowth from fetal mesencephalic grafts was more restricted when implanted into a non denervated striatum, but no attempts to quantify the differences had been made in these early studies.

The question of whether spared host DA neurons may influence the survival and fiber outgrowth of implanted fetal DA neurons is of particular interest in the context of the ongoing transplantation trials in patients with Parkinson’s disease. In this clinical situation the grafted neurons are placed in an environment where the host nigrostriatal system is partially intact. More detailed knowledge on the competition between graft and host DA neurons in the innervation of the host striatum is thus highly warranted.

The aim of the present study was to obtain precise quantitative information on the differences in survival and fiber outgrowth from grafted fetal mesencephalic DA neurons implanted into the intact or DA denervated striatum. For this purpose, the authors have developed further a recently introduced radioautographic method (Doucet et al . , 1986, 1989) for quantifying DA innervation density in extended areas of the rat neostriatum. The authors have also used tyrosine hydroxylase (TH) immunocytochemistry in order to measure the size of the grafts and the number and size of the grafted DA cell bodies.

Materials and methods

Experimental animals

Twenty-five young adult female Sprague-Dawley rats (ALAB, Stockholm, Initial body weight 250*50 g) were used in this study. Twenty-three were subjected to unilateral 6-hydroxydopamine (6-OHDA) lesions of the mesencephalic DA neurons. The drug (6-OHDA-HC1, Sigma, 3 mg/ml in 0.2 m g / d ascorbate-saline) was injected under Equithesin anaesthesia at two sites: (1) 2 pl at 4 mm posterior to bregma, 0.8 mm to the right and 8.0 mm below dura with the tooth bar at 3.4 mm; and (2) 2.5 pl at 4.4 mm posterior to bregma, 1.2 mm to the right and 7.8 mm below dura with the tooth bar at 2.3 mm. Two weeks later, the rats were behaviorally tested in rotometer bowls during 90 rnin following the administration of 5 mg/kg of d-amphetamine (Ungerstedt and Arbuthnott, 1970). Only rats showing a rotational bias of more than seven full turns per minute towards the lesioned side underwent graft surgery after a further 3 weeks.

Nineteen lesioned rats received bilateral intrastriatal injections of ventral mesencephalic cell suspensions from embryonic day 13 - 15 rat fetuses (crown-rump length 10- 14 mm) according to the technique described by Bjorklund et al. (1983). Two aliquots of 1.5 pl (from 50 pl cell suspensions of dissociated mesencephalic tissue taken from 8 - 13 fetuses of pregnant females deeply anaesthetized with Equithesin) were injected under Equithesin anaesthesia on each side of the brain at the following coordinates (tooth bar at 0): 1 .O mm anterior to bregma, 3.0 mm to the right, 5.0 and 4.1 mm below dura. The recipients were retested for amphetamine-induced turning behavior 3 -4 months after transplantation. Twelve rats showed clearly reduced or reversed side bias in their rotational behavior, indicating the presence of a surviving graft at least on the right (lesioned) side. Six other rats showed no modification of their pregrafting rotational behavior and one further rat died some time after grafting. Eight of the behaviorally compensated rats, together with a normal rat, were used for [3H]DA uptake experiments. The brains of the remaining compensated and noncompensated animals were processed for tyrosine hydroxylase (TH) immunocytochemistry .

TH-irnrnunocytochernistry

Six to 12 months after graft surgery, rats were perfused under deep chloral hydrate anaesthesia according to the ‘pH shift’ perfusion technique of Berod et al. (1981) as described by Gerfen and Sawchenko (1984). After perfusion, thick slices including the neostriatum (and the graft) were further fixed overnight in 4 % paraformaldehyde, rinsed in 0.01 M potassium phosphate buffered saline (PBS), and then sectioned at 40 pm on a vibrating microtome. The free floating sections were further rinsed and pre-incubated for 1 h at room temperature with 10% normal swine serum and 0.5% Triton X100. Primary incubation was carried out with a rabbit polyclonal antibody against TH (Eugene Tech Int., 1:2OOO), at 4”C, for 40 h, in the presence of 5% normal swine serum and 0.5% Triton X100. For peroxidase-antiperoxidase immunostaining, we used a swine antibody against rabbit immunoglobulins (Dakopatts, 150) and a rabbit peroxidase- antiperoxidase complex (Dakopatts, 1 : 100). The peroxidase activity was revealed with 0.05% 3,3’-diaminobenzidine (Sigma) and 0.04% hydrogen peroxide. Finally, the reaction product was intensified on flattened sections by successive baths in 0.005% osmium (30 min), 0.05 % thiocarbohydrazide (Electron Microscopy Sciences, USA) (15 rnin), and again in osmium (30 rnin), with 30 min tap water rinsing between baths.

t3HlDA uptake radioautography

One normal and 8 grafted rats were used for [3H]DA uptake and radioautography , 5 -9 months after transplantation surgery. Seven days before the experiments, the grafted rats were anaesthetized with Equithesin and subjected to 6-OHDA injections into the intact (left) side of the ventral midbrain in order to remove the host DA innervation of the neostriatum. Preliminary experiments had indicated that this survival period following 6-OHDA injection was the minimum required in order to obtain a complete disappearance of the specific labeling in radioautographs of the neostriatum after [3H]DA uptake in v i m .

To allow quantification of the DA innervation in large areas of the neostriatum and in several animals, the radioautographic procedure described by Doucet et a/. (1986) was modified by adding an intermediate step of film radioautography to the processing of sections. Briefly, the rats were perfused through the heart under deep chloral hydrate anaesthesia with ice-cold artificial cerebrospinal fluid (CSF). Five or six 200 pm-thick coronal slices of each hemisphere were cut on a vibrating microtome, through the rostral half of the striatum. The first slice was always taken at the rostral tip of the graft, and the whole graft or the major part of it was included in the collected slices. The slices were pre-incubated at 15°C. for 15 min, in CSF containing 0.1 mM pargyline and 5 pM desipramine (5-6 sliced8 ml). They were then incubated for 90 min after adding [3H]DA to a final concentration of 1 pM. The tracer (7,8-[3H]dopamine, 46 Ci/mmol, Amersham) was freshly evaporated from stock solutions and diluted 1:2 with nonradioactive DA-HCI (Sigma) in CSF containing 0.2 mg/ml ascorbic acid. The slices were thereafter fixed in 3.5% glutaraldehyde, post- fixed with osmium vapours and flat-embedded in Epon resin. Four micrometer-thick sections (one section per slice) of the whole hemispheres were obtained on a Polycut microtome (Jung) and flattened in absolute ethanol onto chromalum-gelatin coated slides. These sections were first set in contact with tritium-sensitive film (LKB Ultrofilm3H) which was developed in D-19 after 3 days. Afterwards, they were processed for light microscopic radioautography by dipping in Ilford K-5 nuclear emulsion diluted 1: 1, and developed in D- 19 after 5 days

Fiber outgrowth from dopamine transplants 28 1

of exposure. In addition, 6 series of 6 X 0.5 ym-thick sections, flattened on separate glass slides, were similarly coated with nuclear emulsion but developed after various periods of exposure (3-52 days).

Quantification

Graji size, and TH cell numbers and diameters in immunostained sections One out of every three 40 ym-thick sections sampling the whole grafts were used for these measurements. The size of the grafts was measured on TH-immunostained sections by weighing paper after cutting along the contours of the graft on camera lucida drawings. The volume was calculated from the cross-sectional area and the distance between the sections. The grafted TH neurons were counted visually in the microscope. Their cell body diameter was estimated by measuring under 40 X objective magnification with a calibrated ocular grid the long and short axes of 160 TH-immunostained cell bodies on each side of the brain, sampled randomly from four animals. The number of TH positive cell bodies was then corrected according to Abercrombie (1946) in order to estimate the total population of surviving DA neurons in each graft. The data on the above parameters were compared between the two sides of the brain by a paired Student’s t-test.

DA innervation density and graft size measurements in radioautographs The graft-induced DA re-innervation of the neostriatum and the size of the grafts were quantified on digitized images of the film radioautographs, using a computerized image analysis system (IBAS 11, Kontron).

First, on one section per side in each animal, the density of DA innervation was quantified by measuring grey levels simultaneously in a grid on 525 X 0.04 mm2 squares covering the whole neostriatum (Fig. 1). The background, which was relatively low compared to the specific labeling, was not subtracted. To calibrate grey value measurements with numbers of DA varicosities, the latter were counted in light microscope radioautographs of the same sections by charting silver grain clusters (SGC) in 76 microscopic fields (0.04 mm2 each) selected from both sides of the brain in all eight grafted rats. Counting was performed using a microscope (Leitz) equipped with a 25X objective lens (Leitz, PL APO), a camera lucida and an ocular grid (see Doucet et al., 1986). The exact position of each microscopic field was determined from landmarks in the sections drawn after adjusting the magnification of the light microscope and camera lucida to projection drawings of the film radioautographs including an overlay of the 525 x 0.04 mm2 square grid. Measures of density (on the grey scale, i.e. white=255 and black=O) on film radioautographs, as well as of the number of labeled DA varicosities in light microscope radioautographs, were thus obtained in exactly the same grid squares of the sections. The correlation between the two sets of values could then be computed (Fig. 2), allowing the transformation of the grey values obtained in the remaining areas of the sections into relative numbers of DA varicosities per 0.04 mm2 of section. These measurements were used to compare the density of DA innervation at various distances from the grafts on the two sides of the brain (2-way ANOVA). The counts of SGC in serial semi-thin sections at different exposure times indicated that 20 -25 % of the labeled DA varicosities were detected after 5 days exposure (see Doucet et a t . , 1986).

A second set of measurements was performed with the image analysis system after interactively delineating the contours of the neostriatum and of the grafts on the digitized image of the film radioautographs.

W FIG. I . Semi-schematic projection drawing of the film radioautograph depicted in Figure 4C, with an overlay of a 525 X 0.04 mm2 square grid. The grid served two purposes: to orient the film radioautographs in the image analysis system (which also included a similar overlay) and to precisely localize selected 0.04 m2 microscopic fields in the high resolution radioautographs of the same sections (see Materials and methods). Density measurements of film radioautographs and numbers of labeled DA varicosities in high resolution radioautographs could then be obtained for the same 0.04 mm2 areas of a given section.

Bwl I A 9

600-

ii3 a W I- v)

0 3 400- ‘ I g 200

\r

50 100

10001 \

50 100

GREY VALUES GREY VALUES

FIG. 2. (A) Relationship between the number of SGC (labeled DA varicosities) counted

in high resolution radioautographs and the grey value (white=255; black =0) derived by image analysis of film radioautographic density. Each point represents the SGC number and corresponding grey value for the same 0.04 mm2 areas of neostriatum innervated by DA grafts sampled from both sides of the brain in all grafted animals.

(B) Relationship between the natural logarithm of the number of SGC and the grey value representing the film radioautographic density. This line was used as calibration to transform grey values into numbers of labeled DA varicosities.

(i) The size of the neostriatum and of the grafts was measured by counting in every section the total number of pixels (image points = 100 pm2) in the delineated areas. The mean sectional area of each


graft was then calculated, taking into account the sections containing no graft tissue. This measure was then used as a size index of the grafts for the correlation analysis with overall DA innervation density (see below). (ii) The proportions of neostriatum (excluding the grafts) displaying different densities of DA innervation were also measured. For this, the whole grey scale was subdivided into intervals of ten grey levels. On every section (five to six sections per side in the eight grafted rats and two sections from the normal rat), the number of pixels with density values falling within each grey scale interval was computed as percent of the total neostriatal area. These measurements were used to build histograms in order to test whether larger proportions of the neostriatum were densely innervated by grafted DA neurons on the lesioned side than on the ‘intact’ side. The histograms of each side of the brain were compared by a x2 test. The value obtained in the sections from the normal rat were used merely as reference. (iii) The latter data also served to obtain an estimate of the overall density of DA innervation of the neostriatum (in numbers of varicosities per mm2) on each side of the grafted brains. For this, the calibration described above was used to find the median value, in number of varicosities per mm2, corresponding to each grey scale interval. For

each section, the appropriate median values were then multiplied by the percent neostriatal area covered by pixels within the corresponding interval of grey values and the sum of every product was made for all the intervals of the grey scale, yielding a weighted average density of DA innervation per mm2. These overall density values were used to compare the innervation (two-tailed unpaired Student’s t-test) and to analyse their correlation with the size of the grafts on the lesioned and on the ‘intact’ sides.

Results

Graft size, and number and size of TH-positive cell bodies in the grafts

The general morphology of the grafts (Fig. 3) was that of a well delineated cellular mass of highly variable size, elongated dorsoventrally across the head of the caudate putamen and often extending into the overlying frontal cortex. The graft was obliquely oriented, rostrocaudally , relative to the plane of sectioning. The TH-positive cell bodies were preferentially in the peripheral areas of the grafts.

FIG. 3. Intrastriatal ventral mesencephalic grafts in sections processed for TH-immunocytochemistry. (A) Intact side. (B) Denervated side. The graft and neostriatal tissues can be distinguished by the presence of the fiber fascicles of the internal capsule coursing through the neostriatum (white

profiles). Most of the TH-positive cell bodies are found at the periphery of the graft, and some have migrated for short distances into the neostriatal tissue. The TH-positive staining of neostriatal tissue is in DA fibers of the host as well as of the graft on the intact side (the density of stained fibers was indeed higher near the graft), but only from the graft on the denervated side. Scale bar: 200 pm.

Fiber outgrowth from dopamine transplants 283

The graft volume, as assessed from the TH-immunostained sections, was not significantly different between the side lesioned with 6-OHDA (0.77 &0.22 mm’) and the side left nondenervated (0.73 *0.3 1 mm’) (n=9). The same was found with the grafted brains processed for [ ’HIDA uptake and radioautography : the average cross-sectional area of the grafts was 1.10&0.27 mm2 on the side denervated prior to grafting and 1 .OO & 0.28 mm2 on the side left intact until 7 days before the uptake experiments (n=8).

The number of TH-immunopositive cell bodies was not significantly different between the 6-OHDA denervated side (932 f 338) and the intact side (1 159&432) (n=9), whereas the mean length of the short and long axes of the TH-positive cell bodies was significantly larger on the lesioned side (17 .3 i~0 .6 pm compared to 15.4+0.2 pm on the intact side; piO.05, paired Student’s t-test, 40 cells measured on each side of the brain in each of four apimals). In the radioautographs, the grafted DA cell bodies were not labeled after incubation with [3H]DA in vitro, and hence no assessment of DA neuron survival could be made in the radioautographic experiments.

The graft volume and the cell number in the denervated striata did not differ significantly between rats that showed a reduction in amphetamine-induced rotation (n = 5) and those that displayed no graft effect (n=4) in the rotation test.

Graft-derived DA innervation of the neostriatum General observations The distribution of the DA innervation on the side denervated before grafting and on the side left intact until shortly before the uptake experiments (‘intact’ side) is illustrated in the film radioautographs of Figure 4. In general, DA fiber outgrowth extended in most of the head of the caudate-putamen but was less dense in the periventricular zones. However, very little, if any, DA fiber outgrowth of graft origin was evidenced in nucleus accumbens, even in cases where the ventral part of the graft was almost adjacent to that area. As evident from Figure 4, there was a significant DA fiber outgrowth into the host neostriatum on the ‘intact’ side although the DA innervation by the graft was generally more extensive on the denervated side. In the TH- immunostained specimens, which did not receive a 6-OHDA lesion shortly before the perfusion, it was also clear that the grafts gave rise to a hyperinnervation of the host neostriatum on the nondenervated side. The pattern of this hyperinnervation was consistent with the distribution and density of the DA varicosities found on the ‘intact’ side in the radioautographs.

The appearance of the labeled DA varicosities proximal and distal to the grafts is shown in Figure 5, as visualized in the light microscope radioautographs. There was no apparent difference in the varicosity labeling between the denervated and ‘intact’ sides of the brain. The background was also similar on both sides and from one rat to another, as evidenced by grey value measurements in the cerebral cortex of all rats. However, the background increased in proportion to the density of DA innervation. This is a common feature of the in vitro radioautographic labeling of monoamine innervations, explained by two factors closely linked to the density of innervation: (i) the partial release or mobilization of labeled monoamines at the time of fixation; and (ii) the presence of intervaricose axonal segments within which the radioactivity is less concentrated and does not produce dense SGC in the overlying emulsion. Quantitative data Figure 2 shows the relationship between the film radioautographic density measurements (on the grey scale) given by the image analysis

system and the number of DA varicosities counted in the light microscope radioautographs (i.e. on the emulsion coated slides) of the same sections. The results being similar for both sides of the brain and in different animals, all data were pooled in order to generate the calibration curve. Although the correlation coefficient was relatively good after a simple regression of the data (r=0.832), it was clear that the relation was not linear (Fig. 2a). A better fit was indeed obtained after the data for the numbers of varicosities were transformed into their natural logarithm (Fig. 2b; r=0.954). The resulting curve was thus used to transform the grey values given by the image analysis of the film radioautographs into numbers of varicosities per unit area. Since the data from both sides of the brain fitted well to this single curve, it was concluded that the transformed grey values could be used for comparison of graft derived DA innervation densities between the ‘intact’ and denervated neostriatum.

The overall density of DA innervation by the graft was significantly higher on the lesioned side ([3.55 *0.45] X lo3 varicosities per mm2) than on the ‘intact’ side ([1.93&0.36] X 103/mm2) (2-tailed, unpaired Student’s t-test, p=0.019). In comparison, the measurements similarly performed in a few sections from a normal rat gave a reference value of 9.74 x 103/mm2. Thus, the head of the caudate-putamen was, on average, innervated by DA varicosities of graft origin at 36% (10% -53%) of normal on the denervated side, and at 20% (5% -43%) of normal on the ‘intact’ side.

As expected, the highest density of DA innervation was found in the neostriatal tissue adjoining the graft, with a gradual decrease with distances up to 1.6- 1.8 mm from the graft-host border (or until the ventricle was reached). This is illustrated in Figure 6 where the number of DA varicosities per grid square (0.04 mm2) is plotted as a function of distance from the graft. The local DA innervation density was higher on the denervated side but decreased with distance from the graft in a similar fashion on both the ‘intact’ and the denervated sides (2-way ANOVA, main effects of side [p<O.OOOl] and distance [p<O.OOOI], but non-significant side X distance interaction). Immediately medial to the graft, the mean density of DA varicosities originating from the graft was 4.35 x lo3 per mm2 (range: 2.5-8.8 X lo3) on the ‘intact’ side and 5.35 x lo’ varicosities per mm2 (range: 1.3-7.9 X 10’) on the denervated side. Compared to the overall DA varicosity density measured in the normal neostriatum (see above), the graft-derived reinnervation close to the graft amounted to 45% (25-90%) and 55% (13-80%), respectively, of the normal DA innervation. There was no significant difference between the densities of DA innervation (of graft origin) measured immediately medial, lateral, or ventral to the graft, and this was true for both sides of the brain.

There was a significant difference between the ‘intact’ and the denervated sides in the percent cross-sectional areas of neostriatum covered by different densities of DA innervation from the graft (x2=52.715, 9 degrees of freedom, p=O.OOOl) (Fig. 7). On the ‘intact’ side, on the average nearly 30% of neostriatum was completely devoid of DA varicosities of graft origin as compared to only 7 % on the denervated side. Arbitrarily, taking 13% of the reference value for the overall density of DA innervation in the normal rat to represent a potentially significant innervation, a mean of 76% of the denervated neostriatum was reached by at least this degree of innervation by the graft as opposed to only 45 % for the ‘intact’ neostriatum. Only very small percentages of the neostriatum on either sides were innervated by graft with densities of DA varicosities in the normal range.

Figure 8 shows the relationship between the size of the grafts and the resulting overall density of DA innervation in the neostriatum. On


the ‘intact’ side, the relationship between graft size and innervation density appeared linear and, indeed, the correlation was significant after a simple regression (Pearson’s correlation; r=0.866; p=0.006). However, on the denervated side, there was not a significant correlation after simple regression (r=0.667; p=0.07). A better fit was obtained after transforming the graft size data into 1 /ex and the correlation was then significant (r=0.875; p=0.004). This supported the impression given by the film radioautographs (Fig. 4) that the D A innervation

density reached a plateau with relatively small grafts. Some graft tissue was usually found in the frontal cortex, along the

needle tract. This graft tissue also produced some outgrowth of D A fibers into the frontal cortex for distances of up to 1 mm from the graft (see Fig. 4). D A varicosities were then found in any layers of the adjacent cortex, including the most superficial layers where spots of D A varicosities could reach a maximal density of up to 5 X 103/mm2 in extreme cases.


FIG. 4. Film radioautographs of rat brain hemisphere slices incubated in virro with [3H]DA. The sections from the left side, denervated shortly before the [3H]DA uptake experiments (‘intact’ side) (A-D), and from the right side, denervated before grafting (A‘-D’), were matched according to graft size and, thus, do not necessarily come from the same brains. In many of these sections, some spots of the frontal cortex adjacent to the graft also showed an outgrowth of 13H]DA- labeled varicosities. Note that the 6-OHDA injections into the ventral midbrain usually spared some of the DA innervation of the ventral striatum (nucleus accumbens and olfactory tubercle). This residual innervation from the host DA neurons was always clearly distinct from the innervation originating in the grafts. Scale bars: 2 nun.

Discussion

The radioautographic method used for this study offered several advantages for the quantification of the DA innervation in the

neostriatum. The DA storage sites visualized in the light microscope radioautographs as small clusters of silver grains presumably represent the DA release sites and are thus the most meaningful morphological units to be counted in order to describe the DA terminal network in


FIG. 5 . Light microscope radioautographs of the neostriatum proximal (A and C) and distal (B and D) to the intrastriatal grafts on the ‘intact’ (denervated shortly before sacrifice) (A and B) and predenervated (C and D) sides. A and B and C and D are from the same brains as the film radioautographs shown in Figures 4D and 4C’, respectively. The area shown in A was immediately adjacent to the graft, whereas the area in B was 0.4 mm from it. C was at the distance of 0.05 mm from the graft, and D was 0.8 mm away. Scale bar: 50 pm.

quantitative terms. The specificity of this approach for labeling DA axonal varicosities in normal rat striatum has been discussed at length previously (Doucet era/. , 1986). As mentioned above, the grafted DA cell bodies were not labeled by the uptake process in virro. This was consistent with previous observations on grafted DA neurons (Doucet etal . , 1989), as well as on the normal ventral mesencephalon (unpublished observation, and Berger, B . , personal communication). The same was probably true for DA dendrites since the few SGC visualized inside the grafts were likely to represent DA axonal varicosities, such as those observed by electron microscopy within grafts processed for DA immunocytochemistry (unpublished observations).

Because the present labeling technique made use of fresh brain tissue,

the quantitative results were not affected by factors such as perfusion- fixation and were thus highly reproducible between animals. Moreover, the possibility of processing the same sections for film and high resolution radioautography allowed for macroscopic as well as microscopic analysis. This was particularly valuable since it was possible to transform film density measurements, quickly performed by the image analysis system over large areas of the neostriatum and in several rats, into numbers of DA axonal varicosities. Actual counting of varicosities in the microscope could then be restricted to a limited number of microscopic fields. The computerized image analysis also provided a more objective means of assessing the effects of experimental manipulations.


250-

200- N

E E

9 0 \

[L W I- ul

U

*

*150.

3

5 100 LT U

[L w > =' v,

50

Background

- Denervated side --- "In tact" side - Denervated side --- "In tact" side

1

I I I I I 1 I

1 2 3 4 5 6 3 8 4 GRID SQUARES

I I I I I 1 1 1 I I 0 0 2 0 4 0.6 0.8 l o 1 2 1.4 1.6 1.8

DISTANCE FROM GRAFT (mm)

FIG. 6. Silver grain clusters per grid square (0.04 mm2) as a function of distance from the medial border of the graft on the 'intact' (denervated shortly before sacrifice) (open circles) and predenervated (filled circles) sides (bars: SEMl .

701 Normal striaturn Denervated side

0 "Intact" side

a 9 30 Y

Y u

2 20 3 v)

10

0 14450 9680 6480 4350 2900 1925 1300 815 575 Background

SILVER GRAIN CLUSTERS/mm'

FIG. 7. Percent areas of section with different densities of DA innervation in the normal neostriatum and, in the graftd neostriatum, on the 'intact' (denervated shortly before sacrifice) and predenervated sides. The numbers of SGC given are those corresponding to the median value of each grey scale interval, according to the calibration curve in Figure 2B (the grey scale was subdivided into intervals of 10 grey values, see Materials and methods).

Graft survival

The 6-OHDA denervated striatum did not seem to provide a more favourable environment for graft survival than intact striatum. Thus, there were no differences in graft volume or number of TH- immunopositive neurons between the denervated and 'intact' sides. The size of the grafted TH-positive cell bodies was, however, significantly

6 1 A

O ] I I I 1 a I

0 05 l a 1 5 2 0 2 5 3 0 GRAFT SIZE (mm2 I

B

I I I 1 I I 0 05 1 0 1 s 2 0 2 5 3 0

GRAFT SIZE (mm' 1

FIG. 8. Relationship between the size of the grafts and the overall densit of DA outgrowth into the neostriatum expressed in numbers of SGC per mm on the 'intact side' (denervated shortly before sacrifice) (A) and on the predenervated side (B). The graft size is the mean sectional area of the grafts in 5-6 sections taken at every 200 pm interval.

Y

larger on the side of the 6-OHDA lesion in keeping with the increased graft-derived DA innervation on the same side (see below). It is plausible that DA neurons with a relatively larger axonal network will attain a relatively greater cell body volume simply in order to support increased metabolic demands.

As a whole, the present results indicate that the fetal mesencephalic DA neuronal grafts into the striatum react to denervation in a manner similar to serotonin (5-HT) containing fetal raphe neurons grafted into the hippocampus (Zhou et al., 1987, 1988); the size of the surviving monoaminergic cell bodies and the density of graft-derived innervation in the target tissue were both increased following a specific denervating lesion, but not the size of the graft nor the number of DA or 5-HT cell bodies. In contrast, intrahippocampal grafts of fetal basal forebrain tissue, rich in developing cholinergic neurons have been shown to respond somewhat differently (Gage and Bjorklund, 1986, 1987). Indeed, cholinergic denervation by transection of the fimbria-fornix increased both the overall size of the basal forebrain grafts and the number of surviving cholinergic nerve cell bodies. In the latter case, however, it is worth noting that the denervating lesion did not affect cholinergic projections alone, but probably also involved other septo- hippocampal projecting systems. Additional trophic mechanisms might have been activated in this case.

Graft-derived innervation of the neostriatum

The establishment of an extensive DA innervation from the graft was specific for the neostriatum. In agreement with studies of DA grafts


intentionally (Bjorklund er al . , 1980, 1983; Dunnett et al., 1984; Triarhou et al., 1988) or inadvertently (Brundin et al., 1988; Sladek er al., 1986) placed in the neocortex, some spots of very dense labeling were often encountered in the frontal neocortex, near the needle tract, but these had an irreproducible pattern which did not suggest any specific arrangement of the DA fibers (see Fig. 4). In the 6-OHDA denervated striatum, the ventral part, in particular nucleus accumbens, was poorly innervated by the grafts, while the periventricular areas of neostriatum were innervated less extensively than the remaining caudate-putamen (see Figs. 4). This observation was surprising since when similar grafts were placed directly into the nucleus accumbens, they showed extensive DA fiber outgrowth into the area (Herman et al., 1986; Brundin et al . , 1987). The present pattern was, however, consistent in all animals, even when the graft reached close to the border of the nucleus accumbens. Other mechanisms than the distance between the graft and the nucleus accumbens thus appear to be involved. One might speculate that, within a graft located in the dorsal striatum, i.e. that part of the striatum normally DA innervated exclusively from the substantia nigra, only a subpopulation of DA neurons (those from the substantia nigra proper) would be stimulated by local factors of neostriatal origin to produce axonal outgrowth into the surrounding host tissue. This hypothesis would be supported by a previous report showing that cholecystokinin- (CCK) and DAKCK-positive neurons in solid ventral mesencephalic grafts did not project extensively into the neostriatum (Schultzberg et al . , 1984). These neurons are known to be normally localized in the ventral tegmental area (VTA) and to project to the nucleus accumbens but not the head of the caudate- putamen (Hokfelt er al., 1980). Looking at the outgrowth of DA/CCK- containing axons from ventral mesencephalic transplants placed into the nucleus accumbens could thus be one way to test this hypothesis. If outgrowth from the VTA neurons depends on the area of implantation, one would then expect to find an extensive graft-derived DA/CCK innervation of the ventral striatum in this situation.

Differences in the graft-derived DA innervation of intact and denervated neostriatum

The overall density of neostriatal DA innervation from the graft was between 10 and 53% of normal on the denervated side and between 5 and 43% of normal on the ‘intact’, or acutely denervated, side. These results are in the same range as the measured DA levels restored by cell suspension DA grafts in the 6-OHDA denervated striatum: 6 - 18 % of normal values, with extreme cases reaching 50% (Dunnett et al., 1988; Herman et al . , 1985; Schmidt et al . , 1983).

A principal finding of the present study was the substantial graft- derived DA fiber outgrowth into the neostriatum on the ‘intact’ side representing overall as much as 20% of the normal DA innervation and 56% of the graft-derived innervation on the denervated side. This fiber outgrowth was not the result of the 6-OHDA denervation which was carried out 7 days before the uptake experiments as it was also evident as a marked hyperinnervation in the TH-immunostained sections, which were from brains without such lesions. A similar, but even stronger (80% of normal), graft-derived monoamine hyperinnervation of a nondenervated target was recently reported with raphe transplants into the hippocampus (Zhou et al., 1987, 1988; Tsubokawa et al . , 1988). Raphe grafts placed in the 5-HT denervated hippocampus resulted in three times higher hippocampal 5-HT levels and a six-fold higher hippocampal [‘H]5-HT high affinity uptake than grafts placed into the nondenervated hippocampus (Zhou et al., 1987).

This resulted in an even heavier hyperinnervation of the hippocampus (255% -345% of the normal hippocampal 5-HT content or [jH]5-HT uptake, respectively). In the present study, the graft-derived DA innervation of the striatum was approximately doubled by a pregraft denervating lesion, but the target area remained hypoinnervated (overall 36% of the normal striatal DA innervation, on average). The lesser effect of the denervating lesion on the DA fiber outgrowth from the graft could be due to the fact that, in our experiments, the intrinsic DA system was removed on the ‘intact’ side a week before sacrifice in order to ‘unmask’ the DA fiber outgrowth from the graft. The authors cannot exclude that during this short period of time there was additional boosting of fiber outgrowth, thus lessening the difference between the denervated and ‘intact’ sides. However, it is obvious that the denervating lesion did not result in a DA hyperinnervation of the striatum as it did for 5-HT in the hippocampus. Even with the largest grafts, the overall neostriatal DA innervation density from the graft amounted to only 53% of the normal striatal DA innervation. This suggests that nigral grafts have a limited capacity to fully replace a complete DA denervation of the striatum. It may be related to the very high density of the normal DA innervation in this area (Doucet et al., 1986) which, from the radioautographic data, can be estimated to be as much as 35 times denser than the 5-HT innervation of the normal hippocampus (Oleskevich and Descarries, 1990; see also Descarries et al., 1990).

The differences between the ‘intact’ and denervated sides were seen in three different parameters: the grafted DA neurons were larger, the DA fiber outgrowth was more abundant, and these innervating DA fibers occupied more extensive areas of the neostriatum on the denervated side. The different correlation curves relating graft size and overall density of innervation on each side suggested that there was an increase in the affinity of the neostriatal tissue on the denervated side for DA axons coming from the graft. Previous studies in v i m (Prochiantz et al . , 1979; Di Porzio er al., 1980; Denis-Donini et a l . , 1983) have demonstrated both morphogenetic and growth-stimulating effects of striatal tissue on developing mesencephalic DA neurons. Such effects were produced specifically by DA-innervated target tissue (striatum) and were not seen with tissue obtained from non DA- innervated brain regions (Hemmendinger et al., 1981). While the striatal stimulatory factors detected by Prochiantz et al . , (1981) appeared to be membrane bound, Dal Toso et al. (1988) have reported trophic effects of striatal tissue extracts on cultured mesencephalic DA neurons. It seems likely, therefore, that the specific growth-stimulating properties of striatal tissue on developing DA neurons can be imputed to both membranous and diffusible factors. The present results are consistent with the idea that the target derived DA growth-regulating mechanisms, which are present during ontogenesis of the mesostriatal DA system, may also operate in the adult striatum. These factors appear to be specific for areas normally innervated by DA fibers and to increase after DA deafferentation.

Conclusions

The results show that grafts of fetal ventral mesencephalic tissue are capable of producing a relatively strong outgrowth of DA fibers also in the presence of an intact DA input to the surrounding striatal target. When the striatal target area is completely denervated of its endogenous DA afferents, the DA fiber outgrowth is increased about two-fold. but not the survival of DA neurons nor the overall survival of the graft. The preservation of a residual intrinsic mesostriatal DA system, such


a s occurs in most patients with Parkinson’s disease, should therefore not prevent efficient stimulation o f DA fiber outgrowth from the graft.

Acknowledgements

The expert technical assistance of Mr Stefan Seth, Ms Sylvia Garcia, Ms Gertrude Stridsberg, and Ms Alicja Flasch is gratefully acknowledged as well as the graphic and photographic work of Ms Agneta Persson and Mr Ragnar Mirtensson. This study was supported by grants from the Swedish Medical Research Council (04X-3874). The Bank of Sweden Tri-centenary Fund, the American Parkinson’s Disease Association and the National Institutes of Health (NS 06701). G.D. was supported by a fellowship from the Fonds de la recherche en santC du QuCbec.

Abbreviations

CCK CSF DA 5-HT 6-OHDA PBS SGC TH VTA

cholecystokinin cerebrospinal fluid dopamine 5-hydroxytryptamine (serotonin) 6-hy droxydopamine phosphate buffered saline silver grain cluster tyrosine hydroxylase ventral tegmental area

References

Abercrombie, M. (1946) Estimation of nuclear population from microtome sections. Anat. Rec. 94: 239-247.

Berod, A., Hartman, 9 . K., and Pujol, J. F. (1981) Importance of fixation in immunohistochemistry : use of formaldehyde solutions at variable pH for the localization of tyrosine hydroxylase. J. Histochem. Cytochem. 29:

Bjorklund, A., Schmidt, R. H., and Stenevi, U. (1980) Functional reinnervation of neostriatum in the adult rat by use of intraparenchymal grafting of dissociated cell suspensions from the substantia nigra. Cell Tissue Res. 212: 39-45.

Bjorklund, A., Stenevi, U., Schmidt, R. H., Dunnett, S. B., and Gage, F. H. (1983) Intracerebral grafting of neuronal cell suspensions. Acta Physiol.

A. (1987) Survival growth and function of dopaminergic neurons grafted to the brain. Prog. Brain Res. 71: 293-308.

Brundin, P., Strecker, R. E., Londos, E., and Bjorklund, A. (1987) Dopamine neurons grafted unilaterally to the nucleus accumbens affect drug-induced circling and locomotion. Exp. Brain Res. 69: 183- 194.

Byndin, P., Strecker, R. E., Widner, H., Clarke, D. J., Nilsson, 0. G., Astedt, B., Lindvall, O., and Bjorklund, A. (1988) Human fetal dopamine neurons grafted in a rat model of Parkinson’s disease: immunological aspects, spontaneous and drug-induced behaviour, and dopamine release. Exp. Brain Res. 70: 192-208.

Dal Toso, R., Giorgi, 0.. Soranzo, C., Kirschner, G., Ferrari, G., Favaron, M., Benvegnu, D., Presti, D., Vicini, S., Toffano, G., Azzone, G. F., and Leon, A. (1988) Development and survival of neurons in dissociated fetal mesencephalic serum-free cell cultures: I. Effects of cell density and of an adult mammalian striatal-derived neuronotrophic factor (SDNF). J. Neurosci. 8: 733-745.

Denis-Donini, S., Glowinski, J., and Prochiantz, A. (1983) Specific influence of striatal target neurons on the in vitro outgrowth of dopaminergic neurites: a morphological quantitative study. J. Neurosci. 3: 2292 -2299.

Descarries, L., Audet, M. A., Doucet, G. , Garcia, S., Oleskevich, S., SCguCla, P., Soghomonian, J.-J., and Watkins, K. C. (1990) Morphology of central serotonin neurons: Brief review of quantified aspects of their distribution and ultrastructural relationships. In: Whitaker-Azmitia, P. M. and Peroutka, S. J. The Neuropharmacology of Serotonin. Ann. N.Y. Acad. Sci (In press).

DiPorzio, U. , Daguet, M. C., Glowinski, J., and Prochiantz, A. (1980) Effect of striatal cells on in vitro maturation of mesencephalic dopaminergic neurons grown in serum-free conditions. Nature 288: 370-373.

Doucet, G., Descames, L., and Garcia, S. (1986) Quantification of the dopamine

844-850.

innervation in adult rat neostriatum. Neuroscience 19: 427 -445. Doucet, G . , Brundin, P., Seth, S., Murata, Y. , Strecker, R. E., Triarhou, L. ,

Ghetti, B., and Bjorklund, A. (1989) Regional dopamine fiber degeneration in the weaver mouse neostriatum and its restoration by fetal mesencephalic grafts: assessment by [3H]dopamine uptake radioautography . Exp. Brain Res. 77: 552-568.

Dunnett, S. B., Bunch, S. T., Gage, F. H., and Bjorklund, A. (1984) Dopamine- rich transplants in rats with 6-OHDA lesions of the ventral tegmental area: effects of spontaneous and drug-induced locomotor activity. Behav. Brain Res. 13: 71-82.

Dunnett, S. B., Hernandez, T. D., Summerfield, A, , Jones, G. H., and Arbuthnott, G. (1988) Graft-derived recovery from 6-OHDA lesions: Specificity of ventral mesencephalic graft tissues. Exp. Brain Res. 71: 41 1-424.

Gage, F. H. and Bjorklund, A. (1986) Enhanced graft survival in the hippocampus following selective denervation. Neuroscience 17: 89-98.

Gage, F. H. and Bjorklund, A. (1987) Denervation-induced enhancement of graft survival and growth: A trophic hypothesis. In: Azmitia, E. C. and Bjorklund, A. Cell and Tissue Transplantation Into the Adult Brain pp. 378-395. Ann. N.Y. Acad. Sci. 495.

Gage, F. H., Bjorklund, A,, Stenevi, U., and Dunnett, S. 9 . (1983) Intracerebral grafting of neuronal cell suspensions VIII. Survival and growth of implants of nigral and septal cell suspensions in intact brains of aged rats. Acta Physiol. Scand. Suppl. 522: 67-75.

Gerfen, C. R. and Sawchenko, P. E. (1984) An anterograde neuroanatomical tracing method that shows the detailed morphology of neurons, their axons and terminals: immunohistochemical localization of an axonally transported plant lectin, Phaseolus vulgaris leucoagglutinin (PHA-L). Brain Res. 290: 219-238.

Hemmendinger, I. M., Garber, 9 . B., Hoffman, P. C . , and Heller, A. (1981) Target neuron-specific process formation by embryonic mesencephalic dopamine neurons in vitro. Proc. Natl. Acad. Sci. USA. 78: 1264-1268.

Herman, J.-P., Chouilli, K., and LeMoal, M. (1985) Activation of striatal dopaminergic grafts by haloperidol. Brain Res. Bull. 15: 543 -546.

Herman, J.-P., Choulli, K . , Geffard, M., Nadaud, D., Taghzouti, K . , and LeMoal, M. (1986) Reinnervation of the nucleus accumbens and frontal cortex of the rat by dopaminergic grafts and effect on hoarding behaviour. Brain Res. 372: 210-216.

Hokfelt, T., Skirboll, L., Rehfeld, J. F., Goldstein, M., Markey, K., and Dann, 0. (1980) A subpopulation of mesencephalic dopamine neurons projecting to limbic areas contains a cholecystokinin-like peptide: Evidence from immunohistochemistry combined with retrograde tracing. Neuroscience 5: 2093-2124.

Oleskevich, S. and Descarries, L. (1989) Quantified distribution of the serotonin innervation in adult rat hippocampus. Neuroscience (in press).

Prochiantz, A,, Di Porzio, U., Kato, A,, Berger, B., and Glowinski, J. (1979) In vitro maturation of mesencephalic dopaminergic neurons from mouse embryos is enhanced in presence of their striatal target cells. Proc. Natl. Acad. Sci. USA. 76: 5387-5391.

Prochiantz, A, , Daguet, M. C., Herbet, A, , and Glowinski, J. (1981) Specific stimulation of in vitro maturation of mesencephalic dopaminergic neurons by striatal membranes. Nature 293: 570-572.

Schmidt, R. H., Bjorklund, A,, and Stenevi, U. (1981) Intracerebral grafting of dissociated CNS tissue suspensions: a new approach for neuronal transplantation to deep brain sites. Brain Res. 218: 347-356.

Schmidt, R. H., Bjorklund, A,, Stenevi, U., Dunnett, S . B., and Gage, F. H. ( 1983) Intracerebral grafting of neuronal cell suspension. 111. Activity of intrastriatal nigral suspension implants as assessed by measurements of dopamine synthesis and metabolism. Acta Physiol. Scand. (Suppl.) 522: 19-28.

Schultzberg, M., Dunnett, S. B., Bjorklund, A, , Stenevi, U. , Hokfelt, T., Dockray, G. J., and Goldstein, M. (1984) Dopamine and cholecystokinin immunoreactive neurones in mesencephalic grafts re-innervating the neostriatum: Evidence for selective growth regulation. Neuroscience 12:

Sladek, J. R. Jr, Collier, T. J., Haber, S. N., Roth, R. H., and Redmond, D. E. Jr (1986) Survival and growth of fetal catecholamine neurons grafted into the primate brain. Brain Res. Bull. 17: 809-818.

Trinrhou, L. C., Low, W. C., and Ghetti, 9 . (1988) Layer-specific innervation of the dopamine-deficient frontal cortex in weaver mutant mice by grafted mesencephalic dopaminergic neurones. Cell Tissue Res. 254: 11 - 15.

Tsubokawa, T., Katayama, Y., Miyazaki, S . , Ogawa, H., Iwasaki, M.,

17-32.


Shibanoki, S., and Ishikawa, K. (1988) Supranormal levels of serotonin and its metabolite after raphe cell transplantation in serotonin-denervated rat hippocampus. Brain Res. Bull. 20: 303-306.

Ungerstedt, U. and Arbuthnott, G. W. (1970) Quantitative recording of rotational behavior in rats after 6-hydroxydopamine lesions of the nigrostriatal dopamine system. Brain Res. 24: 484-493.

Zhou, F. C., Auerbach, S., and Azmitia, E. (1987) Denervation of serotonergic

fibers in the hippocampus induced a trophic factor which enhances the maturation of transplanted serotonergic neurons but not norepinephrine neurons. J. Neurosci. Res. 17: 235-246.

Zhou, F. C., Auerbach, S., and Azmitia, E. (1988) Transplanted raphe and hippocampal fetal neurons do not displace afferent inputs to the dorsal hippocampus from serotonergic neurons in the median raphe nucleus of the rat. Brain Res. 450: 51 -59.

Effect of Prior Dopamine Denervation on Survival and Fiber Outgrowth from Intrastriatal Fetal...

Documents

Transcript of Effect of Prior Dopamine Denervation on Survival and Fiber Outgrowth from Intrastriatal Fetal...