GAP-43 and 5B4-CAM immunoreactivity during the development of transplanted fetal mesencephalic neurons

EXPERIMENTAL

NEUROLOGY

114,

l-10

(19%)

GAP-43 and 564-CAM lmmunoreactivity during the Development of Transplanted Fetal Mesencephalic Neurons GERALD H. CLAYTON,THOMAS Department

of Cellular

J. MAHALIK,ANDTHOMAS

and Structural Biology, Box B-l 11, University 4200 East 9th Avenue. Denuer. Colorado

Developing neurons contain high levels of several proteins which are absent or relatively scarce in mature neurons. GAP-43 is a cytoplasmic protein primarily found within neurons; high levels of this protein are correlated with axonal elongation or regeneration. 5B4-CAM, a glycosylated transmembrane protein, is a member of the NCAM family present in growth cones and in plastic CNS structures. Antibodies directed against these two developmentally regulated proteins were used to characterize the time-course of maturation of transplanted fetal mesencephalic neurons. For our experiments unilateral injections of fi-hydroxydopamine were made into the nigrostriatal bundle in Sprague-Dawley rats. The effectiveness of the lesion was verified by apomorphine-induced rotation and by postmortem examination of the substantia nigra. Following behavioral testing, pieces of ventral mesencephalon obtained from El5 fetuses were transplanted into the eaudoputamen ipsilateral to the lesion. Immunocytochemistry revealed high levels of GAP-43 and 5B4CAM at 5, 11, and 15 days post-transplant but relatively lower levels by 3 weeks. At 13 weeks the immunoreactivity present within the transplant tissue was approximately equal to that found within the host striatal neuropil. This time-course of higher GAP-43 and 5B4-CAM immunoreactivities coincides with the timecourse of neuritic outgrowth of dopamine containing cell ~pulations within the ventral mesencephalon in situ as well as within ventral mesencephalic transplants. This implies that axon elongation occurs over a period similar to that which occurs during normal development. These data suggest that the effects of transplantation surgery and the altered environment of the host striatum do not signiileantly affect the time-course of development of ventral mesencephalie neurons. Q 1991

Academic

Press,

Inc.

INTRODUCTION Transplantation of fetal brain tissue is useful in the investigation of neural development and regeneration (for review see (3,548)). For example, grafts of ventral

of Colorado 80262

EFINGER Health

Sciences

Center,

mesencephali~ (VM) dopaminer~c (DA) neurons have been used to reconstruct the damaged nigrostriatal pathway in animal models of Parkinson’s disease (for review see (27,61)). Transplan~d fetal dopamine neurons send out processes over significant distances into the host environment (10,22), receive inputs from host neurons (6, 17, 33), and alter lesion-induced behaviors (2,3,4,18). Additionally, fetal tissue grafts can be used to study basic questions such as how the environment influences the development of specific CNS pathways, e.g., does an adult host environment affect the development of specific neuronal populations? In this study we used antibodies directed against the developmentally regulated proteins GAP-43 and 5B4-CAM (20, 35, 46), putative markers for axon elongation (23, 45, 461, to study the development of VM grafts. During periods of neurite outgrowth, in development or regeneration, neurons express and transport at high levels a number of specific growth-related proteins (GRP’s) including GAP-43 and 5B4-CAM (23, 26, 30, 57,59). GAP-43 is a cytoplasmic phosphoprotein whose level is enhanced within neurons during growth and/or regeneration (for review see (45)). Evidence for a correlation between high levels of GAP-43 and neuronal outgrowth include: (i) increased synthesis during axonal outgrowth (28), (ii) rapid transport to and concentration in actively extending growth cones (35,36,40,44), (iii) increased mRNA levels after injury (54), and (iv) lack of GAP-43 in neurons that do not regenerate following injury (30). In addition, some parts of the CNS known, or suspected of being highly plastic in nature (e.g., hippocampus and olfactory bulb) maintain high levels of GAP-43 during adulthood (34,45,55). Furthermore, in vitro evidence indicates that GAP-43 expression is down-regulated by neurite-cell or cell-cell contact (53). These data suggest that GAP-43 can be used as a marker for axonal outgrowth. The protein 5B4-CAM is a transmembrane glycoprotein that is a member of the neural cell adhesion molecule (NCAM) family (75% homology with rat brain NCAM-ld) (41, 42) known for involvement in cell-cell interactions. 5B4-CAM is expressed on sprouting neu-

Copyright 0 1991 Al1 rights of reproduction

0014~4666/01 $3.00 by Academic Press, Inc. in any form resewed.

2

CLAYTON,

MAHALIK,

rons (59), is found concentrated within growth cone particle fractions (20), and is present in higher amounts within plastic areas of the adult rat CNS such as the hippocampus and olfactory bulb (33, 59). Thus like GAP-43,5B4-CAM may serve as an indicator of neuronal immaturity or plasticity. Ventral mesencephalic grafts contain the developing substantia nigra (SN) cell population which provides a majority of the dopaminergic innervation to the adult striatum. During normal development, axonal outgrowth from these neurons is complete by the second to third week of postnatal life in the rat (47,57). Similarly, tyrosine hydroxylase-like immunoreactive (THLI) neurites from grafted neurons show prolific innervation of the host during the first 2 weeks postgrafting with less vigorous outgrowth continuing up to 6 weeks (1). Thus, in terms of outgrowth of TH neurites, the development of the normal SN and the early development of axonal outgrowth from VM transplants appear similar. Catecholaminergic neurons are, however, not the only cell population contained in the VM grafts. For example, the SN contains GABAergic neurons which do not normally project to the striatum. Target tissues have a significant effect on the growth and development of ingrowing neurons (3, 5, 9). Since the VM grafts are placed in the midst of the target area for dopamine containing nigrostriatal neurons (and others) the close proximity of their target might be expected to reduce the time required for target acquisition and cessation of GRP synthesis (1, 10, 31,49). In addition to the presumptive dopaminergic SN neurons, we and other investigators have found evidence for serotonergic, GABAergic, enkephalinergic, and substance P containing neurons within VM grafts (6,17,32,50). Of these, the dopaminergic and serotonergic (from the mesencephalic raphe) neurons normally send projections to the striatum and readily do so when grafted. By contrast, enkephalinergic and substance P containing neurons, which do not normally innervate the striatum, fail to do so when grafted into host striatal tissues (32). Thus in grafts some neurons may find normal targets readily available within the striatum while others may not. The monoaminergic neurons which grow directly into their target may cease production of GRP’s relatively early. In contrast, those neurons without appropriate targets may continue their process of outgrowth for an extended period and would presumably require continued synthesis of GRP’s. Alternatively, when faced with inappropriate targets, neurons may end neurite outgrowth prematurely, resulting in a premature reduction in GRP expression. There is evidence in favor of this idea in the work of Van der Neut (53) and Bregman (7, 8). Van der Neut’s in vitro work indicates that the expression of GAP-43 in the rat spinal cord is regulated by interneuronal contact. The synthesis of GAP43 remains elevated until contact is made by outgrowing

AND

FINGER

processes on target tissue explants. In addition Bregman (7,8) shows that transplants of fetal “normal targets” into the lesioned neonatal spinal cord were able to extend the time period of developmental plasticity for the corticospinal tract. It therefore seems likely that each different subpopulation of neurons within VM grafts might institute a course of outgrowth whose time span depended on the availability of normal target cells. The time-course of GRP immunoreactivity may therefore be different for each neurochemically identified cell type within the transplants. Another element affecting the outgrowth of neurites from grafts is the availability of appropriate growth factors. The environment of the developing neural tube is replete with general and regionally specific growth factors (e.g., nerve growth factor (NGF), platelet-derived growth factor (PDGF)) (21) as well as a host of cell surface molecules required for axonal pathfinding (e.g., neural cell adhesion molecule (NCAM), neuron-glia cell adhesion molecule (NgCAM)) (15,29,39,49). These developmental agents are known to be expressed differentially so that in the adult their levels and distribution may be completely different from those seen during embryonic development (13, 16, 38). The absence of specific growth factors or cell surface molecules may alter the period of active outgrowth in transplanted cells. Antibodies directed against GAP-43 and 5B4-CAM were used in the present study to assess the differentiation of grafted VM neurons. Our data indicate that the average time-course of GRP expression within heterogeneous neuronal graft populations approximately corresponds to the time-course of vigorous axonal outgrowth from the normally developing substantia nigra (1, 47, 57). METHODS

6-Hydroxydopamine lesions. Thirteen female Sprague-Dawley rats (wt = 175-200 g) were used in these studies [N(<3 wk) = 7; N(>3 wk) = 61. Animals were anesthetized by intraperitoneal (ip) injection of a 4% solution of chloral hydrate (dose = 1.2 cc/100 g). A 4-~1 solution of 6-hydroxydopamine (2 pglpl 6-OHDA made in normal saline with 0.2 mg/ml ascorbate) was injected at a rate of 1 pl/min into the median forebrain bundle on the right side of the brain. The coordinates for injection were (as measured from bregma and the dural surface with the incisor bar at -2.4): AP = -4.4 mm, Lat = -1.2 mm, and DV = -7.8 mm. Animals were allowed to recover with food and water which contained 1 mg/ml acetaminophen as an analgesic. Behavioral testing. Rats were tested for apomorphine-induced turning at 2-week intervals; rats which turned contralaterally (relative to the lesion) more than 400 turns in 60 min after a 0.05 mg/ml dose of apomorphine were used for further study.

GAP-43

AND

5H4-CAM

IMMUNOREACTIVITY

Pretransplantation cavities (60) were made on the lesioned side by inserting a ‘22-gauge needle at AP = -i-1.0, LAT = -3.0, and DV = 5.5 with respect to bregma, 6 days before the grafting of fetal tissues. El5 fetuses, with a crown-rump length of 14-15 mm, were obtained from Sprague-Dawley mothers (Harlan) by cesarian section after an overdose of ether by inhalation. The fetuses were then rinsed in cool tap water and placed on ice in Hanks’ balanced salt solution (HBSS from Sigma Chemical Co., St. Louis, MO). A wedge of the anterior/basal portion of the ventral mesencephalic flexure was excised in a sterile culture dish containing HBSS. The dissected piece of neural tube was then cut into two pieces which were loaded into an l&gauge stainless steel cannula fitted with tight fitting plunger so that the pieces could be injected into the host animal (modified from Stromberg et al. (51)). For transplantation, host animals were reanesthetized (as for the previously described 6-OHDA lesion protocol) and placed into a stereotaxic apparatus where the cannula was lowered to a depth of 5.5 mm from the dura at the coordinates of the prior pretransplantation cavity and kept there for 5 min. The cannula was then raised to a depth of 4.5 mm where the fetal tissue was injected. After withdrawal of the cannula the animal was allowed to recover per the previously described protocol for lesioning. Tissue preparation and immunocytochemistry. Each animal was sacrificed at the appropriate time following transplantation by overdose with sodium pentobarbitol (1 cc ip, 64.8 mglml). The animal was then injected intracardially (left ventricle) with 1 cc of 1% sodium nitrite in normal saline, 0.2 cc of sodium heparin (Invenex, 1000 usp units/ml) and then perfused with 400 ml of a solution of 4% paraformaldehyde (PFA) in 0.1 A4 phosphate buffer (PB) at pH 7.4. The brain was then removed and placed into 4% paraformaldehyde in 0.1 M PB with 20% sucrose. After postfixation overnight, the brain was blocked into smaller sections and placed into 0.1 M PB containing 20% sucrose for l-2 h. Free floating and cryostat sections were taken for immunocytochemistry. For cryostat sectioning, tissue blocks are embedded in Tissue-Tek O.C.T. compound and frozen on cutting pedestals at approximately -20°C. Coronal sections, at a thickness of 15-20 pm, were collected on gelatin-coated slides. Slides were stored at -12°C until used for histology or immunocytochemistry. Alternatively, free floating sections (40-50 pm) were obtained with a freezing-sliding microtome and stored in PB at 4°C until used for immunocytochemistry. Immunocytochemistry was done using monoclonal antibodies to the antigens GAP-43,5B4-CAM, and tyrosine hydroxylase (TH). Cryostat-prepared frozen secTru~p~ntut~n.

IN

TRANSPLANTS

3

tions on slides were brought to room temperature and sections were fixed to the slide by immersion in 4% PFAIO.1 M phosphate buffer. The protocol from this point on was the same for both free-floating and cryostat sections except that all free-floating sections were reacted by immersion in small plastic vials and subsequently mounted after the final reaction. After thorough rinsing in standard phosphate-buffered saline (PBS, pH 7.2) endogenous peroxidase activity was quenched by immersion in a solution of 0.3% H,O, made up in PBS. The slides were again rinsed in PBS and permeabilized in PBS containing 0.3% Triton X-100/ 0.02% sodium azide (PTA) for 15 min. Subsequently each group of slides was immersed for 15 min in a solution of PTA containing 20% normal serum (from the species in which the respective secondary antibody was made) to decrease nonspecific binding. Antibodies against GAP-43 (25) (supplied by P. Skene and D. Schreyer, Stanford Univ.) and 5B4-CAM (20) (supplied by K. Pfenninger, Univ. of Colorado Health Sci. Center) were used at a dilution of l/2000 in PTA/2% normal serum (rat or rabbit). Antibody to TH (58) (supplied by R. Vulliet, Univ. of California/Davis) was used at a dilution of l/1000 in the same diluent as above. A small volume of the antibody was pipetted onto the sections which were then incubated for 48 h at 4°C. For controls some sections were incubated in PTA/B% normal serum while others were incubated with nonimmune sera (pre-immune control) at the same dilution as the primary antibody being controlled for. Neither deletion nor pre-immune controls showed immunoreactivity above very low background levels. Incubated sections were rinsed in PBS 3 X 10 min. The secondary antibodies (FITC rabbit anti-mouse or biotinylated rat anti-mouse) were then applied at a dilution of l/50 for 1 h at room temperature. FZuorescent secondaries were rinsed thoroughly (3 X 10 min) in PBS and coverslipped. After rinsing, the sections treated with biotinylated secondaries were incubated in ABC solution (Elite Vectastain ABC kit, Vector Labs) for 1 h at room temperature. They were then rinsed 3 X 10 min with PBS and then immersed in a solution of 0.02% DAB/O.Ol% H,O, in PBS for 10 min. These sections were again rinsed well in PBS and coverslipped. Image and data collection. Digitized images of DABreacted sections through the transplant were obtained using an image analysis system comprising: an Apple Macintosh 11x equipped with a Data Translation video capture card, a Nikon light microscope with a green filter inserted into the illumination path, a COHU video camera (model 4815ZOOO), and image analysis software developed at NIH (public domain imaging software “Image” is published by the National Technical Information Service). For image acquisition, the “automatic gain” and the “black adjust” controls within the camera

4

CLAYTON,

MAHALIK.

were turned off to prevent the camera from adjusting for variations in overall light intensity within and between sections. Light transmission measurements were obtained from within the transplant tissues and from the contralateral nonlesioned striata as a control comparison. A calibration curve for the camera was created by taking pixel intensity measurements from neutral density filters placed into the light path. This curve was then automatically applied to all measurements taken from imaged sections. ~tat~t~cal a~~ys~. Comparisons of immunoreactive density, as determined by optical density measurements described above, were calculated as a ratio of transplant:control. For statistical analysis these percentages were then pooled into two groups: early (5-15 days) and late (3-13 weeks) time periods. These data were then subjected to one-way ANOVA to determine the significance of any differences between the two groups for GAP-43 and 5B4-CAM, respectively. RESULTS

Early Grafts (5-15 Days) General. The general histology observed after thionin or cresyl violet staining revealed a dense population of darkly stained cells within the graft at 5 days. Many of the cells within the graft could be seen as small, round profiles resulting in an undifferentiated appearance. Within these early grafts cell sizes varied widely (3-11 pm) with the smaller sizes predominating. Some necrotic cells with little cellular staining and distorted shapes also could be found (11). The host brain neurons were larger, less symmetrical, less intensely stained, and more diffise than those within the graft. Cell sizes within the host striatum (lo-15 pm) also were less variable than those seen within the graft. The grafted tissue was often fragile upon sectioning and cavities within the transplant area were common. Graft pieces clung to the walls of the cavities and were interspersed with tissue debris. Vas~ularization was evident in most sections (not shown) and was characterized by medium to large diameter vessels (15-50 pm) with extensive fine branching within the neuropil. By 9-11 days most grafted sections contained well-incorporated grafts with little evidence of cavities. Compared to the 5-day transplants, cell density within the older grafts had decreased and the majority of the cells were larger, asymmetrical, and less darkly stained thus appearing similar to cellular profiles seen within the host. The appearance of 15 day grafts more closely resembled the host striatal neuropil with cell densities and cytochemical staining approaching that seen within the adjacent striatum. All grafts were well incorporated into the host with no evidence of eavities. Tyrosine hydroxykzse-like immunoreactivity. At 5 days post-transplant, several cell bodies showed tyro-

AND

FINGER

sine hydroxylase-like immunoreactivity. By Q-11 days well-developed populations of tyrosine hy~oxylase-lye immunoreactive (THLI) cells were visible primarily as small clusters. A dense plexus of immunoreactive fibers occurred within the graft interconnecting individual clusters of cells (Fig. 1A). THLI processes were also visible near the border of the graft with several neurites extending out for substantial distances (~750 pm) into the host neuropil. Fine branches could be seen within the host in close proximity to the graft. Fifteen-day grafts contained well-differentiated clusters of cells which extended numerous neurites out into the host for distances greater than 1 mm. Growth-related proteins. At early time periods (5-15 days), high intensity GAP-43 immunoreactivity within grafted tissues stood out distinctly against a background of lower level immunoreactivity in the ipsilatera1 host striatum. The immunoreactivity formed a coarse, bundled fiber-like meshwork throughout the graft. Individual fibers could not, however, be distinguished. In contrast, the host tissue contained a very diffuse, fine, punctate immunoreactivity interrupted by the large nonimmunoreactive fiber bundles of the internal capsule and small clear spaces within the neuropil apparently corresponding to neuronal somata. GAP-43 immunoreactivity at Q-11 days post-transplant maintained a coarse appearance along with high intensity immunoreactivity (Fig. 2A). The edges of the graft were easily discernable, but at higher magnification appeared rough and uneven. By 15 days post-transplant, GAP-43 immunoreactivity appeared to have declined slightly although its intensity remained well above that seen within the host striatum. Comparison of adjacent sections showed that areas of dense GRP immunoreactivity included all regions containing clusters of THLI cells and fibers as well as regions of less THLI immunoreactivity. Quantification of data from digitized images of sections within this time group shows that on average the immunoreactivity within transplant tissues was 27% (23% SEM) greater than that seen within the contralateral control striatum. The distribution and appearance of 5B4-CAM immunoreactivity at 5 days closely resembled that of GAP-43. By 9-11 days 5B4-CAM immunoreactivity was still very intense and not noticeably different from that seen in the 5-day grafts (Fig. 3A). The immunoreactivity within the 15-day-old graft also remained very intense in comparison with the adjacent host striatal neuropil. Comparison of adjacent sections show a GRPPfH overlap similar to that seen for GAP-43. The imaging data indicates that 5B4-CAM immunoreactivity within the transplants was on average 33% (44% SEM) greater than that seen within the contralateral control striaturn. Basal levels of GAP-43 and 5B4-CAM immunoreactivity in the contralateral and ipsilateral host striata were equivalent.

GAP-43

AND

5B4CAM

IMMUNOREACTIVITY

Late Grafts

-

.h+

100 vrn

IN

TRANSPLANTS

(3-13

5

Weeks)

General. By 3 weeks cell densities within transplants had declined to adult host tissue levels, graft cell staining was similar to that seen within the mature neurons found within the adult host striatum, and very few round, symmetrical cells were in evidence. The general appearance of the grafts in the late stage showed numerous well-developed cells typical of those found within the adult substantia nigra and related regions. The grafts are characterized by a homogeneous distribution of lightly stained, asymmetrical cells of medium to low densities. As in the previous stage the grafts were well incorporated into the host and had little evidence of previous surgical trauma. Tyrosine hydroxylase-like immunoreactivity. THLI cells were typically found in small to large well-defined clusters (Fig. 1B). Clusters often appeared to be larger and more numerous than in earlier stages. A dense finely branched innervation was often observed near the graft extending in all directions. By 13 weeks most of the immunoreactive fine process remained restricted to the graft although many THLI processes extended into the host neuropil (Fig. 1C). By 3 weeks GAP-43 immuGrowth-related proteins. noreactivity was very adult-like in appearance (Fig. 2B). Grafts at this age exhibited a fine punctate immunoreactivity distributed homogeneously throughout the transplant. Clear spaces, as seen in earlier grafts, were more evident within the graft and host; absence of internal capsule fiber bundles allowed for the identification of the graft. The intensity of immunoreactivity within the graft was slightly above the basal levels seen in the host striatum. By 13 weeks immunoreactivity for GAP-43 appeared to have subsided to near adult levels. Borders of the grafts were difficult to discern due to the lack of differential immunoreactivity; in some cases however, blood cells were evident at the graft/host interface (Fig. 2C). The graft material was virtually indistinguishable from the host in regard to the physical makeup and distribution of GAP-43 immunoreactivity. Image analysis data showed that GAP-43 immunoreactivity within transplants was on average 12% (+-4% SEM) greater than within contralateral control host striata (see Fig. 4). Analysis of these data showed an overall effect due to graft age such that transplants within the late group exhibited, on average, significantly less immunoreactivity than in the early group (ANOVA, FI,I, = 9.51, P -c 0.01) (see Fig. 4).

FIG. 1. Coronal sections through the host striatum showing tyrosine hydroxylase-like immunoreactivity (THLI) in transplanted ventral mesencephalic (VM) neurons. (A) THLI in a g-day-old transplant showing the beginnings of cell clustering, (B) THLI within a 21-day-old graft, (C) 13-week THLI showing large cell clusters with numerous fibers emanating from the graft. Scale bar = 100 mm.

CLAYTON,

MAHALIK,

AND

FINGER

The distribution of 5B4-CAM immunoreactivity was identical to that of GAP-43 during this time period (Figs. 3B and 3C). Quantification of 5B4-CAM immunoreactivity showed that the average immunoreactivity within transplants was 15% (+4% SEM) greater than that found within the contralateral control host striatum (see Fig. 4). Analysis of 5B4-CAM imaging data showed that there was a overall age effect between groups similar to that seen in the GAP-43 data (ANOVA, F,,,, = 10.18, P < 0.01) (see Fig. 4) showing a significant decline in relative immunoreactivity from early to late time periods. DISCUSSION

The neuronal growth-associated proteins GAP-43 and 5B4-CAM are expressed at high levels when neurons are actively extending neurites (45, 59). Thus, GAP-43 and 5B4-CAM are expressed to a much higher degree during embryonic and early postnatal times than during adulthood (20,34,59). The results of our experiments show that these developmentally regulated proteins are present at higher levels in the early life (5-15 days) of fetal VM transplants into the adult striatum but decline by 13 weeks to reach levels near to those found in the host striatum. These data correlate well with descriptions of the development of catecholaminergic cells within similar VM grafts (1) as well as with data regarding the normal development of dopaminergic neurons within the nigrostriatal pathway in situ ;(47,57). The development of the THLI innervation of the lesioned striatum by VM transplants has been suggested to occur over a period of 1 to 6 weeks post-transplant (1) with the majority of growth occurring during the first 2 weeks. Our observations describe a homogeneous distribution of GRP immunoreactivity within grafted tissues. At early stages of graft development (~15 days) GRP immunoreactivity is at its highest intensity, indicative of robust neuritic outgrowth. After this time period (>3 weeks) GRP immunoreactivity is significantly less and approaches levels of immunoreactivity found within the host striatal neuropil. This suggests that neuritic outgrowth from all neuronal populations has subsided within a time frame similar to catecholaminergic outgrowth seen by other investigators (1). Further, our observations as well as those of Abrous et al. (1) describe the formation of THLI clusters early in graft developFIG. 2. GAP-43-like immunoreactivity (GAP-LB within coronal sections through adult brains containing ventral mesencephalic (VM) grafts. (A) GAP-L1 within a 5-day-old graft, immunoreactivity within the graft appears dark against the lighter basal levels found within cavity remnants within the graft; (B) the host; *, pretransplantation 21-day-old GAP-LI: at this point a decrease in immunoreactive intensity within the transplant relative to the host can be seen; (Cl 13-week graft showing near equivalence of graft vs host immunoreactivity, some blood cells can be seen lining the right side of the transplant. Scale bar = 200 mm.

GAP-43

AND

5B4-CAM

~MMUNOREACTIVITY

IN

TRANSPLANTS

7

ment rather than a homogeneous distribution. These clusters appeared well-organized by 15 days, were interconnected with dense, fine fiber networks and were reminiscent of THLI cells within the mature SN in situ. During normal development the dopaminergic innervation of the striatum continues up to approximately 4 weeks (57). The most rapid growth, however, occurs prenatally and continues up through the first postnatal week (equivalent to 2 weeks of transplant growth, El5 P’7). Two weeks of rapid outgrowth with a gradual decline in development is consistent with our observations. The intense GRP immunoreactivity seen in the first 15 days post-transplant and the significant decline seen after 3 weeks suggests that neuritic outgrowth occurring within VM grafts progresses on a time scale similar to that of the dopaminergic innervation on the striatum in situ. GAP-43 and 5B4-CAM immunoreactivity does not begin to decline until approximately 15 days after which levels approach those seen within the host striatal neuropil. It is possible that neurons within the graft tissue “restart” their program of neurite outgrowth because the trauma associated with grafting results in the increased expression of GRPs. Support for this hypothesis appears in work by others which have shown that, in several mammalian and nonmammalian systems, axotomy or neuronal damage induces a cell body reaction within the neuron which causes an increase in the synthesis of proteins necessary for survival and regrowth (40,52). GAP-43 is among the more prominent of these and its presence is maintained at high levels throughout regeneration and declines upon the beginnings of target contact (53). Since some neurons within the VM, especially those comprising the dopaminergic nigrostriatal pathway, have already extended processes into the telencephalon (47) of the El5 neural tube, resection of this area undoubtedly shears off and/or damages numerous axonal processes. The cell body response then may add to or extend the developmental time period of expression of GRP’s within the grafted tissue. Analysis of the time period of expression of these proteins within the grafts lends support to this hypothesis. In addition, it is also possible that the environment into which a neuronal graft is placed would also alter the time course of graft expression of these proteins. Our results demonstrate an overlap of GRP immunoreactivity with THLI in adjacent sections. This suggests FIG. 3. Transplant containing coronal sections through the adult striatum exhibiting 5B4-cam like immunoreactivity (5B4-LI). (A) 5B4-LI at 5 days post-transplant; immunoreactivity within the graft appears dark against the lighter basal levels found within the host; *, pretransplantation cavity remnant, (B) 5B4-LI within a 21day-old graft; note the decline in the relative difference in chromagen intensity between the graft and the host, (C) 13 week transplant showing near equivalence of graft vs host immunoreactivity. Scale bar = 200 mm.

CLAYTON, 150

MAHALIK,

1

I

<3 WK

GAP-43

5B4-CAM

FIG. 4. A comparison of the degree of immunoreactivity (GAP43 and 5B4-cam) of the transplants to that of the host contralateral striatum. A value of 100 would indicate that the transplant is equal to the control in immunoreactivity; higher ratios represent increasing differences in relative immunoreactivity [i.e., transplant/control]. N(<3 wk) = 7 animals; N(>3 wk) = 6 animals.

that during the early phases of outgrowth THLI neurons do express high levels of GAP-43 and 5B4-CAM. There are, however, sections of the graft tissue which do contain high levels of GRP-like immunoreactivity but do not contain THLI cells or fibers. These results imply that other populations of grafted neurons also produce high levels of GRP immunoreactivity during early stages of transplant development. These data therefore fit with the hypothesis that many, if not all, grafted neurons in early stages of growth express GRPs. This does not mean, however, that the developmental time-course of GRP expression is the same for all neuronal phenotypes. The availability of appropriate target populations may significantly affect the outgrowth of grafted neurons and therefore their production of GRPs. Co-localization of specific neurochemically identified neurons with GRP-like immunoreactivity is difficult. At low fiber densities, such as that seen in cell culture studies (25,53), growth-relatedproteins (GRPs) can be localized to individual fibers. In our tissue sections, however, individual GRP-like immunoreactive fibers are not resolvable within the dense fiber networks of the transplant and host neuropil. Given the difficulties of co-localization, inferences from data collected from adjacent sections were considered appropriate for this experimental paradigm. Observations of GRP immunoreactivity made in these experiments may not be entirely of neuronal origin. Recent evidence shows that GAP-43 has been found within various types of developing glia (14, 56), and NCAM epitopes have been found on muscle and glia (43) indicating that GAP-43 and 5B4-CAM may not be neuron specific. Evidence of a 2- to 3-week host derived

AND

FINGER

glial reaction within transplants similar to ours (1) suggests that some portion of GRP immunoreactivity seen within the early time period may be of non-neuronal origin. Future experiments combining in situ hybridization for GRPs with immunocytochemistry for TH or neuropeptides would be useful in establishing the time course of GRP expression of the different neuronal populations within VM grafts. In summary, our data provide evidence to indicate that neural GRPs within grafted neuronal tissues can be used to gauge the average developmental time course of populations of neurons. The time course of production of GAP-43 and 5B4-CAM coincides with descriptions by others of the development of dopamine containing cell populations within similar VM transplants. Comparison of our data with the development of dopamine containing nigrostriatal projections in situ suggests that surgery and transplantation into an abnormal environment doesnot significantly alter the time course of development of this cell population. A heterogeneous population of neurons within the grafts makes it difficult, however, to assess the GRP-like immunoreactivity by specific neurochemically identified groups of neurons. The question then of altered GRP expression within specific neuronal populations remains to be answered. As previously stated, the availability of appropriate target populations may have significant effects. The use of more definitive methods, such as in situ hybridization, may allow for a more accurate determination of the developmental growth periods found within this heterogeneous group of cells. ACKNOWLEDGMENTS These data have been presented in part at the 1990 annual meeting of the Society for Neuroscience (abstract No. 530.9). This work has been supported by a grant from the National Institutes of Health (NIH No. NS09199). The authors thank P. Skene, D. Schreyer, K. Pfenninger, and R. Vulliet for their gifts of antibodies utilized in this study.

REFERENCES ABROUS, N., J. Gw, A. VIGNY, A. CALAS, M. LEMOAL, AND J. P. HERMAN. 1988. Development of intracerebral dopaminergic grafts: A combined immunohistochemical and autoradiographic study of its time course and environmental influences. J. Camp. Neural. 273: 26-41. BJ~RKLUND, A., S. B. DUNNET, AND S. D. IVERSEN. 1980. Reinnervation of the denervated striatum by substantia nigra transplants: Functional consequences as revealed by pharmacological and sensorimotor testing. Brain Res. 199: 307-333. BJ~RKLUND, A., AND U. STENEVI. 1979. Regeneration of monoaminergic and cholinergic neurons in the mammalian central nervous system. Physiol. Reu. 69: 62-100. BJ~RKLUND, A., AND U. STENEVI. 1981. Functional reactivation of the deafferented neostriatum by nigral transplants. Nature 289: 497-499.

GAP-43

AND

5B4CAM

IMMUNOREACTIVITY

5. BJGRKLUND,

A., AND U. STENEVI. 1984. Intracerebral neural plants: Neuronal replacement and reconstruction of damaged cuitries. Annu. Rev. Neurosci. 7: 279-308.

6. BOLAM, J. P., T. F. FREUND, A. D. SMITH. 1987. Synaptic gic neurons in grafts striatum. Exp. Brain.

A. BJGRKLUND, S. B. input and local output that functionally reinnervate Res. 68: 131-146.

imcir-

7. BREGMAN,

B. S., AND E. KUNKEL-BAGDEN. 1988. Effect of target and non-target transplants on neuronal survival and axonal elongation after injury to the developing spinal cord. In Trunsplantation into the Mammalian CNS, Progress in Brain Research (D. M. Gash and J. R. Sladek, Eds.), Vol. 78, pp. 205-211. Elsevier, Amsterdam.

8.

BREGMAN, B. S., AND P. J. REIER. plants rescue rubrospinal neurons Comp. Neurol. 244: 86-95.

9.

BRUNDIN, P., 0. ISACSON, F. H. GAGE, AND A. BJ~RKLUND. 1986. Intrastriatal grafting of dopamine-containing neuronal cell suspensions: Effects of mixing with target or non-target cells. Deu. Brain Res. 24: 77-84.

10.

11.

12.

1986. Neural after neonatal

tissue transaxotomy. J.

BRUNDIN, P., R. E. STRECKER, 0. LINDVALL, 0. ISACSON, 0. G. NILSSON, G. BARBIN, A. PROCHIANTZ, C. FORNI, A. NIEOULLON, H. WIDNER, F. H. GAGE, AND A. BJ~RKLUND. 1987. Intracerebral grafting of dopamine neurons. In Cell and Tissue Transplantation into the Adult Brain (Annals of the New York Academy of Sciences, Vol. 495) (E. C. Azmitia and A. Bjorklund, Eds.), pp. 473-496. The New York Academy of Sciences, New York. CARPENTER, M. B., AND J. SUTIN. 1983. The neuron. Neuroanatomy (M. B. Carpenter and J. Sutin, Eds.), 85-133. Williams & Wilkins, Baltimore/London.

In Human 8th ed., pp.

CLAYTON, G. H., T. J. MAHALIK, AND T. E. FINGER. 1991. GAP43 immunoreactivity during the development of transplanted fetal mesencephalic neurons. Sot. Neurosci. l(2), #530.9: 1284. (Abstract)

13. CROSSIN,

K. L., A. L. PRIETO, S. HOFFMAN, F. S. JONES, AND D. R. FRIEDLANDER. 1990. Expression of adhesion molecules and the establishment of boundaries during embryonic and neural development. Exp. Neurol. 109: 6-18.

14. DA-CUNHA,

L., I. WALLIS, E. Anmu, AND K. H. PFENNINGER. 1985. Nerve growth cones isolated from fetal rat brain. IV. Preparation of a membrane subfraction and identification of a membrane glycoprotein expressed on sprouting neurons. J. Cell Biol. 101: 1977-1989.

21. FALLON,

J. H. 1987. Growth factors in the Basal Ganglia. In The Basal Ganglia II; Adv. in Behavioral Biology. (M. B. Carpenter and A. Jayaraman, Eds.) Vol. 32, pp. 247-260.

22. FREED, W. J., B. J. HOFFER,

L. OLSON, AND R. J. WYATT. 1984. Transplantation of catecholamine-containing tissue to restore the functional capacity of the damaged nigrostriatal system. In Neural Transplants: Development and Function (J. R. Sladek and D. M. Gash, Eds.), pp. 373-402. Plenum Press, New York.

23. GISPEN,

W. H., J. BOONSTRA, P. N. E. DEGRAAN, F. G. I. JENNEKENS, A. B. OESTREICHER, P. SCHOTMAN, L. H. SCHRAMA, J. VERHAAGEN, AND F. L. MARGOLIS. 1990. B-50/GAP-43 in neuronal development and repair. Rest. Neurol. Neurosci. 1: 237-244.

24. GORDON-WEEKS, growth

terning 692-699.

of neuronal

projections

1988. Axon guidance and the patin vertebrates. Science 242:

16. DODD, J., S. B. MORTON, T. M. JESSELL. 1988. expression on subsets 105-116.

D. KARAGOGEOS, M. YAMAMOTO, AND Spatial regulation of axonal glycoprotein of embryonic spinal neurons. Neuron 1:

17. DOUCET,

G., Y. MURATA, P. BRUNDIN, 0. BOSLER, N. MONS, M. GEFFARD, C. C. OUIMET, AND A. BJGRKLUND. 1989. Host afferents into intrastriatal transplants of fetal ventral mesencephalon. Exp. Neural. 106: 1-19.

18. DUBACH,

M., R. H. SCHMIDT, R. MARTIN, D. C. GERMAN, AND D. M. BOWDEN. 1988. Transplant improves hemiparkinsonian syndrome in nonhuman primate: intracerebral injection, rotometry, tyrosine hydroxylase immunohistochemistry. Prog. Bruin Res. 78: 491-496.

19.

DUBACH, R. H., R. H. SCHMLIDT, R. MARTIN, D. C. GERMAN, AND D. M. BOWDEN. 1988. Transplant improves hemiparkinsonian syndrome in nonhuman primate: Intracerebral injection, rotometry, tyrosine hydroxylase immunohistochimistry. In Transplantation into the Mammalian CNS, Progress in Brain Research, (D. M. Gash and J. R. Sladek, Eds.), Vol. 78, pp. 491-506. Elsevier, Amsterdam.

cone?

P. R. 1989. GAP-43-What Trends Neurosci. 12(10):

does it do in the

363-365.

25. GOSLIN,

K., D. J. SCHREYER, J. H. P. SKENEZ, AND G. BANKER. 1988. Development of neuronal polarity: GAP-43 distinguishes axonal from dendritic growth cones. Nature 336: 672-674.

26. GRAFSTEIN,

B., AND I. G. MCQUARRIE. 1978. Role of the nerve cell body in axonal regeneration. In Neuronal Plasticity (C. W. Cotman, Ed.), pp. 155-196. Raven Press, New York.

27. HOFFER,

B. J., A. C. GRANHOLM, J. 0. STEVENS, AND L. OLSON. 1988. Catecholamine-containing grafts in Parkinsonism: Past and Present. Clin. Res. 36(3): 189-195.

28. JACOBSON,

R. D., I. VIRAG, AND J. H. SKENE. 1986. A protein associated with axon growth, GAP-43, is widely distributed and developmentally regulated in rat CNS. J. Neurosci. 6: 18431855.

29. JESSELL, neural

30. KALIL, onally normal

T. M. 1988. Adhesion development. Neuron

molecules

and the hierarchy

of

1: 3-13.

K., AND J. H. SKENE. 1986. Elevated synthesis of an axtransported protein correlates with axon outgrowth in and injured pyramidal tracts. J. Neurosci. 6: 2563-2570.

31. MAHALIK,

T. J., T. E. FINGER, I. STROMBERG, AND L. OLSON. 1985. Substantia nigra transplants into denervated striatum of the rat: Ultrastructure of graft and host interconnections. J. Comp. Neurol. 240: 60-70.

A., AND L. VITKOVIC. 1990. Regulation ofimmunoreactive GAP-43 expression in rat cortical macroglia is cell type specific. J. Cell Biol. 111: 209-215.

15. DODD, J., AND T. M. JESSELL.

9

TRANSPLANTS

20. ELLIS,

DUNNETT, AND of dopaminerthe host neo-

IN

32. MAHALIK,

33.

T. J., AND G. H. CLAYTON. 1991. Specific outgrowth from neurons of ventral mesencephalic grafts to the catecholamine depleted striatum of adult hosts. Exp. Neurol., in press MAHALIK, T. J., K. H. PFENNINGER, AND T. E. FINGER. 1987. Ultrastructural localization of a growth-related antigen (5B4) and its relationship to immunocytochemically identified neurons in the olfactory bulb of the adult rat. Sot. Neurosci. 13(2): 1428. [Abstract]

34. MCGUIRE, microscopic associated Res. 469:

C. B., G. J. SNIPES, AND J. J. NORDEN. 1988. Lightimmunolocalization of the growthand plasticityprotein GAP-43 in the developing rat brain. Bruin 277-291.

35. MEIRI,

K. F., K. H. PFENNINGER, AND M. B. WILLARD. 1986. Growth-associated protein, GAP-43, a polypeptide that is induced when neurons extend axons, is a component of growth cones and corresponds to pp46, a major polypeptide of a subcellular fraction enriched in growth cones [published erratum appears in Proc. Natl. Acad. Sci. USA 1986 Dee; 83(23): 9274). Proc. Natl. Acad. Sci. USA 83: 3537-3541.

36. MEIRI,

K. F., M. WILLARD, tion and phosphorylation GAP-43 in regenerating Neurosci. 8: 2571-2581.

AND M. I. JOHNSON. 1988. Distribuof the growth-associated protein sympathetic neurons in culture. J.

10

CLAYTON,

MAHALIK,

37. MOYA, K. L., L. I. BENOWITZ, S. JHAVERI, AND G. E. SCHNEIDER. 1988. Changes in rapidly transported proteins in developing hamster retinofugal axons. J. Neurosci. 8: 4445-4454. 38. MURRAY, B. A. 1990. Cell-cell adhesion molecules in vertebrate development. Sem. Deu. Biol. 1: 3-14. 39. OPPENHEIM, R. W. 1989. The neurotrophic theory of naturally occurring motoneuron death. Trends Neurosci. 12: 7: 252-255. 40. PERRY, G. W., D. W. BURMEISTER, AND B. GRAFSTEIN. 1987. Fast axonally transported proteins in regenerating goldfish optic axons. J. Neurosci. ‘7: 792-806. 41. RAMOS, P., AND L. ELLIS. 1989. Expression in heterologous cells of the unusual cytoplasmic domain of rat brain 5B4/NCAM-ld. Brain Res. Mol. Brain Res. 6: 61-68. 42. RAMOS, P., R. SAFAEI, C. KAYALAR, AND L. ELLIS. 1989. Isolation and sequence of lambda gtll cDNA clones encoding the 5B4 antigen expressed on sprouting neurons. Brain Res. Mol. Brain Res. 6: 297-303. 43. ROSENBERG, J., L. ELLIS, F. A. TROY, AND C. KAYALAR. 1986. The 5b4 antigen expressed on sprouting neurons contains A-2,8linked polysialic acid. Dev. Brain Res. 30: 262-267. 44. SKENE, J. H., AND M. WILLARD. 1981. Changes in axonally transported proteins during axon regeneration in toad retinal ganglion cells. J. Cell Biol. 89: 86-95. 45. SKENE, J. H. P. 1989. Axonal growth-associated proteins. Annu. Rev. Neurosci. 12: 127-156. 46. SKENE, J. H. P., R. D. JACOBSON, G. J. SNIPES, C. B. MCGUIRE, J. J. NORDEN, AND J. A. FREEMAN. 1986. A protein induced during nerve growth (GAP-43) is a major component of growthcone membranes. Science 233: 783-786. 47. SPECHT, L. A., V. M. PICKEL, T. H. JOH, AND D. J. REIS. 1981. Light-microscopic immunocytochemical localization of tyrosine hydroxylase in prenatal rat brain. II. Late ontogeny. J. Camp. Neurol. 199: 255-276. 48. STENEVI, U., A. BJ~RKLUND, AND L. F. KROMER. 1984. Use of CNS implants to promote regeneration of central axons across denervating lesions in the adult rat brain. In Neural Transplants, Development, and Function (J. R. Sladek and D. M. Gash, Eds.), pp. 325-340. Plenum Press, New York/London. 49. STEWARD, 0.1989. Reorganization of neuronal connections following CNS trauma: Principles and experimental paradigms. J. Neurotrauma 6(2): 99-150. 50. STROMBERG, I., P. ALMQVIST, M. BYGDEMAN, T. E. FINGER, G. GERHARDT, A. C. GRANHOLM, T. J. MAHALIK, A. SEIGER, L. OLSON, AND B. HOFFER. 1989. Human fetal mesencephalic tissue

AND FINGER

51.

52.

53.

54.

55.

56.

57.

58. 59. 60.

61.

grafted to dopamine-denervated striatum of athymic rats: Light- and electron-microscopical histochemistry and in uiuo chronoamperometric studies. J. Neurosci. 9: 614-624. STROMBERG, I., S. JOHNSON, B. HOFFER, AND L. OLSON. 1985. Reinnervation of dopamine-denervated striatum by substantia nigra transplants: Immunohistochemical and electrophysiological correlates. Neuroscience 14: 981-990. TETZLAFF, W., AND M. A. BISBY. 1990. Cytoskeletal protein synthesis and regulation of nerve regeneration in PNS and CNS neurons of the rat. Rest. Neurol. Neurosci. 1: 189-196. VAN DER NEIJT, R., A. B. OESTREICHER, W. H. GISPEN, AND P. R. BAR. 1990. The expression of B-50/GAP-43 during development of rat spinal neurons in culture is regulated by interneuronal contact. Neurosci. Lett. 109: 36-41. VAN DER ZEE, C. E., H. B. NIELANDER, J. P. Vos, S. LOPES-DASILVA, J. VERHAAGEN, A. B. OESTREICHER, L. H. SCHRAMA, P. SCHOTMAN, AND W. H. GISPEN. 1989. Expression of growth-associated protein B-50 (GAP43) in dorsal root ganglia and sciatic nerve during regenerative sprouting. J. Neurosci. 9: 3505-3512. VERHAAGEN, J., A. B. OESTREICHER, M. GRILLO, Y. S. KHEWGOODALL, W. H. GISPEN, AND F. L. MARGOLIS. 1990. Neuroplasticity in the olfactory system: Differential effects of central and peripheral lesions of the primary olfactory pathway on the expression of B-50/GAP43 and the olfactory marker protein. J. Neurosci. Res. 26: 31-44. VITKOVIC, L., AND M. MERSEL. 1989. Growth-associated protein 43 is down-regulated in cultured astrocytes. Metab. Brain Dis. 4: 47-53. VOORN, P., A. KALSBEEK, B. JORRITSMA-BYHAM, AND H. J. GROENEWEGEN. 1988. The pre- and postnatal development of the dopaminergic cell groups in the ventral mesencephalon and the dopaminergic innervation of the striatum of the rat. Neuroscience 25(3): 857-887. VULLIET, P. R., T. A. LANGAN, AND N. WEINER. 1980. Tyrosine hydroxylase: a substrate of cyclic AMP-dependent protein kinase. F’roc. Natl. Acad. Sci. USA 77: 92-96. WALLIS, I., L. ELLIS, S. KYUNGSUN, AND K. H. PFENNINGER. 1985. Immunolocalization of a neuronal growth-dependent membrane glycoprotein. J. Cell Biol. 101: 1990-1998. WHI~EMORE, S. R., M. NIETO-SAMPEDRO, D. L. DEEDELS, AND C. W. COTMAN. 1985. Neuronotrophic factors for mammalian brain neurons: Injury induction in neonatal, adult and aged rat brain. Dev. Brain Res. 20: 169-178. YUREK, D. M., AND J. R. SLADEK. 1990. Dopamine cell replacement: Parkinson’s disease. Annu. Rev. Neurosci. 13: 415-440.