THE JOURNAL OF COMPARATIVE NEUROLOGY 395:91–111 (1998) Remodeling of Lesioned Kitten Visual Cortex After Xenotransplantation of Fetal Mouse Neopallium JITKA OUREDNIK,1* WENZEL OUREDNIK,2 AND DONALD E. MITCHELL1 1Department of Psychology, Life Sciences Center, Dalhousie University, Halifax, Nova Scotia, B3H 4J1 Canada 2Department of Anatomy and Neurobiology, Dalhousie University, Halifax, Nova Scotia, B3H 4H7 Canada ABSTRACT Remodeling of the mechanically injured cerebral cortex of kittens was studied in the presence of a neural xenograft taken from mouse fetuses. Solid neural tissue from the neopallium of a 14-day-old fetus was transferred into a cavity prepared in visual cortical area 18 of 33-day-old kittens. Injections of bromodeoxyuridine (BrdU) were used to monitor postoperative cell proliferation. Two months after transplantation, the presence of graft tissue in the recipient brain was assessed by Thy-1 immunohistochemistry. Antibodies specific for neurons, astrocytes, and oligodendrocytes and hematoxylin staining for endothelial cells were used for the characterization of proliferating (BrdU1) cells. The following were the major observations: 1) Of ten transplanted kittens, four had the cavity completely filled with neural tissue that resembled the intact cerebral cortex in its cytoarchitecture, whereas, in four other kittens, the cavity was partially closed. In two kittens, the cavity remained or became larger, which was also the case with all four sham-operated (lesioned, without graft) animals. 2) A substantial part of the remodeled tissue was of host origin. Only a few donor cells survived and dispersed widely in the host parenchyme. 3) In the remodeled region of transplanted animals, the densities of nerve, glial, and endothelial cells were similar to those in intact animals. 4) Cell proliferation increased after transplantation but only within a limited time, because, 2 months after the operation, the number of mitotic cells in the grafted cerebral cortex did not differ from that in intact controls. Our data suggest that the xenograft evokes repair processes in the kitten visual cortex that lead to structural recovery from a mechanical insult. The regeneration seems to rely on a complex interplay of many different mechanisms, including attenuation of necrosis, cell proliferation, and immigration of host cells into the wounded area. J. Comp. Neurol. 395:91–111, 1998. r 1998 Wiley-Liss, Inc. Indexing terms: neural transplantation; regeneration; plasticity; cerebral cortex; cat The regenerative capacity of the central nervous system (CNS) in mammals has become a major focus of much contemporary research in basic as well as in applied neurobiology. The long-held assumption that repair of the injured CNS is virtually impossible has been called into question by investigations that explore the extent of repair in diverse animal species at different developmental stages and also by studies that attempt to induce regeneration through application of external agents (for reviews, see ffrench-Constant and Mathews, 1994; Logan et al., 1994; Varon and Conner, 1994). Susceptibility of the CNS to injury and its subsequent capacity to heal and regenerate may vary considerably r 1998 WILEY-LISS, INC. through an animal’s lifetime, particularly during the various stages of development. Notably, in this respect, Kolb and coworkers (Kolb, 1987; Kolb and Gibb, 1991; Kolb et al., 1994) reported that in 1–6-day-old rats large lesions of the frontal or parietal neocortex had much more deleteri- Grant sponsor: Natural Sciences and Engineering Research Council of Canada; Grant number: 7660. *Correspondence to: Dr. J. Ourednik, Swiss Federal Institute of Technology, Department of Neurobiology-HPM, Hönggerberg, 8093 Zürich, Switzerland. E-mail: email@example.com Received 30 April 1997; Revised 9 January 1998; Accepted 20 January 1998 92 ous anatomical and functional consequences than similar insults performed in 7–12-day-old animals, in which remarkable functional and structural reorganization was observed. To date, the reasons for the different vulnerability and capacity for recovery in the CNS at different phylogenetic and ontogenetic stages of development are largely unknown. Important factors that may dictate the regenerative capability of the CNS at a defined developmental period include the actual rate of cell proliferation (especially of neuronal precursors) and the degree of cell differentiation that has been attained. To a large extent, these events are under the control of growth-promoting and inhibitory substances, the presence and ratio of which have been shown to vary during development. A recent striking example of the role of growth factors was the demonstration that cells dissociated from the subventricular zone of the adult mouse brain could be stimulated to differentiate in vitro when epidermal growth factor (EGF) or basic fibroblast growth factor (bFGF) were added to the culture medium (Reynolds and Weiss, 1992; Reynolds et al., 1992; Richards et al., 1992; Vescovi et al., 1993; Gritti et al., 1995, 1996). A promising line of research pursued during the last decade has examined the potential of neural grafts to reduce or even eliminate the negative sequelae of injury in the CNS. The transfer of fetal neural tissue into a wounded brain region, under certain circumstances, may lead to at least partial repair and compensation for the induced structural and functional defects (for reviews, see Berry, 1989; Fischer and Gage, 1993; Dunnett and Björklund, 1994). This beneficial effect is virtually always attributed to the proliferation of graft-derived tissue and to its integration into the host’s CNS. Explorations of the contributions from an alternative mechanism, namely, that the host tissue may be stimulated by the presence of the graft to participate actively in the regenerative and proliferative processes, have been attempted only rarely (Lundberg and Møllgård, 1979; Polezhaev et al., 1988). More recently, our own studies have shown that neural tissue isolated from neuroepithelium of the embryonic CNS of the mouse can effectively stimulate repair mechanisms within the lesioned brain of a juvenile host from a genetically closely related mouse strain (Ourednik et al., 1993a,b; Ourednik and Ourednik, 1994). Similar regeneration-inducing and remote effects of grafts were also reported by Valoušková and Gálik (1995), who performed bilateral lesions in the frontal neocortex of adult rats but placed fetal neural grafts into only one of the two cavities. Not only was the lesion on the grafted side refilled with neural tissue, but the contralateral, nongrafted cavity was also significantly reduced in size. The purpose of the present study was to investigate whether the stimulating influence of a fetal neural transplant on CNS regeneration that had been observed previously with two congenic mouse strains as donor and host would also be apparent with grafts between two species that were evolutionary more remote. We chose to produce lesions in the kitten primary visual cortex, a well-known model system for studies of developmental plasticity of the central visual pathways, and to implant the lesion with neural grafts prepared from the fetal mouse telencephalon. The chosen age of the host animals (4–5 weeks) was at a time when neurogenesis in the cat visual system was complete (Luskin and Shatz, 1985) but still before the time of anatomical and functional maturation in the visual J. OUREDNIK ET AL. TABLE 1. Degree of Cavity Remodeling in Transplanted and Sham-Operated Animals1 Animal no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Sex Group Postoperative day of death Remodeling of lesion area Thy-1.21 cells per five sections (level I–V) m f f f f m m m f f f m f m m m m f m m Transplanted Transplanted Transplanted Transplanted Transplanted Transplanted Transplanted Transplanted Transplanted Transplanted Sham Sham Sham Sham Sham Sham Sham Intact Intact Intact 62 62 62 62 63 63 63 61 61 61 61 61 59 59 Immediately Immediately Immediately 62 62 62 1/2 1/2 1 — 1/2 1/2 1 2 11 11 2 2 — — 2 2 2 n.a. n.a. n.a. 98 48 68 179 49 14 154 60 285 59 0 0 0 0 0 0 0 0 0 0 1Five degrees of remodeling of the cavity were defined as follows: complete filling with exuberant tissue, as seen in Figure 5 (11); complete filling without exuberant tissue (1); partial filling (1/2); persistent cavity of about the same size as originally prepared (2) and cavity evidently bigger than the original one (—). Numbers of Thy-1.21 cells reflect the absolute numbers of cells observed at levels I–V (see Fig. 1). m, Male; f, female; n.a., not applicable; level I–V, frontal planes of cutting (see Fig. 1). cortex (Cragg, 1975; Mitchell and Timney, 1984; Shatz and Luskin, 1986; Dyck et al., 1993; Price et al., 1994; Daw, 1995). The age coincided with the time of maximal activitydependent plasticity, when the kitten visual cortex is most susceptible to modification by experiential manipulations of the visual input (Mitchell and Timney, 1984; Daw, 1995). Use of the mouse as donor permitted a more simple and plentiful supply of neural graft tissue than that supplied from cat fetuses, but, more importantly, it allowed the use of anti-Thy-1 immunohistochemistry for later identification of grafted and host tissue (Charlton et al., 1983; Zhou et al., 1985; McKeon et al., 1989; Ourednik et al., 1993a; Ourednik and Ourednik, 1994), because the antigen’s isoforms differ in cat and mouse (Morris and Grosveld, 1989; Saleh and Bartlett, 1989). Although Thy-1 is not fully expressed in all brain areas during early development, in the adult CNS, it is present on the surface of most neuronal cell bodies and their processes and on astroglial cells (for review, see Morris and Grosveld, 1989). To allow a direct comparison with our previous allografting experiments (Ourednik et al., 1993a,b), in which no immunosuppressive drugs were used, we performed the present xenotransplantation in the absence of these agents as well. Pilot experiments on three kittens (Table 1, kittens 1–3) showed that 2-month-old grafted lesions were partially (two kittens) or completely (one kitten) closed and that the repair could occur without immunosuppression. Transplantation of fetal tissue into a cat brain has been reported only rarely (Müller and Best, 1989; Villablanca et al., 1990; Yinon and Gelerstein, 1991; Yinon et al., 1993; Targett and Blakemore, 1994), and these reports have provided only scant descriptions of the operative techniques that are more demanding than those required for grafting into rodent CNS. Therefore, an important secondary aim of the present study was to establish and describe extensively a suitable surgical and transplantation procedure that could be employed in kittens. XENOTRANSPLANTATION INTO KITTEN NEOCORTEX Two sets of data were collected and analyzed. The first set of data, which is documented below in Part A, describes the remodeling of the grafted wound and the extent to which the graft participated in the repair. The second set of data, which is described below in Part B, provides information concerning the contribution of specific cell types (neurons, astrocytes, oligodendrocytes, and endothelial cells) to the remodeling of the lesion (evaluations of changes in related subcortical structures were not included in this study). To this end, densities of these cell populations were determined by using the appropriate cell type-specific immunohistochemical markers, namely, neuron-specific nuclear protein A60 (NeuN) for neurons, glial fibrillary acidic protein (GFAP) for astrocytes, cyclic nucleotide phosphohydrolase (CNP) for oligodendrocytes, and two different isoforms of the Thy-1 antigen for the detection of donor and host cells. Endothelial cells were identified in hematoxylin-stained sections on the basis of their morphology. In combination with these cell type-specific markers, we used postoperative bromodeoxyuridine (BrdU) labeling of dividing cells and a BrdU-specific antibody to monitor the postoperative proliferation of the various cell types of interest. Collectively, these data permitted the evaluation of the relative contribution of two possible processes to the filling of the cavity, namely, rearrangement of existing cells in the vicinity of the lesion vs. postoperative proliferation among either donor or specific host cell populations. MATERIALS AND METHODS Animals Donors. Adult C57BL/6J mice were mated overnight, and the day of vaginal plug detection was defined as embryonic day 0. Transplants were obtained from 14.5-dayold (60.5 day) fetuses that were examined carefully under a dissecting microscope for their overall condition and precise stage of development according to the criteria specified by Kaufman (1992). Only healthy and not developmentally retarded fetuses were used. Recipients. The ten host animals were 33-day-old kittens (61 day) from an outbred, nonspecific, domestic cat lineage. The day of birth was considered to be day 0. At the time of operation, the average weight of the kittens was 470 6 80 g for males and 360 6 50 g for females. Controls. Two types of age-matched control animals were prepared in parallel with the animals that received a graft: 1) sham-operated kittens (n 5 7) that received the lesion but no graft and 2) intact kittens (n 5 3). A total of 20 kittens from seven litters of different queens were used (see Table 1). The experiments conformed to the strict guidelines regarding the care and use of animals specified by the Society for Neuroscience (1991) and the Canadian Council on Animal Care (1993/4). Cat and mouse colonies were held on a standard diet, maintained on a 12-hour light-dark regime, and inspected regularly several times each day. Surgical procedures Operations were performed under sterile conditions with the help of a dissecting microscope. Kittens were injected with xylazine (Rompun; Bayer, Etobicoke, Ontario, Canada; 2.2 mg/kg i.m.) and anesthetized with gaseous halothane (Fluothane-Halothane B.P.; WyethAyerst, Montreal, Quebec, Canada) mixed with oxygen 93 (flow 1,500 mm3/minute) in a Fluotec 3 container (CDMV; Fluotec, St. Hyacinthe, Quebec, Canada). Inhalation started with the gas containing 5% of halothane for about 2 minutes to effect, then the halothane concentration was reduced to 1.5–2.0%, which was maintained during the rest of the operation. Once the animal was sedated, the hair covering the right side of the neurocranium was shaved, and the skin was swabbed with an antiseptic/ antibacterial solution (1:30 dilution of Zephiran chloride; Winthrop Laboratories, Aurora, Ontario, Canada) in distilled water (dH2O) containing 0.08% chlorhexidine gluconate (Hibitan; Ayerst Laboratories, Montreal, Quebec, Canada) and with the surgical microbicidal skin cleanser Betadine (Purdue Frederick Inc., Toronto, Ontario, Canada). Finally, three drops (approximately 150 µl in total) of the local anesthetic Marcaine (Sanofi Winthrop, Markham, Ontario, Canada) were applied to the skin of the scalp before surgery. A 2-cm-long parasagittal incision of skin and subcutaneous tissue was performed 0.5 cm to the right of the midline of the scalp, extending approximately from the external occipital protuberance to the anterior edge of the pinna. The skin was held back with a small retractor, and the pericranium above the chosen brain region was cut in anteroposterior direction, scraped laterally from the bone with a spatula, and then cut away. With a high-speed microdrill, four small holes were drilled (0.9 mm drill bit diameter) that defined the corners of a small rectangular window in the bone (6 3 5 mm) centered 8 mm rostral to lambda and 4 mm to the right of the midline and overlying the area of the lateral sulcus where part of the visual cortical area 18 is represented (see Fig. 1). The bone between the microdrill perforations was cut with a small circular saw (16 mm diameter; FST, North Vancouver, British Columbia, Canada), while the region was moistened constantly and cooled with drops of sterile physiological solution. The removed piece of bone was flipped over and kept beside the prepared cranial opening until it was needed. Preparation of the lesion. The exposed meninges were gently cut off, and, after the bleeding stopped spontaneously (generally within 5–10 minutes), the area was cleaned with small pieces of sterile gauze and sterile saline. A small rectangular block of brain tissue measuring 3 mm in the mediolateral direction and 4 mm in the anteroposterior direction and measuring 1.3–1.5 mm in depth was delineated by four perpendicular cuts with a microsurgical, hook-shaped knife. The block of tissue (18.0 6 1.5 mm3) was then undercut and withdrawn with a custom-made, tiny, spatula-shaped knife. The cavity was carefully cleaned of blood by using dental absorbent paper points and with very gentle aspiration by using Pasteur pipettes drawn out to an outer tip diameter of ca. 0.3 mm and connected to an electrical vacuum pump (negative pressure set to a maximum of 100 mm Hg). Utmost care was taken not to damage the walls of the cavity by the suction. Finally, the cavity was rinsed with Dulbecco’s modified Eagle medium (D-MEM; catalog no. 11965–043; Gibco BRL, Grand Island, NY), and the same was used later during the dissection of the mouse embryos and the preparation of the graft tissue. The position and extent of the cavity in the kitten visual area 18 are shown in Figure 1. Grafting. In parallel with cleaning of the cavity in the host kitten brain, fresh graft tissue was prepared as follows. Mouse dams were deeply anesthetized with an i.p. 94 Fig. 1. Position and extent of the lesion in the kitten visual cortex. Top: Dorsal view of the occipital region of the cerebral hemispheres of a 3-month-old kitten. A schematic representation of visual areas 17, 18, 19, and 21a is projected onto the left hemisphere, whereas the location of the lesion is shown to the right (box). Roman numerals I–V indicate coronal levels from which sections were chosen for analysis. Bottom: Hematoxylin-stained, 20-µm frontal section from level III with the position and extent of the lesion indicated in the right hemisphere (box). Asterisk indicates the Sylvian fissure. Scale bar 5 6 mm. injection (10 µl/g body weight) of a mixture of sodium pentobarbital (11% Somnotol; MTC Standard Pharmaceuticals, Cambridge, Ontario, Canada) diluted with 38% propylene glycol in dH2O, giving an anesthetic dose of 75 mg of sodium pentobarbital per kg body weight. They were then placed on their backs, and their abdomens were swabbed with Zephiran chloride (see above). A 1-cm-long, transverse incision was made in their lower abdomen to allow access to the uterus. Fetuses were removed one by one (starting from the uppermost uterine corner), as needed. Only freshly prepared grafts were used for transplantation, and each fetus served to prepare only one transplant. After a fetus was removed, the incisions of uterus and abdomen were clamped with hemostats, and the abdomen was covered with a piece of sterile gauze. The total time during which the dam was kept under narcosis was about 3–6 hours, depending on the number of transplanted animals operated that day. Additional small doses of the Somnotol mixture were injected as required. The removed fetus was rinsed in D-MEM (see above), liberated from its embryonic envelopes, and killed by a cut through the cervical spinal cord. The subsequent dissection and preparation of the graft was performed with the aid of tiny forceps and microscissors in the same medium in a small, concave, black dish placed under the dissecting microscope. The cartilaginous fetal skull was opened to- J. OUREDNIK ET AL. gether with the overlying skin by one cut in its midsagittal plane and two lateral cuts extending from each eye back to the cervical spinal cord. All meninges covering the cerebral hemispheres were peeled off, and the brain was transferred into fresh medium. A small block of tissue (1.5 6 0.25 mm3), including the whole thickness of the cortical wall, was cut from the right neopallium. The graft was rinsed in fresh medium, placed on the round tip of a glass rod (0.5 mm diameter), and transferred immediately, without any particular orientation, to the bottom of the cleaned recipient cavity. Only one block of graft tissue was placed in each cavity. Conclusion of the operation. The cranial window was closed with the corresponding piece of bone, and the area was cleaned of blood and rinsed with physiological saline. For suturing of the overlying connective tissues and skin, resorbable threads (4–0 Dexon II; Cyanamid Inc., Montreal Quebec, Canada; four to six stitches) and nonresorbable threads (4–0 Novafil; Cyanamid Inc.; six to nine stitches) were used. The sutured skin was cleaned with pieces of gauze soaked in sterile saline and scrubbed with Betadine. One operation lasted approximately 80 6 30 minutes, after which, the kitten was disconnected from the anesthetic apparatus and placed in a plastic cage maintained at 36.5°C. Once it was awake, after about 1.0–1.5 hours, the kitten was returned to its mother. Neither prophylactic antibiotics nor immunosuppressive drugs were ever administered to the animals. Monitoring of postoperatively dividing cells To monitor postoperatively proliferating cells, kittens were injected i.p. with 58-bromodeoxyuridine (BrdU; 50 mg/kg body weight; Sigma, Oakville, Ontario, Canada) administered at a concentration of 10 mg/ml in physiological saline containing 0.007 N NaOH (Takahashi et al., 1992). Because the availability of the injected BrdU is time limited (20–40 minutes; Boswald et al., 1990), the animals were injected repeatedly in order to increase the probability of its incorporation by a high number of dividing cells. The first injection occurred 48 hours after operation, and the following ones were administered every other day, for a total of six injections per animal. Age-matched, shamoperated and intact kittens were injected in parallel. Perfusion and histology All animals were killed 61 6 2 days after operation, when the average weight of the kittens was 1,860 6 450 g for males and 1,510 6 180 g for females. Three additional sham-operated kittens were perfused immediately after the operation in order to observe the extent and variability of the fresh lesions (Table 1, kittens 15–17). The animals were injected intrahepatically with an overdose of sodium pentobarbital (Euthanyl; Canada Packers Ltd., MTC Pharmaceuticals Division, Mississauga, Ontario, Canada; 2 ml/4.5 kg body weight) and, once they were in deep anesthesia, they were perfused through the left heart ventricle for 7 minutes with 0.1 M phosphate-buffered saline (PBS), pH 7.4 (100 ml/minute flow rate maintained with a Masterflex Pump; Cole-Palmer Instrument Co., Chicago, IL). The skull was opened with an electrical circular saw (16 mm diameter) and a rongeur, and the forebrain anterior to the tentorium was removed. Because of the risk of mechanical damage, about 1.5 cm2 of bone were left attached temporarily above the operated area. The frontal lobes XENOTRANSPLANTATION INTO KITTEN NEOCORTEX were cut off, and the rest of the brain was incised midsagittally and immersed into acidic alcohol (5% of acetic acid in 70% ethanol) for 2 days. Following fixation, the remaining bone piece was removed very carefully by using a rongeur, microsurgical scissors, forceps, and a magnifying glass. The brain tissue was dehydrated in successively higher concentrations of ethanol, beginning with an 80% solution (30 hours; two or three baths), followed by immersion in a 95% solution (4 hours; three or four baths), and, finally, absolute ethanol (4 hours; three or four baths). After dehydration, the tissue was embedded in polyester wax (BDH; Poole, Dorset, United Kingdom; 2.5 days; three baths at 38°C). Coronal 20-µm sections were cut on a microtome from anterior level I to posterior level V, as indicated in Figure 1, with a distance of 3–4 mm between each level. These levels correspond approximately to the Horsley-Clarke coronal planes P5.4 to A4.2 of the adult cat brain (Tusa et al., 1981). With the additional help of the atlas by Snider and Niemer (1961), the block of tissue was oriented carefully in the coronal plane to ensure equivalent cutting levels for both hemispheres. Complete series of about 150–300 sections per animal were collected on slides coated with 1% gelatine containing 1% formol or with Vectabond (Vector Laboratories, Burlingame, CA) and were stored at room temperature (RT) until further use. Just before immunostaining, sections were dewaxed by successive immersion for 3 minutes in two baths of absolute ethanol (they were air dried for about 30 seconds after the first bath to improve their attachment to the glass slides), rehydrated, and washed twice for five minutes each in PBS. The sections were used immediately for immunohistochemical staining without being dried. Cell identification and immunohistochemistry To analyze cortical cytoarchitecture, immunostaining with cell type-specific antibodies (Abs) was conducted on sections cut from coronal levels I–V (see Fig. 1). To follow the extent to which the postoperative cell proliferation varied among the different cell types, double immunostaining was conducted on sections cut from coronal level III. The first immunoreaction was always against BrdU, followed by intensive washing in PBS (1–2 hours), and then by incubation with a cell type-specific Ab. The Abs and procedures for the various immunoreactions are described separately below. Negative controls (the primary Ab replaced by the diluent) were always processed in parallel with the immunoreactions for each marker. BrdU Cells labeled with BrdU were detected according to a modified version of the protocol of Takahashi et al. (1992). Sections were hydrolyzed at RT in 2 N hydrochloric acid (HCl). In a test using acridine orange (Von Bertalanffy and Nagy, 1962), we established an optimal denaturing time of 25–35 minutes that provided a balance between maximal production of single-stranded DNA (to improve accessibility of the incorporated BrdU to the Ab) and maintenance of good cytoarchitecture. After hydrolysis, sections were rinsed for 1 minute in PBS (0.1 M), pH 6; blocked for 30 minutes in a solution of 10% normal horse serum (NHS) containing 1% H2O2; and then incubated for 45 minutes at RT with the monoclonal rat anti-BrdU Ab (clone BU1/75; ICR1; 95 Sera-Lab, Crawley Down, United Kingdom) diluted 1:75 in PBS (0.05 M), pH 7.4, containing 0.5% Tween 20. Detection of the bound primary Ab with a biotinylated rabbit anti-rat immunoglobulin (Ig) was conducted in accordance with the instructions of the Vectastain kit (Elite Vectastain; Vector Laboratories). The final dark-blue reaction product was obtained in a nickel/cobalt-3,38-diaminobenzidinetetrahydrochloride (Ni/Co-DAB) solution: 15 ml PBS with 10 mg DAB, 3 mg nickel-ammonium sulphate (NiSO4[NH4]2SO4.6H2O), 3.5 mg cobalt chloride (CoCl2.6H2O), and 1 µl 30% hydrogen peroxide (H202). Some of the immunostained sections were counterstained with Mayer’s hematoxylin (Stevens, 1990). Thy-1 Murine Thy-1. The following protocol was used to detect the Thy-1.2 antigen expressed on the cells xenografted from C57BL/6J mice. After inactivation of endogenous peroxidase in 3% H2O2 for 15 minutes and washing in PBS (0.1 M), pH 7.2, sections were blocked for 30 minutes in 1% fetal calf serum (FCS) and 0.5% bovine serum albumin (BSA). The monoclonal rat anti-mouse Thy-1.2 Ab was diluted 1:100 (clone 30-H12; Becton Dickinson, Mountain View, CA) or 1:50 (Pharmingen, San Diego, CA) in 20% FCS in PBS, and it was left to react overnight at 4°C. After washing, the sections were incubated for 2.5 hours at RT with a horseradish peroxidase (HRP)-conjugated rabbit anti-rat Ig (Dako, Glostrup, Denmark) diluted 1:40 in 20% FCS in PBS. The final color reaction in the presence of DAB was controlled under a microscope and intensified by replacing PBS with an imidazole buffer (Morris, 1991). The reaction was usually stopped within 5–10 minutes in running H2O. To assure specificity of the Thy-1.2 immunostaining, sections from nontransplanted (sham-operated and intact) kitten brains and sections from 2-month-old C57BL/6J mouse brains (Thy-1.21) were assayed as well. Feline Thy-1. To assess the presence of host cells in the remodeled area, after blocking, sections from transplanted animals were incubated with a monoclonal mouse anti-cat Thy-1 Ab diluted 1:1 (Saleh and Bartlett, 1989) in 20% FCS in PBS and left overnight at 4°C. After washing, the sections were incubated at RT with a biotinylated rabbit anti-mouse Ig (dilution 1:150) for 2.5 hours, and visualization occurred according to the protocol provided by the Vectastain kit. To control the specificity of this immunostaining, sections from C57BL/6J mouse brains were left to react with the cat-specific Ab in parallel. NeuN. To detect NeuN1 cells, we used our modification of a protocol by Mullen et al. (1992; personal communication). Sections were incubated overnight at 4°C with the monoclonal mouse Ab A60 (Mullen et al., 1992) diluted 1:100 in 10% NHS in PBS (0.1 M), pH 7.2. Specific binding was detected with a biotinylated horse anti-mouse Ig (dilution 1:150; 60 minutes at RT) by using Vectastain and DAB. GFAP. Sections were reacted with a polyclonal rabbit anti-cow GFAP Ab (Dako) diluted 1:1,000 in 10% NHS supplemented with 0.05 M BSA and 0.1% Triton X-100. After an overnight incubation at 4°C and washing in PBS, an HRP-conjugated Ab against rabbit Ig (Dako; dilution 1:40 in 10% NHS/0.1% Triton X-100 in PBS) was applied for 2.5 hours at RT. After washing, the enzymatic product was visualized by using a routine DAB stain. CNP. Sections were incubated overnight at 4°C with the mouse monoclonal Ab against CNP (Promega, Madi- 96 J. OUREDNIK ET AL. son, WI) diluted 1:300 in 10% NHS. An HRP-conjugated secondary Ab against mouse Ig was applied for 2.5 hours at RT, and the enzymatic product was visualized with DAB. Endothelial cells. Endothelial cells could be identified with ease in sections stained with Mayer’s hematoxylin (Stevens, 1990) on the basis of their nuclear characteristics (dense staining, elongated and irregular shape) and their frequent presence in close vicinity to the lumen of a blood vessel (Fig. 7). Microscopy and photomicrography Sections were mounted in Permount (Fisher Scientific, Halifax, Nova Scotia, Canada) or Entellan (Merck, Darmstadt, Germany) and were analyzed with a Zeiss Axiophot microscope (Thornwood, NY) or with a Nikon Labophot microscope (Tokyo, Japan) equipped with a camera lucida drawing tube. The photomicrographs in Figures 1A, 5A, and 6 were prepared with a Nikon F-601M camera mounted on either a dissecting microscope (Wild M650; Leica, Heidelberg, Germany) or a Nikon Labophot. The photomicrographs in Figures 1B, 2, and 5B,C were taken with a Leica MPS52 camera mounted on a Leica Wild MZ8 microscope equipped with a Leica EB transmission-light stand. Figures 3 and 7 were taken with a Zeiss Axiophot microscope. Photomicrographs were taken on Kodak Technical Pan film (Eastman Kodak, Rochester, NY) at 50 ASA or on Kodak film EPJ 64T. Evaluation Qualitative evaluation. The position and extent of the lesions in transplanted and sham-operated kittens were compared for each animal in hematoxylin-stained sections at all five coronal levels I–V (see Fig. 1). The camera lucida drawings in Figure 4 illustrate examples of both fully and partially filled cavities as well as persisting lesions in the transplanted kittens. For quantitative measurements, these drawings were magnified (approximately 23), equivalent regions of the right and left hemispheres were cut out and weighed, and their weight ratios (operated/intact; O/I) were determined. The cuts followed the cortical circonvolutions and included the representations of areas 17, 18, 19, and 21a on the lateral gyrus and part of the suprasylvian gyrus (compare with Fig. 1; see also Figs. 1.10 and 1.11 in Tusa et al., 1981). To minimize errors caused by possible distortion of the operated cortex, the choice of corresponding coronal levels in the left and right hemispheres was verified according to subcortical structures by using the atlas of Snider and Niemer (1961). Drawings were also made from sections cut through other areas of the transplanted brains to illustrate the dispersion of the Thy-1.21 graft cells (Fig. 6). Quantitative evaluation Thy-1.21 (donor) cells. For every animal, the absolute number of Thy-1.2-labeled cells was determined from five brain sections that were taken from all evaluated levels I–V (Table 1; compare also with Fig. 1A). At each level, the total section area was evaluated, and the average number of donor cells was calculated separately (Fig. 6). Relative cell densities in the operated area. The relative cell density (numbers of NeuN1, GFAP1, Thy-1.21, and CNP1 cells as well as endothelial cells per 1 mm2) was determined in the four animals with completely closed cavities (kittens 3, 7, 9, and 10; see Table 1), and the results were compared with data obtained from intact brains (kittens 18, 19, and 20) and sham-operated brains (kittens 11, 12, 13, and 14). Transplanted animals with only partially closed lesions and sham-operated kittens killed immediately after lesioning were not included in the analysis. Cells were counted in area 18 across the entire thickness of the right sham-operated and transplanted visual cortices, and the results were compared with those obtained in an intact cortex (see schematic drawing in Fig. 8). Two regions were evaluated in sham-operated brains: An area proximal to the cavity (within 100–200 µm) and a distal region located 200–1,000 µm from the cavity walls. For each brain, animal, and cell category, one section from level III (Fig. 1) was analyzed at 4003 or 6303 magnification (Fig. 8). Postoperative cell proliferation. Relative cell densities of double-stained cells (BrdU/NeuN1 BrdU/GFAP1, BrdU/ CNP1, BrdU/Thy-1.21, and BrdU1 endothelial cells) were calculated in a manner similar to that described above (Fig. 9). Statistics For each quantitative evaluation, arithmetic means with standard deviations were calculated, and the statistical significance of the counts with respect to those found in intact animals was assessed by using Student’s t-test (P # 0.05). RESULTS Part A: Morphological aspects of remodeling in transplanted and/or sham-operated kittens and fate of the graft Lesion immediately after operation. Upon opening the skull of sham-operated kittens that had been perfused immediately after preparation of the lesion (kittens 15–17; Table 1, Fig. 2A), an extensive hematoma was observed in the meninges covering the operated hemisphere, accompanied by edema of the neural tissue within 5–10 mm of the wound margin. During subsequent fixation and dehydration of the tissue, the edema deflated and was invisible in mounted sections. Although the position of the lesions varied slightly in anteroposterior and/or mediolateral extent (about 1–3 mm), the size of the lesion was rather uniform in all operated animals, and the cavity was localized, as intended, in the gray matter of the lateral gyrus of the occipital cortex, in visual area 18. Operated region 2 months after surgery General features. All kittens destined for long-term survival recovered quickly and without complication from surgery. After several hours, their behavior was comparable to that of their nonoperated litter mates. No changes of visual perception and/or orientation behavior were evident in either the sham-operated or the transplanted animals. Two months after the operation, the bone piece that had been removed temporarily during surgery and the surrounding skull were well fused together. The lesioned and/or lesioned and transplanted brain regions were covered with an unusually thick layer of meninges (Fig. 3) that firmly attached the brain’s surface to the inner surface of the neurocranium. Sham-operated kittens. Lesions without a transplant always remained present as a cavity filled with cerebrospinal fluid and roofed by heavily vascularized meninges (kittens 11–14; Table 1, Fig. 2B). In most sham-operated XENOTRANSPLANTATION INTO KITTEN NEOCORTEX 97 Fig. 2. Photomicrographs of hematoxylin-stained, 20-µm frontal sections at level III (see Fig. 1) through operated area 18 of three kitten brains. A: The cavity immediately after operation, before the transfer of the graft (animal 15 in Table 1); B,C: The cavity 2 months after operation in a sham-operated (B; kitten 11) and a transplanted (C; kitten 3) animal. Note, that in the grafted cortex in C, the lateral gyrus and the suprasylvian gyrus are fused and that the posterolateral sulcus (arrowhead in A and B), which normally separates them, is missing. Asterisks indicate the Sylvian fissure. w, White matter; g, gray matter. Scale bar 5 3 mm. animals, the wound’s position and dimensions (adjusted for growth) were approximately the same as originally defined. However, in two cases (kittens 13 and 14; Table 1), the lesion was considerably enlarged. The gray matter underneath the wound had apparently degenerated, so that the bottom of the cavity was in close contact with the white matter (Fig. 2B). The walls were lined with a thin layer of nonneuronal tissue that merged gradually with the meninges. A prominent glial scar never developed, although some accumulation of glial cells was observed close to the bottom and the most superficial edges of the cavity. At the time of perfusion, necrosis still appeared to be in progress in the vicinity of the lesion (within 100–200 µm), as evidenced by a reduced cell density and cell pyknosis. Transplanted kittens. Remodeling of cavities in transplanted brains differed from animal to animal, without any linkage to sex. The most obvious variation was in the degree of closure of the cavities. In four of ten grafted kittens (kittens 3, 7, 9, and 10; Table 1), the cavity was completely filled with neural tissue (Figs. 2C, 3, 5C). In two of these (kittens 9 and 10), exuberant tissue connected lateral and suprasylvian gyri in such a way that the sulcus between the gyri was missing (Fig. 2C; Fig. 4, left column, level IV; Fig. 5A,C). In four other cases (kittens 1, 2, 5, and 6), the cavities were closed only partially, persisting as a small hollow covered with meninges at level III and extending variably in the anteroposterior direction but never invading the adjacent gyrus (Fig. 4, middle column). Only two transplanted cavities remained open (kittens 4 and 8; Fig. 4, right column). In the case of kitten 4, the wound extended at least to level II anteriorly and to level IV posteriorly; at level I, the lateral gyrus was diminished compared with the intact contralateral hemisphere, and the sulcus between the lateral and suprasylvian gyri at level V was deeper (compare right column of Fig. 4 with the original position of the cavity around level III in Fig. 1). The remodeled area was covered by meninges which, in this region, appeared thicker than normal (Fig. 3). These qualitative observations were accompanied by the calculations of the O/I indices (see Materials and Methods) in coronal sections of both hemispheres cut at levels I–V (Fig. 4). In the four sham-operated animals, the mean O/I index (0.83) was significantly less than 1.0, reflecting the much smaller area of the lesioned side compared with the contralateral intact side. By contrast, in the four transplanted kittens with completely filled cavities, the hemispheres were of comparable size, as reflected by an O/I index close to 1 (1.074), which was also close to the index observed in intact animals (1.026). Interestingly, the O/I index in animals with partially closed lesions (kittens 1, 2, 5, and 6) was also in this range (1.020), a result that could be attributed to expansion of the tissue surrounding the cavity. In animals in which the grafted lesion failed to close (kittens 1 and 8), the lesioned hemisphere was smaller Figure 3 XENOTRANSPLANTATION INTO KITTEN NEOCORTEX (O/I index, 0.94), which was also the case in the shamoperated kittens (O/I index, 0.83). In the transplanted animals with fully filled cavities, the original borders of the lesion were no longer visible, and the organization of the cerebral cortex, including lamination and presence of both neuronal and nonneuronal cell populations, resembled that of intact kittens (Fig. 3). Only two unusual structures were evident. First, a narrow sheet of nerve fibers lined the remodeled area medially (Figs. 2C, 3; Fig. 4, left and middle columns, levels II–IV). Here, intermingled with glial and endothelial cells, nerve fibers projected in an aberrant way from the white matter up to the pial surface. At the upper cortical layers (I and II), this fiber bundle fanned out slightly and was penetrated by a dense meshwork of vessels originating in the meninges. Second, in the two animals (kittens 9 and 10) with neural tissue protruding above the gyral surface, the tissue was penetrated by a thin fiber tract that also originated in the underlying white matter. In coronal sections, the protruding tissue was reminiscent of a small extra gyrus (Fig. 5A,B). Mitotic figures (Fig. 3) were observed in all animals at the time of death, indicating a weak level of proliferative activity in the neocortical tissue of 3-month-old kitten brains. All mitoses were BrdU2. At this age, the number of mitoses was similar in intact (72.2 6 69.9 mitotic cells per mm2) and transplanted (85.1 6 82.5 mitotic cells per mm2) animals. In most cases, the dividing nuclei appeared to be those of endothelial and glial cells. These observations not only indicate ongoing vasculogenesis in the juvenile cat brain but also suggest that the graft does not influence the rate of cell proliferation at this particular stage (3-monthold kittens) of tissue remodeling. Fate of graft tissue. Immunohistochemical analysis of the tissue replacing the cavity revealed that the vast majority of the constituent cells of this tissue did not express the mouse-specific Thy-1.2 antigen diagnostic of the xenograft (Fig. 5C) but were positive for the feline Thy-1 antigen (Fig. 5D). Although a solid piece of neural tissue had been transplanted initially, no compact graft was observed 2 months later, and just solitary Thy-1.21 cells, which were visible only at high magnification, were found dispersed in the recipient’s brain (Figs. 6, 7). These Thy-1.21 cells were observed not only in the transplanted area but also in subcortical regions and, occasionally, even in the contralateral hemisphere. The majority were situated in or close to the operated cortical region, mostly in upper cortical layers I and II (at cutting level III), where they intermingled with fibers of the aberrant tract deFig. 3. Photomicrographs of 20-µm-thick, hematoxylin-stained sections that show a comparison of the transplanted visual cortical area 18 (Tr; animal 7 in Table 1) 2 months after operation with the same region in an age-matched intact (In) kitten cortex (cortical layers I–VI; animal 18). The schematic drawing in the lower left corner shows the position at cutting level III (compare with Fig. 1) of the photomicrograph shown above. d, Dorsal; l, lateral. The lower middle photomicrograph is a reduced version of the picture from the grafted area and serves for the identification of the relevant structures. The black line indicates the approximate size and shape of the original cavity (compare with Fig. 2A). A narrow sheet of nerve fibers (asterisk) extends from the white matter (wm) to the meninges (m). It fans out as it reaches the upper cortical layers (I and II) before it merges with the meninges. The photomicrograph in the lower right corner shows examples of a neuron (N), a glial cell (G), an endothelial cell (E), and a mitotic cell (M) in the transplanted area. Scale bars 5 1 mm in Tr and ln photomicrographs, 10 µm in lower right corner. 99 scribed above. Graft cells also populated other areas, like the superior colliculus, the lateral geniculate body, and the white matter (optic radiation, optic tract, and corpus callosum); occasionally, they were observed in other regions (e.g., in the mesencephalon) that do not belong to the central visual system. The frequency of Thy-1.21 cells varied from animal to animal (Table 1), and there was no obvious correlation between the degree of tissue remodeling in the lesion area and the number of graft cells present. Thus, for example, a large difference was observed between the numbers of Thy-1.21 cells in animals 9 and 10, even though, in both cases, the cavities were completely filled. On the other hand, in the grafted cavities that persisted and that resembled the cavities observed in sham-operated animals, donor cells could still be detected (kittens 4 and 8; Table 1). Quantitative measurement of the number of donor cells proved their average number to be very low, i.e., only 5–40 cells per brain section. Sections from the center of the transplanted area at level III contained the highest number of Thy-1.21 cells, whereas their numbers decreased with increasing distance from the grafted region (Fig. 6). The Thy-1.21 cells often showed certain abnormalities characteristic of degenerating cells. Such cells either possessed swollen somata, so that they appeared larger than surrounding host cells, or, at the other extreme, they had dark, shrunken nuclei (not shown). Cells of the latter type were often observed in the vicinity of the operated area and close to the meninges, whereas swollen donor cells were found outside of this region. The specificity of the Thy-1.2 immunostaining was verified in several ways. First, in nontransplanted (shamoperated and intact) control animals, no Thy-1.21 cells were detected (Table 1). Second, sections from transplanted animals that were processed without the primary Thy-1.2-specific Ab were negative. Third, brain sections taken from 2-month-old C57BL/6J mice (positive for the Thy-1.2 antigen) were stained heavily, as expected. Finally, immunostaining for the cat-specific Thy-1 antigen indeed showed that the vast majority of the remodeled tissue belonged to the kitten host, whereas control sections from mouse tissue remained negative. Part B: Cell types contributing to the remodeling of the lesion Cell density. Photomicrographs that show examples of staining for different cell types (neurons, astrocytes, oligodendrocytes, endothelial cells, and donor cells) are provided in Figure 7, whereas quantitative data concerning the contribution of these cell types to the remodeling of the lesions are plotted in Figure 8. Intact kitten brains. In the individual graphs of Figure 8, the relative cell densities calculated for the different cell types in intact visual cortical area 18 of 3-month-old kittens are represented by white bars. CNP1 oligodendrocytes (670 6 80 per mm2) and NeuN1 nerve cells (640 6 70 per mm2) were the most numerous. The number of endothelial cells was 470 6 110 per mm2, whereas that of the GFAP1 astrocytes was 170 6 65 per mm2. Together, the overall cell density was 1,950 6 100 cells per mm2. The glial-neuronal index (the ratio of glial [GFAP1 and CNP1] to neuronal [NeuN1] cells) in this area was 0.76. Sham-operated brains. The density of cells around the persisting cavity varied with the distance from the cavity Fig. 4. Top: Camera lucida drawings of the dorsal portions of coronal sections from three transplanted kitten brains taken at anteroposterior levels I–V (see Fig. 1). Each column represents a kitten showing one of the three defined degrees of tissue remodeling observed within the operated region: Fully filled (1), partially filled (1/-) and persisting lesion (-). Arrowheads indicate the most evident differences in relation to the contralateral hemisphere. Bottom: Weight ratios (operated/intact [O/I] indices) of equivalent regions of the right (operated) and left (intact) hemispheres determined by weighing of camera lucida drawings (similar to those shown above). Bars represent means with their standard deviations in sham-operated (Sham) and transplanted (Tr) kitten brains with fully, partially, and unfilled cavities. Statistical significance at P # 0.05 (asterisk) was calculated with respect to the mean value of the O/I index of intact animals (thick horizontal line). XENOTRANSPLANTATION INTO KITTEN NEOCORTEX 101 walls at which the counts were made (see schematic drawing in Fig. 8). Whereas the density distal to the cavity (200–1,000 µm from the cavity walls) was not significantly different from the cell density in the intact animals (Fig. 8, horizontally hatched bars), at locations proximal to the cavity walls (within 100–200 µm), the density was generally lower compared with that of the remaining cortex. The only exception to this general picture was the density of endothelial cells, which was higher by almost 100% (830 6 170 cells per mm2 vs. 480 6 110 cells per mm2) proximal to the cavity walls. This increase in the number of endothelial cells near the cavity is in agreement with the presence of a dense meshwork of blood vessels originating in the meninges that formed around the cavity. Proximal to the cavity (Fig 8., vertically hatched bars), neurons were the sparsest cell type, with a density of only 25 6 10 cells per mm2. The densities of astrocytes and oligodendrocytes were also reduced to 48.5 6 49 cells per mm2 and 450 6 270 cells per mm2, respectively. The glial-neuronal index in this area was very high (19.9). However, because the distribution of the oligodendrocytes was irregular, with the cells often forming clusters of varying sizes, their density fluctuated along the border of the lesion from values as low as 180 cells per mm2 to values approaching those in intact controls. The overall consequence of these changes in the densities of neurons, astrocytes, and oligodendrocytes proximal to the persisting cavity was that their combined density dropped to around 524 cells per mm2, a value that was 35% of that found in the intact visual cortex of age-matched kittens. Transplanted brains. In the transplanted animals with fully closed cavities, the relative numbers of NeuN1, GFAP1, and CNP1 cells as well as of endothelial cells were comparable to those in the intact visual cortex (Fig. 8, cross-hatched bars). The measured values were (in cells per mm2) 650 6 120 for neurons, 160 6 50 for astrocytes, 700 6 90 for oligodendrocytes, and 530 6 50 for endothelial cells, which gave a total of 2,040 6 120 cells per mm2. The glial-neuronal index in the remodeled cortices of transplanted kittens with fully filled cavities was 0.76, a value identical to that calculated in area 18 of intact kittens. Two months after transplantation, as described above, the number of Thy-1.21 donor cells in the kitten brain was very low (see Fig. 6, counts per section). Quantitative counts made only from the center of the remodeled cortex at level III (see Fig. 8, schematic drawing) revealed that the number of Thy-1.21 cells fluctuated around a mean of 10 6 15 cells per mm2 (Fig. 8). The majority of these cells were dispersed near the cortical surface in the vicinity of Fig. 5. Detection of Thy-1 antigens 2 months after operation in grafted visual area 18 of kitten 9 (see Table 1), in which cortical remodeling led to complete filling of the lesion with neural tissue. A: Dorsal view of the brain showing the area of the initial lesion with a ‘‘plug’’ of nervous tissue now linking the lateral (top) and suprasylvian (bottom) gyri. Vertical bars indicate the level at which the sections in B–D were taken. Medial is up, lateral is down. B–D: Details of three adjacent frontal 20-µm sections from the level indicated in A. B: Hematoxylin-stained section. C: Section immunostained for the Thy-1.21 mouse graft tissue indicating that no compact mass of graft tissue forming a distinct boundary with the host brain was found. The relatively few donor cells dispersed in the kitten brain were visible only under high-power magnification, as shown in Figure 6. D: Section immunostained for the cat Thy-1 antigen showing that the vast majority of the tissue filling the cavity is of host origin. For B–D, medial is to the left, and lateral is to the right. Scale bar 5 1.5 mm. Fig. 6. Illustration of the main regions of dispersion of Thy-1.21 mouse cells throughout the cat brain 2 months after grafting. Coronal sections were taken at levels II–IV. Left: Camera lucida drawings showing brain regions where the majority of the immunostained graft cells were localized (solid triangles). Right: Photomicrographs of Thy-1.21 mouse cells observed in transplanted area 18 of the visual cortex (VC; counterstained with hematoxylin) and in subcortical regions, such as the superior colliculus (CS; without counterstaining) or the lateral geniculate body (CGL; without counterstaining). Bottom: Counts of the number of Thy-1.21 cells observed in transplanted kitten brains. Bars depict the means of the absolute numbers of cells (with standard deviations) counted in one section from each of the evaluated levels I–V. Scale bar 5 50 µm. Fig. 7. Photomicrographs showing examples of each of the evaluated cell types as revealed either by immunostaining (Thy-1.2, donor cells; neuron-specific nuclear protein A60 [NeuN], neurons; glial fibrillary acidic protein [GFAP], astrocytes; cyclic nucleotide phosphohydrolase [CNP], oligodendrocytes) or from hematoxylin staining (Hem; endothelial cells). Arrows indicate bromodeoxyuridine-positive (BrdU1) cells of a specific type that incorporated the DNA marker at the time of their last cell cycle close to the time of the BrdU injection. Scale bar 5 100 µm. 104 Fig. 8. Densities of neurons (NeuN1), astrocytes (GFAP1), oligodendrocytes (CNP1), donor cells (Thy-1.21), and endothelial cells (E) per mm2 in visual area 18 of 3-month-old kitten brains. Schematic drawing at top left illustrates the position and size of the region that was analyzed quantitatively in coronal sections at level III (d, dorsal; l, lateral; see also Fig. 1). Two regions were evaluated in shamoperated brains: An area proximal to the cavity (cross-hatched area in J. OUREDNIK ET AL. the schematic drawing located 100–200 µm of the cavity walls) and a distal area (hatched area located 200–1,000 µm of the cavity walls). In, intact animals; Sh-p, sham-operated animals (proximal to cavity); Sh-d, sham-operated animals (distal to cavity); Tr, transplanted kittens. Statistical significance (asterisk) at P # 0.05 was determined with respect to intact controls. XENOTRANSPLANTATION INTO KITTEN NEOCORTEX the broadened, unusual fiber tract described above. In other regions of the remodeled cortex, the distribution of the Thy-1.21 donor cells was more random. Postoperative proliferation of cells as revealed by incorporation of BrdU BrdU/Thy-1.21 cells. The density of Thy-1.21 donor cells that were also BrdU1 (2.7 6 1.9 cells per mm2) represented about 25% of their population (Figs. 7, 9). Most of the double-labeled cells were found in the transplanted area, whereas those Thy-1.21 cells that migrated farther into the host brain parenchyme were usually BrdU2. BrdU/NeuN1 cells. In intact animals, the number of BrdU/NeuN double-labeled cells was very low (about 1.3 6 0.5 cell per mm2; Figs. 7, 9). In sham controls, the density of such cells varied from 0.8 6 0.8 cells per mm2 in areas proximal to the persisting lesion to 1.9 6 2.1 per mm2 in more distal regions, but neither value was significantly different from that observed in intact kittens. In striking contrast to these numbers, the density of BrdU/NeuN double-labeled cells in transplanted animals was higher by almost a factor of ten, at 9.5 6 4 cells per mm2. BrdU/GFAP1 cells. In intact area 18, the density of BrdU/GFAP1 cells was only 4.3 6 1.2 cells per mm2 (Figs. 7, 9). In sham-operated animals, proximal to the border of the cavity, the density was very similar (4.2 6 3 cells per mm2); but, at more distal locations (at 200–1,000 µm from the cavity), the density of BrdU1 astrocytes was nearly five times higher, at 19 6 8 cells per mm2. In transplanted animals, their density (33.9 6 6 cells per mm2) was more than eightfold higher than that observed in intact controls. BrdU/CNP1 cells. The density of BrdU1 oligodendrocytes proximal to the cavity in sham-operated kittens (75 6 55 cells per mm2) was less than half of the value observed in intact controls (165 6 40); but, at more distal locations, the density was similar (Figs. 7, 9). In transplanted kittens, the density of BrdU/CNP1 cells (195 6 20 cells per mm2) was similar to that observed in intact kittens. BrdU-labeled endothelial cells. The density of endothelial cells that stained for BrdU in transplanted animals (155 6 45 cells per mm2) was similar to the value observed in the intact kitten neocortex (155 6 30 cells per mm2; Figs. 7, 9). In sham-operated animals, the density dropped to 40 6 30 cells per mm2 in regions proximal to the lesion and to 85 6 55 mm2 in more distal locations. DISCUSSION The key observation of this study was a remarkable host-mediated recovery of mechanically lesioned kitten visual cortex under the influence of a fetal murine xenograft. In 40% of cases, the original cavity was replaced completely with neural tissue, which, moreover, was not of mouse origin but, instead, was composed almost exclusively of host cells. The cytoarchitecture of the remodeled region resembled that of the intact cerebral cortex. No repair was ever observed in sham-operated animals, in which the lesioning invariably led to permanent damage and to the formation of a necrotic zone that extended 100–200 µm from the cavity borders and within which the densities of neuronal and glial cells were reduced. On the other hand, this zone was characterized by an unusually dense meshwork of blood vessels. The qualitative differences between sham-operated and transplanted cortices 105 were reflected by substantial quantitative differences in cell densities of different cell populations. In sham controls, the densities of most cell types decreased in the degenerating area, with neurons suffering the greatest reduction. Consequently, the glial-neuronal index (the ratio of glial [GFAP1 and CNP1] to neuronal cells) was very high in these animals (19.9). In striking contrast, in the remodeled cortices of transplanted kittens the index was identical to the value in intact control animals (0.76). These various observations suggest that neural transplantation may not rely solely on integration of the graft as a ‘‘patch’’ into the host brain, where it replaces more or less completely the missing tissue in structural and functional terms. Rather, it appears that the graft exerts at least two different effects. First, it seems to reduce both necrotic processes and scar formation, which normally would prevent or seriously impede any repair. In addition, our results suggest that the graft may also serve as a trigger and an initiator of regenerative processes in the recipient tissue that could not occur without its initial presence. The fate of the graft and the nature of the mechanisms of repair it may initiate are considered separately below. Fate of the grafted mouse cells in the kitten CNS Two months after transplantation, only very few donor cells had survived in the host brain. Apparently, the murine graft had been recognized as foreign and was gradually destroyed by the kitten immune system. Although we were always aware of a possible immune response of the host toward the xenograft (see, e.g., Finsen et al., 1991; Pakzaban and Isacson, 1994), we avoided the use of any immunosuppressive drugs, because we did not want to disturb in any way the interaction between graft and host. Such an interaction, although it was much weaker, was also tolerated in our earlier allografts in mice, in which graft tissue was also resorbed (Ourednik et al., 1993a,b; Ourednik and Ourednik, 1994) without any significant detriment to the regenerative process. Finally, experiments by Blakemore et al. (1995) showed that, when administration of immunosuppressive drugs was stopped in animals receiving demyelinating lesions and grafts of myelinating oligodendrocytes, myelination continued by host glia, whereas the graft cells were gradually resorbed. The small number of graft cells did not survive as a compact mass of tissue that could have integrated into the host at the site of the cavity. Instead, the mouse cells became dispersed in the cat brain, reaching regions at significant distances from the graft’s initial placement. Thus, although grafted cells were found mostly in the operated region, additional scattered donor cells were observed in ipsilateral subcortical areas and even in the contralateral hemisphere. A similar migration to the contralateral side was observed with motoneurons grafted into the striatum (Ruiz-Flandes et al., 1993) and with astrocytes that were placed originally into deep neocortical layers (Lund et al., 1993). The ability of grafted nerve and glial cells to migrate and to integrate into the neural network of the recipient has been described in the past for nerve and glial cells (see, e.g., Finsen and Zimmer, 1986; Privat et al., 1986; Goldberg and Bernstein, 1988; McConnell, 1988; Kawamura et al., 1988; Sieradzan and Vrbová, 1989; Demierre et al., 1990; Sotelo et al., 1990; Lund at al., 1993; Warrington et al., 1993; Gumpel, 1994; Lois and Alvarez-Buylla, 1994) for both short- and long-distance 106 J. OUREDNIK ET AL. Fig. 9. Densities of BrdU-stained cells per mm2 in doubleimmunostained sections: Neurons (BrdU/NeuN1), astrocytes (BrdU/ GFAP1), oligodendrocytes (BrdU/CNP1), and donor cells (BrdU/Thy1.21) and endothelial cells in hematoxylin-stained sections (BrdU/E). BrdU was administered to the operated animals every other day during the 12 days immediately following the operation, and the intact animals were injected at corresponding ages. Statistical significance (asterisk) at P # 0.05. For additional details, see the legend to Figure 8. migration. The latter seems to occur mainly after interspecies transplantations, whereas allogenic grafts tend to remain in the vicinity of the region of implantation (Wells et al., 1987). Björklund and coworkers (1982) suggested that, in the case of a xenograft with its stronger immunogenic effect, the capacity to migrate could be crucial for the donor cells in order to escape recognition and attack by immunocompetent cells at the side of a ruptured blood-brain barrier. XENOTRANSPLANTATION INTO KITTEN NEOCORTEX About 25% of the Thy-1.21 (donor) cells were labeled by BrdU. These cells apparently stopped division soon after transplantation, around the time of the BrdU injections, and they were observed mainly at the site of deposit. By contrast, in more distant brain regions, most graft cells were BrdU2, an observation that could mean either that the cells ceased division before the BrdU injections or that the label became diluted in subsequent divisions by newly synthesized, nonlabeled DNA. Because BrdU is available for only a relatively short time (20–40 minutes; Boswald et al., 1990), dilution of the label is the most likely interpretation, because dividing cell precursors have better migratory capabilities than more differentiated cells and, hence, a greater capacity to escape the host’s immune response (mainly in the area where the hematoencephalic barrier was damaged during the operation) and to infiltrate more remote areas of the kitten brain. The migration of the transplanted cells raises many questions. For example, it is not clear how the transplanted cells migrate in the host CNS, what guiding clues they utilize, and the degree to which they migrate to specific targets. We did observe that the grafted cells appeared not only in the operated region and in the various neural structures that form the central visual pathways but also at other sites throughout the brain. It is conceivable that structures like fiber tracts and blood vessels may facilitate cellular movement by serving as scaffolds for mechanical support and by providing molecular sites for attachment and orientation (see, e.g., Zhou et al., 1990; Jacque et al., 1991; for review, see Hynes and Lander, 1992). Cytoarchitecture and cell proliferation in intact vs. sham-operated and transplanted kitten brains Intact brains. Although there have been a number of previous measurements of the density of neurons in the cat visual cortex (see, e.g., Cragg, 1975; Beaulieu and Colonnier, 1983, 1985, 1987; O’Kusky, 1985; Takacs et al., 1992), it is difficult to compare these earlier values with ours because of substantial differences in the ages of the animals, differences in the preparation of the histological specimens that could change the correction factor for shrinkage of the tissue, and in our study the use of cell type-specific markers that were not available in the earlier studies. Indeed, the use of such markers allowed us to provide data on the relative densities of cell types other than neurons, which, to our knowledge, have not been reported previously in the kitten visual cortex. Moreover, because our aim was to compare the densities of various cell types in the different experimental conditions and not to provide definitive data on normal cell densities in the manner of earlier studies, we neither prepared the tissue nor conducted the measurements in a manner that makes meaningful comparisons with earlier studies. The intact kittens received BrdU injections at the same age (between postnatal weeks 5 and 7) as the shamoperated and transplanted kittens. The presence of BrdUlabeled cells implies that, during that particular period of time, many cells still were completing their last cell cycle. Even some mitotic figures were present at the time of perfusion (at the end of 3 months). Double labeling with BrdU and cell type-specific markers or hematoxylin staining revealed that the dividing cells were mostly oligodendrocytes and endothelial cells, whereas, for the most part, the formation of neurons had terminated (with some rare 107 exceptions). These observations were indicative of the fact that, at the particular time of observation, cell proliferation and tissue differentiation linked to vasculogenesis and myelination were still in progress. To our knowledge, quantitative documentations of the time table of nonneuronal cell formation in the cat brain are not available. The genesis of nerve cells in the cat primary visual cortex was described in detail by Luskin and Shatz (1985), who employed [3H]thymidine autoradiography. In agreement with findings in other mammalian brains, in which the period of nerve cell formation in the cerebral cortex has been shown to finish before birth (for reviews, see Korr, 1980; Jacobson, 1991), they demonstrated that proliferation of nerve cell precursors in the kitten visual cortex ceased around the time of birth. On the other hand, there are reports of rare production of single nerve cells during adulthood (see, e.g., Altman, 1962; Kaplan, 1981; Altman and Bayer, 1993). Indeed, we also observed a few postnatally formed cells, which were double labeled with NeuN- and BrdU-specific Abs (Figs. 7, 9). So far, the source and function of postnatally formed neurons are not clear. It was shown recently that the source of these cells may be either the persisting subventricular zone in the forebrain (Lois and Alvarez-Buylla, 1993; Luskin, 1993; Morshead et al., 1994; for reviews, see Gage et al., 1995; Svendsen and Rosser, 1995; Weiss et al., 1996; Stemple and Mahanthappa, 1997) or progenitor cells in the brain parenchyme (Kaplan, 1981, 1985; Okano et al., 1993). It is conceivable that this weak postnatal neurogenesis reflects a slow turn-over of the neuronal population evoked by microinjuries or plastic events accompanying learning and memory. Sham-operated brains. An injury of the CNS is usually followed by a set of reactions that appear to attenuate its structural and functional consequences and that begin with a recruitment of immunocompetent cells (Kerr and Bartlett, 1989; Sloan et al., 1991) and the production of glial (Billingsley et al., 1982) and endothelial (Broadwell et al., 1990) cells, which combine to form a glial scar (Berry et al., 1983). In agreement with this general finding, in the lesioned area, we observed a necrotic zone penetrated by a dense meshwork of blood vessels, which probably guaranteed a continuous supply of macrophages and other immunocompetent cells. From our data, it became evident that the vulnerability to injury is not the same for all brain cell populations. The number of cells proximal to the lesion decreased, with neurons and astrocytes showing the greatest reduction. In addition, the extent and nature of the observed proliferative activity depended on the distance from the walls of the cavity. Proximally, the numbers of BrdU-labeled oligodendrocytes and endothelial cells decreased compared with intact brains (Fig. 9). The substantial reduction of BrdU1 endothelial cells and the simultaneous increase of vascularization and, hence, the density of endothelial cells in this region (compare the graphs for endothelial cells in Figs. 8, 9) could well be attributed to a very active endothelial proliferation and a resulting dilution of the original amount of the BrdU label by newly synthesized, nonlabeled DNA. At more distal locations from the lesion, where cell density was similar to that of the intact neocortex, the numbers of double-labeled neurons and oligodendrocytes were not different from those observed in intact controls, whereas the density of BrdU1 astrocytes was higher. In harmony with recent reports (Huang and Lim, 1990; Skup et al., 1993) describing the rare formation of single 108 neurons in the vicinity of a damaged brain area, we also found individual BrdU1 neurons within 200–1,000 µm of the lesion. However, their number was the same as that found in intact controls, from which we conclude that the degenerative process was overriding any possible attempts to repair the damage. Transplanted animals. The observed variability in the degree of graft-induced closure of the original cavity and in the number of surviving graft cells cannot be explained readily. The regenerative process might be influenced by methodological factors (e.g., variability of the size of the graft at the moment of transplantation, possible damage of the white matter reached during the preparation of the cavity) as well as by unequal dispositions of the individual animals (e.g., degree of the immune response). The vast majority of the tissue that replaced the cavity in the successfully transplanted cortices, as we have shown with Thy-2 immunohistochemistry, was of host origin. The remodeling of the transplanted area was accompanied by quite good preservation of the neocortical cytoarchitecture. The cavity was filled with both neuronal and nonneuronal cells, and morphometric analysis of the representation of individual cell types and their packing densities showed no significant quantitative changes with respect to the tissue in equivalent brain regions of intact controls. No hyperplastic increase in density, like that of endothelial cells in the sham-operated animals, was observed for any specific cell population. The similarity in cell densities of intact and transplanted animals suggested a coordinated increase in all cell classes rather than multiplication and/or immigration of selected cell populations, which occurs in the case of glial cells during postinjury gliosis (for review, see Norenberg, 1994) or in endothelial cells during revascularization (Lawrence et al., 1984). In these animals, the formation of new tissue in the operated region was accompanied by intensification of postoperative cell proliferation. The present data are in agreement with other findings (Lundberg and Møllgård, 1979; Polezhaev et al., 1988; Ourednik et al., 1993a,b; Ourednik and Ourednik, 1994), which indicate that the host brain can be stimulated to cell proliferation by an embryonic neural graft. In the present case, the number of BrdU1 cells (which divided shortly after the operation, i.e., at the time of the BrdU injections) was increased in transplanted animals compared with that in the intact controls. On the other hand, the number of mitotic figures (reflecting actual proliferative activity that was occurring at the time of perfusion, i.e., at 3 months of age) was similar in both groups. Consequently, these data suggest that the graft-dependent activation of cell division was limited in duration and occurred shortly after transplantation. The number of postoperatively labeled dividing cells was variable, depending on the cell type. The density of BrdU/ NeuN1 cells increased nearly ten times, and the density of BrdU/GFAP1 cells increased about eight times. The number of surviving Thy-1.21 (donor) cells in transplanted kitten brains, as mentioned above, was very low, and only 25% of these cells were BrdU1. The overall number of BrdU1 neurons (9.5 6 4 cells per mm2) was significantly higher than the number of BrdU1 donor cells (2.7 6 1.9). After subtraction of the latter from the former, the number of postoperatively dividing neurons derived from the kit- J. OUREDNIK ET AL. ten host can be estimated to have increased from about three in the intact animals to six or seven nerve cells per mm2 in the transplanted animals (see also Fig. 9). Although the cytoarchitecture of transplanted and intact kitten brains and their cell densities were similar, the number of double-labeled (BrdU1) cells per mm2 varied among the different cell types. However, this variability may not necessarily reflect only differences in the extent of cell proliferation among the cell groups but could as well be explained, in part or in whole, by differences in the capacities of surrounding nondividing (BrdU2) cells to immigrate into the transplanted region. Moreover, it should be kept in mind that very intensive cell proliferation can lead to the dilution of the label below the threshold of the immunodetection. To distinguish between these possibilities, it will be necessary to document the total length of cell formation and its exact time table. To accomplish the former, animals would have to be injected repeatedly with BrdU for a longer time; whereas, to address the latter different groups of animals could be administered a single BrdU injection at various postoperative intervals, and the already differentiated tissue could be analyzed at a common, defined time point. Possible mechanisms underlying the closure of the transplanted cavity It has been shown recently that a neural graft may exert both protective and trophic influences on degenerative processes in the host CNS (Bregman and Reier, 1986; Kesslak et al., 1986a,b; Sofroniew et al., 1986; Cunningham et al., 1987; Sørensen et al., 1989, 1990; Dunnett, 1990; Gage and Fischer, 1991) and may reduce formation of a glial scar, which normally would hinder or prevent repair (Krüger et al., 1986; Houlé and Reier, 1988). Along the same lines, Valoušková and Gálik (1995) observed a positive influence of a graft on the closure of a nontransplanted cortical cavity prepared in the cortex contralateral to the transplanted cavity. In an injured brain region, reactive astrocytes and microglia are known to produce a plethora of factors, some of which may induce pathological changes, but others may also support recovery (Giulian, 1993; Giulian et al., 1993). It has been documented (Altar et al., 1992) that administration of exogenous factors (in this case, nerve growth factor) may enhance the neuroprotective effect of the endogenously elevated levels of the same molecule. Therefore, it is conceivable that, in our experiments, we counteracted injury-induced necrosis and stimulated recovery by supplying the site of injury with a mixture of trophic factors present in the fetal tissue. These factors, together with the endogenous ones, would shift the balance toward a host-based recovery and would override the influence of molecules interfering with adult CNS plasticity. The possible presence of graft-derived trophic factors that might assist the host to recover from brain injury receives support from other studies of neurotransplantation (Stein et al., 1985; Bregman and Reier, 1986; Kesslak et al., 1986a,b; Cunningham et al., 1987; Valoušková and Gálik, 1995) as well as from important recent reports of multipotent stem cells persisting in the subependymal layer of the adult mammalian brain (see, e.g., Lois and Alvarez-Buylla, 1993, 1994; Morshead et al., 1994; Weiss et al., 1996), which can be stimulated to proliferation and differentiation by neurotrophins, such as EGF (Reynolds at al., 1992) or bFGF (Gritti et al., 1996). Along these lines, XENOTRANSPLANTATION INTO KITTEN NEOCORTEX Craig et al. (1996) showed that infusion of EGF into adult mouse forebrain for 6 days resulted in a dramatic increase in the proliferation and total numbers of subependymal cells and induced their migration away from the walls of the lateral ventricle into the adjacent cortical, striatal, and septal parenchyme. Seven weeks after EGF withdrawal, 3% of the newly formed cells were of neuronal origin. This last report is an example of how, in vivo, a single signaling molecule can exert a dramatic effect on progenitor and stem cells in the mammalian CNS. Several mechanisms could account for successful remodeling of the transplanted cavity. At one extreme, the graft could induce a reactivation of host cell proliferation, and the lesion could be filled by resulting new cell populations. At the other extreme, discrete cell populations within the host tissue could be stimulated to migrate toward the area of repair, or a morphogenic movement of whole blocks of host tissue could lead to a convergence of the cavityforming walls toward each other in the parenchyme (a ‘‘zipping-up’’ effect). The absence of a sulcus between the suprasylvian and medial gyri in the operated area provides partial evidence for the latter process. These unusual gyral patterns clearly exceed the reported interindividual and interhemispheric variability of gyrification described earlier in the intact cat brain (Webster, 1981). However, the hyperplastic thickening of the parenchyme in this region, which, in two cases, led to the appearance of protruding tissue rising above the surrounding brain surface and having the appearance of a small extra gyrus, could just as easily be reconciled with the graft’s activation of reconstruction of the operated area through immigration and/or proliferation of cells originating elsewhere in the host brain parenchyme. Both of these processes would imply a communication of graft and host cells through both direct contact and the release of soluble substances by the graft cells during the early postoperative period, before their rejection by the host’s immune system. Whether the postoperative activation of cell proliferation was sufficient by itself to compensate for the structural defects caused by the injury is uncertain. However, it seems more likely that additional mechanisms, such as immigration of cells from the surrounding parenchyme, make major contributions to the process of regeneration. CONCLUSIONS The presented data provide evidence that murine fetal neural tissue has the potential to stimulate regeneration by injured kitten neocortex. This unexplored role of a neural graft in CNS repair will require further investigation. Thus, the description of early postoperative stages and the analysis of the degree of damage and remodeling in subcortical visual centers as well as experiments assessing possible functional recovery are the next logical steps in the study. The resulting morphological and physiological data will have to be complemented by molecular and biochemical analyses aimed at the characterization of the factor(s) that mediate the graft/host interactions necessary for the observed regeneration. ACKNOWLEDGMENTS We thank Drs. Mullen and Bartlett for kindly providing the anti-NeuN and anti-cat Thy-1 antibodies; Drs. P.E. Neumann and A. Fine for their constructive comments on 109 the paper; Dr. C.J. Beaver, P. Ritchie, and V. Baker for assistance with the operations; and V.R. Fraser and R. Burgess for their dedicated care of the mouse and cat colonies. LITERATURE CITED Altar, C.A., M. Armanini, M. Dugich-Djordjevic, G.L. Bennett, R. Williams, S. Feinglass, V. Anicetti, D. Sinicropi, and C. Bakhit (1992) Recovery of cholinergic phenotype in the injured rat neostriatum: Roles for endogenous and exogenous nerve growth factor. J. Neurochem. 59:2167– 2177. Altman, J. (1962) Are new neurons formed in the brains of adult mammals? Science 135:1127–1128. Altman, J. and S. Bayer (1993) Are new neurons formed in the brains of adult mammals? A progress report, 1962–1992. In A.C. Cuello (ed): Neuronal Cell Death and Repair. New York: Elsevier, pp. 203–225. Beaulieu, C. and M. Colonnier (1983) The number of neurons in the different laminae of the binocular and monocular regions of area 17 in the cat. J. Comp. Neurol. 217:337–344. Beaulieu, C. and M. Colonnier (1985) A comparison of the number of area 17, 18 and posteromedial suprasylvian (PMLS) area in the cat. Brain Res. 339:166–170. Beaulieu, C. and M. Colonnier (1987) Effect of richness of the environment on the cat visual cortex. J. Comp. Neurol. 226:478–494. Berry, M. (1989) Transplantation and regeneration of neural tissue in the central nervous system. Curr. Opin. Neurol. Neurosurg. 2:946–952. Berry, M., W.L. Maxwell, A. Logan, A. Mathewson, P. McConnell, D.E. Ashhurst, and G.H. Thomas (1983) Deposition of scar tissue in the central nervous system. Trauma 3:31–53. Billingsley, M.L., J.A. Straw, and H.G. Mandel (1982) Glial DNA synthesis and cell proliferation in the lesioned frontal cortex. Brain Res. 247:325– 334. Björklund, A., U. Stenevi, S.B. Dunnett, and F.H. Gage (1982) Cross-species neural grafting in a at model of Parkinson’s disease. Nature 298:652– 654. Blakemore, W.F., A.J. Crang, R.J. Franklin, K. Tang, and S. Ryder (1995) Glial cell transplants that are subsequently rejected can be used to influence regeneration of glial cell environments in the CNS. Glia 13:79–91. Boswald, M., S. Harasim, and B. Maurer-Schultze (1990) Tracer dose and availability of thymidine and bromodeoxyuridine: Application of bromodeoxyuridine in cell kinetic studies. Cell Tissue Kinet. 23:169–181. Bregman, B. and P.J. Reier (1986) Neural tissue transplants rescue axotomized rubrospinal cells from retrograde death. J. Comp. Neurol. 224:86–95. Broadwell, R.D., H.M. Charlton, P. Ebert, W.F. Hickey, J.C. Villegas, and A.A. Wolf (1990) Angiogenesis and the blood-brain barrier in solid and dissociated cell grafts within the CNS. Progr. Brain Res. 82:95–101. Canadian Council of Animal Care (1993/4) E.D. Olfert (ed): Guide to the Care and Use of Experimental Animals, Vol. 2. Ottawa, Ontario: Brada Printing Services, Inc. Charlton, H.M., A. Barclay, and A.F. Williams (1983) Detection of neuronal tissue from brain grafts with anti-Thy-1.1 antibody. Nature 305:825– 827. Cragg, B.G. (1975) The development of synapses in the visual system of the cat. J. Comp. Neurol. 160:147–166. Craig, C.G., V. Tropepe, C.M. Morshead, B.A. Reynolds, and S. Weiss (1996) In vivo growth factor expansion of endogenous subependymal neural precursor cell populations in the adult mouse brain. J. Neurosci. 16:2649–2658. Cunningham, T.J., C.B. Sutilla, and F. Haun (1987) Trophic effect of transplants following damage to the cerebral cortex. Ann. NY Acad. Sci. 495:153–168. Daw, N.W. (1995) Visual Development. New York: Plenum Press. Demierre, B., J.-C Martinou, and A.C. Kato (1990) Embryonic motoneurons grafted into the adult CNS can differentiate and migrate. Brain Res. 510:355–359. Dunnett, S.B. (1990) Neural transplantation in animal models of dementia. Eur. J. Neurosci. 2:657–587. Dunnett, S. and A. Björklund (1994) Functional Neuronal Transplantation. New York: Raven Press. Dyck, R.H., L.J. Van-Eldik, and M.S. Cynader (1993) Immunohistochemical localization of the S-100 beta protein in postnatal cat visual cortex: 110 Spatial and temporal patterns of expression in cortical and subcortical glia. Dev. Brain Res. 72:181–192. ffrench-Constant, C. and G.A. Mathews (1994) Brain repair lessons from developmental biology. J. Neurol. 242:S29–S32. Finsen, B. and J. Zimmer (1986) Timm staining of hippocampal nerve cell bodies in the Koyoto rat. A cell marker in allo- and xenografting of rat and mouse brain tissue, revealing neuronal migration. Dev. Brain Res. 29:51–59. Finsen, B., T. Sørensen, B. Castellano, E.B. Pedersen, and J. Zimmer (1991) Leucocyte infiltration and glial reactions in xenografts of mouse brain tissue undergoing rejection in the adult rat brain. J. Neuroimmunol. 32:159–183. Fischer, L.J. and F.H. Gage (1993) Grafting in the mammalian central nervous system. Physiol. Rev. 73:583–616. Gage, F.H. and L.J. Fisher (1991) Intracerebral grafting: A tool for the neurobiologist. Neuron 6:1–12. Gage, F.H., J. Ray, and L.J. Fisher (1995) Isolation, characterization, and use of stem cells from the CNS. Annu. Rev. Neurosci. 18:159–192. Giulian, D. (1993) Reactive glia as rivals in regulating neuronal survival. Glia 7:102–110. Giulian, D., K. Vaca, and M. Corpuz (1993) Brain glia release factors with opposing actions upon neuronal survival. J. Neurosci. 13:29–37. Goldberg, W.J. and J.J. Bernstein (1988) Fetal cortical astrocytes migrate from cortical homografts throughout the host brain and over the glia limitans. J. Neurosci. Res. 20:38–45. Gritti, A., L. Cova, E.A. Parati, R. Galli, and A.L. Vescovi (1995) Basic fibroblast growth factor supports the proliferation of epidermal growth factor-generated neuronal precursor cells of the adult mouse CNS. Neurosci. Lett. 185:151–154. Gritti, A., E.A. Parati, L. Cova, P. Frolichsthal, R. Galli, E. Wanke, L. Faravelli, D.J. Morassutti, F. Roisen, D.D. Nickel, and A.L. Vescovi (1996) Multipotential stem cells from the adult mouse brain proliferate and self–renew in response to basic fibroblast growth factor. J. Neurosci. 16:1091–1100. Gumpel, M. (1994) Transplanted transgenically marked oligodendrocytes survive, migrate and myelinate in the normal mouse brain as they do in the shiverer mouse brain. Eur. J. Neurosci. 6:814–824. Houlé, J.D. and P.J. Reier (1988) Transplantation of fetal spinal cord into the chronically injured adult rat spinal cord. J. Comp. Neurol. 269:535– 547. Huang, L. and R. Lim (1990) Identification of injury-induced mitotic cells in adult rat cerebral cortex by neuron-specific markers. Dev. Brain Res. 51:123–127. Hynes, R.O. and A.D. Lander (1992) Contact and adhesive specificities in the associations, migrations, and targeting of cells and axons. Cell 68:303–322. Jacobson, M. (1991) Developmental Neurobiology, 3rd Ed. New York: Plenum Press. Jacque, C., I. Suard, P.V. Collins, and N. Baumann (1991) Migration patterns of donor astrocytes after reciprocal striatum-cerebellum transplantation into newborn hosts. J. Neurosci. Res. 29:421–428. Kaplan, M.S. (1981) Neurogenesis in the 3-month-old rat visual cortex. J. Comp. Neurol. 195:323–338. Kaplan, M.S. (1985) Formation and turnover of neurons in young and senescent animals, an electron microscopic and morphometric analysis. Ann. N.Y. Acad. Sci. 457:173–192. Kaufman, M.H. (1992) The Atlas of Mouse Development. London: Academic Press, Inc. Kawamura, K., T. Nanami, Y. Kikuchi, and A. Kitakami (1988) Grafted granule and Purkinje cells can migrate into the mature cerebellum of normal adult rats. Exp. Brain Res. 70:477–484. Kerr, R.S.C. and P.F. Bartlett (1989) The immune response to intraparenchymal fetal CNS transplants. Transplant. Proc. 21:3166–3168. Kesslak, J.P., M. Nieto-Sampedro, J. Globus, and C.W. Cotman (1986a) Transplants of purified astrocytes promote behavioral recovery after frontal cortex ablation. Exp. Neurol. 92:377–390. Kesslak, J.P., L. Brown, C. Steichen, and C.W. Cotman (1986b) Adult and embryonic frontal cortex transplants after frontal cortex ablation enhance recovery on a reinforced alternation task. Exp. Neurol. 94:615– 626. Kolb, B. (1987) Recovery from early cortical damage in rats. I. Differential behavioral and anatomical effects of frontal lesions at different ages of neural maturation. Behav. Brain Res. 25:205–220. J. OUREDNIK ET AL. Kolb, B. and R. Gibb (1991) Environmental enrichment and cortical injury: Behavioral and anatomical consequences of frontal cortex lesions. Cerebral Cortex 1:189–198. Kolb, B., R. Gibb, and D. Van der Kooy (1994) Neonatal frontal cortical lesions in rats alter cortical structure and connectivity. Brain Res. 645:85–97. Korr, H. (1980) Proliferation of different cell types in the brain. Ado. Anat. Embryol. Cell Biol. 61:1–72. Krüger, S., J. Sievers, C. Hansen, M. Sadler, and M. Berry (1986) Three morphologically distinct types of interface develop between adult host and fetal brain transplants: Implications for scar formation in the adult central nervous system. J. Comp. Neurol. 249:103–116. Lawrence, J.M., S.K. Huang, and G. Raisman (1984) Vascular and astrocytic reactions during establishment of hippocampal transplants in adult host brain. Neuroscience 12:745–760. Logan, A., J.J. Oliver, and M. Berry (1994) Growth factors in CNS repair and regeneration. Progr. Growth Factor Res. 5:379–405. Lois, C. and A. Alvarez-Buylla (1993) Proliferating subventricular zone cells in the adult mammalian forebrain can differentiate into neurons and glia. Proc. Natl. Acad. Sci. USA 90:2074–2077. Lois, C. and A. Alvarez-Buylla (1994) Long-distance neuronal migration in the adult mammalian brain. Science 264:1145–1148. Lund, R.D., H.-F. Zhou, and K.T. Yee (1993) The migration of astrocytes after implantation to immature brains. Int. J. Neurosci. 11:595–601. Lundberg, J.J. and K. Møllgård (1979) Mitotic activity in adult rat brain induced by implantation of pieces of fetal rat brain and liver. Neurosci. Lett. 13:265–270. Luskin, M.B. (1993) Restricted proliferation and migration of postnatally generated neurons derived from the forebrain subventricular zone. Neuron 11:173–189. Luskin, M.B. and C.J. Shatz (1985) Neurogenesis of the cat’s primary visual cortex. J. Comp. Neurol. 242:611–631. McConnell, S.K. (1988) Fates of visual cortical neurons in the ferret after isochronic and heterochronic transplantation. J. Neurosci. 8:945–974. McKeon, R.J., B.P. Vietje, and J. Wells (1989) Integrations between donor and host tissue following cross-species septohippocampal transplants. Exp. Neurol. 103:213–221. Mitchell, D.E. and B. Timney (1984) Postnatal development of function in the mammalian visual system. In I. Darian-Smith (ed): Handbook of Physiology, Section I: The Nervous System, Vol. 3, Part 1. Sensory Processes. Bethesda: American Physiological Society, pp. 507–555. Morris, R. (1991) Immunoperoxidase staining of gene products in cultured cells using monoclonal antibodies. Methods Mol. Biol. 7:339–359. Morris, R. and P. Grosveld (1989) Expression of Thy-1 in the nervous system of the rat and mouse. In A.E. Reif and M. Schlesinger (eds): Cell Surface Antigens. New York: Marcel Dekker Inc., pp. 121–147. Morshead, C.M., B.A. Reynolds, C.G. Craig, M.W. McBurney, W.A. Staines, D. Morassutti, S. Weiss, and D. Van der Kooy (1994) Neural stem cells in the adult mammalian forebrain: A relatively quiescent subpopulation of subependymal cells. Neuron 13:1071–1082. Mullen, R.J., C.R. Buck, and A.M. Smith (1992) NeuN, a neuronal specific nuclear protein in vertebrates. Development 116:201–211. Müller, C.M. and J. Best (1989) Ocular dominance plasticity in adult cat visual cortex after transplantation of cultured astrocytes. Nature 342:427–430. Norenberg, M.D. (1994) Astrocyte responses to CNS injury. J. Neuropathol. Exp. Neurol. 53:213–220. Okano, H.J., D.W. Pfaff, and R.B. Gibbs (1993) RB and Cdc2 expression in brain: Correlation with 3H-thymidine incorporation and neurogenesis. J. Neurosci. 13:2930–2938. O’Kusky J. (1985) Synapse elimination in the developing visual cortex: A morphometric analysis in normal and dark-reared cats. Dev. Brain Res. 22:81–91. Ourednik, J., W. Ourednik, and H. Van der Loos (1993a) Do foetal neural grafts induce repair by the injured juvenile neocortex? Neuroreport 5:133–136. Ourednik, J., W. Ourednik, H. Van der Loos, and B.M. Riederer (1993b) Do embryonic neural grafts induce repair by the injured juvenile neocortex? Soc. Neurosci. Abstr. 19:623.9. Ourednik, W. and J. Ourednik (1994) Newly formed host cells in grafted juvenile neocortex express neuron-specific marker proteins. Neuroreport 5:1073–1076. XENOTRANSPLANTATION INTO KITTEN NEOCORTEX Pakzaban, P. and O. Isacson (1994) Neural xenotransplantation: Reconstruction of neuronal circuitry across species barriers. Neuroscience 62:989– 1001. Polezhaev, L.V., M.A. Alexandrova, and V.N. Kleshchinov (1988) DNA synthesis and mitotic division of cortical neurons in adult rats upon intrabrain transplantation of embryonic nervous tissue. J. Hirnforsch 29:673–681. Price, D.J., J.M.R. Ferrer, C. Blakemore, and N. Kato (1994) Postnatal development and plasticity of corticocortical projections from area 17 to area 18 in the cat’s visual cortex. J. Neurosci. 14:2747–2762. Privat, A., H. Mansour, A. Pavy, M. Geffard, and F. Sandillon (1986) Transplantation of dissociated foetal serotonin neurons into the transected spinal cord of adult rats. Neurosci. Lett. 66:61–66. Reynolds, B.A. and S. Weiss (1992) Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 255:1707–1710. Reynolds, B.A., W. Tetzlaff, and S. Weiss (1992) A multipotent EGFresponsive striatal embryonic progenitor cell produces neurons and astrocytes. J. Neurosci. 12:4565–4574. Richards, L.J., T.J. Kilpatrick, and P.F. Bartlett (1992) De novo generation of neuronal cells from the adult mouse brain. Proc. Natl. Acad. Sci. USA 89:8591–8595. Ruiz-Flandes, P., B. Demierre, L. Mattenberger, and A.C. Kato (1993) Migration of neurons after transplantation. Migration of purified embryonic motoneurons grafted into adult mouse CNS. Int. J. Dev. Neurosci. 11:525–533. Saleh, M. and P.F. Bartlett (1989) Isolation of a Thy-1-like glycoprotein from cat brain: Distribution in retina and brain defined by monoclonal antibodies. J. Neurosci. 23:152–161. Shatz, C.J. and M.B. Luskin (1986) The relationship between the geniculocortical afferents and their cortical target cells during development of the cat’s primary visual cortex. J. Neurosci. 6:3655–3668. Sieradzan, K. and G. Vrbová (1989) Replacement of missing motoneurons by embryonic grafts in the rat spinal cord. Neuroscience 31:115–130. Skup, M., D.G. Herrera, A. Ribeiro-da-Silva, and A.C. Cuello (1993) Neurogenesis in the adult rat hippocampus: Lesion stimulated phenomenon? In A.C. Cuello (ed): Neuronal Cell Death and Repair. New York: Elsevier, pp. 267–281. Sloan, D.J., M.J. Wood, and H.M. Charlton (1991) The immune response to intracerebral neural grafts. Trends Neurosci. 14:341–346. Snider, R.S. and W.T. Niemer (1961) A Stereotaxic Atlas of the Cat Brain. Chicago: The University Press. Society for Neuroscience (1991) Handbook for the Use of Animals in Neuroscience Research. Washington, DC: Society for Neuroscience. Sofroniew, M.V., O. Isacson, and A. Björklund (1986) Cortical grafts prevent atrophy of cholinergic basal nucleus neurons induced by excitotoxic cortical damage. Brain Res. 378:409–415. Srensen, J.C., J. Zimmer, and A.J. Castro (1989) Fetal cortical transplants reduce thalamic atrophy induced by frontal cortical lesions in newborn rats. Neurosci. Lett. 98:33–38. Srensen, J.C., A.J. Castro, E.J. Neafsey, and J. Zimmer (1990) Reduction of postlesion atrophy by transplants of fetal cerebral tissue: Hosttransplant nerve connections and trophic factors. In A. Björklund, A.J. Aguayo, and D. Ottoson (eds): Brain Repair, Vol. 56: Wenner-Gren Center International Symposium Series. New York: M-Stockton Press, pp. 122–144. Sotelo, C., R.M. Alvarado-Mallart, R. Gardette, and F. Crapel (1990) Fate of grafted embryonic Purkinje cells in the cerebellum of the adult ‘‘Purkinje cell degeneration’’ mutant mouse. I. Development of reciprocal graft-host interactions. J. Comp. Neurol. 295:165–167. Stein, D.G., R. Labbe, M.J. Attella, and H.A. Rakowsky (1985) Fetal brain tissue transplants reduce visual deficits in adult rats with bilateral lesions of the occipital cortex. Behav. Neural Biol. 44:266–277. 111 Stemple, D.L. and N.K. Mahanthappa (1997) Neural stem cells are blasting off. Neuron 18:1–4. Stevens, A. (1990) The hematoxylins. In J.D. Bancroft and A. Stevens (eds): Theory and Practice of Histological Techniques, 3rd Ed. Edinburg: Churchill Livingstone, pp. 107–118. Svendsen, C.N. and A.E. Rosser (1995) Neurones from stem cells. TINS 18:465–467. Takacs, J., P. Saillour, M. Imbert, M. Bogner, and J. Hamori (1992) Effect of dark rearing on the volume of visual cortex (areas 17 and 18) and number of visual cortical cells in young kittens. J. Neurosci. Res. 32:449–459. Takahashi, T., R.S. Nowakowski, and V.S. Caviness, Jr. (1992) BUdR as an S-phase marker for quantitative studies of cytokinetic behaviour in the murine cerebral ventricular zone. J. Neurocytol. 21:185–197. Targett, M.P. and W.F. Blakemore (1994) The use of xenografting to evaluate the remyelinating potential of glial cell cultures. Eye 8:238– 244. Tusa, R.J., L.A. Palmer, and A.C. Rosenquis (1981) Multiple cortical visual areas: Visual field topography in the cat. In C.N. Woolsey (ed): Cortical Sensory Organisation. Multiple Visual Areas, Vol. 2. Clifton, NJ: Humana Press, pp. 1–31. Valoušková, V. and J. Gálik (1995) Unilateral grafting of fetal neocortex into a cortical cavity improves healing of a symmetric lesion in the contralateral cortex of adult rats. Neurosci. Lett. 186:103–106. Varon, S. and J.M. Conner (1994) Nerve growth factor in CNS repair. J. Neurotrauma 11:473–486. Vescovi, A.L., B.A. Reynolds, D.D. Fraser, and S. Weiss (1993) bFGF regulates the proliferative fate of unipotent (neuronal) and bipotent (neuronal/astroglial) EGF-generated CNS progenitor cells. Neuron 11:951–966. Villablanca, J.R., B.L. Shook, D.A. Hovda, and R.L. Sutton (1990) Transplantation of fetal frontal cortex onto degenerating thalamus of cats and kittens. Dev. Neurosci. 12:1–10. Von Bertalanffy, F.D. and K.P. Nagy (1962) Fluorescence microscopy and photomicrography with acridine orange. Med. Radiogr. Photogr. 38: 82–52. Warrington, A.E., E. Barbarese, and S.E. Pfeifer (1993) Differential myelinogenic capacity of specific developmental stages of the oligodendrocyte lineage upon transplantation into hypomyelinating hosts. J. Neurosci. Res. 34:1–13. Webster, W.G. (1981) Morphological asymmetries of the cat brain. Brain Behav. Evol. 18:72–79. Weiss, S., B.A. Reynolds, A.L. Vescovi, C. Morshead, C. Craig, and D. Van der Koy (1996) Is there a neural stem cell in the mammalian forebrain? TINS 19:387–393. Wells, J., B.P. Vietje, and D.G. Wells (1987) Migration and differentiation of xenogeneic and homogeneic brain cells transplanted into the adult rat hippocampus. Ann. N.Y. Acad. Sci. 495:804–806. Yinon, U. and S. Gelerstein (1991) Isochronic transplantation of neonatal grafts in the visual cortex of cats: Responsiveness, ocular dominance and specificity of cortical cells to visual stimulation. Exp. Brain Res. 87:181–192. Yinon, U., R. Shemesh, H. Arda, G. Dobin, and P.P. Jaros (1993) Physiological studies in deafferented visual cortex cells of cats following transplantation of fetal xenografts from the rat’s cortex. Exp. Neurol. 122:335– 341. Zhou, C.-F., G. Raisman, and R.J. Morris (1985) Specific patterns of fibre outgrowth from transplants to host mice hippocampi, shown immunohistochemically by the use of allelic forms of Thy-1. Neuroscience 16:819–833. Zhou, H.F., L.H.-C. Lee, and R.D. Lund (1990) Timing and pattern of astrocytic migration from xenogeneic transplants of the cortex and corpus callosum. J. Comp. Neurol. 292:320–330.