Survival of transplanted neural progenitor cells enhanced ... · the tumor cell suspension at a point 2.0 mm rostral to bregma, 2.0 mm to the right of the midline, and 3.0 mm ventral

HE discovery of NSCs within the adult CNS has pro-voked an intense and rapidly growing interest in thepotential use of such cells for tissue repair and gene

product delivery.2,4,6,10,28–31,38 The interest was generated be-cause of two important features of NSCs. First, they aremultipotent and can therefore generate neurons, oligoden-drocytes, and astrocytes to replace damaged tissue. Second,they are capable of self-renewal so that they can be contin-ually expanded as a source of neurons and glia. Althoughthere are a number of existing reports of the use of trans-planted NSCs or their descendants (NPCs), authors of moststudies have used either embryonic or immortalized cells.

In the present study we investigate the behavior of adultrat–derived, nontransformed NPCs when implanted into al-logeneic hosts. We presume that such cells will also easilyimplant and integrate into the brain and even track growingtumors. There have been only a few previous studies oftransplantation of this form.24,26 We also investigate the im-pact of various mechanical factors (such as route and meth-od of administration) and biological factors (for exam-ple, internal milieu of host and graft tissue) on cell survival.Defining these factors and their impact is critical to the suc-cessful clinical use of cell therapy. Despite numerous re-ports of the amenability of the CNS to transplanted cells, noclinical trial investigating feasibility and efficacy of stemcell transplantation into the human brain has been reported.In this report, we describe the impact of cell type, deliverymethod, delivery site, and immunological status on the sur-vival and integration of neural progenitors after transplan-tation. In addition, we report that brain irradiation beforetransplantation can enhance NPC integration.

J. Neurosurg. / Volume 107 / August, 2007

J Neurosurg 107:383–391, 2007

Survival of transplanted neural progenitor cellsenhanced by brain irradiation

AJAY NIRANJAN, M.B.B.S., M.S., M.CH.,1 WENDY FELLOWS, PH.D.,1

WILLIAM STAUFFER, B.S.,1 EDWARD A. BURTON, M.D., D.PHIL.,2 CHANG-SOOK HONG, PH.D.,2

L. DADE LUNSFORD, M.D.,1 DOUGLAS KONDZIOLKA, M.D.,1 JOSEPH C. GLORIOSO, PH.D.,2

AND GLENN T. GOBBEL, D.V.M., PH.D.1

Departments of 1Neurological Surgery and 2Molecular Genetics and Biochemistry, University ofPittsburgh, Pennsylvania

Object. Authors of previous studies have reported that adult transplanted neural progenitor cells (NPCs) are suitablefor brain cell replacement or gene delivery. In this study, the authors evaluated survival and integration of adult rat–derived NPCs after transplantation and explored the potential impact on transplant survival of various mechanical andbiological factors of clinical importance.

Methods. Adult female Fischer 344 rats were used both as a source and recipient of transplanted NPCs. Both 9L andRG2 rat glioma cells were used to generate in vivo brain tumor models. On the 5th day after tumor implantation, NPCsexpressing green fluorescent protein (GFP) were administered either intravenously (3.5 3 107 cells) or by stereotacticinjection (1 3 104–1 3 106 cells) into normal or tumor-bearing brain. The authors evaluated the effect of delivery meth-od (sharp compared with blunt needles, normal compared with zero-volume needles, phosphate-buffered saline com-pared with medium as vehicle), delivery sites (intravenous compared with intratumoral compared with intraparenchy-mal), and pretreatment with an immunosuppressive agent (cyclosporin) or brain irradiation (20–40 Gy) on survival andintegration of transplanted NPCs.

Results. Very few cells survived when less than 105 cells were transplanted. When 105 cells or more were trans-planted, only previously administered brain irradiation significantly affected survival and integration of NPCs. Al-though GFP-containing NPCs could be readily detected 1 day after injection, few cells survived 4 days to 1 week unlesspreceded by whole-brain radiation (20 or 40 Gy in a single fraction), which increased the number of GFP-containingNPCs within the tissue more than fivefold.

Conclusions. The authors’ findings indicate that most NPCs, including those from a syngeneic autologous source,do not survive at the site of implantation, but that brain irradiation can facilitate subsequent survival in both normal andtumor-bearing brain. An understanding of the mechanisms of this effect could lead to improved survival and clinicalutility of transplanted NPCs. (DOI: 10.3171/JNS-07/08/0383)

KEY WORDS • neural progenitor cell • neural stem cell • radiation therapy • rat

T

383

Abbreviations used in this paper: CNS = central nervous system;GFP = green fluorescent protein; HBSS = Hanks balanced salt solu-tion; MASC = multipotent astrocytic stem cell; NPC = neural pro-genitor cell; NSC = neural stem cell; PBS = phosphate-bufferedsaline; SEZ = subependymal zone.

Materials and Methods

Animal Preparation

All in vivo experiments were performed in adult female Fischer344 rats, which also served as the source of rat NPCs. A total of 85rats was used as hosts for NPC transplantation. In addition, C57BL/6mice were used as the source of mouse NPCs, which were used insome of our initial studies. All animals were housed within the De-partment of Laboratory Animal Resources at the University of Pitts-burgh Medical Center, and all experiments were conducted in ac-cordance with the rules of the Institutional Animal Care and UseCommittee.

Brain Tumor Model

Both 9L and RG2 rat glioma cells were used in tumor experi-ments. The RG2 cells were grown in medium consisting of RPMI1640 containing 5% fetal bovine serum and 2 mM glutamine. The9L cells were grown in Dulbecco modified Eagle medium with 5%fetal bovine serum. Penicillin (100 IU/ml) and streptomycin (100mg/ml) were added to all media. Cells were replated twice weekly bytrypsinization followed by a 1:5 dilution in fresh medium. For trans-plantation, cells were collected by trypsinization, centrifuged at 185G for 5 minutes, and resuspended at a final concentration of 1.3 3104 cells/ml in HBSS. The cells were stored on ice until use. The ani-mals were anesthetized and then stereotactically injected with 3 ml ofthe tumor cell suspension at a point 2.0 mm rostral to bregma, 2.0mm to the right of the midline, and 3.0 mm ventral to the brain sur-face.

Neural Progenitor Isolation and Cultivation

The NPCs were prepared as previously described.14 Briefly, theanimals were killed by CO2 inhalation followed by decapitation. Thestriatum was aseptically removed bilaterally and digested for 30 min-utes at 37˚C in medium consisting of 0.5 mM Na2–ethylenediamine-tetraacetic acid, 1.0 mM cysteine, 0.9 mg/ml papain (Sigma-Ald-rich), and 1 mg/ml DNase (Worthington) dissolved in Earle balancedsalt solution (Gibco BRL). The cells were plated into 96-well plates(Nalge Nunc International) and grown in medium consisting of thefollowing: NS-A with an additional 2 mM L-glutamine, 3 mM D-glu-cose, 1.2 mM sodium bicarbonate, 0.46 mM HEPES, 2% B27 (Gib-co BRL), 0.1 mg/ml apotransferrin, 23 mg/ml insulin, 55 mM pu-trescine, 20 nM progesterone, 30 nM sodium selenite, 100 U/ml–100mg/ml penicillin/streptomycin, 20 ng/ml human recombinant epider-mal growth factor (Gibco BRL), and 20 ng/ml human recombinantbasic fibroblast growth factor (Gibco BRL). The B27 was excludedfor cultivation of mouse NPCs. Cells were cultivated at 37˚C in 5%CO2 and 95% air with 100% humidity and replated weekly using me-chanical dissociation.

Some neural progenitor cultures were transfected with a lentiviralconstruct to create cell lines that would express GFP, thereby allow-ing detection following transplantation. The cells were incubated for5 hours with replication-defective vesicular stomatitis virus–G pseu-dotyped lentiviral vectors at a multiplicity of infection of 10. Thevector contains the genes encoding enhanced GFP and puromycin-resistance under control of the cytomegalovirus promoter. Bothgenes were connected by the internal ribosome entry site, which per-mits the translation of two open reading frames from one strand ofmRNA. At 48 hours postinfection, neural progenitors were cultivat-ed in the presence of 200 ng/ml puromycin for selection of trans-duced cells.

To demonstrate that cells were multipotent and capable of gener-ating neurons, astrocytes, and oligodendroglia, cultures were immu-nocytochemically stained with antibodies to b-tubulin III, glial fibril-lary acidic protein, and 2,3-cyclic nucleotide 3-phosphohydrolase,respectively, using previously described methods.14 Binding of theprimary antibodies was detected using a red fluorescent secondaryantibody. Photographs (Fig. 1) of the stained cells were obtained un-der epifluorescence using an inverted Zeiss Axiovert microscope andan image analysis system (KS300, Carl Zeiss Microimaging).

Neural Progenitor Cell Transplantation

Neural progenitor cells were used between passages 20 and 40 fol-

lowing isolation. The progenitors were dislodged from the plate bygentle flushing of medium repeatedly across the cells at 7 days afterpassaging. The dislodged cell clusters were dissociated into singlecells by trituration with a fire-polished pipette. To ensure that break-down or lack of expression of GFP was not interfering with de-tection, cells were also labeled with Cell Tracker Green (MolecularProbes) in some of the experiments. To accomplish this, the cellswere suspended in NS-A medium containing 20 mM Cell TrackerGreen for 30 minutes at 37˚C. The cells were then centrifuged at30 G for 5 minutes, filtered through a 70-mm mesh to remove cellclumps, and incubated for another 30 minutes at 37˚C. Just beforetransplantation, cells were centrifuged at 30 G for 5 minutes and re-suspended in PBS at a final concentration of 2 3 104 or 2 3 105 cells/ml and stored on ice until transplantation. The NPCs were adminis-tered either intravenously or by direct stereotactic injection into thebrain on the 5th day after tumor cell implantation using a variety ofHamilton syringes or a glass pipette (see Results). The total volumeinjected was 5 ml.

A. Niranjan et al.

384 J. Neurosurg. / Volume 107 / August, 2007

FIG. 1. Photomicrographs demonstrating the presence of neuro-nal, astrocytic, and oligodendroglial markers in cultures of NPCsthat express GFP. When NPCs were cultivated on poly-D-lysine inmedium containing fetal calf serum, the cells differentiated intomultiple phenotypes and expressed markers of neurons (b-tubulinIII, B), astrocytes (glial fibrillary acidic protein, D), and oligoden-droglia (2,3-cyclic nucleotide 3-phosphohydrolase; F), which weredetected by immunocytochemical analysis and fluorescent micros-copy. Differences between the nonspecific, GFP expression to theleft (A, C, and E) of each of the fluorescent, immunocytochemicalstains (bright white areas) demonstrate that only subsets of the pop-ulation expressed a particular phenotypic marker. Bar = 50 mm.

Enhancement of Implant Survival

In our initial studies, we found very few cells at the surgical sitebeyond 1 day after implantation. Because of this finding, severaldifferent factors were systematically evaluated for their impact ontransplant survival. The factors examined included donor cell type,injection technique, and host tissue microenvironment following im-munosuppression or prior irradiation (Fig. 2).

Immunosuppression. Cyclosporin A (10 mg/kg) was administeredintramuscularly to a small group of animals to determine if immuno-suppression could improve implant survival. The drug was started 2days after implantation of the RG2 brain tumor cell and 3 days beforeNPC transplantation. Daily administration was continued until death1 and 4 days after NPC transplant.

Brain Irradiation. Treatment with doses ranging from 20 to 40 Gyof a single fraction of whole-brain irradiation was performed using a

linear accelerator at the Department of Radiation Oncology at theUniversity of Pittsburgh Medical Center. Animals were anesthetizedwith an intraperitoneal injection of ketamine (44 mg/kg) and acepro-mazine (4.4 mg/kg) and placed in lateral recumbency beneath theradiation source. A 0.5-cm tissue-equivalent bolus was placed overthe irradiated region to reduce the skin dose. The radiation beam wascollimated to exclude tissues outside the anterior and posterior fossa.After delivering half of the total dose, the animal was turned over,and the remaining dose was delivered to the contralateral side. Thecontrol animals received sham irradiation.

Immunohistochemical Analysis

To detect GFP-containing neural progenitors in the tissue, the ani-mals were deeply anesthetized with ketamine and acepromazine andperfused with PBS followed by 10% buffered formalin. The brains


Survival of transplanted neural progenitor cells

385

FIG. 2. Flow chart showing the experimental plan to test factors that may affect transplant survival including cell type(A), injection technique (B), and recipient tissue microenvironment (C). Hanks = HBSS; IV = intravenous; NPGM = neu-ral progenitor growth medium.

were removed and embedded in paraffin. Transverse brain sections(6 mm thick) were then collected and placed on glass slides. One sec-tion was collected for every 150 mm of tissue throughout the lengthof the brain. The sections were deparaffinized by passing themthrough a series of graded alcohol solutions and xylene, and theywere boiled in a 1:100 dilution of antigen retrieval solution (VectorLaboratories) for 10 minutes. Next, they were rinsed in PBS andtreated with proteinase K (20 mg/ml) for 10 minutes at room tem-perature followed by incubation in 0.3% H2O2 for 30 minutes toquench endogenous peroxidase activity. Nonspecific antibody bind-ing was blocked by incubating the sections in PBS with 2% horseserum and 0.2% Triton X-100 in PBS for 30 minutes at room tem-perature. The sections were rinsed and incubated for 1 hour at roomtemperature with mouse anti-GFP antibody (Santa Cruz Biotech) di-luted 1:100 in PBS with 2% horse serum and 0.05% Tween-20. Af-terward, they were rinsed in PBS and incubated in a 1:200 dilutionof biotinylated anti–mouse immunoglobulin G (dilution 1:100, Vec-tor Laboratories). Antibody binding was detected using the Vecta-stain avidin–biotin complex kit (Vector Laboratories) and diaminobenzadine. Finally, sections were counterstained in Mayer hema-toxylin and mounted with Permount (Sigma Chemical Co).

ResultsDelivery Method

Initial experiments were designed to optimize delivery ofNPCs. Mouse NPCs were used in initial experiments be-cause they are readily available. They were implanted intonormal rat brain using glass pipettes. When 104, 105, or 106

NPCs were injected and the tissues examined 1 day later,very few cells were detected in the tissue after injection ofless than 105 cells. All subsequent experiments were there-fore performed using at least 105 cells. To evaluate howlong the cells would survive after such an injection, eightanimals were injected with mouse NPCs and killed 1, 2, 3,or 4 days later. Cells were visible at the site of injection af-ter 1 and 2 days, but none were present beyond that time.

To determine whether mechanical factors, such as dam-age or loss of the cells during injection, might diminish cellsurvival, several alternative administration methods wereevaluated. One concern was whether plugs of tissue enter-ing into the tip during insertion might obstruct the relative-ly small diameter of the glass pipettes (80–100 mm). Be-cause of this, larger bore Hamilton syringes were evaluatedfor cell administration. Two types of syringes were used.One was open-ended, and the other had zero dead space.For the second type of syringe, the plunger extended to theend of the needle tip and thus insured that both NPCs andtissue were expelled from the syringe. Additionally, bothblunt and beveled needle tips were used (Fig. 3) because ofconcern that NPCs could be sheared by the sharp, bevelededge of the pipette and needle tips. For each of the three sy-ringe types, a single animal was injected bilaterally andkilled 1 day later. The GFP-positive NPCs could be seen inall animals at both injection sites, and there were no appar-ent differences in cell number. As a result of this finding, aHamilton syringe with a beveled, open-ended needle wasused in all ensuing studies.

Effect of Tumor and NPC Type

Next, we tested whether cellular syngeneic allografts,rather than the mouse xenografts described earlier, could beimplanted into the brain and prove useful for cell therapy ofbrain tumors. To test whether NPCs might serve as ef-fective agents for the delivery of gene products to tumors,adult rats bearing 9L tumors were injected with 105 rat

NPCs. In six of the animals, the cells were injected direct-ly into the tumor at the site of the original implant. In an ad-ditional six animals, cells were injected directly contralat-eral to the tumor, 4 mm away from the original site. In thefinal group, cells were injected 4 mm caudal to the tumor.The NPC survival in the tissue was assessed in two animalsfrom each group at 1, 4, or 7 days after administration. TheGFP-positive cells were visible in the tissue on Day 1 afterinjection but only at the site of administration, and nonewere present anywhere in the brain, including at the site ofinitial injection and within tumor on Day 4 or 7.

Based on a previous report that mouse NPCs would sur-vive in RG2 tumors1 we changed from the 9L to the RG2tumor model, and we compared the survival of mouse andrat NPCs in this model. A total of eight rats with RG2tumors were injected with NPCs of rat or mouse origin.When the animals were killed 4 days later and survival wasassessed, a small number of mouse NPCs could be identi-fied within the tumor. No NPCs were found in animals in-jected with rat cells, suggesting that mouse NPCs remainedviable longer following injection. To determine if tumorcells themselves might enhance survival, eight normal andeight tumor-bearing animals were injected intratumorallywith mouse NPCs; survival was assessed in two animalsper group 1 to 4 days later. Initially, there did appear to begreater cell survival in the tumor-bearing brain than in nor-mal brain, but this apparent difference was short-lived. ByDay 2 after injection, only a few cells could be identified inanimals without tumors. In RG2 tumor-bearing animals,there were substantially more cells visible, with more than100 detected on Day 1 and approximately 10 to 50 on Day2. However, unlike in the previous experiment, no mouseNPCs were visible in the RG2 tumors by Day 3 or 4.

Effect of Environmental Variables

Although the xenografted mouse NPCs appeared to sur-vive better in tumor-bearing rat brain than the allografts,there are several reasons that allografts are more likely tobe used clinically. We wanted to model the clinical situa-tion as closely as possible and therefore used allografts inour studies. However, our initial experimental results indi-cated that, at least in using syngeneic cells in the RG2 tu-mor model, methods to enhance rat NPC survival follow-ing transplant were needed. An immunosuppressive agent,cyclosporin, was administered to determine if such an ap-proach could enhance cell survival. The agent was givencontinuously starting 7 days before NPC administration.However, no cells were detected 4 days after NPC trans-plantation, which suggests that immunosuppression is notan effective strategy for enhancing survival.

Next, eight animals were treated with ionizing radiationbased on the report that spinal cord irradiation could im-prove the survival of oligodendroglial progenitors.11 A sin-gle dose of 20 or 40 Gy of whole-brain irradiation wasdelivered either 7 days before NPC transplantation into oth-erwise normal brain or 12 days before transplant into tu-mor-bearing brain (7 days before RG2 cell administration).Unlike the previous experiments in which cells were notdetected beyond 1 day after NPC transplantation into non-irradiated tissues, GFP-positive NPCs were present in bothnormal (Fig. 4) and tumor-bearing (Fig. 5) brain when irra-diated animals were killed 1 or 4 days later. An additional

A. Niranjan et al.


eight animals were then treated with 20 Gy followed by in-jection of RG2 tumor cells and NPCs, and the presence ofGFP-positive NPCs was assessed both 4 and 8 days later.In accordance with our previous results, GFP-positive cellswere detected at 4 days, and they were also detected at 8days after implantation (Fig. 6).

Finally, an experiment was performed to determine if ir-radiation might enhance NPC migration to tumor sites.Eight rats were treated with 20 Gy whole-brain irradiationfollowed 1 week later by RG2 tumor cell administration.The NPCs were then administered to these rats after an ad-ditional 5 days. Four of the animals received 3.5 3 107

NPCs given intravenously via the tail vein. In the remain-ing four animals, NPCs were administered at a site 2 mmdirectly caudal to the tumor site. When the animals werekilled and the tissues examined either 4 or 8 days after NPCadministration, no GFP-positive NPCs could be detectedwithin the tumor although the cells could be detected at theoriginal site of administration. Thus, brain irradiation wasable to enhance survival but not migration of NPCs.

Discussion

Adult mammalian brain, including that of primates andhumans, contains populations of stem cells. Such cells havean inherent ability to proliferate and self-renew indefinite-ly. They are also multipotent, and differentiated progeny ofNSCs will express the principal cellular phenotypes of themature CNS: neurons, astrocytes, and oligodendrocytes.Because the cells can be expanded ex vivo and because oftheir apparent migratory ability in the CNS, NSCs as wellas NPCs, which have more limited capabilities in terms ofself-renewal and differentiation, might be useful clinicallyfor cell replacement or as vectors for widespread gene de-livery and the expression of therapeutic proteins.

Despite a multitude of reports that NSCs and/or NPCscan be successfully transplanted into the brain, our initialstudies suggested that survival of such cells was far fromcertain. Indeed, we undertook the current studies based onthe hypothesis that NPCs would readily incorporate in-to the growing tumors and hence prove to be ideal vec-tors for gene delivery.1 Authors of several animal stud-ies have shown that transplanted NPCs incorporate into thebrain.15,16,20 Aboody et al.,1 using immortalized mouseNPCs,33 demonstrated that NPCs, when implanted into ex-perimental intracranial gliomas in vivo in adult rodents,distribute themselves extensively throughout the tumor bedand migrate in juxtaposition to the expanding and advanc-ing tumor cells. When implanted intracranially at distantsites from the tumor, these cells migrated through normaltissue, targeting the tumor cells.1 In contrast, in our initialstudies, the cells did not survive beyond 1 or 2 days aftertransplantation, which indicated to us that a number of fac-tors need to be considered prior to transplantation. We areunaware of any studies to date in which the authors havesystematically addressed these issues, and translating basicstudies into successful clinical use of these cells will bechallenging unless such factors are carefully identified andevaluated. In the present study we have begun to addressthe issues that should be taken into consideration.

Some of the factors that may affect transplant successare mechanical and include delivery device, placement,

timing of administration, and number of cells to deliver. In-traparenchymal injection was commonly used in initialstudies.8,9,15,32,34 Authors of additional studies have indicat-ed that intrathecal5,18 or intravenous7 administration can al-so be used to successfully deliver NPCs and bone marrow



387

FIG. 3. Illustrations (upper) and photomicrographs (lower)showing the impact of syringe type on intracerebral implantation ofNPCs. Both open needle (upper left) and positive-displacement (up-per center and right) syringes were used as well as beveled (left andcenter) and blunt (upper right) needle tips. At 1 day after admin-istration, plugs of bright white, GFP-positive cells could be seen byepifluorescent microscopy in the tissue directly adjacent to the in-jection track for all of the three needle types (lower). Although theblunt-ended needle did appear to produce larger injection tracks inthe tissue (lower right), there were no clear differences among thethree syringe types in terms of the number or viability of NPCsdelivered. Each photomicrograph demonstrates a region of the ratcaudate nucleus that was injected with NPCs by using the syringetype shown directly above (upper). Original magnification 3 400.

stromal cells to CNS lesions. The intravenous route is at-tractive because of its minimally invasive nature, but it isnot clear that this method of administration will be usefulfor all of the types of CNS disorders that are treated or forall of the cell types that might be used for therapy. In thepresent study, we delivered cells intraparenchymally by us-ing standard needles, zero-volume needles with plungersextending to the end of the needle to ensure extrusion intothe tissue, and needles with openings directed to the sideto reduce backflow of cells up the needle track (Fig. 3).However, no needle showed a significant advantage overthe others with respect to cell delivery. Placement of thecells was a critical factor because, although the number ofcells surviving after injection into tumor or normal tissuewas similar (Fig. 6), there was no detectable migrationaway from the site of injection to the tumor target. Also, nocells could be detected within the tumor or any other partof the brain following intravenous administration. In ourhands, 105 cells appeared to be the optimal number for de-livery via a single injection in that no or few cells survivedafter transplantation of 104 cells. Cells may secrete auto-crine survival factors,23 so injection of fewer than a criticalnumber of cells may result in too low a concentration ofsuch factors to support cell viability. One possible way tocircumvent this would be to include the necessary factorswithin the medium used to inject the cells. However, in our

A. Niranjan et al.


TABLE 1Factors evaluated for their impact

on NPC survival following transplantation

Variable Conditions Examined Result

no. of cells 104 vs 105 vs 106 no cells apparent after104; more cells after106 than 105

injection device glass pipette vs blunt tip no impact of injection vs zero-volume syringe device

species of origin mouse vs rat NPCs slightly enhanced mouseNPC survival

tumor model 9L vs RG2 no apparent role in NPCsurvival

injection site intratumorally vs adjacent no cells after IV admin-to tumor vs IV istration

immunosuppression none vs cyclosporin A no improvement due tocyclosporin A

suspension medium NPGM vs PBS vs HBSS no apparent role in NPC survival

time of injection concurrent w/ tumor in- no apparent role in NPC jection vs 5 days post- survivaltumor injection

ionizing radiation 0 vs 20 vs 40 radiation enhanced sur-(Gy) vival of NPCs; no im-

pact on migration

FIG. 4. Photomicrographs demonstrating the presence of GFP-positive NPCs in the normal brain tissue at 1 day (upper)and 1 week (center and lower) after administration. In previously nonirradiated brain, no cells were visible beyond 1 dayafter NPC administration (upper and center). In contrast, numerous cells could be found at the site of injection up to 1 weekafter 20 Gy of whole-brain irradiation (lower).

experiments, there was no difference in survival whetherthe cells were suspended in straight, nonnutritive PBS, inHBSS, or in cultivation medium (Table 1). The necessarysurvival factors may have been either absent or at insuffi-cient concentrations to support intracerebral survival. An-other reason that we consider 105 cells as the optimal num-ber for transplant is that injection of 106 rather than 105 cellsdid not produce a comparable 10-fold increase in cell num-ber in the tissue. This result suggests that the host tissuemay be limited in its ability to support the survival of trans-planted cells. Implanting the cells at multiple sites mighthelp to overcome this limitation.

Biological factors may also influence the success oftransplantation; these factors include cell type, such aswhether the cells are NPCs, bone marrow stromal cells, orembryonic stem cell–derived; and origin, such as whetherthe cells constitute a xenograft, allograft, isograft, or auto-graft. The status of the recipient tissue, which can stimulatean immune response or secrete factors to support or inhib-it growth, may also have an impact. We did not comparesurvival of various cell types in our experiments, but we didevaluate survival of cells from mice compared with that ofcells from rats. Somewhat surprisingly, in our initial stud-ies done without the use of prior irradiation the mouseNPCs survived up to 4 days after implantation, whereas rat

NPCs did not survive in the rat brain for more than 1 day.This finding is not unprecedented; there have been a num-ber of previous reports of successful xenotransplantationsinto rat brain.26,32,34 Although it might be possible to sub-stitute mouse NPCs for rat NPCs for experimental studies,such xenotransplants would probably not be optimal in aclinical situation. Allogenic grafts would have less poten-tial for rejection by the immune system or transmission ofzoonotic pathogens. In our experiments, we wanted to sim-ulate the likely clinical situation as closely as possible, sowe focused our studies on the use of allografts of syngene-ic rat cells.

The environment of the host tissue can have a significantimpact on transplant survival and the tissue milieu may bealtered by prior or concurrent therapy. In the case of braintumors, the tissue may have been treated with surgery, radi-ation therapy, and/or chemotherapy before transplantation.Based on published studies on the spinal cord by Hinkset al.17 and Franklin et al.,11 we explored whether a sin-gle fraction of radiation might increase the survival of ratNSCs. In the cited studies, high-dose irradiation promotedthe engraftment and migration of oligodendrocyte pro-genitors in the spinal cord. Franklin et al. transplanted CG4oligodendrocyte progenitor cell lines into the spinal cordsof adult rats that had undergone 40-Gy irradiation. Hinks et



389

FIG. 5. Photomicrographs demonstrating the presence of GFP-positive NPCs in intracerebral tumors 8 days after im-plantation. In previously nonirradiated brain, as in normal brain, very few cells could be detected at 1 week after implan-tation, even after injecting as many as 106 NPCs (upper). In contrast, numerous cells could be found within and around theintracerebral tumor in brains pretreated with 20 Gy of whole-brain irradiation followed 7 days later by injection of 105 (cen-ter) or 106 (lower) NPCs.

al.17 irradiated the spinal cords of rats with 40 Gy and in-jected PVG rat–derived oligodendrocyte progenitors. In thepresent study, we irradiated brain with a lower dose (20 Gy)and implanted NPCs derived from adult rats rather than oli-godendrocyte progenitors. The clinical significance of thisfinding is that the patients treated with x-ray therapy forprimary brain tumors can be expected to have enhancedsurvival of transplanted, allogenic NPCs. Our findings areconsistent with these previous studies and suggest that ir-radiation can also facilitate survival of NPCs transplantedinto brain. Our results are also in agreement with those in arecent report by Marshall et al.,22 who noted that whole-body irradiation enhanced the engraftment of MASCs in-to the lateral ventricles of mice. Although a single, lethaldose of ionizing radiation significantly depleted the SEZand rendered it unreceptive to transplanted MASCs, ex-posure to milder levels of radiation resulted in a transientdecrease in mitotic SEZ neuroblasts, and substantially in-creased MASC engraftment.22 Unlike these previous stud-ies in which either spinal cord or whole-body irradiationwas used, in the present study we used brain irradiationspecifically. This is a significant difference because, eventhough whole-body irradiation includes brain irradiation,the irradiation of extracranial tissues could lead to the re-lease of cytokines and clastogenic factors that might mi-grate to the brain to exert their effects. In our study we havedemonstrated that radiation targeted to brain alone can en-hance survival of transplanted cells. It would be useful toknow if lower doses would have the same effect. Thiscould help us to further understand the mechanism respon-sible for the effect and also might make radiation a usefultool in transplantation.

At present, the mechanism by which radiation exerts itseffect on NPC survival is not clear. Similar observations

with respect to the impact of radiation on transplant sur-vival have been noted in other organ systems. For example,Malhi et al.21 found that pretreatment of liver tumors withirradiation enhanced proliferation of transplanted hepato-cytes, which thus increased repopulation of liver by trans-planted cells, and radiation is commonly used before hema-tological transplants to “condition” the bone marrow andmake it receptive to the cells of the donor. It is thought thatradiation, by ablating the endogenous cells, creates spacefor transplanted cells.37 With regard to the effect of radia-tion on cells transplanted into the CNS, various hypotheseshave been suggested.17 Radiation may increase the avail-ability of NPC survival factors, allowing the CNS to sup-port a greater number of progenitors. Another possibility isthat radiation may alter the local immune response mediat-ed through microglia, and studies have shown that radiationcan cause a decrease in the number of microglia.12,13,19 Wehave observed that radiation can alter immune surveillanceas indicated by upregulation of CD68-positive microglia(unpublished observations). However, in our studies, animmunosuppressive agent was ineffective at improvingNPC survival, which argues against this hypothesis. Never-theless, a recent report indicates that even allogeneic neu-ral progenitors can induce an immune reaction when trans-planted into the CNS, and the use of autologous cells candecrease immunoreactivity and enhance cell survival.26 Ra-diation may also deplete the endogenous NPC population,thereby providing vacant niches available to transplantedNPCs. In fact Hinks et al.17 found no detectable increase inCNS growth factors but confirmed a decrease in en-dogenous NPCs, a result consistent with the creation ofa vacant niche. This hypothesis might also account for theresults of our studies, in which a 10-fold increase in thenumber of cells implanted produced only an approximate-ly twofold increase in the number of surviving transplantedcells. These data suggest that the “size” of the vacant nicheis not unlimited and that there is little advantage derivedfrom implanting many more than 105 cells at a single injec-tion site. Depletion of the NSC niche has been previouslyachieved using exposure to ionizing irradiation,3,25,27,35,36 re-sulting in transient and long-term depletion of NSCs andNPCs in the hippocampus and SEZ. Authors of previousstudies investigating the effects of focused exposure of ra-diation to the brain in adult rats on SEZ neurogenesis havereported ablation of NSCs in the SEZ,3,36 and our own pre-liminary investigations (data not shown) have shown thin-ning of and cell loss within the SEZ after irradiation.

Conclusions

Our results show that pretreatment with radiation canenhance survival of rat NPCs after transplantation into tu-mor-bearing rat brain. It will be important in future stud-ies to understand the mechanism behind this effect, whichcould lead to other, nonradiation-based methods of enhanc-ing transplant survival. One approach to identifying factorsthat might contribute to survival would be an assessment ofmRNA and protein levels of known cytokines and growthfactors after irradiation. Alterations in immune surveillancecould be delineated by immunohistochemical evaluationsof radiation-induced alterations in microglial and lympho-cytic cell types. The results of such studies could lead to

A. Niranjan et al.


FIG. 6. Results of a quantitative analysis of the impact of irradi-ation on the number of NPCs detected following implantation intonormal (left) or tumor-bearing (right) brain. No or extremely fewNPCs could be located following implantation of either 1 3 105

(small symbols) or 1 3 106 (large symbols) cells into the four ani-mals that had not been treated with previous brain irradiation (opensymbols). In contrast, GFP-positive (GFP1), NPCs could readily bedetected in the six animals treated with 20 Gy of whole-brain irra-diation (closed symbols). Each symbol represents one animal.

clinically useful methods to facilitate transplantation of hu-man NPCs into human brain.

References

1. Aboody KS, Brown A, Rainov NG, Bower KA, Liu S, Yang W,et al: Neural stem cells display extensive tropism for pathology inadult brain: evidence from intracranial gliomas. Proc Natl AcadSci U S A 97:12846–12851, 2000

2. Ahmed S, Reynolds BA, Weiss S: BDNF enhances the differenti-ation but not the survival of CNS stem cell-derived neuronal pre-cursors. J Neurosci 15:5765–5778, 1995

3. Amano T, Inamura T, Wu CM, Kura S, Nakamizo A, Inoha S, etal: Effects of single low dose irradiation on subventricular zonecells in juvenile rat brain. Neurol Res 24:809–816, 2002

4. Bailey KA, Drago J, Bartlett PF: Neuronal progenitors identifiedby their inability to express class I histocompatibility antigens inresponse to interferon-gamma. J Neurosci Res 39:166–177, 1994

5. Bakshi A, Hunter C, Swanger S, Lepore A, Fischer I: Minimallyinvasive delivery of stem cells for spinal cord injury: advantag-es of the lumbar puncture technique. J Neurosurg Spine 1:330–337, 2004

6. Bartlett PF, Brooker GJ, Faux CH, Dutton R, Murphy M, TurnleyA, et al: Regulation of neural stem cell differentiation in the fore-brain. Immunol Cell Biol 76:414–418, 1998

7. Brown AB, Yang W, Schmidt NO, Carroll R, Leishear KK,Rainov NG, et al: Intravascular delivery of neural stem cell linesto target intracranial and extracranial tumors of neural and non-neural origin. Hum Gene Ther 14:1777–1785, 2003

8. Catapano LA, Sheen VL, Leavitt BR, Macklis JD: Differentiationof transplanted neural precursors varies regionally in adult stria-tum. Neuroreport 10:3971–3977, 1999

9. Chow SY, Moul J, Tobias CA, Himes BT, Liu Y, Obrocka M, etal: Characterization and intraspinal grafting of EGF/bFGF-depen-dent neurospheres derived from embryonic rat spinal cord. BrainRes 874:87–106, 2000

10. Craig CG, Tropepe V, Morshead CM, Reynolds BA, Weiss S, vander Kooy D: In vivo growth factor expansion of endogenous sub-ependymal neural precursor cell populations in the adult mousebrain. J Neurosci 16:2649–2658, 1996

11. Franklin RJ, Bayley SA, Blakemore WF: Transplanted CG4 cells(an oligodendrocyte progenitor cell line) survive, migrate, andcontribute to repair of areas of demyelination in X-irradiated anddamaged spinal cord but not in normal spinal cord. Exp Neurol137:263–276, 1996

12. Gilmore SA, Phillips N, Liu KM, Houle JD: Radiation-inducedmodulation of the microglial population in the normal and injuredmature spinal cord. Exp Neurol 182:169–179, 2003

13. Gilmore SA, Sims TJ, Davies DL, Durgun MB: Microglial devel-opment is altered in immature spinal cord by exposure to radia-tion. Int J Dev Neurosci 15:1–14, 1997

14. Gobbel GT, Choi SJ, Beier S, Niranjan A: Long-term cultivationof multipotential neural stem cells from adult rat subependyma.Brain Res 980:221–232, 2003

15. Groves AK, Barnett SC, Franklin RJ, Crang AJ, Mayer M, Blake-more WF, et al: Repair of demyelinated lesions by transplantationof purified O-2A progenitor cells. Nature 362:453–455, 1993

16. Herrera DG, Garcia-Verdugo JM, Alvarez-Buylla A: Adult-de-rived neural precursors transplanted into multiple regions in theadult brain. Ann Neurol 46:867–877, 1999

17. Hinks GL, Chari DM, O’Leary MT, Zhao C, Keirstead HS, Blake-more WF, et al: Depletion of endogenous oligodendrocyte pro-genitors rather than increased availability of survival factors is alikely explanation for enhanced survival of transplanted oligoden-drocyte progenitors in X-irradiated compared to normal CNS.Neuropathol Appl Neurobiol 27:59–67, 2001

18. Jin K, Sun Y, Xie L, Mao XO, Childs J, Peel A, et al: Comparisonof ischemia-directed migration of neural precursor cells after in-trastriatal, intraventricular, or intravenous transplantation in therat. Neurobiol Dis 18:366–374, 2005

19. Kreutzberg GW: Microglia, the first line of defence in brain pa-thologies. Arzneimittelforschung 45:357–360, 1995

20. Lacza Z, Horvath E, Busija DW: Neural stem cell transplantationin cold lesion: a novel approach for the investigation of brain trau-ma and repair. Brain Res Brain Res Protoc 11:145–154, 2003

21. Malhi H, Gorla GR, Irani AN, Annamaneni P, Gupta S: Cell trans-plantation after oxidative hepatic preconditioning with radiationand ischemia-reperfusion leads to extensive liver repopulation.Proc Natl Acad Sci U S A 99:13114–13119, 2002

22. Marshall GP II, Scott EW, Zheng T, Laywell ED, Steindler DA:Ionizing radiation enhances the engraftment of transplanted invitro-derived multipotent astrocytic stem cells. Stem Cells 23:1276–1285, 2005

23. Menter DG, Herrmann JL, Nicolson GL: The role of trophic fac-tors and autocrine/paracrine growth factors in brain metastasis.Clin Exp Metastasis 13:67–88, 1995

24. Mligiliche NL, Xu Y, Matsumoto N, Idel C: Survival of neuralprogenitor cells from the subventricular zone of the adult rat aftertransplantation into the host spinal cord of the same strain of adultrat. Anat Sci Int 80:229–234, 2005

25. Monje ML, Mizumatsu S, Fike JR, Palmer TD: Irradiation inducesneural precursor-cell dysfunction. Nat Med 8:955–962, 2002

26. Muraoka K, Shingo T, Yasuhara T, Kameda M, Yuan W, HayaseH, et al: The high integration and differentiation potential of autol-ogous neural stem cell transplantation compared with alloge-neic transplantation in adult rat hippocampus. Exp Neurol 199:311–327, 2006

27. Parent JM, Tada E, Fike JR, Lowenstein DH: Inhibition of dentategranule cell neurogenesis with brain irradiation does not preventseizure-induced mossy fiber synaptic reorganization in the rat. JNeurosci 19:4508–4519, 1999

28. Reynolds BA, Rietze RL: Neural stem cells and neurospheres—re-evaluating the relationship. Nat Methods 2:333–336, 2005

29. Reynolds BA, Tetzlaff W, Weiss S: A multipotent EGF-respon-sive striatal embryonic progenitor cell produces neurons and as-trocytes. J Neurosci 12:4565–4574, 1992

30. Reynolds BA, Weiss S: Clonal and population analyses demon-strate that an EGF-responsive mammalian embryonic CNS pre-cursor is a stem cell. Dev Biol 175:1–13, 1996

31. Reynolds BA, Weiss S: Generation of neurons and astrocytes fromisolated cells of the adult mammalian central nervous system.Science 255:1707–1710, 1992

32. Sabate O, Horellou P, Vigne E, Colin P, Perricaudet M, Buc-Caron MH, et al: Transplantation to the rat brain of human neuralprogenitors that were genetically modified using adenoviruses.Nat Genet 9:256–260, 1995

33. Snyder EY, Deitcher DL, Walsh C, Arnold-Aldea S, HartwiegEA, Cepko CL: Multipotent neural cell lines can engraft andparticipate in development of mouse cerebellum. Cell 68:33–51,1992

34. Svendsen CN, Clarke DJ, Rosser AE, Dunnett SB: Survival anddifferentiation of rat and human epidermal growth factor-respon-sive precursor cells following grafting into the lesioned adult cen-tral nervous system. Exp Neurol 137:376–388, 1996

35. Tada E, Parent JM, Lowenstein DH, Fike JR: X-irradiation caus-es a prolonged reduction in cell proliferation in the dentate gyrusof adult rats. Neuroscience 99:33–41, 2000

36. Tada E, Yang C, Gobbel GT, Lamborn KR, Fike JR: Long-termimpairment of subependymal repopulation following damage byionizing irradiation. Exp Neurol 160:66–77, 1999

37. Vriesendorp HM: Aims of conditioning. Exp Hematol 31:844–854, 2003

38. Weiss S, Reynolds BA, Vescovi AL, Morshead C, Craig CG, vander Kooy D: Is there a neural stem cell in the mammalian fore-brain? Trends Neurosci 19:387–393, 1996

Manuscript submitted August 31, 2006.Accepted February 15, 2007.Address reprint requests to: Glenn T. Gobbel, D.V.M., Ph.D., De-

partment of Neurological Surgery, B-400, University of PittsburghMedical Center-Presbyterian, 200 Lothrop Street, Pittsburgh, Penn-sylvania 15213. email: [email protected].



391

Documents

Survival of transplanted neural progenitor cells enhanced ... · the tumor cell suspension at a point 2.0 mm rostral to bregma, 2.0 mm to the right of the midline, and 3.0 mm ventral