Work has shown that stem cell transplantation can rescue or replace neurons in models of retinal degenerative disease. Neural progenitor cells (NPCs) modified to overexpress neurotrophic factors are one means of providing sustained delivery of therapeutic gene products in vivo. To develop a nonrodent animal model of this therapeutic strategy, we previously derived NPCs from the fetal cat brain (cNPCs). Here we use bicistronic feline lentiviral vectors to transduce cNPCs with glial cell-derived neurotrophic factor (GDNF) together with a GFP reporter gene. Transduction efficacy is assessed, together with transgene expression level and stability during induction of cellular differentiation, together with the influence of GDNF transduction on growth and gene expression profile. We show that GDNF overexpressing cNPCs expand in vitro, coexpress GFP, and secrete high levels of GDNF protein—before and after differentiation—all qualities advantageous for use as a cell-based approach in feline models of neural degenerative disease.
The retina is susceptible to a variety of degenerative diseases, including age-related macular degeneration (AMD), retinitis pigmentosa (RP) and other inherited photoreceptor degenerations, photoreceptor loss following retinal detachment, ganglion cell loss in glaucoma and optic neuropathies, as well as the loss of retinal neurons associated with nondegenerative conditions such as diabetic retinopathy (DR), macular edema and ischemia, vascular occlusions, trauma, and inflammatory diseases. Any of these can lead to debilitating visual deficits. AMD is a particularly prevalent cause of blindness among elderly persons, affecting more than 30 million people globally. That number is expected to double over the next decade in association with demographic shifts towards an older population, particularly in developed countries [
A large body of research has shown that the use of exogenous neurotrophic factors can reproducibly promote the survival of specific neurons in various parts of the central nervous system (CNS), including the retina [
Stem and progenitor cell transplantation has also shown considerable promise in animal models of neural degeneration. Subretinal transplantation of neural progenitor cells (NPCs) has yielded intriguing evidence of cellular repopulation of damaged retinas, growth of neurites into the optic nerve head and retardation of ongoing retinal degeneration [
The visual system of the cat is quite sophisticated and one of the most extensively studied among higher mammals. There are many similarities to the human retina although that of the cat has a tapetum and is generally optimized for performance under scotopic conditions [
Previously, we showed that it is possible to derive NPCs from the developing cat brain and that these cells are capable of integration into the retina of dystrophic feline recipients [
Cat neural progenitor cells (cNPCs) were originally isolated from 47 day cat fetuses as previously described [
The lentiviral vector used in this study was an FIV-based bicistronic vector (GeneCopoeia, Germantown, Maryland) designated as lenti-GDNF-GFP, which carries a human GDNF gene driven by the cytomegalovirus (CMV) immediate-early promoter as well as an enhanced green fluorescent protein (GFP) reporter gene with an internal ribosome entry site (IRES). Lenti-GDNF-GFP vectors were prepared by transient transfection of 293T cells using a standard calcium phosphate precipitation protocol (Clontech, Mountain View, CA). Briefly, 293T cells cultured in 10 cm tissue culture dishes (BD Biosciences, San Jose, CA) were transfected with 2
Cat neural progenitor cells were transduced with lenti-GDNF-GFP vectors at a MOI of 10 following the standard procedure. Briefly, cNPCs were seeded at a density that allowed them to grow to 90% confluency on the day of transduction. The cells were then transduced by 6–24 hours of exposure to virus-containing supernatant in the presence of 5–8
Cells were harvested using TrypLE Express (Invitrogen) and filtered through cell strainer caps (35
The growth properties of transduced and nontransduced cNPCs were assessed by culturing both types of cells under proliferation conditions in Ultraculture-based medium (UM). Cells of identical passage number (p17) were seeded in four T25 culture flasks at a density of 0.25 million cells/flask. One flask of each cell type were trypsinized and counted daily. Cell numbers were graphed at each time point to compare the growth properties of transduced versus nontransduced cells.
Transduced and nontransduced cNPCs of identical passage number were seeded in T25 culture flasks (0.25 million/flask). Following attachment of cells (approx. 4 hours), the original media were replaced with 3 mL of fresh media. Subsequently, 3 mL of conditioned media were collected and replaced with fresh media at 24 hour intervals and conditioned samples were saved at
Total RNA was extracted from each sample using the RNeasy Mini Kit (Qiagen, Valencia, CA). DNaseI was used to eliminate the possibility of genomic DNA contamination. RNA concentration was measured at a wavelength of 260 nm (A260) for each sample, and the purity of isolated total RNA was determined by the A260/A280 ratio. Quantitative RT-PCR analyses were only performed on samples with A260/A280 ratios between 1.9 and 2.1. Two micrograms of total RNA in a 20
Cat-specific primers for quantitative RT-PCR (GDNF = human).
Gene | Forward primer ( | Reverse primer ( |
---|---|---|
GCCGTCTTCCCTTCCATC | CTTCTCCATGTCGTCCCAGT | |
Nestin | CTGGAGCAGGAGAAGGAGAG | GAAGCTGAGGGAAGCCTTG |
Sox2 | ACCAGCTCGCAGACCTACAT | TGGAGTGGGAGGAAGAGGTA |
Vimentin | ATCCAGGAGCTACAGGCTCA | GGACCTGTCTCCGGTACTCA |
Pax6 | AGGAGGGGGAGAGAATACCA | CTTTCTCGGGCAAACACATC |
Hes1 | GCCAGCAGATATAATGGGAGA | GCATCCAAAATCAGTGTTTTCA |
Hes5 | CTCAGCCCCAAAGAGAAGAA | AGGTAGCTGACGGCCATCTC |
Notch1 | CAGTGTCTGCAGGGCTACAC | CTCGCACAGAAACTCGTTGA |
Mash1 | CATCTCCCCCAACTACTCCA | CCAACATCGCTGACAAGAAA |
Ki-67 | TCGTCTGAAGCCGGAGTTAT | TCTTCTTTTCCCGATGGTTG |
DCX | GGCTGACCTGACTCGATCTC | GCTTTCATATTGGCGGATGT |
CATTCTCGTGGACCTTGAGC | GCAGTCGCAATTCTCACATT | |
Map2 | ACCTAAGCCATGTGACATCCA | CTCCAGGTACATGGTGAGCA |
PKC-alpha | TTCACAAGAGGTGCCATGAA | CCATACAGCAATGACCCACA |
GFAP | CGGTTTTTGAGGAAGATCCA | TTGGACCGATACCACTCCTC |
Lhx2 | GATCTGGCGGCCTACAAC | AGGACCCGTTTGGTGAGG |
CD81 | CCACAGACCACCAACCTTCT | CAGGCACTGGGACTCCTG |
CD133 | AGGAAGTGCTTTGCGGTCT | TGCCAGTTTCCGAGTCTTTT |
NCAM (CD56) | AGAACAAGGCTGGAGAGCAG | TTTCGGGTAGAAGTCCTCCA |
EGFR | AACTGTGAGGTGGTCCTTGG | CGCAGTCCGGTTTTATTTGT |
NagoA | TTTGCAGTGTTGATGTGGGTA | TAACAGGAACGCTGAAGAGTGA |
SDF1 | ACAGATGTCCTTGCCGATTC | CCACTTCAATTTCGGGTCAA |
CXCR4 | TCTGTGGCAGACCTCCTCTT | TTTCAGCCAACAGCTTCCTT |
Cyclin D2 | CAAGATCACCAACACGGATG | ATATCCCGCACGTCTGTAGG |
Pbx1 | CTCCGATTACAGAGCCAAGC | GCTGACCATACGCTCGATCT |
FABP7 | TGGAGGCTTTCTGTGCTACC | TGCTTTGTGTCCTGATCACC |
AQP4 | TACACTGGTGCCAGCATGA | CACCAGCGAGGACAGCTC |
Nucleostemin | CAGTGGTGTTCAGAGCCTCA | CCGAATGGCTTTGCTGTAA |
Synapsin1 | ACGACGTACCCTGTGGTTGT | CGTCATATTTGGCGTCAATG |
Caspase 3 | ATGGAGAACAGTGAAAACTCAGTGG | AATTATTATACATAAACCCATTTCAGG |
Bax 4 | CTGAGCAGATCATGAAGACAGG | GTCCAGTTCATCTCCGATGC |
hGDNF* | TGGGCTATGAAACCAAGGAG | CAACATGCCTGCCCTACTTT |
*Human GDNF gene.
Quantitative PCR was performed using an Applied Biosystems 7500 Fast Real-Time PCR Detection System (Applied Biosystems, Foster, CA). Triplicate wells were used for each gene. A total volume of 20
Transduced cNPCs were differentiated in UM without added EGF or bFGF and containing 10% FBS (UM-FBS). Cells (0.2 million) in UM were seeded in T25 culture flasks and allowed to attach, then culture medium was aspirated and replaced with either UM-FBS for differentiation or fresh UM for comparison. Conditioned media were collected and replaced with fresh media every 24 hours for 4 days and frozen for ELISA analysis. At the end of day 4, cells were trypsinized, counted, and ELISA analysis was performed on lysates as well as thawed media samples. For FACS analysis, transduced cNPCs were cultured in either UM-FBS or UM for 10 days prior to processing.
Transduced and nontransduced cNPCs were seeded in 4-well chamber slides (Nalge Nunc International, Rochester, NY) and allowed to grow for 3–5 days. Cells were re-fed every 2 days and fixed with freshly prepared 4% paraformaldehyde (Invitrogen) in 0.1 M phosphate-buffered saline (PBS) for 20 minutes at room temperature and washed with PBS. Cells were then incubated in antibody blocking buffer consisting of PBS containing 10% (v/v) normal goat serum (NGS) (Biosource, Camarillo, CA), 0.3% Triton X-100, 0.1% NaN3 (Sigma-Aldrich, Saint Louis, MO) for 1 hour at room temperature. Slides were incubated in primary antibodies (Table
Primary antibodies used for immunocytochemistry on cNPCs.
Target | Antibody type | Reactivity in retina | Source | Dilution |
---|---|---|---|---|
Nestin | Mouse monoclonal | Progenitors, reactive glia | BD | 1 : 200 |
Vimentin | Mouse monoclonal | Progenitors, reactive glia | Sigma | 1 : 200 |
Ki-67 | Mouse monoclonal | Proliferating cells | BD | 1 : 200 |
GFAP | Mouse monoclonal | Astrocytes, reactive glia | Chemicon | 1 : 200 |
Mouse monoclonal | Immature neurons | Chemicon | 1 : 200 | |
GDNF | Rabbit polyclonal | Growth factor | SCBT | 1 : 200 |
Currently, there are relatively few molecular tools with enhanced specificity for feline cells. Recent development of feline immunodeficiency virus- (FIV-) based vectors could present a means for improved delivery of transgenes into cells of this species. Here, we employed an FIV-based bicistronic vector for delivery of glial cell line-derived neurotrophic factor (GDNF) to cat neural progenitor cells (cNPCs). Forty eight hours after lenti-GDNF-GFP viral vector transduction, approximately 50% of cNPCs expressed the GFP reporter gene based on direct observation via fluorescence microscopy. To enrich for transgene-expressing cells, cNPCs were trypsinized at 72 hours postviral vector incubation and sorted by FACS based on GFP expression. The GFP-enriched population was subsequently cultured in Ultraculture-based proliferation medium (UM) for more than 60 days. High levels of GFP expression were sustained throughout this time period (Figure
GDNF-transduced cNPCs: morphology and reporter gene expression. Feline NPCs transduced using a bicistronic lenti-GDNF-GFP vector and cultured under proliferation conditions (UM) for 60 days (p9–p26). Cellular growth, morphology, and GFP expression were monitored over this time period. In this figure, paired phase contrast ((a), (c), (e), (g), (i), (k)) and fluorescence ((b), (d), (f), (h), (j), (l)) micrographs of the same field are presented for each of 6 sequential time points, as indicated. Transduced cNPCs exhibited consistent mophologies, continued growth, and sustained GFP expression throughout the period examined. Bars = 100
GDNF is known to have a range of biological activities in the context of the nervous system and cultured neural cell populations. Because this activity might extend to neural progenitors, we examined the effect of GDNF transduction on cNPC behavior, specifically the ability to proliferate. Proliferation is an important consideration for large-scale expansion of modified donor cell populations for use in transplantation studies. Transduced cNPCs continued to proliferate in a logarithmic manner, similar to but slightly slower than the nontransduced cNPCs (Figure
Growth properties of transduced versus nontransduced cNPCs. The growth of lenti-GDNF-GFP vector transduced cNPCs was compared to nontransduced cNPCs under proliferation conditions (UM). One flask of each type of cells was harvested and counted daily for 3 consecutive days. From this data it can be seen that the transduced cNPCs continued to proliferate despite overexpression of GDNF and that growth was similar to that of nontransduced cells out to day 2, after which the nontransduced cells exhibited relatively greater growth at the day 3 time point.
Neuronal differentiation has been implicated in gene silencing; therefore FACS analysis was performed to evaluate the effects of cell differentiation on GDNF transgene expression using the GFP reporter. Approximately 95% of transduced cNPCs expressed GFP, either when cultured in UM (proliferation conditions) or 10% FBS-containing UM (differentiation conditions). Among the cells expressing GFP, approximately 70% expressed GFP at high levels. There was no evidence of diminished GFP expression by the cells grown in the presence of FBS, thereby demonstrating maintained transgene expression was under differentiation conditions (Figure
Flow cytometric analysis of GFP expression after induction of differentiation. Nontransduced cNPCs and lenti-GDNF-GFP vector transduced cNPCs cultured under proliferation conditions (UM) were compared to transduced cNPCs cultured for 10 days in Ultraculture-based medium without EGF or bFGF and containing 10% FBS in order to induce differentiation (UM-FBS). Curve A: nontransduced cNPCs as negative controls; curve B: lenti-GDNF-GFP transduced cNPCs and curve C: lenti-GDNF-GFP transduced cNPCs in UM-FBS. Induction of differentiation did not attenuate expression of the GFP reporter gene.
The levels of GDNF produced by transduced cNPCs, as present in conditioned culture medium and collected cell lysates, were analyzed by ELISA and compared to nontransduced controls. High levels of secreted GDNF were present in the culture medium of transduced cNPCs, measured on days 28, 33, and 38 posttransduction (Figure
ELISA analysis of GDNF production by transduced cNPCs. (a) Lenti-GDNF-GFP vector transduced and nontransduced cells at passage 17 (cNPCp17) were seeded equally, under identical conditions, and allowed to grow for 15 days in UM, over which period the cells were passaged 3 times. At the time of each passage, culture media conditioned over the prior 48 hours was collected for ELISA assay. The conditioned media from transduced cNPCs was substantially enriched for GDNF compared to nontransduced cells. (b) Lenti-GDNF-GFP transduced and nontransduced cNPCp17 cells were trypsinized, lysed, and subjected to ELISA. GDNF was markedly elevated in lysates of transduced cells.
Having shown above that expression of the GFP reporter was sustained when transduced cNPCs were subjected to differentiation conditions, and that the transduced cells overexpress GDNF, we next verified that GDNF expression was sustained during cNPC differentiation (Figure
Effect of cell differentiation on transgene GDNF expression by ELISA. Lenti-GDNF-GFP vector transduced cNPCp24 cells were seeded equally in either UM (proliferation conditions) or Ultraculture-based medium without additional growth factors but containing 10% FBS (differentiation conditions, UM-FBS). Cultures were fed 24 hours prior to collecting GDNF conditioned media for ELISA assay at which time the cells were counted. ELISA data is presented as GDNF (ng) per million cells per day in order to further evaluate whether differentiation of transduced cNPCs had an influence on transgene expression. These data are consistent with sustained GDNF overexpression, confirming the flow cytometric data (Figure
Neural progenitor cells have shown great promise as a source of neural cell types in transplantation studies. We therefore investigated whether genetically modified cNPCs retained their neural progenitor phenotype in the presence of high levels of GDNF expression, as assessed by a gene expression profile (Figure
Expression profiles of cNPCs before and after transduction. The relative impact of GDNF overexpression on transcript expression levels was evaluated using qPCR analyses for a profile of 32 genes, which included
Immunocytochemical analysis demonstrated that cNPCs produced low levels of GDNF protein at baseline (Figure
GDNF expression by cNPCs before and after transduction and differentiation. Immunocytochemistry (ICC) was performed on cNPCs using a rabbit anti-human GDNF antibody to evaluate expression of GDNF at the protein level, before and after transduction and before and after exposure to growth factor deprived/FBS-containing differentiation conditions (UM-FBS). (a) Nontransduced cNPCp20 cultured in UM (proliferation conditions) exhibit baseline cytoplasmic labeling for GDNF (red). (b) Lenti-GDNF-GFP vector transduced cNPCs cultured in UM show increased intensity of GDNF (red) labeling. (c) Transduced cNPCs cultured in UM-FBS (differentiation conditions) are larger in size and show persistent overexpression of GDNF (red), that is, heterogeneously distributed among the profiles. Nuclear labeling = DAPI (blue), scale bar = 50
The expression of progenitor and lineage markers was also examined at the protein level, for both transduced and control cells, before and after induction of differentiation (Figure
Expression of NPC and lineage markers before and after transduction and differentiation. The effects of passage number, induction of differentiation and GDNF transgene expression on the expression of 5 markers was evaluated using ICC. Nontransduced and lenti-GDNF-GFP vector transduced cNPCp20 were cultured in UM or UM-FBS, then immunolabeled with specific antibodies. The changes in expression patterns seen predominantly reflected exposure to differentiation conditions (alternating columns), with little that might be attributable to passage number or lenti-GDNF-GFP transduction. Scale bar = 50
Among mammals, the highly developed visual system of the domestic cat has been studied in particular detail, owing in part to greater similarities with the human visual system as compared to laboratory rodents. This body of work, combined with the availability of naturally occurring retinal dystrophic mutants, would serve to recommend the cat as a powerful model for retinal regeneration research. A major limiting factor to regenerative research in this species is the paucity of available donor cells of the type suitable for such work, including stem, progenitor, or precursor cells of allogeneic origin. Furthermore, the use of these cells in transplantation studies would benefit from the inclusion of a reporter gene and, in some cases, additional transgenes of potential therapeutic value.
Here we demonstrate the feasibility of using feline lentiviral vectors to genetically modify cNPCs for sustained delivery of GDNF. These cells possess multiple desirable features for use in transplantation studies including ease of expansion in vitro, coexpression of a green fluorescence protein (GFP) reporter gene serving to both confirm GDNF expression as well as allowing easy tracking of donor cells after transplantation, and sustained transgene expression following differentiation. In addition, they are allogeneic with respect to the targeted host species and therefore likely to be well tolerated without for the need of exogenous immune suppression [
The ability of a progenitor cell to sustain proliferation is important in order to avoid the necessity of repeated rederivation of the modified cell type. Importantly, the GDNF-GFP overexpressing cNPCs continued to exhibit log growth characteristics, indicating that neither the genetic modification process nor GDNF overexpression presents a major barrier to continued proliferation of these cells. Nevertheless, the growth of the GDNF-transduced cNPCs was less rapid than that of unmodified controls. This slower growth rate is also reflected in the lower number of cells that were Ki-67 positive following introduction of the transgene construct. Since we have recently shown that exogenous GDNF tends to promote, rather than hinder, the growth of murine RPCs [
Another consideration in terms of clinical application of transduced cells is the regulation of transgene expression. Sustained overexpression might result in undesired effects such as decreased sensitivity to the gene product, as might result from down-regulation of the corresponding growth factor receptor or, alternatively, toxic responses to high levels of the cytokine, either within the eye or systemically. Titrating the dose of transplanted cells should set an upper limit on GDNF delivery, since the progenitor cells tend to cease proliferation in vivo, however, a more sophisticated approach would be the use of inducible promoters which allow for the dynamic regulation of transgene expression levels.
Looking forward, the GDNF-GFP overexpressing cNPCs developed here are suitable for allogeneic transplantation to the vitreous cavity or subretinal space of cats with retinal disease. Of particular interest is the application of these cells to existing animals with photoreceptor dystrophy, such as the Swedish Abyssinian breed with the CEP290 mutation [
The authors are grateful to Victor David for providing the cat-specific primer sequences used in this study and to Kristina Narfstrom for her longstanding involvement in cat models of retinal dystrophy as well as the provenance of fetal feline tissue used for our original derivation of feline neural progenitor cells. In addition, the authors would like to thank the Lincy Foundation, the Discovery Eye Foundation, the Andrei Olenicoff Memorial Foundation, and the Polly and Michael Smith Foundation for their generous financial support of this work, as well as a research grant from the Science and Technology Commission of Shanghai (09PJ1407200, to the second author). Dr. Joann You and Dr. Ping Gu contributed equally to this work.