|Molecular Vision 2000;
Received 24 January 2000 | Accepted 16 June 2000 | Published 24 June 2000
Retinal degeneration in cone photoreceptor cell-ablated transgenic mice
Saixia Ying,1,2 Heiko T. Jansen,2 Michael N.
Lehman,2 Shao-Ling Fong,3
Winston W. Y. Kao1
1Departments of Ophthalmology and 2Cell Biology, Neurobiology and Anatomy, University of Cincinnati, Cincinnati, OH; 3Department of Ophthalmology, Indiana University, Indianapolis, IN
Correspondence to: Winston W.-Y. Kao, Ph.D., Department of Ophthalmology, University of Cincinnati, Eden and Bethesda Ave, M. L. 527, Cincinnati, OH, 45267-0527; Phone: (513) 558-2802; FAX: (513) 558-3108; email: Winston.Kao@UC.edu
Purpose: To examine the effect of loss of cone photoreceptor cells on retinal degeneration.
Methods: We previously identified a cone photoreceptor cell-specific promoter of human cone transducin a-subunit (GNAT2) gene. In this report, a minigene, Trc-Tox176, that contains the GNAT2 promoter, an attenuated diphtheria toxin A-chain gene, and an enhancer element from human interphotoreceptor retinoid-binding protein (IRBP) was used to generate coneless transgenic mice. Transgenic mice were identified by PCR and the copy number of the transgene was determined by Southern hybridization, and examined by histology.
Results: The results of immunostaining with anti-mouse GNAT2 antibodies and reverse transcription-PCR (RT-PCR) analysis with mRNA from the retinas of transgenic mice showed that cone photoreceptor cells were ablated in one of four transgenic mouse lines. The ablation of cone cells began at postnatal day 8, at the same time as the expression of endogenous GNAT2. An age-related rod degeneration was also found in this cone-ablated mouse line, beginning at postnatal day 9, proceeding from the central retina to the peripheral retina.
Conclusions: Cone photoreceptor cells may play an important role in the survival of rod photoreceptor cells during mouse retina development.
During mouse retina development, cones are among the earliest postmitotic cells in the retina, while rods are the last [1,2]. Lineage tracing studies have shown that both rods and cones as well as other retinal cells can develop from a single precursor cell [3-7]. Therefore, local environmental cues and cell-cell interactions are thought to play an important role in photoreceptor development. Photoreceptors may also depend on each other for survival. In rd mouse, rod loss is followed by cone cell loss , and in retinitis pigmentosa (RP), the initial loss of rod function is usually followed by loss of cone-mediated vision [9-11]. In human macular dystrophies and cone/cone-rod dystrophies, the opposite holds true, loss of cone-mediated vision is followed by subsequent loss of rod-mediated vision [12-14]. However, there was no cone degenerated mouse model that demonstrates the influence of cone cells on the survival of rod cells.
In this report, we used a transgenic approach to selectively ablate cone photoreceptor cells by diphtheria toxin via the expression of a minigene containing the cone cell-specific GNAT2 promoter  and an attenuated diphtheria toxin A-chain gene [16,17]. This coneless transgenic mouse was used to study the consequences of cone cell loss on the survival of rod photoreceptor cells by examining the retina morphological changes.
Construction of Trc-Tox176 minigene
The 277 bp GNAT2 promoter and the 214 bp IRBP enhancer were released from plasmid pTc-151/+126 CAT (3' Enh)  by Hind III/Hinc II, and BamH I digestion, respectively. The 680 bp attenuated diphtheria toxin A cDNA (Tox176) was released from plasmid pIBI30-Tox176 (a generous gift from Dr. Maxwell)  by Hind III/Hinc II digestion. These three fragments were inserted into plasmid pb-gal-Basic (Clontech, Palo Alto, CA) by replacing the LacZ gene with Tox176, placing GNAT2 promoter upstream of Tox176 and placing the IRBP enhancer downstream of the SV40 intron and poly A site to create a new plasmid, pGNAT2-Tox176-IRBP. The 2.4 kb Trc-Tox176 minigene was then isolated from pGNAT2-Tox176-IRBP by Hind III/Sal I digestion.
Generation of transgenic mice
Animal experiments were performed in compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The 2.4 kb Trc-Tox176 minigene (Figure 1) released from pGNAT2-Tox176-IRBP was gel purified and microinjected into single-cell stage FVB/N embryos by the transgenic core facility in the Department of Molecular Genetics, University of Cincinnati. Transgenic mice were identified by PCR of mouse tail DNA with a primer pair chosen from within Tox176 gene, 5'-AACTTTTCTTCGTACCACGG-3' and 5'-ACTCATACATCGCATCT TGG-3'. Since FVB/N mice have a recessive mutation at the rd locus, the founders were bred with C57BL/6J mice to establish stable Trc-Tox176 transgenic mouse lines. Both C57BL/6J mice and C57BL/6J x FVB/N F1 hybrids have normal retinas.
Southern blot analysis was performed using DNA isolated from mouse tails to determine the copy number of the transgene. 10 mg of genomic DNA obtained from individual transgenic mouse tail was digested with Nco I to release the transgene fragments. Four sets of 10 mg genomic DNA from wild-type mouse were also digested with Nco I and mixed with 0, 1, 10 and 100 genome equivalents of the 2.4 kb Trc-Tox176 fragment to serve as copy number control. DNA was resolved on an agarose gel and transferred to a nylon membrane (MSI, Westborough, MA). The Southern blot was probed with the 680 bp Tox176 fragment and analyzed by PhosphoImager (Molecular Dynamics, Sunnyvale, CA).
RT-PCR was used to examine the relative levels of endogenous rod transducin a-subunit (GNAT1) and endogenous cone transducin a-subunit (GNAT2) messages. Retinas were dissected under RNase-free conditions from Trc-Tox176 transgenic mice and their wild-type littermates at P7, P8, P30 and P180 (postnatal days). Total RNA was isolated using TRI Reagent (Molecular Research Center, Inc., Cincinnati, OH) following the manufacturer's protocol. Reverse transcription was performed under the following condition: 10 mg of total RNA, reverse transcription buffer (Promega, Madison, WI), 10 mM dithiothreitol(DTT), 1 mM dNTPs, 2 units Rnasin (Promega), 2.5 mM random hexamers (Pharmacia Biotech, Piscataway, NJ) and 100 units AMV reverse transcriptase (Promega) in a volume of 200 ml and incubated at 42 °C for 90 min. The primer pairs used to detect GNAT1 cDNA and GNAT2 cDNA are, RT1, 5'-GAGGATGCTGAGAAGGATGC-3' and RT2, 5'-TGAAG GCTCTCGTGCATTCG-3'; CT1, 5'-GGACAAAGAACTTGCCAGGAGG-3' and CT2, 5'-CAAAGCAGGCTTGAACTCCACC-3', respectively. PCR for GAPDH was used as a positive control for the quality of the reverse transcription and the amount of cDNA added to each PCR reaction. The mouse GAPDH primer pairs were, 5'-ACAGCCGCAT CTTCTTGTGCAGTG-3' and 5'-GGCCTTGACTGTGCCGTTGAATTT-3'. The reaction conditions for all of the PCR experiments were: 2 ml 10X PCR buffer (Promega), 4 ml cDNA of reverse transcription product, 2 mM MgCl2, 1 mM each primer, 1 unit Taq DNA polymerase (Fisher Scientific, Pittsburgh, PA) in a total volume of 20 ml. The PCR cycling conditions were: denaturation at 94 °C for 1 min, annealing and elongation at 72 °C for 1 min, for 30 cycles.
Hematoxylin and Eosin staining was used to examine the morphological changes of the retinas in transgenic mice. Mouse eyes were removed from transgenic mice and their wild-type littermates at P8, P9, P30, P60, and P180. The eyes were fixed in 4% paraformaldehyde at 4 °C overnight, then dehydrated in a graded series of alcohol and embedded in paraffin. Five mm horizontal sections were cut through the entire eye. Every tenth section was mounted on a glass slide and stained with hematoxylin and eosin following the routine procedure.
Anti-Ga t2 rabbit polyclonal antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was used in immunostaining of mouse GNAT2 to confirm the presence or absence of cone photoreceptor cells in different transgenic mouse lines. Mouse eyes were removed from adult transgenic mice and their wild-type littermates and fixed in 4% paraformaldehyde at 4 °C overnight, then dehydrated in a graded series of alcohol and embedded in paraffin. Five mm horizontal sections were cut through the entire eye and every tenth section was mounted on a glass slide. After air drying, the tissue sections were deparaffinized in xylene and hydrated through a graded series of alcohols, rinsed with water and finally in 0.01 M PBS (10 mM NaPO4, 150 mM NaCl, pH 7.4). Sections were then incubated with 0.2% hydrogen peroxide in PBS for 10 min at room temperature, followed by incubation with 0.01 M PBS containing 5% normal goat serum and 3% milk at room temperature for 1 h. Sections were then incubated with primary antibody diluted (1:100) in 0.01 M PBS containing 3% milk and 0.25% Triton X-100 at 4 °C overnight followed by further incubation with a 1:250 dilution of goat anti- rabbit IgG horseradish peroxidase conjugate (Southern Biotechnology Associates, Inc., Birmingham, AL). The antibody reactions were visualized using diaminobenzidine (DAB) solution (Pierce, Rockford, IL). Sections were counterstained with hematoxylin. Normal rabbit IgG was used as negative control for immunohistochemistry.
Generation of Trc-Tox176 transgenic mice
Four transgenic founders (F0), three females (C8, C12, and C13) and one male (C98) were generated and identified by PCR and Southern blot analysis (data not shown). They were all established as stable transgenic mouse lines by breeding with C57BL/6J mice. The integrity of the transgene DNA and its copy number were determined by Southern blot analysis with Nco I that makes a single cut at 345 bp of the minigene. A 2.4 kb genomic DNA fragment was hybridized by the 32P-labeled cDNA of diphtheria toxin. The copy number of the transgene varied from 2 to 10 among the four founder lines. Transgenic mouse line 98 had approximately 10 copies of the minigene inserted in tandem into the genome of transgenic mouse line Tg98. Overexposure of the Southern blot hybridization did not yield evidence to support multiple insertion sites in the genome of Tg98 mice (data not shown). There were approximately 2 copies of transgene integrated in tandem in the other three lines, C8, C12 and C13 (Figure 2).
Characterization of the Trc-Tox176 transgenic mice
We examined the phenotypes of transgenic mice expressing diphtheria toxin encoded by the Trc-Tox176 transgene using RT-PCR analysis, H& E staining and immunostaining. In non-transgenic littermates, the expression of endogenous rod transducin a-subunit, GNAT1 and cone transducin a-subunit, GNAT2 began to be detected by RT-PCR at postnatal day 8, signal increased thereafter and remained at a constant level after postnatal day 30. RT-PCR also confirmed the expression of GNAT1 and GNAT2 by transgenic mouse lines C8, C12 and C13; even at a late stage, P180, the transducin messages could still be detected. There are no significant differences in either the time course or the intensity of the expression levels among these three transgenic lines and the wild-type mice (data not shown). Histological examination of H & E stained sections at higher magnification did not reveal morphological changes on retinas of these three lines as compared to the non-transgenic age-matched litter mates. The density of the cone photoreceptor cells was the same among these three lines and non-transgenic mice of P21 to P180 (data not shown). These animals were not examined further.
The transgenic mouse line 98 (Tg98) exhibited a marked phenotype following Trc-Tox176 transgene expression. In this mouse line, GNAT2 mRNA could not be detected by RT-PCR at P8 and afterwards. GNAT1 mRNA was dramatically reduced at the age of 6 months (Figure 3). Northern hybridization failed to detect the presence of Trc-Tox176 mRNA in the retinas of Tg98 transgenic mice of various ages (data not shown). The observations are consistent with the notion that ablation of cone cells in Tg98 mice coincides with the expression of the endogenous GNAT2, at a time the diphtheria toxin was synthesized from the transcription of Trc-Tox176 minigene.
To examine the morphological changes during retinal development in Tg98 mice, the retinas of Tg98 mice and non-transgenic littermates were subjected to histological examination. The results indicated that there was an age-dependent rod degeneration associated with the loss of cone photoreceptor cells in these mice. Rod degeneration began at postnatal day 9 (P9), proceeding from the central retina to the peripheral retina (Figure 4), and occurred in a sequence that was quadrant-specific, such degeneration proceeding more rapidly in the inferior quadrants than in the superior quadrants (data not shown). No significant morphological differences were noted between the retinas of Tg98 mice and wild-type littermates at P8. However, at P9, the overall lamination pattern seen in the adult retina is already evident, both the outer segments and the outer nuclear layer (ONL) of the central retina in Tg98 mice were thinner than those seen in the wild-type littermates. The ONL of Tg98 mice was only 3~4 layers of cell bodies, whereas the wild-type ONL was approximately 9 or 10 cell bodies thick. At this age (P9), the peripheral retina of Tg98 mice appeared normal. At P30, the ONL of the central retina in Tg98 mice was reduced to a single layer of photoreceptor cell bodies, in the peripheral retina, the ONL contained 2~3 layers of cell bodies. At P60 the entire ONL at the very central part of Tg98 mouse retina disappeared, in most of the peripheral parts, only one single cell layer remained. At P180, in many parts of the retina (i.e., central and peripheral regions especially in the ventral portion) the entire ONL was absent (Figure 4).
During development, maturation of the wild-type mouse retina is essentially complete by P21. At this age, the two classes of photoreceptor cells can be distinguished based on morphological criteria and their laminate location. At high magnification, the nuclei of rod photoreceptors that represent about 97% of the total photoreceptor population appear smaller and round with a single large clump of heterochromatin surrounded by very little euchromatin. In contrast, the nuclei of cone photoreceptor cells that is about 3% of the total photoreceptors are oval with one to several apparent clumps of heterochromatin surrounded by pale stained euchromatin. Rod cell bodies are distributed evenly throughout the ONL, while cone cell bodies are found only in the outer one-third of the ONL; most of them lie adjacent to the outer limiting membrane. At P21, cone cells were not observed in the remaining photoreceptor cells in the ONL of Tg98 mice (Figure 5).
Immunostaining using a polyclonal antibody to cone transducin a-subunit showed that the expressions of GNAT2 in transgenic mouse line 8, 12 and 13 were similar to that of wild-type mice (data not shown). However, GNAT2 could not be detected in transgenic mouse line 98 at the age of one month (Figure 6).
The absence of GNAT2 mRNA in Tg98 mice identified by RT-PCR is concurrent to the onset of expression of the endogenous GNAT2 in wild type mouse, suggesting that the cone photoreceptor cells were ablated as a result of expressed toxin. The absence of cone cells in Tg98 mice was confirmed by immunostaining for GNAT2 and the morphological criteria based on the heterochromatin pattern of the cone cells and their laminate location in the outer nuclear layer of the retina. The failure of cone cell ablation in the other three transgenic lines may have been caused by the integration of the transgene into an inactive region of the genome. Alternatively, level of diphtheria toxin synthesized may be very low due to low copy number of transgene in these mouse lines and is insufficient to ablate cone cells. This is consistent with the fact that Tg98 mice had the highest copy number of the transgene (approximately 10 copies per mouse genome) among the four transgenic mouse lines established (Figure 2).
The rod cell degeneration could be seen in the central retina of Tg98 mouse at P9, following the loss of cone cells as exemplified by the absence of GNAT2 in the transgenic mouse at P8 (Figure 3). Consistent with the cone cell distribution in mouse retina, the rod cell degeneration in Tg98 mice proceeding from the central retina to the peripheral retina. These temporal and spatial characteristics indicate that the rod cell degeneration observed in Tg98 mice was secondary to the loss of cone photoreceptors. A possible leakage effect of toxin seems untenable. Although DT-A could be released into the extracellular fluid after cone death, this form of the toxin could not be assimilated by adjacent cells because DT-A itself could not cross the cell membrane without the B subunit . DT-A is an effective cellular toxin . It kills cells by blocking protein synthesis via ADP-ribosylating elongation factor 2 . In our study, we used an attenuated diphtheria toxin A chain (Tox176) with a mutation of a Gly628 to Asp substitution . This attenuated form of the diphtheria toxin is useful for transgenic ablation studies because it causes less overall embryonic lethality . The observation is consistent with the notion that it may require a higher dose of the attenuated toxin to cause cell death. This can in part explain why the other three transgenic lines having fewer copy number of the transgene did not show retinal degeneration.
It is likely that the rod degeneration observed in present studies may be due to disruption of the normal environment attributed by cone photoreceptor loss. The potential mechanisms include the liberation of endotoxins by the killed cone cells or deprivation of cone-derived trophic factors. With respect to the latter, it is possible that cones produce a diffusible survival-promoting factor for rods. For example, neuron-specific enolase (NSE), a distinct form of the glycolytic enzyme enolase, is expressed in most terminally differentiated neurons and neuroendocrine cells . In mouse retina, immunocytochemistry revealed that NSE preferentially labeled cone photoreceptor cells in the outer nuclear layer . It was detected in the entire cone cell body as well as inner segment, both in adult and in developing retina, whereas, the rods are only weakly labeled at the level of the inner segments from around P8 to P10 . NSE has recently been implicated as an extracellular survival factor in both bovine brain and retina [24,25].
The maintenance of normal relationships between photoreceptor cells, the retinal pigment epithelium, and the interphotoreceptor matrix appears to be essential for photoreceptor viability. The cone-specific matrix has been implicated in the development and maintenance of photoreceptors, and serves as a major factor in retinal adhesion . Cone matrix sheath glycoconjugates likely play a major role in mediating retinal adhesion by forming a molecular bridge between the neural retina and the retinal pigment epithelium [26,27], which is involved in both the development and function of photoreceptors .
The underlying causes for the regional differences in rod cell death observed in our cone-ablated mouse are unknown. But, the more pronounced rod degeneration in the inferior retina may correlate with higher levels of incident light in this region. The superior to inferior gradients of intrinsic retinal molecules, such as retinoic acid, bFGF, etc. that are either protective or harmful to photoreceptors may also be involved in this phenotype [25,29,30].
Our observations on the rod photoreceptor cell degeneration in the cone-ablated transgenic mouse differ from those of Soucy et al., who used the attenuated diphtheria toxin-A gene under the control of 6.5 kb of 5' flanking sequences derived from the human red pigment gene (h.red DT-A) to create the coneless transgenic mouse. The retinas of these transgenic mice had normal numbers of rod photoreceptors. It is possible that the timing of onset and completeness of cone cell ablation may influence the survival of rod cells. For example, cone pigment expression is first detectable at postnatal days 0-3; however, the time course of cone cell death in the h.red DT-A transgenic mouse has not yet been determined. It should be noted that the pattern of transgene expression driven by human red opsin promoter did not transpire in all murine green and/or uv cone cells. In fact, a small number of cone cells survive in h.red DT-A transgenic mice . Likewise, the small number of surviving cone cells may produce sufficient factors for the maintenance of rod cells in the transgenic mice. It is also possible that the different phenotypes of these two coneless transgenic mouse models may be due to different mouse genetic backgrounds. Soucy et al. generated the coneless mouse from DNX backcrossed to C57BL/6J, and ours are hybrids from FVB/N crossed with C57BL/6J.
Other explanations such as leaky expression of DT-A by rod cells and the disruption of rod survival related gene by the transgene insertion may account for the rod degeneration in Tg98 mice. The former proposition seems unlikely because of the following reasons: In our previous study, the expression of CAT activities by GNAT2-CAT minigene is restricted to cone cells of the transgenic mice . In this study, the rod degeneration proceeded from the central retina to the peripheral retina, a pattern similar to the distribution of cone cells in mouse retina, implicating that the rod degeneration could be caused by the loss of the support from cone cells. However, one cannot rule out the latter possibility that in line Tg98, insertion of transgene may interrupt rod survival related genes. The founders of our transgenic mouse lines were created with the FVB/N mouse strain that has a recessive mutation of the Pdeb rd-1 gene (from Mouse Genome Informatics database). The transgenic mice analyzed in present studies were offspring of FVB/N and C57BL/6J strains. The transgenic mice and their littermates have one allele of functional Pdeb rd-1 gene. The non-transgenic littermates do not have pathological changes in retina. These observations are consistent with the notion that interruption of the Pdeb rd-1 locus by the transgene insertion could not account for the rod degeneration observed in Tg98. However, mutation caused by insertion of the transgene at other loci important for rod cell survival is still possible. Further studies, by FISH (Fluorescent In Situ Hybridization), 5'- or 3'-RACE, will be needed to localize the insertion site(s) of the transgene in line Tg98 to examine this possibility.
In summary, our finding that the degeneration of rod photoreceptor cells in the cone cell-ablated transgenic mouse is consistent with the notion that cone cells may play a role in the survival of rod cells during mouse retina development.
We thank Dr. Ian H. Maxwell for supplying pIBI30-Tox176 DNA; Dr. Michael Behbehani, Dr. Chia-Yang Liu, Dr. Maureen A. McCall and Dr. Linda Parysek for their valuable advice; Mr. Richard Converse and Ms. Candace Kao for their excellent technical assistance. This study was supported by NIH grants EY 10556, Ohio Lions Eye Research Foundation (Columbus, OH) and Quest of Vision, Department of Ophthalmology, University of Cincinnati (WWY Kao).
1. Sidman RL. Histologenesis of the mouse retina studied with thymidine-3H. In: Smelser GK, editor. The structure of the eye. Proceedings of the 7th International Congress of Anatomists; 1960 Apr 11-13; New York, New York. New York: Academic Press; 1961. p. 487-506.
2. Robinson SR. Development of the mammalian retina. In: Dreher B, Robinson SR, editors. Neuroanatomy of the visual pathways and their development. Boca Raton: CRC Press; 1991. p. 69-128.
3. Turner DL, Cepko CL. A common progenitor for neurons and glia persists in rat retina late in development. Nature 1987; 328:131-6.
4. Wetts R, Fraser SE. Multipotent precursors can give rise to all major cell types of the frog retina. Science 1988; 239:1142-5.
5. Holt CE, Bertsch TW, Ellis HM, Harris WA. Cellular determination in the Xenopus retina is independent of lineage and birth date. Neuron 1988; 1:15-26.
6. Turner DL, Snyder EY, Cepko CL. Lineage-independent determination of cell type in the embryonic mouse retina. Neuron 1990; 4:833-45.
7. Goldowitz D, Rice DS, Williams RW. Clonal architecture of the mouse retina. Prog Brain Res 1996; 108:3-15.
8. Jimenez AJ, Garcia-Fernandez JM, Gonzalez B, Foster RG. The spatio-temporal pattern of photoreceptor degeneration in the aged rd/rd mouse retina. Cell Tissue Res 1996; 284:193-202.
9. Massof RW, Finkelstein D. A two-stage hypothesis for the natural course of renitis pigmentosa. In: Zrenner E, Krastel H, Goebel HH, editors. Research in retinitis pigmentosa. Proceedings of the 4th Congress of the International Retinitis Pigmentosa Association; 1986 May 10-13; Bad Nauheim, West Germany. New York: Pergamon; 1987. p. 29-58.
10. Dryja TP, Li T. Molecular genetics of retinitis pigmentosa. Hum Mol Genet 1995; 4 Spec No:1739-43.
11. Petters RM, Alexander CA, Wells KD, Collins EB, Sommer JR, Blanton MR, Rojas G, Hao Y, Flowers WL, Banin E, Cideciyan AV, Jacobson SG, Wong F. Genetically engineered large animal model for studying cone photoreceptor survival and degeneration in retinitis pigmentosa. Nat Biotechnol 1997; 15:965-70.
12. Berson EL, Gouras P, Gunkel RD. Progressive cone-rod degeneration. Arch Ophthalmol 1968; 80:68-76.
13. Sadowski B, Zrenner E. Cone and rod function in cone degenerations. Vision Res 1997; 37:2303-14.
14. Szlyk JP, Fishman GA, Alexander KR, Peachey NS, Derlacki DJ. Clinical subtypes of cone-rod dystrophy. Arch Ophthalmol 1993; 111:781-8.
15. Ying S, Fong SL, Fong WB, Kao CW, Converse RL, Kao WW. A CAT reporter construct containing 277bp GNAT2 promoter and 214bp IRBP enhancer is specifically expressed by cone photoreceptor cells in transgenic mice. Curr Eye Res 1998; 17:777-82.
16. Maxwell F, Maxwell IH, Glode LM. Cloning, sequence determination, and expression in transfected cells of the coding sequence for the tox 176 attenuated diphtheria toxin A chain. Mol Cell Biol 1987; 7:1576-9.
17. Harrison GS, Maxwell F, Long CJ, Rosen CA, Glode LM, Maxwell IH. Activation of a diphtheria toxin A gene by expression of human immunodeficiency virus-1 Tat and Rev proteins in transfected cells. Hum Gene Ther 1991; 2:53-60.
18. Maxwell IH, Maxwell F, Glode LM. Regulated expression of a diphtheria toxin A-chain gene transfected into human cells: possible strategy for inducing cancer cell suicide. Cancer Res 1986; 46:4660-4.
19. Yamaizumi M, Mekada E, Uchida T, Okada Y. One molecule of diphtheria toxin fragment A introduced into a cell can kill the cell. Cell 1978; 15:245-50.
20. Collier RJ. Diphtheria toxin: mode of action and structure. Bacteriol Rev 1975; 39:54-85.
21. Burrows HL, Birkmeier TS, Seasholtz AF, Camper SA. Targeted ablation of cells in the pituitary primordia of transgenic mice. Mol Endocrinol 1996; 10:1467-77.
22. Marangos PJ, Schmechel DE. Neuron specific enolase, a clinically useful marker for neurons and neuroendocrine cells. Annu Rev Neurosci 1987; 10:269-95.
23. Rich KA, Zhan Y, Blanks JC. Migration and synaptogenesis of cone photoreceptors in the developing mouse retina. J Comp Neurol 1997; 388:47-63.
24. Takei N, Kondo J, Nagaike K, Ohsawa K, Kato K, Kohsaka S. Neuronal survival factor from bovine brain is identical to neuron-specific enolase. J Neurochem 1991; 57:1178-84.
25. Li A, Lane WS, Johnson LV, Chader GJ, Tombran-Tink J. Neuron-specific enolase: a neuronal survival factor in the retinal extracellular matrix? J Neurosci 1995; 15:385-93.
26. Hageman GS, Marmor MF, Yao XY, Johnson LV. The interphotoreceptor matrix mediates primate retinal adhesion. Arch Ophthalmol 1995; 113:655-60.
27. Fleming PA, Braekevelt CR, Harman AM, Beazley LD. Retinal pigment epithelium and photoreceptor maturation in a wallaby, the quokka. J Comp Neurol 1996; 370:47-60.
28. Hollyfield JG, Witkovsky P. Pigmented retinal epithelium involvement in photoreceptor development and function. J Exp Zool 1974; 189:357-78.
29. Sanes JR. Topographic maps and molecular gradients. Curr Opin Neurobiol 1993; 3:67-74.
30. Hyatt GA, Schmitt EA, Fadool JM, Dowling JE. Retinoic acid alters photoreceptor development in vivo. Proc Natl Acad Sci U S A 1996; 93:13298-303.
31. Soucy E, Wang Y, Nirenberg S, Nathans J, Meister M. A novel signaling pathway from rod photoreceptors to ganglion cells in mammalian retina. Neuron 1998; 21:481-93.