- Email: [email protected]
A new procedure for multiple intraretinal transplantation into mammalian eyes
Journal of Neuroscience Methods, 43 (1992) 157-169
157
© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-0270/92/$05.00 NSM 01382
A new procedure for multiple intraretinal transplantation into mammalian eyes 1 Eliot Lazar and Manuel del Cerro Departments of Neurobiology and Anatomy, and Ophthalmology, University of Rochester Medical School, Rochester, New York 14642 (USA) (Received 4 October 1991) (Revised versions received 15 January 1992 and 2 April 1992) (Accepted 3 April 1992)
Key words: Transplantation; Retina; Neural grafts; Fetal tissue; Intra-ocular grafting; Retinoblastoma A new method of intraretinal grafting is described which avoids opening the globe and allows direct visual control of the transplant by observing placement of the graft through the host pupil. A microsyringe with a bevelled needle partially ensheathed in plastic tubing in order to limit penetration was used to inject dissociated donor cells into 2 to 4 pre-selected points of the sub-retinal space of adult rats. The needle tip was inserted through the scleral and choroidal tissue, and observed through the animal's pupil as it reached into the sub-retinal space. Donor cells were then injected into several sites of the retina. Results, using ophthalmoscopy as well as light and electron microscopy, confirmed the atraumatic nature of this technique and the survival of grafted cells. Intraretinal injection of colloidal carbon visually illustrated the effectiveness of this method in covering an extended portion of the host retina. The procedure, which is virtually free of complications has enabled us to perform multisite intra-ocular grafting in a safe, easy, and reliable fashion, under direct visual control.
Introduction
Since research on intra-ocular retinal transplantation between unrelated individuals began (del Cerro et al., 1984; Gouras et al., 1984; Turner and Blair, 1985), the field of retinal transplantation and the techniques to perform this operation have been undergoing constant evolution. Although successful for seeding small areas of the host retina with donor cells, these procedures suffer from definite limitations. Features corn-
Correspondence: Manuel del Cerro, M.D., Departments of Neurobiology and Anatomy, and Ophthalmology, University of Rochester Medical School, Rochester, New York, 14642, USA. Tel.: (716) 275-4027; Fax: (716) 442-8766. t Supported by NEI Grant No. 05262, Rochester Eye Bank, and generous private gifts. Some results from this work have been presented in abstract form (Lazar and del Cerro, 1990).
mon to the retinal transplantation methods thus far described are that they necessitate surgically opening the eye and suturing the incision. These requirements carry the potential for surgical complications, such as vitreous loss, retinal detachment, and ocular hemorrhage. These inherent risks may have discouraged attempts at multiple implantation since, to date, none have been reported. A further limitation of existing procedures is that, when the sub-retinal space is reached for cell delivery, it is done in a blind fashion with the needle tip being out of view of the observer at this critical moment. Thus realtime documentation of the transplants while they are being performed has not been provided in the literature. We describe here an injection technique for intra-ocular transplantation which does not require a surgical incision, while allowing us to retain visual control of the grafting procedure
158
throughout. Results obtained by transplanting suspensions of fetal neuroretinal cells or Y79 human retinoblastoma cells using this technique are presented, as are sub-retinal injections of colloidal carbon made in similar fashion into control animals. The method has already been used successfully to achieve functional visual recovery in blinded rats (del Cerro et al., 1991).
Materials and methods
Donor cells Several different types of donor tissue have already been used for intraretinal transplantation applying this technique. Here, we illustrate the use of human retinoblastoma Y79 cells and human embryonic neuroretinal cells. For demonstration purposes, the injection of a suspension of colloidal carbon using the same procedure is shown. Cell preparation Human retinoblastoma Y79 ceils obtained from the American Type Culture Collection were grown under standard conditions in suspension in RPMI medium supplemented with 10% fetal calf serum (FCS) (heat inactivated, Hyclone) 0.03% glutamine and 5 0 / z g / m l gentamycin at 37°C with humid 95% air, 5% CO 2. For transplantation studies, Y79 cells were transferred to coated, 100-mm tissue culture dishes for attachment and differentiation. Dishes were first treated with 100 izg/ml polylysine D for 30 min at 37°C followed by exposure to fibronectin 5 / x g / m l for 30 rain at 37°C to generate an adhesive surface. Y79 cells (2 x 106/dish) were cultured overnight in standard medium for adhering cells. Monolayers were
A
C
J
F Fig. 1A. Diagram illustrating the intact globe method of retinal transplantation. U n d e r direct visualization, using a posterior, external approach, a needle is inserted at a predetermined depth into the subretinal space. The graft is then injected at the desired site.
rinsed with CMF buffer and with E D T A and treated with 0.125% trypsin (Gibco) for 10 min at 37°C to remove cells from the adherent surface. Trypsin was quenched with 10% calf serum, the cells were centrifuged at 900 rpm and then rinsed with buffer. After pelleting, cells were suspended in cold human plasma at a concentration of 4000 cells/iz 1. Second-trimester human embryonic retinal cells obtained from electively aborted embryos aged 13-17 weeks were used as donor tissue. Procurement of the donor cells was in strict accordance with scientific and ethical guidelines, which included institutional review and approval of the experimental protocol. In all cases, maternal consent was given only after the decision to have an elective abortion was made. The eyes
Fig. lB. Macrophotograph showing the needle tip, as viewed by the operator, following penetration of the sclera, as it enters the sub-retinal space. Note that the presence of the needle in the sub-retinal space has caused an elevation of a retinal vessel. No hemorrhage has occurred at the penetration point. The atraumatic nature of the procedure permits us to repeat the injection at more than one location in the same eye. Magnification = 7.5 x . Fig. 2. A: this schematic indicates how the material immediately spreads out upon completion of the injection. As the retina was not perforated, the injected fluid was dispersed through the shallow sub-retinal space, and resulted in extended coverage being obtained by injecting only a small volume of material. B: macrophotograpb obtained within seconds of injecting 2 /xl of colloidal carbon. This relatively small volume of fluid covered approximately 120 ° of retinal surface. Magnification = 7.5 x .
159
160 were obtained within an hour of abortion and collected in either calcium-magnesium free medium or in human plasma at 4°C. Operating with the aid of a surgical microscope, the eyes were dissected open. The retinas were cut away, free of contamination from either the vitreous or the retinal pigment epithelium. Isolated retinas were trimmed into small fragments and then placed into ice-cold medium. Mechanical dissociation was used to obtain suspensions of retinal cells and cell clusters. The retinal fragments were dissociated by aspiration and injection through a 27-30-ga hypodermic needle.
Hosts and anesthesia Two to three groups of 6 - 8 male adult Fischer 344 rats were used per experiment. A total of 120 hosts were used for the experiments reported here. The animals were anesthetized with a mixture of ketamine (60 m g / k g ) and sodium pentobarbital (21 mg/kg). Topical 1% Alcaine drops (proparacaine hydrochloride, Alcon, Fort Worth, TX) were also used. The eyes were dilated with 1% neosynephrine and 1% mydriacyl (Alcon, Fort Worth, TX). Cyclosporine A, at a dose of 10 m g / k g / d a y , was given to the rats after transplantation of human cells. Delivery system and transplantation procedure A 27-ga Butterfly needle with a 15 ° bevel, tightly sheathed in plastic, with 1.1-1.4 mm of the needle tip left exposed, was connected to a 10 or 50/xl syringe (Series 1700, Hamilton, Reno, NV) prior to the procedure. The plastic sheath placed on the needle served as a stop to limit the depth of penetration. Its length was set so as to leave only enough of the needle tip exposed to reach the sub-retinal space without actually penetrating the retina. This proved highly effective in preventing the accidental formation of retinal holes or tears. The microsyringe was preloaded with the cell suspension. Collibri forceps (Storz, St. Louis, MO) were used to grasp the sclera and rotate the globe anteriorly. Then, using a stereomicroscope fitted with a 35-mm photographic camera and videotaping apparatus for direct visualization, the needle was manually inserted through the sclera with the
bevel facing the operator in order to afford the best view. It was then gently rotated without changing its angle until the tip, viewed through the pupil, could be seen in the sub-retinal space (Fig. 1B). Then, the tip was advanced further so as to elevate the retina slightly (Fig. 2B). At this point, an injection of 2 - 4 ~1 of cell suspension was made. The procedure was repeated at a point 180° opposite to the first injection site in the same eye. Typically, 2 micro-injections were made into the equatorial region of each eye. One was made superiorly at the 12 o'clock position and the other at the 6 o'clock position inferiorly. However, as many as 4 penetrations were performed in the eyes of some animals. The needle was quickly withdrawn following each injection, with no obvious reflux. After the experiment was completed, a topical lubricant (Tobrex: tobramycin ophthalmic ointment; Alcon, Ft. Worth, TX) was placed on the cornea in order to prevent drying.
Control injections In order to obtain instant, permanent, and multilevel visualization of the spread of the injected fluid, 12 retinas were injected with colloidal carbon (particle size from 20 to 30 nm). The colloidal carbon (Pelikan, Hannover, Germany) was prepared in a 1 : 4 dilution with saline and then used to perform control injections (del Cerro et al., 1985; Triarhou and del Cerro, 1985) in order to evaluate and demonstrate the extent of dispersion of the injected particles over the host retina. Clinical observations All the animals were observed at several points during the course of the experiment. Routinely, animals were examined prior to any procedure in order to obtain a baseline status. Following surgical intervention, they were examined at postoperative day 1, and then weekly for the remainder of the experiment. Clinical exams included in every case possible, the anterior segment, Schiotz tonometry, and a full funduscopic exam. Surviual times and histological procedures Survival times for hosts receiving cell grafts ranged from 3 to 90 post-transplantation days
161 (PTD). Control animals which received injections of 2 - 4 /xl of colloidal carbon were sacrificed 4 days following intra-ocular injections, because at this point no free colloidal carbon can be found in the sub-retinal space, which may confuse the results. The animals were enucleated and sacrificed under deep anesthesia using intramuscular injections of ketamine at 90 m g / k g and Rompun (xylazine, a sedative and analgesic) at a dose of 8 mg/kg. The eyes were fixed in 6% glutaraldehyde and then split along a sagittal axis. The hemisected eyes were examined and photographed under a stereomicroscope and were then embedded in plastic (Eponate 12; Ted Pella, Redding, CA). One micrometer-thick sections were cut and stained with Stevenel Blue (del Cerro et al., 1980) for light microscopic study. Ultrathin sections were cut for electron microscopic studies. The retinas from eyes injected with colloidal carbon were dissected free and prepared as flat-mounts.
Results
Fig. 1A illustrates the 2 basic principles of this procedure: (1) predetermined, limited penetration, and (2) direct visualization of the needle tip within the sub-retinal space. In fact the entire sequence, from the needle penetrating the sclera, though the release of material, could be viewed directly under the operating microscope (Figs. 1A and 2A). It was possible to observe and to photograph the needle as it pushed the retina upwards, once it had reached the sub-retinal space (Figs. 1B-2B). At this point, the bevel was rotated under direct visual control in order to place it in direct apposition to the retina. Then, the injection was carried out (Fig. 2B). When colloidal carbon was injected for demonstration purposes, it could be seen to spread out readily over the retinal surface. It quickly fanned out and covered up to 12 mm 2 or 120° of retinal surface per injection (Fig. 3), dramatically illustrating the wide
Fig. 3. Photograph of a hemisected eye which received colloidal carbon 4 days previously, as shown in Fig. 2B. In this particular case, a single injection of India ink covered a surface area of the retina as large as 12 mm2. Magnification= 40 x.
163
Fig. 5. A: h u m a n donor retinal cells ( H R C ) were transplanted into the host subretinai space using the procedure described in the text. This micrograph shows that the donor cells are well preserved into the rat eye 33 days following transplantation. T h e nuclear characteristics of the H R C contrast sharply with those of the host photoreceptors. Neither the host neural retina (RET) nor its pigment epithelium (RPE) demonstrate any visible signs of t r a u m a or an inflammatory response. Ultrastructural features of the junction areas between the host neural retina and RPE, and the transplant are seen in B and C. Magnification = 600 x . Fig. 5B and C: electron micrographs of the graft illustrated in A. B: the junction between the grafted h u m a n retinal cells (HRC) and host neural retina (RET). The outer tips of the outer segments, now separated from the pigment epithelium, form whorls of m e m b r a n o u s material (arrows). Magnification = 5150 x . C: the junction between the host retinal pigment epithelium (RPE) and the grafted h u m a n retinal cells (HRC). Even at this level of magnification there is a close contract between the graft and host tissues. There is also a remarkable absence of any signs of trauma or inflammation. Magnification = 16 000 x .
diffusion of the injected fluid over the surface of the host retina. Access to any portion of the retina, even as far posteriorly as the optic nerve head was possible. Intra-operative results, from both control experiments using colloidal carbon and actual cell grafts, clearly demonstrated that this material was delivered precisely into the sub-retinal space (Fig. 4B and C).
Clinical observations have confirmed the atraumatic nature of this technique. The injection tracks were self-sealing and thus required no surgical closure. There was no vitreous loss at the time of injection and postoperative examinations showed no corneal opacities, lenticular changes, or optic nerve head damage. Intra-ocular pressure, measured using a Schiotz tonometer, re-
Fig. 4. A: this low-power light micrograph shows grafts of Y79 h u m a n retinoblastoma cells at 2 points in the same rat eye. Of the 2 grafts (GR), the one to the left is shown at higher magnification in Fig. 4B, and the one at the right is shown in Fig. 4C. Magnification = 25 x . B and C: illustration of the junction areas between the host neuroretina and the sub-retinal masses of transplanted Y79 retinoblastoma cells (GR). In B, the growth of the grafted cells displaces the host retina (RET) upwards, whereas C shows the mass of the tumor growing into the subretinal space. In both cases, the lack of hemorrhage, macrophage invasion, or pigment epithelium (RPE) reaction is evident. Magnification = 600 x .
164
Fig. 5. (continued).
mained stable throughout. There were no pre- to postoperative elevations in pressure measurements at any time. The only changes seen by
direct and indirect ophthalmoscopy were 7 subretinal hemorrhages, 5 of them petechial in size, all of which resolved within 2 weeks. No other
165
Fig. 5. (continued). clinical complications were noted. In 120 rats of all ages which have been transplanted, we have had no additional adverse side effects of complications secondary to our method.
Histological results correlated quite well with the clinical observations just described. Sections of all the eyes injected with the tracer, as well as from those injected with Y79 retinoblastoma cells
166 (Figs. 4A-C) or fetal retinal ceils showed considerable dispersion of the injected material. This was seen to spread onto the outer retinal surface from the sub-retinal injection point. In histological specimens involving colloidal carbon injections o f 2-4 /zl, there was clearly intraretinal colloidal carbon material scattered throughout, at distances of more than 2000/xm from the penetration point. The frequency of successful transplantation compared favorably with the open approach previously used in this laboratory. The results of studies involving transplantation of human fetal retinal cells or Y79 retinoblastoma using this procedure are the subject of 2 separate reports (Del Cerro et al., 1992). A summary description shows that typically the penetration point was marked by the presence of a comparatively large cluster of cells, while smaller cell clusters were found distributed mostly throughout the hemisphere where the injection had occurred. The human fetal retinal cells survived well and accumulated into clusters, with rosette formation being a common occurrence (Fig. 5A and B). The ceils bordering the lumen of the rosettes had a chromatin distribution similar to that of in situ developing rod ceils. No appreciable apical differentiation, i.e., inner or outer segments, was detected. Their nature as photoreceptor cells was nonetheless established by the development of presynaptic differentiation in many of the cells, including the formation of presynaptic ribbons (data not shown) with associated clusters of vesicles. Actual ribbon synapses were found in the band of neuropile which surrounded these ceils. In keeping with the protracted development of the human neural retina, mitotic activity continued within the grafted cell population for periods of up to 3 months post-transplantation, the longest duration studied. Grafts of Y79 human retinoblastoma ceils initially developed within the confines of the subretinal space. Growth of the grafted cells induced the host retina to become progressively detached from the retinal pigment epithelium. The separation, however, was strictly limited to those areas occupied by the grafted cells. Yet, even at those points of detachment, the host neural retina ad-
hered closely to the tumoral cells (Figs. 4B, 5A). The host retina, including the retinal pigment epithelium, showed a remarkable lack of reaction to the mechanical disturbance from the surgical manipulation or the presence of the grafted cells (Fig. 5A-C). The tips of the detached photoreceptor cells, deprived from the phagocytic activity of the pigment epithelium, exhibited anomalous accumulations of degenerating cisternal material (Fig. 5B). As a rule, a minute disruption of Bruch's membrane was the only visible histologically indication of the penetration point. In the numerous transplants performed using this technique, the method has been complication-free. In 120 cases, the vitreous has been entered twice as seen by direct visualization and documented microscopically. In those two instances histologically demonstrable lacerations of the neural retina were observed.
Discussion
This study reports on a new method for performing intra-ocular transplantation which, unlike previously described methods can be performed under direct visual control and does not require surgically opening the eye. In the following paragraphs we will first discuss features of techniques previously used for retinal transplantation and then illustrate some of the experimental results obtained while using the present method. From its inception, the techniques used for intra-ocular retinal transplantation cited in the literature have uniformly required a surgical incision. A survey of the published methods underlines their features. Our initial retinal transplantation experiments were performed using glass micropipettes. A scleral incision was made, and the transplant was then performed. This technique, although effective in producing viable grafts, was difficult to perform and time consuming. Suturing of the minute incision often resulted in complications such as retinal detachments, sub-retinal hemorrhages, or formation of intravitreal membranes. A modified approach using self-sealing incisions (del Cerro et al., 1988, 1989),
167 decreased the formation of preretinal membranes and sharply reduced the incidence of hemorrhage. It did not, however, overcome the limitations common to all the procedures thus far described, of being unable to visualize the location of the needle tip at the moment of transplantation and of being able to consistently place the transplants at the proper depth. The method used by Turner et al. (Blair and Turner, 1987; Li and Turner, 1989; Sheedlo et al., 1989), for transplantation of R P E cells, required a preliminary excision of the superior rectus muscle, followed by an incision with a blade with penetrated the sclera and choroid to expose a bleeding surface. Surgical closure with suture material was required. The visual disturbance caused by a detached superior rectus muscle, which was not reportedly re-attached, was not addressed. Also, it was not made clear how the blade was stopped after cutting through the retinal pigment epithelium, but before it had reached the adjacent neural retina. Silverman and Hughes (1989) described a transplantation technique based on a transcorneal approach to the sub-retinal space. This required a wide incision through the cornea and then traversing the entire globe with a carrier 2.5mm-wide and 0.5-mm-thick, which was made to reach the posterior pole prior to delivering a tissue slab mounted on the carrier. Lopez et al. (1987) used a technique which required a conjunctival flap, a pars plana incision, and the insertion of a micropipette into the subretinal space. They noted that "there are a number of potential pitfalls associated with this technique", and pointed out that, "it may be difficult to be certain that no cells enter the vitreous cavity". The comments by Lopez and collaborators summarize the limitations of the methods used for achieving retinal transplantation, including those originally used in our own work. This prompted us to devise a more precise means of intra- or sub-retinal transplantation based on the principles of limited pre-set penetration of the needle used to deliver the donor cells and direct visual control through the dilated pupil of the host eye. Such a method would offer the added
advantage of allowing real time photography or video to document the transplantation procedure. It would also make multi-site grafting not only possible but reliable and nearly free of complications. Most intra-ocular grafts of neural retinal cells or retinal pigment epithelium, including ours, have been made using a transcleral or posterior approach. Some workers have instead used a transvitreal or anterior approach. Wongpichedchai et al. (1990; personal communication, 1992) compared the results obtained by using these two routes to implant retinal epithelium. They found that the external approach resulted in better grafts, although the lack of direct visualization using the external approach was a limitation. Our experience with the external approach matches that of Wongpichedchai et al. (1990). The desire to overcome the lack of visual control while using the external approach was one of our main reasons for developing the present method. No neovascularization, uveitis, vitreous hemorrhage, or ocular infections were seen in the eyes of 120 rats transplanted using this procedure. The microscopic findings correlate perfectly with these resuits. Histological examination has shown neither retinal detachments nor any evidence of transplant rejection. Comparison of results obtained using this procedure and those obtained by other laboratories are made difficult by the fact that most of the published studies have been made using donor cell-host combinations which differ from each other. For example, Aramant et al. (1990) and Ehinger et al. (1991) have transplanted human fetal retinal cells into rat retinas, but their donor tissues are first trimester versus second trimester (del Cerro et al. 1990). This fact may account for the extensive layer differentiation exhibited in their transplants. On the other hand, this effect may be related to the fact that as their technique involves incision of the neural retina, many of the grafts extend into the vitreal cavity, while ours remain confined to the subretinal space. The results obtained by Gouras et al. (1991) transplanting tritiated labeled photoreceptors into the subretinal space of RCS rats are more comparable to those obtained here by transplanting
168
fetal rat retinal cells into light blinded animals by means of the present procedure (del Cerro et al., 1991, 1992). In both cases, photoreceptor cells are seen to be distributed along the outer surface of the neural retina. The method used by Gouras involves the very delicate procedure of transchoroidally inserting a glass micropipette through a scleral incision. No mention is made of the surface extension covered by these single site grafts or of the steps required to stop the advancement of the micropipette precisely within the sub-retinal space. Silverman and Hughes (1989) transplanted tissue slabs containing photoreceptor cells into the sub-retinal space of light-damaged eyes with reported success. Their donor cells were taken from post-natal donors, that is an older age than that of the donors used in any of our studies (del Cerro et al., 1989, 1991). Since it has not been possible to duplicate their published procedure in our laboratory, we can not directly compare the 2 methods. The results of actual retinal transplantation using human fetal neuroretinal cells, Y79 human retinoblastoma cells, or intraretinal injections of colloidal carbon, into control animals, show that a considerable portion of the host retinal surface is covered by the explant. Although different in size than the carbon particles, the cells spread throughout the sub-retinal space in a like fashion, driven by the hydrostatic pressure of the injected fluid. The shallowness of the sub-retinal space forced the cells to extend over a considerable area, thus vastly improving the odds for functional success. The procedure has already been used with success in multisite transplantation of fetal retinal cells into the retinas of blind adult rats, which regained a measure of visual function (del Cerro et al., 1991, 1992). In summary it can be said that this method offers a previously unattainable degree of simplicity and control in the performance of retinal transplantation.
Acknowledgements The authors wish to thank Lazaros C. Triarhou, M.D., Ph.D., and David Diloreto, Jr., for their critical reviews of the manuscript. Drs. Gaff
Seigel and Mary Notter (deceased) provided valuable assistance with the Y79 cells. We would also like to thank Anita Matthews for her graphics contribution.
References Arament, R. Seller, M., Ehinger, B., Bergstrom, A., Gustavii, B., Brundin, P. and Adolph, A. (1990) Transplantation of human embryonic retina to adult rat retina, J. Restor. Neurol. Neurosci., 2: 9. Blair, J.R. and Turner, J.E. (1987) Optimum conditions for successful transplantation of immature transplantation of immature rat retina to the lesioned adult retina, Dev. Brain Res., 36: 257. Del Cerro, M., Cogen, M.J. and del Cerro, C. (1980) Stevenel Blue, an excellent stain for optical microscopy study of plastic embedding tissues, Microscop. Acta, 83: 217-220. Del Cerro, M., Gash, D.M., Notter, M., Rao, G.N., Wiegand, S. and Ishida, N. (1984) Intraocular retinal transplants, Invest. Ophthalmol. Vis. Sci., 25:62 (abst.). Del Cerro, M., Grover, D.A., Dematte, J.E. and Williams, W.M. (1985) Colloidal carbon as a combined ophthalmoscopic and microscopic probe of the retinal-blood barrier integrity, Phtbal. Res., 17: 34. Del Cerro, M., Notter, M., Wiegand, S., Jiang, L. and del Cerro, C. (1988) Intraretinal transplantation of fluorescently labeled retinal cell suspensions, Neurosci. Lett., 92: 21. Del Cerro, M., Notter, M., Del Cerro, C., Wiegand, S., Grover, D. and Lazar, E. (1989) Intraretinal transplantation for rod-cell replacement in light-damaged retinas, J. Neur. Transplant., 1: 1. Del Cerro, M., Lazar, E., Grover, D., Gallagher, M., Sladek, C., Chu, J. and del Cerro, C. (1990) lntraocular transplantation and culture of human embryonic retinal cells. ARVO meeting, Sarasota FL, May 1990, Invest. Ophthalmol. Vis. Sci., 31: 593. Del Cerro, M., Ison, J., Bowen, P., Lazar, E. and Del Cerro, C. (1991) Intraretinal grafting restores visual function in light-blinded rats, Neuroreport, 2: 529. Del Cerro, M., lson, J., Bowen, P., Collier, R., Lazar, E., Grover, D. and del Cerro, C. (1992) Retinal transplants improve vision in light-blinded rats. In: D. Lam and G. Bray (Eds.), Regeneration and Plasticity in the Visual System, Proc. of the Retina Research Symposia, Vol. 4, MIT Press, in press. Ehinger, B., Bergstrom, A., Seiler, M., Arament, R., Zucker, C., Gustavii, B. and Adolph, A. (1991) Ultrastructure of human retinal cell transplants with long survival times in rats, Exp. Eye Res., 53: 447. Gouras, P., Flood, M. and Kjeldbye, H. (1984) Transplantation of human retinal cells to monkey retina, An. Acad. Bras. Cienc., 56: 431.
169 Gouras, P., Du, J., Gelanze, M., Kwun, R., Kjeldbye, H. and Lopez, R. (1991) Transplantation of photoreceptors labeled with tritiated thymidine into RCS rats, Invest. Ophthalmol. Vis. Sci., 32: 1704. Lazar, E. and del Cerro, M. (1990) A new procedure for retinal transplantation, Invest. Ophthalmol. Vis. Sci., 31: 593 (abst.). Li, L. and Turner, J. (1988) Transplantation of retinal pigment epithelial cells to immature and adult rat hosts: short and long-term survival characteristics, Exp. Eye Res., 47: 771. Lopez, R., Gouras, P., Brittis, M. and Kjeldbye, H. (1987) Transplantation of cultured rabbit retinal epithelium to rabbit retina using a closed eye method, Invest. Ophthalmol. Vis. Sci., 28: 1131. Sheedlo, H., Li, L. and Turner, J. (1989) Functional and
structural characteristics of photoreceptor cells rescued in RPE-cell grafted retinas of RCS dystrophic rats, Exp. Eye Res., 48: 841. Silverman, M. and Hughes, S. (1989) Transplantation of photoreceptors to light-damaged retina, Invest. Ophthalmol. Vis. Sci., 30: 1684. Triarhou, L.C. and del Cerro, M. (1985) Colloidal carbon as a multilevel marker for experimental lesions, Experientia, 41: 620. Turner, J. and Blair, J. (1985) Neonatal rat retinal grafts integrate into the lesion site of adult host retina, Invest. Ophthalmol. Vis. Sci., 26: 336. Wongpichedchai, S., Weber, P., Dorey, K. and Weiter, J. (1990) Comparison of internal and external approaches for transplantation of autologous retinal pigment epithelium, Invest. Ophthalmol. Vis. Sci., 31: 593.
157
© 1992 Elsevier Science Publishers B.V. All rights reserved 0165-0270/92/$05.00 NSM 01382
A new procedure for multiple intraretinal transplantation into mammalian eyes 1 Eliot Lazar and Manuel del Cerro Departments of Neurobiology and Anatomy, and Ophthalmology, University of Rochester Medical School, Rochester, New York 14642 (USA) (Received 4 October 1991) (Revised versions received 15 January 1992 and 2 April 1992) (Accepted 3 April 1992)
Key words: Transplantation; Retina; Neural grafts; Fetal tissue; Intra-ocular grafting; Retinoblastoma A new method of intraretinal grafting is described which avoids opening the globe and allows direct visual control of the transplant by observing placement of the graft through the host pupil. A microsyringe with a bevelled needle partially ensheathed in plastic tubing in order to limit penetration was used to inject dissociated donor cells into 2 to 4 pre-selected points of the sub-retinal space of adult rats. The needle tip was inserted through the scleral and choroidal tissue, and observed through the animal's pupil as it reached into the sub-retinal space. Donor cells were then injected into several sites of the retina. Results, using ophthalmoscopy as well as light and electron microscopy, confirmed the atraumatic nature of this technique and the survival of grafted cells. Intraretinal injection of colloidal carbon visually illustrated the effectiveness of this method in covering an extended portion of the host retina. The procedure, which is virtually free of complications has enabled us to perform multisite intra-ocular grafting in a safe, easy, and reliable fashion, under direct visual control.
Introduction
Since research on intra-ocular retinal transplantation between unrelated individuals began (del Cerro et al., 1984; Gouras et al., 1984; Turner and Blair, 1985), the field of retinal transplantation and the techniques to perform this operation have been undergoing constant evolution. Although successful for seeding small areas of the host retina with donor cells, these procedures suffer from definite limitations. Features corn-
Correspondence: Manuel del Cerro, M.D., Departments of Neurobiology and Anatomy, and Ophthalmology, University of Rochester Medical School, Rochester, New York, 14642, USA. Tel.: (716) 275-4027; Fax: (716) 442-8766. t Supported by NEI Grant No. 05262, Rochester Eye Bank, and generous private gifts. Some results from this work have been presented in abstract form (Lazar and del Cerro, 1990).
mon to the retinal transplantation methods thus far described are that they necessitate surgically opening the eye and suturing the incision. These requirements carry the potential for surgical complications, such as vitreous loss, retinal detachment, and ocular hemorrhage. These inherent risks may have discouraged attempts at multiple implantation since, to date, none have been reported. A further limitation of existing procedures is that, when the sub-retinal space is reached for cell delivery, it is done in a blind fashion with the needle tip being out of view of the observer at this critical moment. Thus realtime documentation of the transplants while they are being performed has not been provided in the literature. We describe here an injection technique for intra-ocular transplantation which does not require a surgical incision, while allowing us to retain visual control of the grafting procedure
158
throughout. Results obtained by transplanting suspensions of fetal neuroretinal cells or Y79 human retinoblastoma cells using this technique are presented, as are sub-retinal injections of colloidal carbon made in similar fashion into control animals. The method has already been used successfully to achieve functional visual recovery in blinded rats (del Cerro et al., 1991).
Materials and methods
Donor cells Several different types of donor tissue have already been used for intraretinal transplantation applying this technique. Here, we illustrate the use of human retinoblastoma Y79 cells and human embryonic neuroretinal cells. For demonstration purposes, the injection of a suspension of colloidal carbon using the same procedure is shown. Cell preparation Human retinoblastoma Y79 ceils obtained from the American Type Culture Collection were grown under standard conditions in suspension in RPMI medium supplemented with 10% fetal calf serum (FCS) (heat inactivated, Hyclone) 0.03% glutamine and 5 0 / z g / m l gentamycin at 37°C with humid 95% air, 5% CO 2. For transplantation studies, Y79 cells were transferred to coated, 100-mm tissue culture dishes for attachment and differentiation. Dishes were first treated with 100 izg/ml polylysine D for 30 min at 37°C followed by exposure to fibronectin 5 / x g / m l for 30 rain at 37°C to generate an adhesive surface. Y79 cells (2 x 106/dish) were cultured overnight in standard medium for adhering cells. Monolayers were
A
C
J
F Fig. 1A. Diagram illustrating the intact globe method of retinal transplantation. U n d e r direct visualization, using a posterior, external approach, a needle is inserted at a predetermined depth into the subretinal space. The graft is then injected at the desired site.
rinsed with CMF buffer and with E D T A and treated with 0.125% trypsin (Gibco) for 10 min at 37°C to remove cells from the adherent surface. Trypsin was quenched with 10% calf serum, the cells were centrifuged at 900 rpm and then rinsed with buffer. After pelleting, cells were suspended in cold human plasma at a concentration of 4000 cells/iz 1. Second-trimester human embryonic retinal cells obtained from electively aborted embryos aged 13-17 weeks were used as donor tissue. Procurement of the donor cells was in strict accordance with scientific and ethical guidelines, which included institutional review and approval of the experimental protocol. In all cases, maternal consent was given only after the decision to have an elective abortion was made. The eyes
Fig. lB. Macrophotograph showing the needle tip, as viewed by the operator, following penetration of the sclera, as it enters the sub-retinal space. Note that the presence of the needle in the sub-retinal space has caused an elevation of a retinal vessel. No hemorrhage has occurred at the penetration point. The atraumatic nature of the procedure permits us to repeat the injection at more than one location in the same eye. Magnification = 7.5 x . Fig. 2. A: this schematic indicates how the material immediately spreads out upon completion of the injection. As the retina was not perforated, the injected fluid was dispersed through the shallow sub-retinal space, and resulted in extended coverage being obtained by injecting only a small volume of material. B: macrophotograpb obtained within seconds of injecting 2 /xl of colloidal carbon. This relatively small volume of fluid covered approximately 120 ° of retinal surface. Magnification = 7.5 x .
159
160 were obtained within an hour of abortion and collected in either calcium-magnesium free medium or in human plasma at 4°C. Operating with the aid of a surgical microscope, the eyes were dissected open. The retinas were cut away, free of contamination from either the vitreous or the retinal pigment epithelium. Isolated retinas were trimmed into small fragments and then placed into ice-cold medium. Mechanical dissociation was used to obtain suspensions of retinal cells and cell clusters. The retinal fragments were dissociated by aspiration and injection through a 27-30-ga hypodermic needle.
Hosts and anesthesia Two to three groups of 6 - 8 male adult Fischer 344 rats were used per experiment. A total of 120 hosts were used for the experiments reported here. The animals were anesthetized with a mixture of ketamine (60 m g / k g ) and sodium pentobarbital (21 mg/kg). Topical 1% Alcaine drops (proparacaine hydrochloride, Alcon, Fort Worth, TX) were also used. The eyes were dilated with 1% neosynephrine and 1% mydriacyl (Alcon, Fort Worth, TX). Cyclosporine A, at a dose of 10 m g / k g / d a y , was given to the rats after transplantation of human cells. Delivery system and transplantation procedure A 27-ga Butterfly needle with a 15 ° bevel, tightly sheathed in plastic, with 1.1-1.4 mm of the needle tip left exposed, was connected to a 10 or 50/xl syringe (Series 1700, Hamilton, Reno, NV) prior to the procedure. The plastic sheath placed on the needle served as a stop to limit the depth of penetration. Its length was set so as to leave only enough of the needle tip exposed to reach the sub-retinal space without actually penetrating the retina. This proved highly effective in preventing the accidental formation of retinal holes or tears. The microsyringe was preloaded with the cell suspension. Collibri forceps (Storz, St. Louis, MO) were used to grasp the sclera and rotate the globe anteriorly. Then, using a stereomicroscope fitted with a 35-mm photographic camera and videotaping apparatus for direct visualization, the needle was manually inserted through the sclera with the
bevel facing the operator in order to afford the best view. It was then gently rotated without changing its angle until the tip, viewed through the pupil, could be seen in the sub-retinal space (Fig. 1B). Then, the tip was advanced further so as to elevate the retina slightly (Fig. 2B). At this point, an injection of 2 - 4 ~1 of cell suspension was made. The procedure was repeated at a point 180° opposite to the first injection site in the same eye. Typically, 2 micro-injections were made into the equatorial region of each eye. One was made superiorly at the 12 o'clock position and the other at the 6 o'clock position inferiorly. However, as many as 4 penetrations were performed in the eyes of some animals. The needle was quickly withdrawn following each injection, with no obvious reflux. After the experiment was completed, a topical lubricant (Tobrex: tobramycin ophthalmic ointment; Alcon, Ft. Worth, TX) was placed on the cornea in order to prevent drying.
Control injections In order to obtain instant, permanent, and multilevel visualization of the spread of the injected fluid, 12 retinas were injected with colloidal carbon (particle size from 20 to 30 nm). The colloidal carbon (Pelikan, Hannover, Germany) was prepared in a 1 : 4 dilution with saline and then used to perform control injections (del Cerro et al., 1985; Triarhou and del Cerro, 1985) in order to evaluate and demonstrate the extent of dispersion of the injected particles over the host retina. Clinical observations All the animals were observed at several points during the course of the experiment. Routinely, animals were examined prior to any procedure in order to obtain a baseline status. Following surgical intervention, they were examined at postoperative day 1, and then weekly for the remainder of the experiment. Clinical exams included in every case possible, the anterior segment, Schiotz tonometry, and a full funduscopic exam. Surviual times and histological procedures Survival times for hosts receiving cell grafts ranged from 3 to 90 post-transplantation days
161 (PTD). Control animals which received injections of 2 - 4 /xl of colloidal carbon were sacrificed 4 days following intra-ocular injections, because at this point no free colloidal carbon can be found in the sub-retinal space, which may confuse the results. The animals were enucleated and sacrificed under deep anesthesia using intramuscular injections of ketamine at 90 m g / k g and Rompun (xylazine, a sedative and analgesic) at a dose of 8 mg/kg. The eyes were fixed in 6% glutaraldehyde and then split along a sagittal axis. The hemisected eyes were examined and photographed under a stereomicroscope and were then embedded in plastic (Eponate 12; Ted Pella, Redding, CA). One micrometer-thick sections were cut and stained with Stevenel Blue (del Cerro et al., 1980) for light microscopic study. Ultrathin sections were cut for electron microscopic studies. The retinas from eyes injected with colloidal carbon were dissected free and prepared as flat-mounts.
Results
Fig. 1A illustrates the 2 basic principles of this procedure: (1) predetermined, limited penetration, and (2) direct visualization of the needle tip within the sub-retinal space. In fact the entire sequence, from the needle penetrating the sclera, though the release of material, could be viewed directly under the operating microscope (Figs. 1A and 2A). It was possible to observe and to photograph the needle as it pushed the retina upwards, once it had reached the sub-retinal space (Figs. 1B-2B). At this point, the bevel was rotated under direct visual control in order to place it in direct apposition to the retina. Then, the injection was carried out (Fig. 2B). When colloidal carbon was injected for demonstration purposes, it could be seen to spread out readily over the retinal surface. It quickly fanned out and covered up to 12 mm 2 or 120° of retinal surface per injection (Fig. 3), dramatically illustrating the wide
Fig. 3. Photograph of a hemisected eye which received colloidal carbon 4 days previously, as shown in Fig. 2B. In this particular case, a single injection of India ink covered a surface area of the retina as large as 12 mm2. Magnification= 40 x.
163
Fig. 5. A: h u m a n donor retinal cells ( H R C ) were transplanted into the host subretinai space using the procedure described in the text. This micrograph shows that the donor cells are well preserved into the rat eye 33 days following transplantation. T h e nuclear characteristics of the H R C contrast sharply with those of the host photoreceptors. Neither the host neural retina (RET) nor its pigment epithelium (RPE) demonstrate any visible signs of t r a u m a or an inflammatory response. Ultrastructural features of the junction areas between the host neural retina and RPE, and the transplant are seen in B and C. Magnification = 600 x . Fig. 5B and C: electron micrographs of the graft illustrated in A. B: the junction between the grafted h u m a n retinal cells (HRC) and host neural retina (RET). The outer tips of the outer segments, now separated from the pigment epithelium, form whorls of m e m b r a n o u s material (arrows). Magnification = 5150 x . C: the junction between the host retinal pigment epithelium (RPE) and the grafted h u m a n retinal cells (HRC). Even at this level of magnification there is a close contract between the graft and host tissues. There is also a remarkable absence of any signs of trauma or inflammation. Magnification = 16 000 x .
diffusion of the injected fluid over the surface of the host retina. Access to any portion of the retina, even as far posteriorly as the optic nerve head was possible. Intra-operative results, from both control experiments using colloidal carbon and actual cell grafts, clearly demonstrated that this material was delivered precisely into the sub-retinal space (Fig. 4B and C).
Clinical observations have confirmed the atraumatic nature of this technique. The injection tracks were self-sealing and thus required no surgical closure. There was no vitreous loss at the time of injection and postoperative examinations showed no corneal opacities, lenticular changes, or optic nerve head damage. Intra-ocular pressure, measured using a Schiotz tonometer, re-
Fig. 4. A: this low-power light micrograph shows grafts of Y79 h u m a n retinoblastoma cells at 2 points in the same rat eye. Of the 2 grafts (GR), the one to the left is shown at higher magnification in Fig. 4B, and the one at the right is shown in Fig. 4C. Magnification = 25 x . B and C: illustration of the junction areas between the host neuroretina and the sub-retinal masses of transplanted Y79 retinoblastoma cells (GR). In B, the growth of the grafted cells displaces the host retina (RET) upwards, whereas C shows the mass of the tumor growing into the subretinal space. In both cases, the lack of hemorrhage, macrophage invasion, or pigment epithelium (RPE) reaction is evident. Magnification = 600 x .
164
Fig. 5. (continued).
mained stable throughout. There were no pre- to postoperative elevations in pressure measurements at any time. The only changes seen by
direct and indirect ophthalmoscopy were 7 subretinal hemorrhages, 5 of them petechial in size, all of which resolved within 2 weeks. No other
165
Fig. 5. (continued). clinical complications were noted. In 120 rats of all ages which have been transplanted, we have had no additional adverse side effects of complications secondary to our method.
Histological results correlated quite well with the clinical observations just described. Sections of all the eyes injected with the tracer, as well as from those injected with Y79 retinoblastoma cells
166 (Figs. 4A-C) or fetal retinal ceils showed considerable dispersion of the injected material. This was seen to spread onto the outer retinal surface from the sub-retinal injection point. In histological specimens involving colloidal carbon injections o f 2-4 /zl, there was clearly intraretinal colloidal carbon material scattered throughout, at distances of more than 2000/xm from the penetration point. The frequency of successful transplantation compared favorably with the open approach previously used in this laboratory. The results of studies involving transplantation of human fetal retinal cells or Y79 retinoblastoma using this procedure are the subject of 2 separate reports (Del Cerro et al., 1992). A summary description shows that typically the penetration point was marked by the presence of a comparatively large cluster of cells, while smaller cell clusters were found distributed mostly throughout the hemisphere where the injection had occurred. The human fetal retinal cells survived well and accumulated into clusters, with rosette formation being a common occurrence (Fig. 5A and B). The ceils bordering the lumen of the rosettes had a chromatin distribution similar to that of in situ developing rod ceils. No appreciable apical differentiation, i.e., inner or outer segments, was detected. Their nature as photoreceptor cells was nonetheless established by the development of presynaptic differentiation in many of the cells, including the formation of presynaptic ribbons (data not shown) with associated clusters of vesicles. Actual ribbon synapses were found in the band of neuropile which surrounded these ceils. In keeping with the protracted development of the human neural retina, mitotic activity continued within the grafted cell population for periods of up to 3 months post-transplantation, the longest duration studied. Grafts of Y79 human retinoblastoma ceils initially developed within the confines of the subretinal space. Growth of the grafted cells induced the host retina to become progressively detached from the retinal pigment epithelium. The separation, however, was strictly limited to those areas occupied by the grafted cells. Yet, even at those points of detachment, the host neural retina ad-
hered closely to the tumoral cells (Figs. 4B, 5A). The host retina, including the retinal pigment epithelium, showed a remarkable lack of reaction to the mechanical disturbance from the surgical manipulation or the presence of the grafted cells (Fig. 5A-C). The tips of the detached photoreceptor cells, deprived from the phagocytic activity of the pigment epithelium, exhibited anomalous accumulations of degenerating cisternal material (Fig. 5B). As a rule, a minute disruption of Bruch's membrane was the only visible histologically indication of the penetration point. In the numerous transplants performed using this technique, the method has been complication-free. In 120 cases, the vitreous has been entered twice as seen by direct visualization and documented microscopically. In those two instances histologically demonstrable lacerations of the neural retina were observed.
Discussion
This study reports on a new method for performing intra-ocular transplantation which, unlike previously described methods can be performed under direct visual control and does not require surgically opening the eye. In the following paragraphs we will first discuss features of techniques previously used for retinal transplantation and then illustrate some of the experimental results obtained while using the present method. From its inception, the techniques used for intra-ocular retinal transplantation cited in the literature have uniformly required a surgical incision. A survey of the published methods underlines their features. Our initial retinal transplantation experiments were performed using glass micropipettes. A scleral incision was made, and the transplant was then performed. This technique, although effective in producing viable grafts, was difficult to perform and time consuming. Suturing of the minute incision often resulted in complications such as retinal detachments, sub-retinal hemorrhages, or formation of intravitreal membranes. A modified approach using self-sealing incisions (del Cerro et al., 1988, 1989),
167 decreased the formation of preretinal membranes and sharply reduced the incidence of hemorrhage. It did not, however, overcome the limitations common to all the procedures thus far described, of being unable to visualize the location of the needle tip at the moment of transplantation and of being able to consistently place the transplants at the proper depth. The method used by Turner et al. (Blair and Turner, 1987; Li and Turner, 1989; Sheedlo et al., 1989), for transplantation of R P E cells, required a preliminary excision of the superior rectus muscle, followed by an incision with a blade with penetrated the sclera and choroid to expose a bleeding surface. Surgical closure with suture material was required. The visual disturbance caused by a detached superior rectus muscle, which was not reportedly re-attached, was not addressed. Also, it was not made clear how the blade was stopped after cutting through the retinal pigment epithelium, but before it had reached the adjacent neural retina. Silverman and Hughes (1989) described a transplantation technique based on a transcorneal approach to the sub-retinal space. This required a wide incision through the cornea and then traversing the entire globe with a carrier 2.5mm-wide and 0.5-mm-thick, which was made to reach the posterior pole prior to delivering a tissue slab mounted on the carrier. Lopez et al. (1987) used a technique which required a conjunctival flap, a pars plana incision, and the insertion of a micropipette into the subretinal space. They noted that "there are a number of potential pitfalls associated with this technique", and pointed out that, "it may be difficult to be certain that no cells enter the vitreous cavity". The comments by Lopez and collaborators summarize the limitations of the methods used for achieving retinal transplantation, including those originally used in our own work. This prompted us to devise a more precise means of intra- or sub-retinal transplantation based on the principles of limited pre-set penetration of the needle used to deliver the donor cells and direct visual control through the dilated pupil of the host eye. Such a method would offer the added
advantage of allowing real time photography or video to document the transplantation procedure. It would also make multi-site grafting not only possible but reliable and nearly free of complications. Most intra-ocular grafts of neural retinal cells or retinal pigment epithelium, including ours, have been made using a transcleral or posterior approach. Some workers have instead used a transvitreal or anterior approach. Wongpichedchai et al. (1990; personal communication, 1992) compared the results obtained by using these two routes to implant retinal epithelium. They found that the external approach resulted in better grafts, although the lack of direct visualization using the external approach was a limitation. Our experience with the external approach matches that of Wongpichedchai et al. (1990). The desire to overcome the lack of visual control while using the external approach was one of our main reasons for developing the present method. No neovascularization, uveitis, vitreous hemorrhage, or ocular infections were seen in the eyes of 120 rats transplanted using this procedure. The microscopic findings correlate perfectly with these resuits. Histological examination has shown neither retinal detachments nor any evidence of transplant rejection. Comparison of results obtained using this procedure and those obtained by other laboratories are made difficult by the fact that most of the published studies have been made using donor cell-host combinations which differ from each other. For example, Aramant et al. (1990) and Ehinger et al. (1991) have transplanted human fetal retinal cells into rat retinas, but their donor tissues are first trimester versus second trimester (del Cerro et al. 1990). This fact may account for the extensive layer differentiation exhibited in their transplants. On the other hand, this effect may be related to the fact that as their technique involves incision of the neural retina, many of the grafts extend into the vitreal cavity, while ours remain confined to the subretinal space. The results obtained by Gouras et al. (1991) transplanting tritiated labeled photoreceptors into the subretinal space of RCS rats are more comparable to those obtained here by transplanting
168
fetal rat retinal cells into light blinded animals by means of the present procedure (del Cerro et al., 1991, 1992). In both cases, photoreceptor cells are seen to be distributed along the outer surface of the neural retina. The method used by Gouras involves the very delicate procedure of transchoroidally inserting a glass micropipette through a scleral incision. No mention is made of the surface extension covered by these single site grafts or of the steps required to stop the advancement of the micropipette precisely within the sub-retinal space. Silverman and Hughes (1989) transplanted tissue slabs containing photoreceptor cells into the sub-retinal space of light-damaged eyes with reported success. Their donor cells were taken from post-natal donors, that is an older age than that of the donors used in any of our studies (del Cerro et al., 1989, 1991). Since it has not been possible to duplicate their published procedure in our laboratory, we can not directly compare the 2 methods. The results of actual retinal transplantation using human fetal neuroretinal cells, Y79 human retinoblastoma cells, or intraretinal injections of colloidal carbon, into control animals, show that a considerable portion of the host retinal surface is covered by the explant. Although different in size than the carbon particles, the cells spread throughout the sub-retinal space in a like fashion, driven by the hydrostatic pressure of the injected fluid. The shallowness of the sub-retinal space forced the cells to extend over a considerable area, thus vastly improving the odds for functional success. The procedure has already been used with success in multisite transplantation of fetal retinal cells into the retinas of blind adult rats, which regained a measure of visual function (del Cerro et al., 1991, 1992). In summary it can be said that this method offers a previously unattainable degree of simplicity and control in the performance of retinal transplantation.
Acknowledgements The authors wish to thank Lazaros C. Triarhou, M.D., Ph.D., and David Diloreto, Jr., for their critical reviews of the manuscript. Drs. Gaff
Seigel and Mary Notter (deceased) provided valuable assistance with the Y79 cells. We would also like to thank Anita Matthews for her graphics contribution.
References Arament, R. Seller, M., Ehinger, B., Bergstrom, A., Gustavii, B., Brundin, P. and Adolph, A. (1990) Transplantation of human embryonic retina to adult rat retina, J. Restor. Neurol. Neurosci., 2: 9. Blair, J.R. and Turner, J.E. (1987) Optimum conditions for successful transplantation of immature transplantation of immature rat retina to the lesioned adult retina, Dev. Brain Res., 36: 257. Del Cerro, M., Cogen, M.J. and del Cerro, C. (1980) Stevenel Blue, an excellent stain for optical microscopy study of plastic embedding tissues, Microscop. Acta, 83: 217-220. Del Cerro, M., Gash, D.M., Notter, M., Rao, G.N., Wiegand, S. and Ishida, N. (1984) Intraocular retinal transplants, Invest. Ophthalmol. Vis. Sci., 25:62 (abst.). Del Cerro, M., Grover, D.A., Dematte, J.E. and Williams, W.M. (1985) Colloidal carbon as a combined ophthalmoscopic and microscopic probe of the retinal-blood barrier integrity, Phtbal. Res., 17: 34. Del Cerro, M., Notter, M., Wiegand, S., Jiang, L. and del Cerro, C. (1988) Intraretinal transplantation of fluorescently labeled retinal cell suspensions, Neurosci. Lett., 92: 21. Del Cerro, M., Notter, M., Del Cerro, C., Wiegand, S., Grover, D. and Lazar, E. (1989) Intraretinal transplantation for rod-cell replacement in light-damaged retinas, J. Neur. Transplant., 1: 1. Del Cerro, M., Lazar, E., Grover, D., Gallagher, M., Sladek, C., Chu, J. and del Cerro, C. (1990) lntraocular transplantation and culture of human embryonic retinal cells. ARVO meeting, Sarasota FL, May 1990, Invest. Ophthalmol. Vis. Sci., 31: 593. Del Cerro, M., Ison, J., Bowen, P., Lazar, E. and Del Cerro, C. (1991) Intraretinal grafting restores visual function in light-blinded rats, Neuroreport, 2: 529. Del Cerro, M., lson, J., Bowen, P., Collier, R., Lazar, E., Grover, D. and del Cerro, C. (1992) Retinal transplants improve vision in light-blinded rats. In: D. Lam and G. Bray (Eds.), Regeneration and Plasticity in the Visual System, Proc. of the Retina Research Symposia, Vol. 4, MIT Press, in press. Ehinger, B., Bergstrom, A., Seiler, M., Arament, R., Zucker, C., Gustavii, B. and Adolph, A. (1991) Ultrastructure of human retinal cell transplants with long survival times in rats, Exp. Eye Res., 53: 447. Gouras, P., Flood, M. and Kjeldbye, H. (1984) Transplantation of human retinal cells to monkey retina, An. Acad. Bras. Cienc., 56: 431.
169 Gouras, P., Du, J., Gelanze, M., Kwun, R., Kjeldbye, H. and Lopez, R. (1991) Transplantation of photoreceptors labeled with tritiated thymidine into RCS rats, Invest. Ophthalmol. Vis. Sci., 32: 1704. Lazar, E. and del Cerro, M. (1990) A new procedure for retinal transplantation, Invest. Ophthalmol. Vis. Sci., 31: 593 (abst.). Li, L. and Turner, J. (1988) Transplantation of retinal pigment epithelial cells to immature and adult rat hosts: short and long-term survival characteristics, Exp. Eye Res., 47: 771. Lopez, R., Gouras, P., Brittis, M. and Kjeldbye, H. (1987) Transplantation of cultured rabbit retinal epithelium to rabbit retina using a closed eye method, Invest. Ophthalmol. Vis. Sci., 28: 1131. Sheedlo, H., Li, L. and Turner, J. (1989) Functional and
structural characteristics of photoreceptor cells rescued in RPE-cell grafted retinas of RCS dystrophic rats, Exp. Eye Res., 48: 841. Silverman, M. and Hughes, S. (1989) Transplantation of photoreceptors to light-damaged retina, Invest. Ophthalmol. Vis. Sci., 30: 1684. Triarhou, L.C. and del Cerro, M. (1985) Colloidal carbon as a multilevel marker for experimental lesions, Experientia, 41: 620. Turner, J. and Blair, J. (1985) Neonatal rat retinal grafts integrate into the lesion site of adult host retina, Invest. Ophthalmol. Vis. Sci., 26: 336. Wongpichedchai, S., Weber, P., Dorey, K. and Weiter, J. (1990) Comparison of internal and external approaches for transplantation of autologous retinal pigment epithelium, Invest. Ophthalmol. Vis. Sci., 31: 593.