Molecular Vision 2005; 11:256-262 <>
Received 15 March 2004 | Accepted 24 March 2005 | Published 15 April 2005

Use of superparamagnetic microbeads in tracking subretinal injections

Jing Wen,1,2 Kyle C. McKenna,1,2 Brent C. Barron,1,2 Heather P. Langston,1,2 Judith A. Kapp1,2

Departments of 1Ophthalmology and 2Pathology, Emory University School of Medicine, Atlanta, GA

Correspondence to: Judith A. Kapp, PhD, Department of Ophthalmology, University of Alabama at Birmingham, 615 18th Street South, Birmingham, AL, 35205; Phone: (205) 975-7081; FAX: (205) 996-2435; email:


Purpose: The purpose of these studies was to develop a method to track intraocular injections.

Methods: Retinal pigment epithelial (RPE) cells, purified from adult mouse eyes, were incubated with superparamagnetic microbeads (Dynabeads, 4.5 μm) coated with bovine serum albumin to verify that they could phagocytose the microbeads. For in vivo tracking studies, mice were anesthetized and a small incision was made at the pars plana and 2 μl of microbeads (around 105 microbeads) or RPE that had taken up the microbeads were injected into the subretinal space (SRS). Mice were sacrificed at various times after injection. The eyes were enucleated, fixed in formalin, and embedded in paraffin. Sections were stained with H&E, visualized by light microscopy. Some eyes were digested with collagenase and inflammatory cells determined by flow cytometry.

Results: Cultured adult RPE phagocytosed the magnetic microbeads. One day after injection into the SRS, a retinal detachment was observed at the injection point and free microbeads were easily detected at this site. One week later, the host RPE cells had phagocytized the microbeads and the retina had reattached. No inflammatory response was detected in the eyes that were injected with microbeads in the SRS at any time examined. Histology showed normal morphology of all retinal layers around the injection site. The microbeads remained in situ throughout the study.

Conclusions: Protein coated magnetic microbeads are non-inflammatory after injection into the SRS. Host RPE cells phagocytized the microbeads and the retina maintained a healthy morphology after reattachment. This technique proved not only to be a good training tool to determine the precise location of injection, but also provided a noninflammatory method for long term marking of delivery into the SRS. However, the microbeads can not be used as a tracer of injected RPE because the microbeads were readily transferred to the endogenous RPE.


Age related macular degeneration (AMD) is a major cause of blindness among elderly people in the United States [1]. AMD is characterized by the loss of photoreceptor cells in the macula that is preceded by the death of the supporting retinal pigment epithelial (RPE) cells and changes in Bruch's membrane [2]. Consequently, replacement of dead or damaged cells with healthy retinal cells has been thought to be a promising approach to the treatment for AMD. This hypothesis is supported by the observations that transplantation of RPE prolongs vision in the Royal College of Surgeons (RCS) rat [3].

Our long term goal is to make transplantation of RPE cells a viable treatment for AMD. With the exception of identical twins, however, development of an immune response to the foreign tissue usually leads to immunological rejection unless some form of clinical intervention is employed. We have initiated studies to identify the mechanisms of RPE rejection using the mouse as a model because numerous immunologically relevant strains (inbred, Tg, and knockout) are available, which provides a wealth of tools not available in any other experimental animal.

During our initial studies of transplanted RPE, we found that identifying graft rejection based on the failure to find transplanted RPE in histological sections of transplant recipients was complicated by the fact that a technical failure in the injection of cells into the subretinal space (SRS) could present the same histological picture. RPE from Tg mice expressing green fluorescent protein (GFP) [4] or β-galactosidase [5] have been used for tracking the injection of cells into the proper compartment. However, GFP [6,7] and β-galactosidase [8] are antigenic in mice, which could potentially complicate more longterm studies of rejection. Moreover, cells labeled with GFP and β-galactosidase suffer from the original problem that there is no way to determine whether the lack of labeled cells is due to a technical failure or that they were rejected. Thus, we have developed a new tracking method to verify that injected materials are actually delivered to the appropriate compartment using biologically inert, nondegradable, superparamagnetic Dynabeads (Dynal, Inc.), which is described herein.


Experimental animals

Adult C57BL/6 and Balb/c mice (NCI Frederic Cancer Facility, Frederick, MD) were housed under standard conditions and maintained in accordance with the Association for Research in Vision and Ophthalmology resolution for the use of animals in research. All procedures on animals were conducted according to the principles in the guidelines of the Committee on Care and Use of Laboratory Animals, Institute of Laboratory Animal Resources, National Research Council (NIH publication number 85-23, revised 1985) and in adherence to the provisions of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

Isolation and culturing of RPE from adult mice

Mouse RPE cells were isolated from pigmented (C57BL/6) mice and albino (Balb/c) mice using a modification of the methods for human RPE [9,10]. Eyes were removed aseptically and were incubated at 37 °C with collagenase and hyaluronidase for 40 min followed by an additional 50 min with trypsin. The sclera and choroid were dissected from the eye and discarded; the RPE were gently rubbed from the neurosensory retina, which remained attached to the vitreous. Approximately 105 RPE cells/eye were routinely obtained. RPE were cultured as described by Sheedlo et al. [11] and used between passages 2 and 5.

Identification of cultured cells as RPE

Cultured RPE were stained with an isotype control or PCK-26, a pancytokeratin specific antibody (Abcam limited, Cambridge, UK) that recognizes cytokeratin 6 and cytokeratin 8, to verify that they were epithelial cells [12]. Bound anti-cytokeratin antibody was detected using a fluorescently (Alexa 488) labeled goat anti-mouse IgG (Molecular Probes, Eugene, OR). The nuclei were counter stained with DAPI and the slides examined by fluorescent microscopy.

Preparation of serum coated microbeads

Superparamagnetic microbeads (4.5 μm, Dynabeads® M-450 Tosyl activated, Dynal Inc., Lake Success, NY) were washed in 0.1 M phosphate buffer (pH 7.4). The microbeads were resuspended in PBS (phosphate buffered saline) containing 0.1% (w/v) BSA (bovine serum albumin Fraction V; Sigma-Aldrich Corp., St. Louis, MO) and incubated for 18 h at 37 °C to block the activated tyosyl groups. The microbeads were then washed twice in PBS containing 0.1% (w/v) BSA, once in 0.2 M Tris containing 0.1% (w/v) BSA, and stored in PBS containing 0.1% (w/v) BSA. The microbeads were not coupled to any other antibodies or selecting markers.

Uptake of microbeads by cultured RPE

Microbeads were added to cultures of RPE at a ratio of 3:1 and incubated for 2 to 4 days. Free microbeads were removed from the RPE by thoroughly washing the monolayer. RPE were removed from the tissue culture wells by treatment with 0.5% trypsin in EDTA (Mediatech Inc., Herndon, VA) and enriched for cells that had phagocytosed the microbeads by passage through a magnetic field according to the manufacturer's directions.

Injection of microbeads or RPE pulsed with microbeads into the murine subretinal space

Mice were anesthetized with an intramuscular injection of ketamine (Sigma-Aldrich Corp., St. Louis, MO) and xylazine (Phoenix Scientific, Inc., St. Louis, MO). The pupils were dilated with 1% cyclopentolate HCl and 2.5% phenylephrine HCl (Wilson Ophthalmic Corp., Mustang, OK) and 0.5% proparacaine HCl was applied for topical anesthesia. Under a Weck surgical microscope, the eye was proptosed with slight bilateral pressure from forceps and a small sclerotomy was made at the ora serrata with a 30 gauge needle. A Hamilton syringe was loaded with the microbeads and fitted with a Lambert 33 gauge blunt end subretinal needle (bent at a 45° angle). The needle was inserted through the sclerotomy site at about 11 o'clock and angled past the crystalline lens until the sensory retina/RPE interface was reached. Vehicle (Hank's Balanced Salt Solution (HBSS) without phenol red; 2 μl) containing 105 microbeads or 105 purified RPE that had taken up the microbeads in vitro were injected into the SRS. The retinal detachment at the injection site was confirmed by gently pressing a microscope coverslip on the cornea and viewing through the surgical microscope. The position and size of the detachment was documented to aid in histological sectioning. Mice were sacrificed by cervical dislocation and a 6-0 polypropylene suture was placed through the limbal conjunctiva (as a marker for proper orientation during paraffin embedding). The eyes were enucleated and placed in 10% buffered formalin for 4 h.

Histological sectioning

Formalin fixed eyes were mechanically processed, properly oriented in a histological cassette, and embedded in paraffin. Eyes were sectioned (5 μm) followed by routine hematoxylin and eosin (H&E) staining. Sections were examined with a Nikon Eclipse E800 microscope (Melville, NY) equipped with an Optronics Magnafire S99800 camera head (Goleta, CA).

Flow cytometric analysis of collagenase digested eyes

Eyes were dissected, washed in PBS, digested with 58.5 units/ml collagenase Type IV (Sigma, St. Louis, MO), in RPMI 1640 containing 1% volume/volume FBS by incubation for about 2 h at 37 °C in a 5% CO2 atmosphere. Single cell suspensions, made by pressing the eye between frosted glass slides, were stained with allophycocyanin labeled anti-CD45, phycoerythrin labeled anti-CD90.2 (Thy1.2), 7-AAD (Viaprobe) and fluorescein isothiocyanate labeled anti-GR-1 antibodies. Flow cytometric analysis was performed using a FACSCalibur (Becton Dickinson, San Jose, CA) flow cytometer. From a live (7-AAD negative) CD45+ cell gate, the percentage of live CD45 that were Thy 1.2 positive, GR-1 positive, or Thy 1.2 negative and GR-1 negative was determined by analysis using FlowJo (Treestar, Inc., San Carlos, CA) version 5.3 data analysis software. The number of live CD45 cells collected was determined by multiplying the percentage of live CD45+ cells by the total number of collected cells. The absolute number of CD45 cells within the entire sample was determined by multiplying the number of CD45 cells collected by the total volume of the sample, then dividing by the duration of the collection interval and by the flow rate of the cytometer (1 μl/s). Absolute numbers of Thy 1.2 positive, GR-1 positive, or Thy 1.2 negative and GR-1 negative cells were determined by multiplying the percentage of the indicated CD45+ cell population by the absolute number of CD45 cells. Statistical significance was calculated by a Student's t-test.


Isolation and characterization of RPE

Freshly isolated RPE were shown to be typical, hexagonal epithelial cells; RPE from C57BL/6 (Figure 1A) were visualized by light microscopy whereas those from Balb/c (Figure 1B) mice were visualized by phase contrast microscopy. RPE were cultured for two to five passages and stained with anti-pancytokeratin to verify that they were epithelial cells. Most, if not all, of the cells from C57BL/6 mice reacted with the cytokeratin specific antibody (Figure 1D) but not an isotype control (Figure 1C). Similar results were obtained with Balb/c RPE (not shown).

Uptake of microbeads by RPE

Various cell types have been purified from complex mixtures of cells by incubating them with superparamagnetic microbeads coated with an antibody to a cell surface antigen differentially expressed by the cells. Cells that bound the microbeads are separated from those that did not by passage through a powerful magnetic field [13,14]. Because RPE are phagocytic, we tested whether donor RPE might take up BSA coated microbeads (without bound antibody). BSA coated microbeads are uniformly sized, round, and refractile under phase contrast microscopy (Figure 2A). These microbeads were added to RPE and incubated for 2 to 3 days and then the free microbeads were washed from the adherent RPE. Most of the adherent cells had taken up one to several microbeads, which are detectable as refractile spheres in the cytoplasm by phase contrast microscopy (Figure 2B).

Use of microbeads to track intraocular injections

Subretinal injections in this study were modeled after the pars plana vitrectomy approach that is used in humans. This technique avoids damage to the iris and crystalline lens with subsequent cataract formation [15] when an anterior chamber approach is used through the cornea. A sclerotomy site at the ora serrata also avoids injection through the cornea with displacement of the lens to reach the subretinal space [16]. Inserting the needle at avascular locations near the optic nerve head reduced the risk of hemorrhage (see schematic in Figure 3). Animals that hemorrhaged into the vitreous cavity were discarded from the study. The retinal detachment was visualized via a fundus exam. No complications or endophthalmitis were observed in any of the mice at any time after injection.

Histology showed a retinal detachment at the subretinal injection site (RPE/PRC interface) in BALB/c mice after injection of 2 μl of microbeads. The photoreceptor cells were displaced from the RPE and the microbeads (orange colored spheres) are easily seen in the detachment zone immediately after injection (not shown) and at 24 h after injection (Figure 4A). Photoreceptor cells were not lost and the morphology of the neurosensory retina above the detachment was normal. One week later, the outer segments of the photoreceptor cells were opposed to the recipient RPE and the overall structural morphology and integrity of the retina was restored (Figure 4B). Moreover, the RPE had phagocytosed virtually all of the microbeads. By 3 weeks, the microbeads were still detected within the cytoplasm of the RPE and normal retinal morphology was observed (Figure 4C). Similar results were found in pigmented eyes from C57BL/6 mice but the microbeads in the RPE were partially obscured by the pigment (Figure 4D).

The microbeads generally spread out as a single layer of beads in a wide area of the retina (Figure 5) and could be detected in several serial sections of the eye (not shown). The injected microbeads served as long lasting markers since they were readily detected in the SRS of rabbits that had been injected six months earlier (not shown). Microbeads also proved to be a useful teaching tool as early in the training period, microbeads were occasionally found in the choroid or the vitreous (not shown).

Careful examination of the histological sections failed to reveal any inflammatory infiltration in eyes of mice one day, or one, two, and three weeks after receiving the microbeads. The lack of an inflammatory response was further verified by quantifying bone marrow derived inflammatory cells by flow cytometry of single cell suspensions of collagenase digested eyes 14 days after injection of the microbeads (Figure 6). Small numbers of bone marrow derived cells were found in the digests of normal eyes, which are attributed to blood borne cells in transit through the extensive vasculature of the eye. No significant increases in Thy1.2 positive (T cells) or GR-1 positive cells (neutrophils) were found in treated eyes compared to untreated eyes. A very small, but significant, increase was observed in the Thy1.2 negative, GR-1 negative (presumably myeloid) cells in eyes of mice that received the microbeads. However, the number of Thy1.2 negative, GR-1 negative cells in mice injected with microbeads was not significantly different from that in eyes challenged with HBSS (data not shown). Thus, the small increase in the myeloid cell population most likely results from the injury caused by the SRS injection and not in response to the microbeads.

Transplantation of RPE containing microbeads

Next, we wanted to test whether the RPE that had taken up microbeads in vitro could be tracked after they were injected into the subretinal space. Thus, RPE monolayers were incubated with the microbeads and nonadherent microbeads were removed by washing with medium. RPE were recovered by trypsinization and those that phagocytosed the superparamagnetic microbeads were enriched from RPE that had not taken up the microbeads by passage through a magnet. The enriched RPE were injected into the SRS and their eyes examined histologically. Several instances in which the donor RPE were found as a monolayer on top of the host RPE showed that the microbeads could be detected in both layers of RPE (Figure 7). This suggests that free microbeads, which were presumably released from the donor RPE, could be transferred to the host RPE. Thus, this method is not capable of tracking donor RPE.


The purpose of this study was to develop a tracking method to evaluate the success of subretinal injections. These experiments showed that the superparamagnetic microbeads were phagocytosed by RPE both in vitro and in vivo. Microbeads injected into the SRS were easily detected in H&E sections by light microscopy as round, orange objects either within the SRS or within the cytoplasm of the RPE. The BSA coated microbeads caused no deleterious effects as detected by inflammatory reactions. Such microbeads potentially can be used to deliver antibodies or other proteins to various compartments of the eye.

No alterations were detected in the morphology of the photoreceptors or other neurological cell layers above the recipient RPE that had phagocytosed the microbeads suggesting that the function of the RPE was not noticeably compromised by the phagocytosis of the microbeads. In addition, the microbeads were detected six months after injection into the SRS of rabbit eyes suggesting that they can serve as long term markers of the successful injection of RPE even if the RPE are lost to attrition or rejection. Thus, the microbeads can provide objective evidence that the RPE were injected into the appropriate compartment even if the RPE are rejected. Unfortunately, phagocytosed microbeads cannot serve as markers of donor RPE because the recipient RPE readily phagocytosed microbeads.


This work was supported by a center grant from the Foundation Fighting Blindness, an NEI core grant (P30 EYO06360), research grants from the NEI (EY 13459 and EY14877), and a gift from Malcolm and Musette Powell. National Research Service Award F32-07079 supported K. C. McKenna. Training Grant T32-EY07092 from the National Eye Institute, NIH, supported H. P. Langston. J. A. Kapp is the recipient of the Jules and Doris Stein Professorship in Ophthalmology awarded by Research to Prevent Blindness.


1. Fine SL, Berger JW, Maguire MG, Ho AC. Age-related macular degeneration. N Engl J Med 2000; 342:483-92.

2. Zarbin MA. Age-related macular degeneration: review of pathogenesis. Eur J Ophthalmol 1998; 8:199-206.

3. Lund RD, Kwan AS, Keegan DJ, Sauve Y, Coffey PJ, Lawrence JM. Cell transplantation as a treatment for retinal disease. Prog Retin Eye Res 2001; 20:415-49.

4. Lai CC, Gouras P, Doi K, Lu F, Kjeldbye H, Goff SP, Pawliuk R, Leboulch P, Tsang SH. Tracking RPE transplants labeled by retroviral gene transfer with green fluorescent protein. Invest Ophthalmol Vis Sci 1999; 40:2141-6.

5. Dunaief JL, Kwun RC, Bhardwaj N, Lopez R, Gouras P, Goff SP. Retroviral gene transfer into retinal pigment epithelial cells followed by transplantation into rat retina. Hum Gene Ther 1995; 6:1225-9.

6. Stripecke R, Carmen Villacres M, Skelton D, Satake N, Halene S, Kohn D. Immune response to green fluorescent protein: implications for gene therapy. Gene Ther 1999; 6:1305-12.

7. Rosenzweig M, Connole M, Glickman R, Yue SP, Noren B, DeMaria M, Johnson RP. Induction of cytotoxic T lymphocyte and antibody responses to enhanced green fluorescent protein following transplantation of transduced CD34(+) hematopoietic cells. Blood 2001; 97:1951-9.

8. Krzych U, Fowler AV, Miller A, Sercarz EE. Repertoires of T cells directed against a large protein antigen, beta-galactosidase. I. Helper cells have a more restricted specificity repertoire than proliferative cells. J Immunol 1982; 128:1529-34.

9. Castillo BV Jr, del Cerro M, White RM, Cox C, Wyatt J, Nadiga G, del Cerro C. Efficacy of nonfetal human RPE for photoreceptor rescue: a study in dystrophic RCS rats. Exp Neurol 1997; 146:1-9.

10. Tezel TH, Del Priore LV. Repopulation of different layers of host human Bruch's membrane by retinal pigment epithelial cell grafts. Invest Ophthalmol Vis Sci 1999; 40:767-74.

11. Sheedlo HJ, Li L, Turner JE. Effects of RPE age and culture conditions on support of photoreceptor cell survival in transplanted RCS dystrophic rats. Exp Eye Res 1993; 57:753-61.

12. Moll R, Franke WW, Schiller DL, Geiger B, Krepler R. The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 1982; 31:11-24.

13. Hewett PW, Murray JC. Human microvessel endothelial cells: isolation, culture and characterization. In Vitro Cell Dev Biol Anim 1993; 29A:823-30.

14. Su X, Sorenson CM, Sheibani N. Isolation and characterization of murine retinal endothelial cells. Mol Vis 2003; 9:171-8 <>.

15. Price J, Turner D, Cepko C. Lineage analysis in the vertebrate nervous system by retrovirus-mediated gene transfer. Proc Natl Acad Sci U S A 1987; 84:156-60.

16. Timmers AM, Zhang H, Squitieri A, Gonzalez-Pola C. Subretinal injections in rodent eyes: effects on electrophysiology and histology of rat retina. Mol Vis 2001; 7:131-7 <>.

Wen, Mol Vis 2005; 11:256-262 <>
©2005 Molecular Vision <>
ISSN 1090-0535