• A). Representative Retinal Sections Stained For Macbook Pro

Nov 22, 2018 - Representative OCT horizontal volume scans of B-scans moving from. H&E-stained retinal sections from Asah1+/+. IHC staining for Mac-2 on sections from Asah1+/+ mice and Asah1P361R/P361R mice (H-J).

Our goal was to define a clinically significant population of cells by utilizing a single‐step selection process to enrich hematopoietic cells capable of regenerating the retinal pigment epithelium (RPE). Utilizing intravitreal injection of bone marrow cells from a mouse with pigment (C57BL6: gfp) into albino recipient mice (C57BL6:Tyr ‐), we show that hematopoietic progenitor cells (HPCs) enriched for CD133 can regenerate RPE cells and improve retinal function. The chemokine CXCL12 (stromal cell‐derived factor 1α) is essential for migration, incorporation, and RPE regeneration by CD133 + HPCs.

Once incorporated, CD133 + HPCs become pigmented, adopt an RPE morphology, and express RPE‐specific proteins, leading to partial functional recovery by electroretinogram. Human CD133 + HPCs also incorporate in the retina and assume RPE morphology in nonobese diabetic/severe combined immunodeficient mice xenografts. These data show that a clinically accessible CD133 + hematopoietic cell can home to an injured RPE layer, differentiate into cells with significant RPE morphology, and provide therapeutic functional recovery of the visual cycle. S TEM C ELLS 2009;27:457–466.

INTRODUCTION Retinal pigment epithelium (RPE) dysfunction has been linked to many devastating eye disorders, including age‐related macular degeneration (AMD) as well as hereditary disorders such as Stargardt's disease and retinitis pigmentosa. AMD is the leading cause of irreversible blindness and visual disability with a progressive loss of central vision in patients 60 years old in the western hemisphere. Current treatments for AMD can only slow progression of vision loss, primarily benefiting patients with advanced stages of the disease. RPE integrity is an essential component for retinal function and visual health. The RPE consists of a monolayer of cuboidal cells that separates the photoreceptors and the choroid , forming the blood–retina barrier via tight junctions. The RPE serves an integral function in the visual cycle by phagocytosis of rod and cone outer segments following shedding into the subretinal space and the production of paracrine factors for retinal health.

The RPE is responsible for the movement of ions and water to maintain a proper state of dehydration for visual clarity, and its pigmentation absorbs stray light that would otherwise degrade the visual image. Pigmentation is due to the presence of melanosomes, which are organelles containing the light‐absorbing pigment melanin. Current dogma posits that once the RPE is terminally differentiated, it does not renew itself by cell division. Sodium iodate has been used extensively in rodents and causes selective damage of the RPE layer in a dose‐dependant manner -.

It also destroys the blood–retina barrier in a patchy pattern ,. The complete mechanism is not fully understood, but sodium iodate has been shown to inhibit lysosomal enzyme activities within RPE cells, particularly acid phosphatase activity, which is essential for RPE function and cell survival. A moderate dosage of sodium iodate was selected for our studies to create significant damage of the RPE layer with measurable loss of visual function, while maintaining a RPE microenvironment intact enough to potentially instruct transplanted progenitor cells during regeneration. We had previously improved the established sodium iodate RPE injury mouse model to allow tracking of pigmented bone marrow‐derived cells following transplantation into albino mice. This model would allow us to definitively establish the potential for bone marrow hematopoietic cells to repair a damaged RPE layer. Our initial study showed that hematopoietic stem cell (HSC)‐derived cells could home to a damaged RPE layer and adopt an RPE‐like morphology at very low levels, thereby suggesting a potential therapeutic effect.

In this manuscript, we demonstrate the therapeutic efficacy of intravitreal injections of CD133 + hematopoietic progenitor cells (HPCs) in improving visual function. CD133 was chosen because it is an enrichment marker for multipotent HPCs. It is also expressed on a variety of tissue‐specific stem/progenitor cells, and functional loss of CD133/prominin‐1 leads to retinal degeneration in humans. Transplanted CD133 + HPCs resulted in a 10‐fold increase in pigmented cells from our previous experiments utilizing a sodium iodate‐damaged RPE layer. This was achieved by assuming RPE morphology, expressing pigment, and expressing the RPE‐specific genes RPE65 and cellular retinaldehyde‐binding protein ( CRALBP). Human CD133 + HPCs also regenerated RPE cells in a xenograft model.

These results suggest that an easily enriched population of bone marrow progenitor cells can provide an effective means to repair damage to the RPE layer in debilitating blindness disorders such as AMD. Isolation of CD133 + HPCs Full‐grown (8–10 weeks of age) GFP male mice were euthanized and sacrificed. The long bones in the legs (femurs and tibias) were immediately excised. Bone marrow was flushed from the long bones of male gfp + STOCK Tg(GFPU)5Nagy/J (Jackson Laboratory, Bar Harbor, ME; ) mice. Bone marrow cells were incubated with rat anti‐mouse CD133 (eBiosciences Inc, San Diego, ). The bone marrow cells were enriched by magnetic‐activated cell sorting (MACS) via goat anti‐rat IgG microbeads (Miltenyi MACS; Miltenyi Biotec, Auburn, CA, ) and passing cells twice through a column.

Typical enrichment yielded approximately 90% enrichment for CD133 + cells. Flow Cytometry Whole bone marrow isolates were isolated and incubated with antibodies of interest for 30 minutes on ice. The samples were incubated with CD133 (eBiosciences, clone 13A4) at a concentration of 0.2 μg per sample.

The following antibodies were used in addition to CD133: CD45.2 (0.1 μg, clone 104; eBiosciences), Sca‐1 (0.1 μg, clone D7; BD Pharmingen, San Diego, ), CD117 (0.1 μg, clone 2B8; BD Pharmingen), CD59 (0.1 μg, clone ER‐MP 20; Abcam, Cambridge, U.K., ), CD11b (0.1 μg, clone M1/70; BD Pharmingen), F4/80 (0.1 μg, clone BM8; Caltag Laboratories, Burlingame, CA, ), Gr‐1 (0.1 μg, clone RB6–8C5; BD Pharmingen), and CD184 (0.1 μg, clone cxcr4; BD Pharmingen). Retina Immunohistochemistry Murine eyes were enucleated by sliding Botvin serrated‐tip microdissecting forceps (Roboz Surgical Instrument Co, Gaithersburg, MD, ) underneath the eyeball and gently pulling on the optic nerve. For retinal flatmounts, eyes were punctured with a needle to allow complete perfusion, and incubated in 4% paraformaldehyde (PFA) for 1 hour at 4°C. The neural retina was dissected from the posterior cup (RPE–choroid–sclera complex).

The posterior cup was flat‐mounted with four to seven radial cuts and mounted with Aqua Poly/Mount (Polysciences, Inc., Warrington, PA, ). For retinal cross‐sections, the intact eyes were embedded into frozen blocks using optimal cutting temperature (OCT) (Sakura Tissue Tek; IMEB, San Marcos, CA, ) in a 2‐methylbutane (Sigma) and dry ice slurry. Prepared eyes were permitted to equilibrate to cutting temperature 3 hours prior to sectioning at 6 μm and thaw‐mounting onto plus‐charged slides (Fisher Scientific, Suwanee, GA, ).

Cut slides were air‐dried overnight at room temperature, fixed for 5 minutes in ice‐cold acetone, and air‐dried again before they were stained. Sections were Fc receptor blocked (Biogenex Inc., San Ramon, CA, ), normal horse serum blocked, and then incubated overnight at 4°C with 1:25 rat anti‐mouse CD133 (clone 13A4; eBiosciences) or 1:150 goat anti‐mouse matrix metalloproteinase (MMP)9 (R&D Systems Inc., Minneapolis, ). Immunoreactivity was detected using the species‐appropriate donkey AlexaFluor 647 or 488 (Molecular Probes Invitrogen, Carlsbad, CA, ) secondary antibodies applied at 1:500 for 45 minutes in the dark. Slides were briefly postfixed in 4% PFA before rinsing in 1× phosphate‐buffered saline and mounting with Vectashield with 4′,6‐diamidino‐2‐phenylindole (DAPI) (Vector Laboratories, Burlingame, CA, ). Concentration‐ and isotype‐matched Ig slides were included as controls. For RPE and CRALBP staining, a modification to the ARK (DAKO Cytomation, Glostrup, Denmark, ) was used. Mouse anti‐RPE (1:100, Novus Biologicals, Littleton, CO, ) and mouse anti‐CRALBP (1:50, Affinity BioReagents, Golden, CO, ) were biotinylated according to the ARK kit instructions.

Briefly, sections were sequentially blocked with serum‐free protein block (DAKO) and avidin and biotin (Vector Laboratories) prior to a 15‐minute incubation in the prepared primary antibodies. Slides were washed, followed by detection using 1:500 streptavidin AlexaFluor 647 (Molecular Probes) for 45 minutes at room temperature in the dark. Once again, in order to preserve morphology, the postfixation was completed before mounting in Vectashield with DAPI. For imaging, the retinas were immediately documented by confocal microscopy using a Leica TCS SP2 laser scanning spectral confocal microscope (Leica Microsystems Heidelberg GmbH, Wetzlar, Germany, ) or light microscopy using a Leica DM5500B (Leica Microsystems) upright microscope with differential interference contrast (DIC) microscopy. Quantitation and Blocking Experiments Entire globes were flash‐frozen in OCT and serially sectioned at 6 μm per section on a cryostat as described above. Sections were always collected 100 μm into the eye for an additional 240 μm (eight 6‐μm sections per slide, for five total slides).

Slides were fixed as described above in acetone, mounted with Vectashield with DAPI (Vector Labs), and observed. Pigmented cells were scored every three sections to ensure that they were not counted twice. For blocking experiments, mice were anesthetized, and antibodies were injected into the contralateral left eye in a volume of 2 μl.

The specific antibodies used were: CD184 (0.1 μg/μl, clone 2B11; BD Pharmingen), CXCL12/stromal cell‐derived factor (SDF)‐1 (10 μg/μl, clone 79014; R&D Systems), and CCL‐2/JE/monocyte chemoattractant protein (MCP)‐1 (10 μg/μl, clone 123616; R&D Systems). IgG controls were concentration matched. Enzyme‐Linked Immunosorbent Assay Approximately 1 ml of blood was collected via cardiac puncture from mice ( n = 3 per time point) into microtubes fitted with a heparin‐coated ring (Sarstedt, Numbrecht, Germany, ). Blood was spun down and serum was collected and flash frozen on dry ice in a volume of 60 μl in six total tubes (the enzyme‐linked immunosorbent assay ELISA kit suggests 50 μl of serum per sample; samples were run in triplicate). The R&D Systems ELISA kit for CXCL12 (Quantikine Mouse SDF‐1/CXCL12, MCX120) and CCL2 (Quantikine Mouse JE/MCP‐1 Immunoassay, MJE00) was used following the manufacturer's instructions. Concentrations of protein were calculated using the standard curve. The CXCL12 standard curve equation was y = 0.182 x + 0.0104 with an r 2 value of 0.9975 and the CCL2 standard curve equation was y = 0.002 x + 0.0257 with an r 2 value of 0.9984.

Human Cord Blood Isolation and Transplantation Non‐obese diabetic/severe combined immunodeficient (NOD/SCID) mice (Jackson Laboratory) that were 8–10 weeks old were transplanted with 1 × 10 6 CD133 + human cord blood (HCB) cells. HCB cells were isolated from the University of Florida Life Cord Bank in Shands Hospital. HCB cells were transplanted intravitreally after sodium iodate administration, as discussed previously.

The HCB cells were isolated as follows. HCB was acquired from Life Cord and brought to the lab on ice. A 20% volume of Hextend (6% Hetastarch in Lactated Electrolyte Injection; Hospira, Inc., Lake Forest, IL, ) was added to the bag of HCB. The mixture (of HCB and Hextend) was removed from the bag. The suspension was then spun at 500 rpm for 5 minutes at 4°C and the top layer was transferred into a new, single tube (the bottom layer containing red cells was discarded).

The solution was then spun at 1,300 rpm for 10 minutes at 4°C. The top layer of plasma was then discarded, and the sample was filled to a 20‐ml volume with Plasma‐Lyte (Baxter Worldwide, Deerfield, IL, ), underlayed with 15 ml of Ficoll‐Paque (Amersham Biosciences, Piscataway, NJ, ), and spun at 600 rpm for 30 minutes at 4°C.

The buffy coat was isolated and cells were counted using a hemacytometer and trypan blue dilution. Once cells were isolated, they were sorted using the MACS anti‐human CD133 MicroBead Kit (Miltenyi) according to the manufacturer's instructions. The cell suspension was passed over two separate MACS columns and counted by a hemacytometer. Retinal flatmounts were analyzed for pigment. Phenotypic Analysis of the Murine CD133 + Bone Marrow Population Previous experiments have shown that CD133 cells within the bone marrow have stem or progenitor cell activity in humans , and CD133 expression has been described for many tissue‐specific stem and progenitor cell populations.

As a first step toward understanding the functional significance of the CD133 + population in the murine bone marrow, a phenotypic analysis was carried out using known hematopoietic stem and progenitor cell surface markers. CD133 + cells comprised approximately 10% of the bone marrow cells ( n = 5) in Figure A. The CD133 + bone marrow cells were positive for the hematopoietic lineage marker CD45.2 (Fig. C) and cKit (CD117) (Fig. D) were expressed on the majority of the cells and are found on hematopoietic stem and progenitor cells. CXCR4 (CD184) (Fig.

E), the receptor for CXCL12, was expressed on 94% of CD133 + HPCs, which indicates that this population is very responsive to damage. Phenotypic expression profile of CD133 + hematopoietic progenitor cells. ( A): Whole bone marrow was analyzed for CD133 ( n = 5) expression (red) compared with IgG control (purple). CD133 + cells were analyzed for their expression of the panhematopoietic marker CD45.2 ( B) ( n = 4), the hematopoietic stem cell markers Sca‐1 ( C) ( n = 4) and CD117 ( D) ( n = 4), the migration receptor CD184 ( E) ( n = 4), the myeloid cell markers CD11b ( F) ( n = 4), F4/80 ( G) ( n = 3), and Gr‐1 ( H) ( n = 3), and CD59 ( I) ( n = 4). The IgG control is represented by blue, and a shift is represented by the white area. A shift is represented by the average percentage ± the standard error.

Phenotypic expression profile of CD133+ hematopoietic progenitor cells. (A): Whole bone marrow was analyzed for CD133 (n = 5) expression (red) compared with IgG control (purple). CD133+ cells were analyzed for their expression of the panhematopoietic marker CD45.2 (B) (n = 4), the hematopoietic stem cell markers Sca‐1 (C) (n = 4) and CD117 (D) (n = 4), the migration receptor CD184 (E) (n = 4), the myeloid cell markers CD11b (F) (n = 4), F4/80 (G) (n = 3), and Gr‐1 (H) (n = 3), and CD59 (I) (n = 4). The IgG control is represented by blue, and a shift is represented by the white area. A shift is represented by the average percentage ± the standard error.

We also examined the CD133 + population for the expression of known myeloid commitment markers. Mature myeloid lineage markers, such as Mac‐1 (CD11b) (Fig. F), F4/80 (Fig. G), and Gr‐1 (Fig.

A). Representative Retinal Sections Stained For Macbook Pro

H), were expressed on 85% of CD133 + HPCs. I) is the sole inhibitor of complement membrane attack complex formation and is widely distributed throughout the body. CD59 was originally identified as a monocyte precursor marker and represented about 92% of the CD133 + population.

Isolated CD133 + HPCs transplanted into lethally irradiated recipient mice did not achieve host radioprotection (data not shown). Collectively, these data demonstrate that the CD133 + bone marrow population is enriched for myelomonocytic hematopoietic progenitors that express the chemokine receptor CXCR4. Given the functional similarities between the RPE layer of the retina and tissue macrophages in other organs, we concluded that the CD133 + population would be an ideal population to test for RPE regeneration ability. CD133 + Progenitor Cells Incorporate into the RPE Layer To investigate the potential of CD133 + HPCs for mediating RPE repair, we modified our previous model of long‐term gfp‐C57BL6‐HSC transplant into a short‐term passive transfer model into nonirradiated, sodium iodate‐treated, albino BL6 recipients via intravitreal injection. This modification allowed the defined bone marrow populations to be functionally characterized, and resulted in improved efficiency of the transplanted bone marrow cells migrating to the site of injury. CD133 + HPCs were isolated from adult bone marrow via MACS. Albino mice (C57BL/6J‐ Tyr c‐2J/J) were given 40 mg/kg sodium iodate, and 4 days postinjection, 1 × 10 4 CD133 + HPCs from gfp + mice were adoptively transferred via direct vitreal injection into the right eye of albino mice ( n = 5).

Animals were analyzed short term at 1 month post‐transplant. RPE/sclera flatmounts and retinal cross‐sections were prepared from both experimental and control eyes.

Figure A shows a representative 40× DIC image of an RPE/sclera flatmount from an untreated control. The DIC imaging allows us to better utilize contrast of the pigmented cells in the nonpigmented host tissue. Figure B and C are representative 40× DIC images of pigmented cells in the right and left experimental eyes, respectively.

Approximately 10% of the RPE/sclera area of both eyes was seeded with pigmented, donor‐derived cells as quantified by Volocity Image analysis software (PerkinElmer, Waltham, MA, ). CD133 + hematopoietic progenitor cells were incorporated into the retinal pigment epithelium layer.

(A–C): Representative 40× DIC images of retinal pigment epithelium/sclera flatmounts of normal albino mice (A) and the right eye (B) and left eye (C) of albino mice treated with sodium iodate and systemically administered CD133 + hematopoietic progenitor cells. Size bar = 32.0 μm.

(D–G): Representative 63× confocal microscopic images of retinal cross‐sections of pigmented (donor‐derived) cells in an albino host. (D): Nomarsky technique used to show pigmented cells along BM (arrows). (E): DAPI nuclear stain overlay on Nomarsky image showing nuclei within pigmented cells. (F): 488 nm green laser autofluorescence overlay on Nomarsky image indicating autofluorescence in the same position as pigment (arrows). (G): Merge of Nomarsky, DAPI, and 488 nm green laser. Size bar = 47.62 μm. Abbreviations: BM, Bruch's membrane; DAPI, 4′,6‐diamidino‐2‐phenylindole; DIC, differential interference contrast; INL, inner nuclear layer; ONL, outer nuclear layer.

CD133+ hematopoietic progenitor cells were incorporated into the retinal pigment epithelium layer. (A–C): Representative 40× DIC images of retinal pigment epithelium/sclera flatmounts of normal albino mice (A) and the right eye (B) and left eye (C) of albino mice treated with sodium iodate and systemically administered CD133+ hematopoietic progenitor cells. Size bar = 32.0 μm. (D–G): Representative 63× confocal microscopic images of retinal cross‐sections of pigmented (donor‐derived) cells in an albino host. (D): Nomarsky technique used to show pigmented cells along BM (arrows). (E): DAPI nuclear stain overlay on Nomarsky image showing nuclei within pigmented cells. (F): 488 nm green laser autofluorescence overlay on Nomarsky image indicating autofluorescence in the same position as pigment (arrows).

(G): Merge of Nomarsky, DAPI, and 488 nm green laser. Size bar = 47.62 μm.

Abbreviations: BM, Bruch's membrane; DAPI, 4′,6‐diamidino‐2‐phenylindole; DIC, differential interference contrast; INL, inner nuclear layer; ONL, outer nuclear layer. To determine if the pigmented bone marrow‐derived cells had localized to the RPE layer, retinal cross‐sections were examined. Figure D–2G display confocal microscopic images from a retinal cross‐section demonstrating incorporation of the bone marrow‐derived cells into the RPE layer. The pigmented CD133 + bone marrow‐derived cells in the RPE layer are shown in Figure D using a 63× DIC image along Bruch's membrane (indicated by arrows). Figure E is a DAPI nuclear‐stained retinal section superimposed onto the corresponding DIC image. This shows the colocalization of DAPI nuclei and pigmented cells. The DAPI staining also highlights the retinal architecture of the outer nuclear layer (ONL) and inner nuclear layer (INL), which lack sufficient contrast to be easily visualized in noncounterstained tissue.

Figure F shows the colocalization of green fluorescence expression from the pigmented cells. Figure G is a merge of DAPI and gfp fluorescence as confirmed by spectral analysis (supporting information Fig. 1), and DIC images displaying pigmented cells with DAPI‐stained nuclei. Transplanted CD133 + HPCs Express RPE‐Specific Proteins Following Incorporation To confirm that the incorporated, pigmented cells in the RPE layer possess RPE‐associated characteristics, the expression of RPE‐specific proteins was examined. Retinal cross‐sections from normal and treated animals were analyzed by immunohistochemistry for expression of RPE‐specific proteins: RPE65 and CRALBP.

To further characterize the RPE layer before and after sodium iodate injury and transplantation, the expression of CD133 was also examined, which to our knowledge has never been reported in the murine retina. Figure A–3D are 63× confocal microscopic images of normal albino retinal cross‐sections stained with Alexa‐Fluor 647 (A, B, C red) or 488 (D green) secondary antibody with DAPI counterstain for nuclei (blue). As expected, both RPE65 and CRALBP are expressed in the RPE layer in the normal retina (Fig. A, B; IgG‐negative control inlay). CD133 expression (Fig.

C; IgG‐negative control inlay) is limited to the apical surface of the RPE layer and the outer segments of the ONL. MMP9, which has been shown to be involved in hematopoietic cell migration , is not expressed in the normal retina (Fig. D; IgG‐negative control inlay).