November 2009
Volume 50, Issue 11
Free
Retinal Cell Biology  |   November 2009
Thrombospondin-1–Mediated Regulation of Microglia Activation after Retinal Injury
Author Notes
  • From the Schepens Eye Research Institute, Department of Ophthalmology, Harvard Medical School, Boston, Massachusetts. 
  • Corresponding author: Sharmila Masli, Schepens Eye Research Institute, 20 Staniford Street, Boston, MA 02114; sharmila.masli@schepens.harvard.edu
Investigative Ophthalmology & Visual Science November 2009, Vol.50, 5472-5478. doi:10.1167/iovs.08-2877
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Tat Fong Ng, Bruce Turpie, Sharmila Masli; Thrombospondin-1–Mediated Regulation of Microglia Activation after Retinal Injury. Invest. Ophthalmol. Vis. Sci. 2009;50(11):5472-5478. doi: 10.1167/iovs.08-2877.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: Thrombospondin (TSP)-1 has been demonstrated to play a vital role in immune privilege. The functional phenotype of ocular antigen-presenting cells that contributes to the immune privilege status of the eye is dependent on their expression of TSP-1. Microglia, the local antigen-presenting cells in the retina, undergo rapid activation in response to injury and have the ability to produce both proinflammatory and regenerative neurotrophic factors. In this study, the authors examined TSP-1 as a potential regulator of these phenotype of microglia activated in response to retinal injury.

Methods.: Expression of markers associated with activated microglia were examined by immunofluorescent staining and semiquantitative real-time PCR analysis of retina derived from WT or TSP-1 null mice at various time intervals after light- or laser-induced retinal injury.

Results.: In the absence of TSP-1, microglia in uninjured retina express major histocompatibility complex class II and migrate to the outer layers of the retina. Constitutively increased expression of activated microglia-derived inflammatory molecules such as TNF-α and iNOS is detectable in TSP-1 null retina compared with WT controls. After both light-induced and laser-induced retinal injury, enhanced migration of microglia is detected in TSP-1 null retina, and these microglia express markers associated with a proinflammatory phenotype. Compared with WT retina, TSP-1 null retina fails to recover from the laser-induced injury, resulting in irreversible damage.

Conclusions.: TSP-1 supports an anti-inflammatory phenotype of microglia in the retina and promotes recovery from injury.

Microglia in the retina are considered resident immune cells that resemble antigen-presenting cells such as macrophages. 13 These two cell types share many cell surface markers and phagocytic activity on their activation. In the adult retina, microglia are located in the inner retinal layer, such as the ganglion cell layer, and under normal conditions are not detected in the outer nuclear layers. Morphologically resting microglia are highly branched (ramified) cells, with little cytoplasm, capable of monitoring their surroundings. 4 In response to injury, ischemia, neurodegeneration, infection, or any pathologic insult, resting microglia are rapidly activated. 58 On activation microglia change their morphology, acquire an amoeboid form, and migrate to the site of injury. Activated microglia exhibit many other changes such as proliferation, phagocytosis, and secretion of bioactive molecules. 
In the developing retina one of the main functions of microglia is to phagocytose and eliminate cellular debris from apoptotic neurons. Both in their resting and activated states, microglia are known to produce several neurotrophic factors such as BDNF, CNTF, GDNF, neurotrophin-3, and bFGF, which are known to protect and regulate photoreceptor survival. 8,9 Besides these, additional factors capable of stimulating the regeneration of retinal ganglion cells after nerve injury have been reported. 10 Thus microglia are known to sense their microenvironment to clear metabolic products and tissue debris and to facilitate neuronal tissue regeneration. However, microglia are also capable of producing proinflammatory molecules, such as the inflammatory cytokines TNF-α and IL-1β, reactive oxygen species (ROS), nitric oxide (NO), neurotoxic matrix metalloproteinases, and glutamate. 5,11,12 Such activated microglia are linked to many pathologic processes in retinal diseases because their abnormal accumulation and altered morphologies are reported in animal models of diseases such as retinitis pigmentosa, 13 age-related macular degeneration, 14 and diabetic retinopathy. 15  
Immunologically, microglia are considered comparable to tissue macrophages. In their resting state, microglia express low levels of costimulatory molecules and are weakly phagocytic. 16 It has been reported that the immunosuppressive cytokine TGF-β contributes to reduced expression of antigen-presenting molecules such as major histocompatibility complex (MHC) class II and costimulatory CD80 and CD86 on microglia. 8 The retinal pigment epithelium is known to produce this cytokine, which can also block inflammatory gene expression in activated microglia. 17 We have reported previously that the activation of latent TGF-β produced by RPE to its biologically active form is mediated primarily by the large extracellular matrix protein thrombospondin (TSP)-1. 18 By producing immunosuppressive cytokines such as TGF-β and IL-10, RPE cells are known to contribute to the local regulation of an inflammatory immunologic response and immune privileged status of the retina. Expression of TSP-1 was demonstrated as essential for maintaining this immune privilege. 
TSP-1 was originally reported as produced by platelets and is now known to be produced by various cell types including activated microglia. 19 Expression of TSP-1 by macrophages supports their anti-inflammatory function, 20,21 but its significance in microglia is unknown. Studies report strong immunoreactivity to TSP in the normal retina in blood vessels and Bruch's membrane. 22 Such immunoreactivity to TSP was also reported in the brain on striatal microglia during their transformation to phagocytic macrophages. 23 Furthermore, it has been reported that TSP expression is induced, in response to injury, in neurons and activated microglia in a facial nerve axotomy model 19 and in RPE cells adjoining the edges of laser lesions in laser photocoagulation–treated eyes. 24 Functionally TSP-1 is known to exert its antiangiogenic role in the retina 25 ; however, its role in the modulation of the proinflammatory phenotype of activated microglia is unknown. We hypothesize that TSP-1 plays an important role in the recovery from retinal injury by supporting a neuroprotective rather than an inflammatory phenotype of microglia, thereby facilitating the wound-healing process. 
In this study using TSP-1–deficient mice, we show that in the absence of TSP-1 the microenvironment in the retina, presumably devoid of biologically active TGF-β, is proinflammatory and supports enhanced migration and activation of retinal microglia in response to injury; eventually, no wound healing is achieved. We also demonstrate that exacerbated expression of microglia-associated inflammatory genes in the TSP-1–deficient retina for an extended period correlates with poor recovery from the injury. 
Materials and Methods
Mice
C57BL/6 (H-2b) wild-type (WT) mice, 6 to 8 weeks old, were purchased from the Jackson Laboratory (Bar Harbor, ME). TSP-1 null mice (C57BL/6 background), originally received from John Lawler (Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA), were bred in-house in a pathogen-free facility at Schepens Eye Research Institute (Boston, MA). All animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 
Light-Induced Retinal Injury
Light intensity conditions described previously were adopted. 26 Briefly, mice were exposed to either 500 or 50 lux white light in 12-hour light cycle for a period of 28 days. 
Laser-Induced Retinal Injury
Mice were anesthetized with ketamine/xylazine (Phoenix Pharmaceutical, St. Joseph, MO) followed by pupil dilation with tropicamine (Akorn, Inc., Buffalo Grove, IL). Local anesthetic proparacaine (Akorn, Inc.) was applied to the cornea. The retina, viewed through a slit lamp, was divided into quadrants. Each quadrant was treated with a thermal diode laser (Oculight SLx; IRIS Medical, Mountain View, CA) with wavelength of 810 nm, spot size of 350 μm, duration of 150 msec, and energy of 150 mW (3 spots/quadrant). Animals were killed at 1, 3, 14, and 120 days after laser injury to enucleate carefully while avoiding any further damage to the injured posterior segment. 
Histology
The eyes were fixed in 4% paraformaldehyde (frozen sections) or 10% formalin (methacrylate sections). For frozen sections, the eyes were exposed to sucrose gradient (5%–30%) in PBS before embedding in optimal cutting temperature (OCT) medium and sectioning (10-μm-thick sections). For methacrylate sectioning, the eyes were dehydrated before embedding in the glycol methacrylate resin (Technovit 7100; Heraeus Kulzer GmbH, Hanau, Germany), sectioned (3-μm-thick sections), and stained in hematoxylin and eosin (H&E). To determine photoreceptor thickness, the numbers of rows of outer nuclear layer were counted in the center of the injury site and on the edge of the intact area adjacent to the injury site on both sides. Average counts from three sites were determined in five serial sections from each animal (n = 5/group). Student's t-test was used to determine statistical significance. 
Immunohistochemistry
Cryosections were used for immunohistochemistry. The primary antibodies were KSPG (5D4; Associates of Cape Cod, Falmouth, MA), MHC class II (I-Ab), CD40, CD45, and CD86 (BD Biosciences, PharMingen; San Diego, CA), and Iba1 (Wako Chemicals USA., Inc., Richmond, VA). The target antigens in stained sections were then visualized with Cy3- or Cy2-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, PA). Positively stained cells for KSPG and I-Ab in normal and bright light-exposed retina were counted in the whole retina. In laser-injured retina, positively stained cells for I-Ab, CD40, and CD86 were counted around the laser-induced injury site in five serial sections in each animal within a 200 × 200-μm grid placed around the injury site. One-way ANOVA was used to determine statistical significance. 
Real-Time PCR
Total RNA was isolated from the retinal tissue harvested at different time intervals using RNA STAT-60 kit (Tel-Test, Inc., Friendswood, TX) and were reverse transcribed to generate cDNA. Real-time PCR was performed using a standard SYBR green protocol using the following primers: TNF-α, F-5′-GGCCTCCCTCTCATCAGTTCTATG, R-5′-GTTTGCTACGACGTGGGCTACA; iNOS, F-5′-ATTCAGATCCCGAAACGCTTCA, R-5′-GCTGAGGGCTCTGTTGAGGTCTAA; CD80, F-5′-GAATTACCTGGCATCAATACGACA, R-5′-CTTAATGGTGTGGTTGCGAGTC; MHC class II (I-Ab ), F-5′-AGGGCATTTCGTGTACCAGTT, R-5′-GTACTCCTCCCGGTTGTAGATGTA; MCP-1, F-5′-AACTGCATCTGCCCTAAGGTCTT, R-5′-GCTTCAGATTTACGGGTCAACTTC; MIP-2, F-5′-TACTGAACAAAGGCAAGGCTAACT, R-5′-CGAGGCACATCAGGTACGA; GAPDH, F-5′-CGAGAATGGGAAGCTTGTCA, R-5′-AGACACCAGTAGACTCCACGACAT. Amplification reactions were set up using mastermix (SurePRIME-&GO; QBiogene, Solon, OH) in triplicate with the following thermal profile: 50°C for 2 minutes; 1 cycle, 95°C for 15 minutes; 1 cycle, 52°C to 55°C for 1 minute, 40 cycles, 72°C for 30 seconds, 1 cycle on analyzer (ABI Prism; Applied Biosystems Inc., Foster City, CA). Fluorescence signal generated at each cycle was analyzed using system software. The threshold cycle (Ct) values were used to determine relative quantitation of gene expression with GAPDH as a reference gene. 
Results
Detection of Enhanced Expression of Microglia-Associated Proinflammatory Molecules in Uninjured TSP-1–Deficient Retina
In a normal retina, resting microglia are typically detectable in the ganglion cell layer and outer plexiform layer (OPL) but not in the outer nuclear layer (ONL) or the subretinal space. On activation, these cells become phagocytic and are considered equivalent to tissue macrophages with antigen-presenting capabilities. Resting microglia express very low levels of MHC class II, a molecule important in antigen presentation. 2,27 Many inflammatory mediators are also known to activate microglia and induce MHC class II expression, 12,28 and such mediators may accumulate after any form of retinal injury or insult. Previously, we demonstrated that in the absence of TSP-1, the immune privilege status of the retina is compromised, and this was associated with the ability of TSP-1 to activate TGF-β, an immunosuppressive cytokine. 18 To determine whether the TSP-1–deficient retina, which is incapable of generating immunosuppressive active TGF-β, represents a more proinflammatory environment, we examined the expression of microglia-associated inflammatory molecules such as TNF-α, iNOS, and MHC class II. Expression of MHC class II was determined by immunohistochemical staining of uninjured WT and TSP-1 null retina. As shown in Figure 1a, microglial cells expressing MHC class II were detected in the outer layers of the TSP-1 null retina but not in WT retina. The overall number of MHC class II–expressing cells was significantly increased in TSP-1 null retina compared with the WT controls (Fig. 1b). Furthermore, comparison of message levels for TNF-α and iNOS in TSP-1 null retina with those in WT retina indicated significantly increased levels of these proinflammatory molecules in the absence of TSP-1 (Fig. 1c). Thus, in the absence of TSP-1, the retinal microenvironment appears to be more proinflammatory, supporting an activated state of microglia. 
Figure 1.
 
Constitutive expression of MHC class II, TNF-α, and iNOS in TSP-1 null retina. (a) Frozen sections of WT or TSP-1 null retina were stained with anti-IAb (MHC class II) antibodies and counterstained with fluorescence-labeled conjugated secondary antibody. Nuclei were stained with DAPI (blue). Arrows show detectable positively stained cells. (b) Number of MHC class II–positive cells in the whole section were counted in WT and TSP-1 null retina (n = 3 each). (c) RNA isolated from either WT or TSP-1 null retina was subjected to semiquantitative real-time PCR analysis. Expression of TNF-α and iNOS relative to GAPDH is presented. *P < 0.05.
Figure 1.
 
Constitutive expression of MHC class II, TNF-α, and iNOS in TSP-1 null retina. (a) Frozen sections of WT or TSP-1 null retina were stained with anti-IAb (MHC class II) antibodies and counterstained with fluorescence-labeled conjugated secondary antibody. Nuclei were stained with DAPI (blue). Arrows show detectable positively stained cells. (b) Number of MHC class II–positive cells in the whole section were counted in WT and TSP-1 null retina (n = 3 each). (c) RNA isolated from either WT or TSP-1 null retina was subjected to semiquantitative real-time PCR analysis. Expression of TNF-α and iNOS relative to GAPDH is presented. *P < 0.05.
Enhanced Migration of Microglia to the Outer Nuclear Layers of TSP-1–Deficient Retina in Response to Light-Induced Damage
Microglia in the retina undergo activation rapidly in response to injury. Previously, we demonstrated that in albino mice, compared with pigmented mice, a significantly increased number of microglia migrate to the subretinal space (SR) in response to light-induced damage. 26 Because uninjured TSP-1 null retina appeared to support the presence of activated microglia, we investigated whether this retina was more susceptible to light-induced migration of microglia to the SR in spite of the pigment. Both TSP-1 null and WT mice were placed in two different light-intensity conditions and were killed 28 days later. Retinas harvested from these mice were stained with the microglia markers keratan sulfate proteoglycan (KSPG) and Iba1 (a calcium-binding protein). Both KSPG- and Iba1-expressing microglia were found in the SR in TSP-1 null mice (Figs. 2a–f). Moreover, in the wholemounts of TSP-1 null retina, an uneven distribution of KSPG+ microglia in the SR was detectable with more accumulation in certain areas (Fig. 2b). In H&E-stained sections of photoreceptors in some areas appeared disorganized and extended out in OPL-forming small clusters (Figs. 2g, 2h) and disrupting inner nuclear layer (INL). Staining for the photoreceptor marker recoverin, together with KSPG (Fig. 2c), demonstrated that indeed the clusters were composed of photoreceptors and KSPG and Iba1+ microglia. Such changes were not detectable in TSP-1 null mice exposed to lower light intensities (data not shown), nor did they appear in WT mice exposed to bright light (Figs. 2h, i). Although the numbers of KSPG+ cells in the subretinal space were significantly increased in TSP-1 null mice exposed to both ambient and bright light conditions compared with the WT controls (Fig. 2j), we also detected a significant decline in INL and ONL in bright light-exposed TSP-1 null retina (Fig. 2k). These results indicate that in the absence of TSP-1, the retina is vulnerable to light-induced damage in spite of the pigment and that the damaged areas of the retina attract activated microglia. Therefore, enhanced migration of activated microglia is detectable in the retina of TSP-1 null mice in response to light-induced injury. 
Figure 2.
 
Light-induced damage in TSP-1 null retina induces migration of microglia. Retina isolated from TSP-1 null (ag) or WT (h, i) mice exposed to bright or ambient light intensity either were subjected to wholemount staining with fluorescence-conjugated antibodies or were sectioned and stained with H&E. (a) Bright light-exposed retina: KSPG+ microglia (red) detectable in the subretinal space. Nuclei were visualized with DAPI (blue). (b) Cluster of KSPG+ microglia detectable in the subretinal space of bright light-exposed TSP-1 null retina. (c) Recoverin (green)–expressing photoreceptors extended from the OPL, forming a rosette-like structure infiltrated with KSPG+ microglia (red). (df) KSPG and Iba1 staining of microglia within the rosette-like structures. (g) H&E staining of TSP-1 null retina exposed to bright light forming rosette-like structures. (h) No detectable KSPG+ cells in the subretinal space of WT control retina exposed to bright light and no detectable structural damage in H&E-stained sections (i). (j) Quantitative assessment of number of KSPG+ cells in the subretinal space of WT or TSP-1 null mice exposed to two light intensities. (k) INL count in bright light-exposed WT and TSP-1 null retina. n = 5 in each group. *P < 0.05. Scale bars: (a, c, d, f) 100 μm; (b) 40 μm; (e) 25 μm.
Figure 2.
 
Light-induced damage in TSP-1 null retina induces migration of microglia. Retina isolated from TSP-1 null (ag) or WT (h, i) mice exposed to bright or ambient light intensity either were subjected to wholemount staining with fluorescence-conjugated antibodies or were sectioned and stained with H&E. (a) Bright light-exposed retina: KSPG+ microglia (red) detectable in the subretinal space. Nuclei were visualized with DAPI (blue). (b) Cluster of KSPG+ microglia detectable in the subretinal space of bright light-exposed TSP-1 null retina. (c) Recoverin (green)–expressing photoreceptors extended from the OPL, forming a rosette-like structure infiltrated with KSPG+ microglia (red). (df) KSPG and Iba1 staining of microglia within the rosette-like structures. (g) H&E staining of TSP-1 null retina exposed to bright light forming rosette-like structures. (h) No detectable KSPG+ cells in the subretinal space of WT control retina exposed to bright light and no detectable structural damage in H&E-stained sections (i). (j) Quantitative assessment of number of KSPG+ cells in the subretinal space of WT or TSP-1 null mice exposed to two light intensities. (k) INL count in bright light-exposed WT and TSP-1 null retina. n = 5 in each group. *P < 0.05. Scale bars: (a, c, d, f) 100 μm; (b) 40 μm; (e) 25 μm.
Persistence of MHC Class II–Expressing Cells in TSP-1–Deficient Retina for an Extended Period after Laser Injury
Although light-induced damage represents a chronic process, we next examined the migration of microglia and their phenotype in TSP-1–deficient retina in a laser-induced injury model in which RPE cells are damaged. Expression of MHC class II in activated microglia is associated with their antigen-presenting capabilities typically induced by inflammatory conditions. 29,30 We observed the expression of this marker in the retina after laser injury. One day after laser injury, the expression of MHC class II (I-Ab ) was detectable in both WT and TSP-1 null retina (Figs. 3a, b). Although Iba1 staining was detectable in ramified microglia in WT retina, it colocalized with I-Ab in cells with rounded morphology in TSP-1 null retina (Figs. 3c, d). After the initial increase in the number of I-Ab+ cells within 1 day after laser injury, there was a significant decline in these cells in WT retina and no detectable expression by 14 days after injury (Figs. 3e, h). However, I-Ab immunoreactivity remained detectable in the TSP-1 null retina 14 days after injury (Figs. 3f, h). Similarly, though the expression of costimulatory molecules relevant to antigen-presenting properties of activated microglia, such as CD40 and CD86, was barely detectable in the WT retina at this later time point after laser injury, both these markers remained detectable in TSP-1 null retina (data not shown). Thus, our results suggest that in the absence of TSP-1, inflammatory conditions persist in the retina for an extended period after laser injury. These results are consistent with the significant contribution of TSP-1 to TGF-β–dependent immunoregulation in the retinal microenvironment. 
Figure 3.
 
In the absence of TSP-1, laser-induced retinal injury results in prolonged persistence of MHC class II–expressing cells at the site of injury. Frozen sections of retina harvested from WT (a, c, e) or TSP-1 null (b, d, f) mice after laser injury were stained with fluorescence-conjugated antibodies to detect microglia marker Iba1 and MHC class II (I-Ab). Retina were harvested either 1 day (ad) or 14 days (e, f) after laser injury. Scale bar, 50 μm. (g) Negative control shows staining background. (h) Quantitative estimates of MHC class II–expressing cells in the WT or TSP-1 null retina on indicated days after laser injury. n = 5 for each time point. N.D., not detectable. *P < 0.05).
Figure 3.
 
In the absence of TSP-1, laser-induced retinal injury results in prolonged persistence of MHC class II–expressing cells at the site of injury. Frozen sections of retina harvested from WT (a, c, e) or TSP-1 null (b, d, f) mice after laser injury were stained with fluorescence-conjugated antibodies to detect microglia marker Iba1 and MHC class II (I-Ab). Retina were harvested either 1 day (ad) or 14 days (e, f) after laser injury. Scale bar, 50 μm. (g) Negative control shows staining background. (h) Quantitative estimates of MHC class II–expressing cells in the WT or TSP-1 null retina on indicated days after laser injury. n = 5 for each time point. N.D., not detectable. *P < 0.05).
Prolonged Expression of Proinflammatory Molecules in the Absence of TSP-1 in Laser-Injured Retina
To confirm our observations from immunohistochemistry, we opted to compare transcript levels for various proinflammatory molecules also typically associated with activated microglia. These included cell surface molecules related to antigen presentation MHC class II and CD80, chemokines that attract more inflammatory cells such as MCP-1 and MIP-2, the enzyme iNOS known to generate the toxic free radical nitric oxide, and the inflammatory cytokine TNF-α. Retinas harvested from both WT and TSP-1 null mice 14 days after injury were processed to isolate RNA, and a semiquantitative real-time PCR assay was performed on the cDNA derived from this RNA using appropriate primer sets, as described in Materials and Methods. Message levels of each selected molecule were determined relative to the housekeeping gene GAPDH. Significantly increased levels of transcripts were detected in the TSP-1 null retina compared with WT control retina (Fig. 4) for all the selected genes. These results support the existence of a more proinflammatory microenvironment in TSP-1–deficient retina after laser injury compared with WT retina and therefore suggest an immunoregulatory role of TSP-1 in the normal retina. 
Figure 4.
 
Retina deficient in TSP-1 expresses increased levels of microglia-associated proinflammatory molecules 14 days after laser-induced injury. Real-time PCR analysis was performed using primer sets for the indicated proinflammatory molecules on cDNA prepared from RNA harvested from the retina derived from either WT or TSP-1 null mice (n = 5 each) 14 days after laser injury. Gene expression relative to GAPDH is presented. *P < 0.05.
Figure 4.
 
Retina deficient in TSP-1 expresses increased levels of microglia-associated proinflammatory molecules 14 days after laser-induced injury. Real-time PCR analysis was performed using primer sets for the indicated proinflammatory molecules on cDNA prepared from RNA harvested from the retina derived from either WT or TSP-1 null mice (n = 5 each) 14 days after laser injury. Gene expression relative to GAPDH is presented. *P < 0.05.
Correlation between the TSP-1 Deficiency of the Retina and Its Failure to Recover from Laser-Induced Injury
In a quiescent retina, activated microglia are reported to produce various neurotrophic factors that protect and regulate the survival of photoreceptors. 9 There are reports highlighting the neuroprotective contribution of microglial activation that permits tissue repair after injury. 31 Both TGF-β and TSP-1 are known to play a significant role in wound healing. 32,33 We hypothesized that TSP-1 in the retina may also promote recovery from retinal injury. To test this possibility we compared recovery from laser-induced injury in WT and TSP-1 null retina by evaluating histology findings before and several days after laser injury. As seen in Figures 5a and 5b, the histologic structure of retina in WT and TSP-1 null mice before laser injury is comparable. In WT retina, as expected, by 14 days after laser injury, signs of recovery were detectable as RPE cells were found to cover the injury site (Fig. 5c), and 120 days after the laser injury a contiguous RPE layer was detected (Fig. 5e). However, such recovery was not seen in TSP-1 null retina. In the absence of TSP-1, though most photoreceptors in the ONL adjacent to the injury site were damaged, this site was also detached from the RPE layer (Fig. 5d). Even as long as 120 days after laser injury, RPE cells failed to cover the injured sites in TSP-1 null retina (Fig. 5f). In fact, by this time, significantly reduced ONL was detected in the TSP-1 null retina compared with WT control retina (Fig. 5g), and pigmented cells appeared to extend into the adjacent disorganized areas of the photoreceptor layer. In some injury sites, RPE also migrated to the ganglion cell layer. These results suggest that in the absence of TSP-1, the retina fails to recover from retinal injury; this coincides with the more proinflammatory microenvironment that presumably prevents the neuroprotective phenotype of microglia. 
Figure 5.
 
Retina deficient in TSP-1 fails to recover from the laser-induced injury. Retina histology (H&E staining) of WT (a, c, e) and TSP-1 null (b, d, f) mice. Normal uninjured (a, b), 14 days (c, d), and 120 days (e, f) after laser injury. Arrows: RPE cells or layer. (g) Quantitative estimate of photoreceptor layer thickness in WT and TSP-1 null retina on indicated day after injury. n = 3–5/group. Scale bar, (af) 50 μm.
Figure 5.
 
Retina deficient in TSP-1 fails to recover from the laser-induced injury. Retina histology (H&E staining) of WT (a, c, e) and TSP-1 null (b, d, f) mice. Normal uninjured (a, b), 14 days (c, d), and 120 days (e, f) after laser injury. Arrows: RPE cells or layer. (g) Quantitative estimate of photoreceptor layer thickness in WT and TSP-1 null retina on indicated day after injury. n = 3–5/group. Scale bar, (af) 50 μm.
Discussion
In the normal retina, TSP-1 immunoreactivity is noted primarily in the vascular cells, OPL, and some photoreceptors. 22 It was also reported in RPE 34 and microglia in a facial nerve axotomy model. 19 In a laser photocoagulation study, TSP-1 immunoreactivity was detected in the laser lesions and the RPE cells adjoining the edges of these lesions. 24 Previously, we reported that TSP-1 derived from RPE cells makes biologically active TGF-β available in the SR, allowing local antigen-presenting cells to induce an anti-inflammatory immune response. 18 Microglia, capable of presenting antigen, 30 in the normal retina on TGF-β exposure are believed to help maintain immune privilege. However, the ability of activated microglia to express inflammatory mediators is clearly implicated in the pathogenesis of various retinal diseases, leading to a hypothesis that activated microglia may actually participate in the initiation and perpetuation of the degenerative process. 8 Regulation of the functional phenotype of activated microglia still remains unclear. In this study we have examined the role of TSP-1 as a potential regulator of microglia function. 
In our previous studies, nonpigmented albino mice were found to be more susceptible than pigmented mice to light-induced retinal damage, 26 and, as such, induced enhanced migration of microglia to their SR. Such migrating microglia are considered activated and capable of producing proinflammatory TNF-α. 12 Similar migration of microglia was also reported by others in response to ROS generated by phototoxicity. 35,36 Nitric oxide (NO) is a ROS that is generated by enzyme iNOS in macrophages, and increased levels of this enzyme in the retina are linked to photoreceptor death. 11,37 Therefore, constitutively increased expression of iNOS and TNF-α in TSP-1 null retina not only implied photoreceptor damage but supported detection of activated MHC class II–expressing microglia near the SR. Thus retinal microenvironment devoid of TSP-1 appeared conducive to the activation of microglia and their migration toward the SR. Therefore, we hypothesized that such retina is likely to make TSP-1 null mice more susceptible to any injury. Consistent with this hypothesis, we detected significant light-induced damage in TSP-1 null retina in spite of their pigment. The damaged areas in these retinas were represented by nonproliferating, Ki-67–negative (data not shown) photoreceptor clusters extending in the OPL. The presence of microglia in these clusters further confirms the damaged nature of these areas. Therefore, we conclude that in normal retina, TSP-1 is likely to play a protective role, possibly avoiding damage by regulating microglia activation. 
Besides chronic form of injury such as light exposure, we examined microglia activation in response to an acute form of retinal injury induced by a thermal laser. Detection of MHC class II–expressing cells of bone marrow origin (CD45+) that included most antigen-presenting cells 1 day after laser-induced injury in both WT and TSP-1 null retina is consistent with recently reported observations by others that within 2 days of laser injury, a mixed population of antigen-presenting cells (macrophages, dendritic cells, microglia) accumulates at the site of the injury. 38 However, in the absence of TSP-1, we found such cells to persist until 14 days after injury. At this time, significantly increased expression of microglia-associated proinflammatory molecules and microglia-derived chemokines in the TSP-1–deficient retina compared with that in the WT retina correlates with the persistence of activated microglia. 
Although the migration of microglia to the site of injury is reported to permit efficient clearance of dead cells and debris, 58 continued activation during this process is typically avoided in normal retina to promote healing rather than inflammation. Accordingly, it has been demonstrated that after phagocytosis, microglia suppress proinflammatory cytokine production. 39,40 Consistent with these reports in WT retina, healing of the injured area was detectable by 14 days after injury with a restored RPE layer. Yet such recovery was not detected in the TSP-1–deficient retina, even up to 120 days after injury, and was accompanied by significantly decreased ONL compared with WT controls. The prolonged detection of activated microglia-associated inflammatory markers in the TSP-1–deficient retina correlates with the enhanced damage and the lack of recovery from it. These observations are consistent with the reports that chronic overactivation of microglia eventually causes retinal degeneration. 8 Therefore, our study strongly supports the possibility that TSP-1 regulates microglia activation, possibly by its ability to activate latent TGF-β. 18  
In conclusion, our data present a significant role played by the extracellular matrix protein TSP-1 in the regulation of activation of microglia to support their neuroprotective and regenerative properties while avoiding prolonged activation and a proinflammatory phenotype. 
Footnotes
 Supported by Department of Defense Grant W81XWH-07–2-0038 and National Institutes of Health Grant EY015472.
Footnotes
 Disclosure: T.F. Ng, None; B. Turpie, None; S. Masli, None
Footnotes
 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
References
Roque RS Caldwell RB . Isolation and culture of retinal microglia. Curr Eye Res. 1993;12:285–290. [CrossRef] [PubMed]
Sedgwick JD Schwender S Gregersen R Dorries R ter Meulen V . Resident macrophages (ramified microglia) of the adult brown Norway rat central nervous system are constitutively major histocompatibility complex class II positive. J Exp Med. 1993;177:1145–1152. [CrossRef] [PubMed]
Thanos S . Sick photoreceptors attract activated microglia from the ganglion cell layer: a model to study the inflammatory cascades in rats with inherited retinal dystrophy. Brain Res. 1992;588:21–28. [CrossRef] [PubMed]
Ling EA Wong WC . The origin and nature of ramified and amoeboid microglia: a historical review and current concepts. Glia. 1993;7:9–18. [CrossRef] [PubMed]
Zeng HY Zhu XA Zhang C Yang LP Wu LM Tso MO . Identification of sequential events and factors associated with microglial activation, migration, and cytotoxicity in retinal degeneration in rd mice. Invest Ophthalmol Vis Sci. 2005;46:2992–2999. [CrossRef] [PubMed]
Zhang C Shen JK Lam TT . Activation of microglia and chemokines in light-induced retinal degeneration. Mol Vis. 2005;11:887–895. [PubMed]
Lee JE Liang KJ Fariss R Wong WT . Ex vivo dynamic imaging of retinal microglia using time-lapse confocal microscopy. Invest Ophthalmol Vis Sci. 2008;49:4169–4176. [CrossRef] [PubMed]
Langmann T . Microglia activation in retinal degeneration. J Leukoc Biol. 2007;81:1345–1351. [CrossRef] [PubMed]
Carwile ME Culbert RB Sturdivant RL Kraft TW . Rod outer segment maintenance is enhanced in the presence of bFGF, CNTF and GDNF. Exp Eye Res. 1998;66:791–805. [CrossRef] [PubMed]
Yin Y Cui Q Li Y . Macrophage-derived factors stimulate optic nerve regeneration. J Neurosci. 2003;23:2284–2293. [PubMed]
Boje KM Arora PK . Microglial-produced nitric oxide and reactive nitrogen oxides mediate neuronal cell death. Brain Res. 1992;587:250–256. [CrossRef] [PubMed]
Kreutzberg GW . Microglia: a sensor for pathological events in the CNS. Trends Neurosci. 1996;19:312–318. [CrossRef] [PubMed]
Zeiss CJ Johnson EA . Proliferation of microglia, but not photoreceptors, in the outer nuclear layer of the rd-1 mouse. Invest Ophthalmol Vis Sci. 2004;45:971–976. [CrossRef] [PubMed]
Combadiere C Feumi C Raoul W . CX3CR1-dependent subretinal microglia cell accumulation is associated with cardinal features of age-related macular degeneration. J Clin Invest. 2007;117:2920–2928. [CrossRef] [PubMed]
Gaucher D Chiappore JA Paques M . Microglial changes occur without neural cell death in diabetic retinopathy. Vision Res. 2007;47:612–623. [CrossRef] [PubMed]
Broderick C Duncan L Taylor N Dick AD . IFN-gamma and LPS-mediated IL-10-dependent suppression of retinal microglial activation. Invest Ophthalmol Vis Sci. 2000;41:2613–2622. [PubMed]
Paglinawan R Malipiero U Schlapbach R Frei K Reith W Fontana A . TGFbeta directs gene expression of activated microglia to an anti-inflammatory phenotype strongly focusing on chemokine genes and cell migratory genes. Glia. 2003;44:219–231. [CrossRef] [PubMed]
Zamiri P Masli S Kitaichi N Taylor AW Streilein JW . Thrombospondin plays a vital role in the immune privilege of the eye. Invest Ophthalmol Vis Sci. 2005;46:908–919. [CrossRef] [PubMed]
Moller JC Klein MA Haas S Jones LL Kreutzberg GW Raivich G . Regulation of thrombospondin in the regenerating mouse facial motor nucleus. Glia. 1996;17:121–132. [CrossRef] [PubMed]
Masli S Turpie B Hecker KH Streilein JW . Expression of thrombospondin in TGFbeta-treated APCs and its relevance to their immune deviation-promoting properties. J Immunol. 2002;168:2264–2273. [CrossRef] [PubMed]
Masli S Turpie B Streilein JW . Thrombospondin orchestrates the tolerance-promoting properties of TGFbeta-treated antigen-presenting cells. Int Immunol. 2006;18:689–699. [CrossRef] [PubMed]
Suzuma K Takagi H Otani A Oh H Honda Y . Expression of thrombospondin-1 in ischemia-induced retinal neovascularization. Am J Pathol. 1999;154:343–354. [CrossRef] [PubMed]
Chamak B Morandi V Mallat M . Brain macrophages stimulate neurite growth and regeneration by secreting thrombospondin. J Neurosci Res. 1994;38:221–233. [CrossRef] [PubMed]
Binz N Graham CE Simpson K . Long-term effect of therapeutic laser photocoagulation on gene expression in the eye. FASEB J. 2006;20:383–385. [PubMed]
Chan CK Pham LN Zhou J Spee C Ryan SJ Hinton DR . Differential expression of pro- and antiangiogenic factors in mouse strain-dependent hypoxia-induced retinal neovascularization. Lab Invest. 2005;85:721–733. [CrossRef] [PubMed]
Ng TF Streilein JW . Light-induced migration of retinal microglia into the subretinal space. Invest Ophthalmol Vis Sci. 2001;42:3301–3310. [PubMed]
Matsubara T Pararajasegaram G Wu GS Rao NA . Retinal microglia differentially express phenotypic markers of antigen-presenting cells in vitro. Invest Ophthalmol Vis Sci. 1999;40:3186–3193. [PubMed]
Zhang J Wu GS Ishimoto S Pararajasegaram G Rao NA . Expression of major histocompatibility complex molecules in rodent retina: immunohistochemical study. Invest Ophthalmol Vis Sci. 1997;38:1848–1857. [PubMed]
Rao NA Kimoto T Zamir E . Pathogenic role of retinal microglia in experimental uveoretinitis. Invest Ophthalmol Vis Sci. 2003;44:22–31. [CrossRef] [PubMed]
Gregerson DS Sam TN McPherson SW . The antigen-presenting activity of fresh, adult parenchymal microglia and perivascular cells from retina. J Immunol. 2004;172:6587–6597. [CrossRef] [PubMed]
Neumann J Gunzer M Gutzeit HO Ullrich O Reymann KG Dinkel K . Microglia provide neuroprotection after ischemia. FASEB J. 2006;20:714–716. [PubMed]
Cordeiro MF . Beyond mitomycin: TGF-beta and wound healing. Prog Retin Eye Res. 2002;21:75–89. [CrossRef] [PubMed]
Uno K Hayashi H Kuroki M Uchida H Yamauchi Y Oshima K . Thrombospondin-1 accelerates wound healing of corneal epithelia. Biochem Biophys Res Commun. 2004;315:928–934. [CrossRef] [PubMed]
Uchida H Hayashi H Kuroki M . Vitamin A up-regulates the expression of thrombospondin-1 and pigment epithelium-derived factor in retinal pigment epithelial cells. Exp Eye Res. 2005;80:23–30. [CrossRef] [PubMed]
Boulton M Rozanowska M Rozanowski B Wess T . The photoreactivity of ocular lipofuscin. Photochem Photobiol Sci. 2004;3:759–764. [CrossRef] [PubMed]
Saraswathy S Rao NA . Photoreceptor mitochondrial oxidative stress in experimental autoimmune uveitis. Ophthalmic Res. 2008;40:160–164. [CrossRef] [PubMed]
Yang LP Li Y Zhu XA Tso MO . Minocycline delayed photoreceptor death in rds mice through iNOS-dependent mechanism. Mol Vis. 2007;13:1073–1082. [PubMed]
Eter N Engel DR Meyer L . In vivo visualization of dendritic cells, macrophages, and microglial cells responding to laser-induced damage in the fundus of the eye. Invest Ophthalmol Vis Sci. 2008;49:3649–3658. [CrossRef] [PubMed]
Magnus T Chan A Grauer O Toyka KV Gold R . Microglial phagocytosis of apoptotic inflammatory T cells leads to down-regulation of microglial immune activation. J Immunol. 2001;167:5004–5010. [CrossRef] [PubMed]
Chan A Magnus T Gold R . Phagocytosis of apoptotic inflammatory cells by microglia and modulation by different cytokines: mechanism for removal of apoptotic cells in the inflamed nervous system. Glia. 2001;33:87–95. [CrossRef] [PubMed]
Figure 1.
 
Constitutive expression of MHC class II, TNF-α, and iNOS in TSP-1 null retina. (a) Frozen sections of WT or TSP-1 null retina were stained with anti-IAb (MHC class II) antibodies and counterstained with fluorescence-labeled conjugated secondary antibody. Nuclei were stained with DAPI (blue). Arrows show detectable positively stained cells. (b) Number of MHC class II–positive cells in the whole section were counted in WT and TSP-1 null retina (n = 3 each). (c) RNA isolated from either WT or TSP-1 null retina was subjected to semiquantitative real-time PCR analysis. Expression of TNF-α and iNOS relative to GAPDH is presented. *P < 0.05.
Figure 1.
 
Constitutive expression of MHC class II, TNF-α, and iNOS in TSP-1 null retina. (a) Frozen sections of WT or TSP-1 null retina were stained with anti-IAb (MHC class II) antibodies and counterstained with fluorescence-labeled conjugated secondary antibody. Nuclei were stained with DAPI (blue). Arrows show detectable positively stained cells. (b) Number of MHC class II–positive cells in the whole section were counted in WT and TSP-1 null retina (n = 3 each). (c) RNA isolated from either WT or TSP-1 null retina was subjected to semiquantitative real-time PCR analysis. Expression of TNF-α and iNOS relative to GAPDH is presented. *P < 0.05.
Figure 2.
 
Light-induced damage in TSP-1 null retina induces migration of microglia. Retina isolated from TSP-1 null (ag) or WT (h, i) mice exposed to bright or ambient light intensity either were subjected to wholemount staining with fluorescence-conjugated antibodies or were sectioned and stained with H&E. (a) Bright light-exposed retina: KSPG+ microglia (red) detectable in the subretinal space. Nuclei were visualized with DAPI (blue). (b) Cluster of KSPG+ microglia detectable in the subretinal space of bright light-exposed TSP-1 null retina. (c) Recoverin (green)–expressing photoreceptors extended from the OPL, forming a rosette-like structure infiltrated with KSPG+ microglia (red). (df) KSPG and Iba1 staining of microglia within the rosette-like structures. (g) H&E staining of TSP-1 null retina exposed to bright light forming rosette-like structures. (h) No detectable KSPG+ cells in the subretinal space of WT control retina exposed to bright light and no detectable structural damage in H&E-stained sections (i). (j) Quantitative assessment of number of KSPG+ cells in the subretinal space of WT or TSP-1 null mice exposed to two light intensities. (k) INL count in bright light-exposed WT and TSP-1 null retina. n = 5 in each group. *P < 0.05. Scale bars: (a, c, d, f) 100 μm; (b) 40 μm; (e) 25 μm.
Figure 2.
 
Light-induced damage in TSP-1 null retina induces migration of microglia. Retina isolated from TSP-1 null (ag) or WT (h, i) mice exposed to bright or ambient light intensity either were subjected to wholemount staining with fluorescence-conjugated antibodies or were sectioned and stained with H&E. (a) Bright light-exposed retina: KSPG+ microglia (red) detectable in the subretinal space. Nuclei were visualized with DAPI (blue). (b) Cluster of KSPG+ microglia detectable in the subretinal space of bright light-exposed TSP-1 null retina. (c) Recoverin (green)–expressing photoreceptors extended from the OPL, forming a rosette-like structure infiltrated with KSPG+ microglia (red). (df) KSPG and Iba1 staining of microglia within the rosette-like structures. (g) H&E staining of TSP-1 null retina exposed to bright light forming rosette-like structures. (h) No detectable KSPG+ cells in the subretinal space of WT control retina exposed to bright light and no detectable structural damage in H&E-stained sections (i). (j) Quantitative assessment of number of KSPG+ cells in the subretinal space of WT or TSP-1 null mice exposed to two light intensities. (k) INL count in bright light-exposed WT and TSP-1 null retina. n = 5 in each group. *P < 0.05. Scale bars: (a, c, d, f) 100 μm; (b) 40 μm; (e) 25 μm.
Figure 3.
 
In the absence of TSP-1, laser-induced retinal injury results in prolonged persistence of MHC class II–expressing cells at the site of injury. Frozen sections of retina harvested from WT (a, c, e) or TSP-1 null (b, d, f) mice after laser injury were stained with fluorescence-conjugated antibodies to detect microglia marker Iba1 and MHC class II (I-Ab). Retina were harvested either 1 day (ad) or 14 days (e, f) after laser injury. Scale bar, 50 μm. (g) Negative control shows staining background. (h) Quantitative estimates of MHC class II–expressing cells in the WT or TSP-1 null retina on indicated days after laser injury. n = 5 for each time point. N.D., not detectable. *P < 0.05).
Figure 3.
 
In the absence of TSP-1, laser-induced retinal injury results in prolonged persistence of MHC class II–expressing cells at the site of injury. Frozen sections of retina harvested from WT (a, c, e) or TSP-1 null (b, d, f) mice after laser injury were stained with fluorescence-conjugated antibodies to detect microglia marker Iba1 and MHC class II (I-Ab). Retina were harvested either 1 day (ad) or 14 days (e, f) after laser injury. Scale bar, 50 μm. (g) Negative control shows staining background. (h) Quantitative estimates of MHC class II–expressing cells in the WT or TSP-1 null retina on indicated days after laser injury. n = 5 for each time point. N.D., not detectable. *P < 0.05).
Figure 4.
 
Retina deficient in TSP-1 expresses increased levels of microglia-associated proinflammatory molecules 14 days after laser-induced injury. Real-time PCR analysis was performed using primer sets for the indicated proinflammatory molecules on cDNA prepared from RNA harvested from the retina derived from either WT or TSP-1 null mice (n = 5 each) 14 days after laser injury. Gene expression relative to GAPDH is presented. *P < 0.05.
Figure 4.
 
Retina deficient in TSP-1 expresses increased levels of microglia-associated proinflammatory molecules 14 days after laser-induced injury. Real-time PCR analysis was performed using primer sets for the indicated proinflammatory molecules on cDNA prepared from RNA harvested from the retina derived from either WT or TSP-1 null mice (n = 5 each) 14 days after laser injury. Gene expression relative to GAPDH is presented. *P < 0.05.
Figure 5.
 
Retina deficient in TSP-1 fails to recover from the laser-induced injury. Retina histology (H&E staining) of WT (a, c, e) and TSP-1 null (b, d, f) mice. Normal uninjured (a, b), 14 days (c, d), and 120 days (e, f) after laser injury. Arrows: RPE cells or layer. (g) Quantitative estimate of photoreceptor layer thickness in WT and TSP-1 null retina on indicated day after injury. n = 3–5/group. Scale bar, (af) 50 μm.
Figure 5.
 
Retina deficient in TSP-1 fails to recover from the laser-induced injury. Retina histology (H&E staining) of WT (a, c, e) and TSP-1 null (b, d, f) mice. Normal uninjured (a, b), 14 days (c, d), and 120 days (e, f) after laser injury. Arrows: RPE cells or layer. (g) Quantitative estimate of photoreceptor layer thickness in WT and TSP-1 null retina on indicated day after injury. n = 3–5/group. Scale bar, (af) 50 μm.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×