| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print June 22, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Laboratories of Immunology and
Molecular and Developmental Biology, National Eye Institute, National Institutes of Health, Bethesda, Maryland;
Laboratory of Molecular Immunoregulation, National Cancer Institute, National Institutes of Health, Frederick, Maryland; and
Department of Ophthalmology and Visual Science, Tokyo Medical and Dental University Graduate School, Japan
1 Correspondence: Laboratory of Immunology, National Eye Institute, NIH, Building 10, Room 10N208, 10 Center Drive, Bethesda, MD 20892-1857. E-mail: geryi{at}nei.nih.gov
ABSTRACT
Immune cell-mediated inflammatory responses are triggered by TCR engagement with the target antigen, the initial event that brings about the complex sequence of events of the inflammatory process. Another form of inflammation is induced by local expression of certain cytokines. Unlike the former form of inflammation, little is known about the basic features of the cytokine-induced responses. Here, we analyzed tissue morphology, the infiltrating cells, and up-regulated, inflammation-related genes in mouse eyes in which inflammation is triggered by local transgenic (Tg) expression of cytokines and compared these features with those in eyes with experimental autoimmune uveitis (EAU), in which inflammation is initiated by engagement of TCR on sensitized T cells with their target antigen, followed by the well-defined, subsequent cytokine production. Eyes of IFN-
Tg mice exhibited severe, morphological changes but essentially no inflammation, and intense inflammation was found in eyes of interleukin (IL)-1 or IL-7 Tg mice. The cellular infiltration in eyes of these latter two lines of Tg mice resembled that in eyes with EAU by including many CD4 cells, but unlike in EAU, the infiltration in Tg eyes contained large proportions of B cells and only small numbers of macrophages. Real-time PCR analysis of eye RNA revealed differences among the disease models in the expression profiles of various inflammation-related genes. It is interesting that a bias toward T helper cell type 1 immunity (high IFN-, RANTES/CCL5, MIG/CXCL9, and T-bet but low IL-4, IL-5, and GATA-3 transcripts) was found in EAU eyes but not in eyes of IL-1 and IL-7 Tg mice. The results thus show that similar to TCR engagement, local expression of certain cytokines triggers a complex, subsequent production of numerous inflammation-related molecules, but features of the ensued inflammatory process are determined by the triggering mechanism.
Key Words: immune-mediated inflammation • eye • experimental autoimmune uveitis (EAU) • transgenic mice
INTRODUCTION
Experimental autoimmune diseases serve as models for localized, inflammatory conditions that affect organs such as the central nervous system, thyroid, testes, or eye. These experimental diseases are induced by immunization with antigens specific to the corresponding tissue and are triggered by the engagement of TCR on sensitized T cells with the specific target antigen. This event initiates activation of the lymphocyte and the release of cytokines that brings about the complex sequence of events of the inflammatory process, which culminates with the migration of different lymphoid cell populations into the affected tissue with further amplification of inflammatory molecule production. Extensive investigation in recent years of the inflammatory process has dissected its different components and identified a large number of molecules that are involved in this process. These molecules belong to four major families: cytokines, chemokines, chemokine receptors, and adhesion molecules, and well-defined function(s) are attributed to each of these molecules [1 2 3 ].
Transgenic (Tg) technology has made it possible to express biologically active molecules locally, including cytokines. In some of these experimental systems, cytokine production initiated local inflammatory processes characterized by various forms of lymphoid cell infiltration and tissue damage [4 5 6 7 8 ]. Little or no information has been collected, however, concerning the molecular and cellular mechanisms involved in these inflammatory processes.
In previous studies, we described the phenotype and certain features of eyes of Tg mice expressing IFN- or IL-1 under control of the lens 
A-crystallin promoter [9
10
11
]. More recently, we generated a new line of Tg mice expressing IL-7 under control of the same promoter. To learn about the mechanisms and molecules involved in the pathogenic processes in eyes of these three lines of Tg mice, we compared them in this study with eyes with experimental autoimmune uveitis (EAU), induced by immunization with a retinal antigen. The comparison was made for three parameters of inflammation, i.e., the typical morphological changes, the populations of the infiltrating cells, and the profiles of the inflammation-related gene expression. Severe, morphological changes but almost no inflammation were found in eyes of the IFN- Tg mice. Conversely, eyes of IL-1 or IL-7 Tg mice exhibited intense, inflammatory changes with heavy cellular infiltration and a remarkable up-regulation of most tested inflammation-related genes. Notable differences were found among eyes of the Tg mice and eyes with EAU in the profile of their cellular infiltration and the expression levels of certain major genes.
MATERIALS AND METHODS
Mice
Six-week-old female B10.RIII mice, purchased from Jackson Laboratory (Bar Harbor, ME), were used for EAU induction. Tg mice expressing IFN- or IL-1ß under control of the A-crystallin promoter were generated as described elsewhere ([9
, 11
], respectively), and IL-7 Tg mice were generated as described below. IL-1 and IL-7 Tg mice were on the FVB/N background, whereas IFN- Tg mice were on the BALB/c background. The specificity of transgene expression in the three Tg mouse lines was confirmed by real-time analysis of lens RNA, using the method detailed below. As shown in Figure 1
, only the corresponding transcripts were detected in these lens RNA samples. All Tg mice were used at the age of 6–10 weeks.
|
Generation of IL-7 Tg mice
The "DH1" transgene construct was made with the mouse A-crystallin promoter, driving expression of mouse IL-7 (mIL-7). Splicing and polyadenylation signals were derived from the mouse ßA3/A1-crystallin gene.
A plasmid vector containing the mIL-7 cDNA (pcDNA3-mIL-7) was a kind gift from Jay Bream, National Institute of Arthritis and Muskuloskeletal and Skin Diseases, NIH. An 
660-base pair (bp) fragment containing the IL-7 open-reading frame was excised by cutting with HindIII and ApaI. The plasmid pCBB1, containing the chicken ßB1-crystallin promoter and the mouse ßA3/A1-crystallin splice and polyadenylation signals, was a kind gift from James F. Hejtmancik, National Eye Institute (NEI), NIH. An 1150-bp fragment, containing the mouse ßA3/A1-crystallin splice (exon 5, intron, and exon 6)/polyadenylation signals, was excised by cutting with ApaI and AccI. These two DNA segments were then cloned between the HindIII and AccI sites of pEW30, which contains the mouse A-crystallin promoter (–366/+46). The completed plasmid was cut with SacII and AscI to liberate the 5.2-kb transgene used for microinjection.
The DH1 (A-crystallin/IL-7/poly A) construct was microinjected into fertilized, one-celled FVB/N embryos by conventional methods. Twenty mice resulting from these injections were screened by PCR for the presence of the transgene, using oligonucleotide primers that amplify a 541-bp DNA segment spanning the IL-7/polyadenylation signal ligation junction. A single Tg founder was identified and confirmed by Southern blot analysis. This female founder was bred to a male FVB/N mouse to establish the Tg line.
EAU induction
EAU was induced by immunization with human interphotoreceptor retinoid-binding protein (IRBP) peptide 161–180 (SGIPYIISYLHPGNTILHVD) [12
]. The peptide was administered subcutaneously in emulsion with complete Freund’s adjuvant (CFA) at 20 µg per mouse in a volume of 200 µl, given in the tail base (100 µl) and the two thighs (50 µl each). Eyes were collected 14 days postimmunization.
Flow cytometric analysis of the ocular-infiltrating cells
Eyes collected from the experimental mice were put into RPMI-1640 medium (Mediatech, Inc., Herndon, VA) and dissected under microscope to remove the lens. The remaining portion of the eye was minced with scissors and shaken in 10 ml RPMI-1640 medium supplemented with 0.5 mg/ml collagenase type D (Sigma-Aldrich, St. Louis, MO) at 37°C for 60 min. Released cells were washed twice with RPMI 1640, suspended with 100 µl fluorescein-activated cell sorter (FACS) staining buffer (BD Bioscience, San Diego, CA) containing 10 µg/ml anti-mouse CD16/CD32 (BD Bioscience), and incubated at 4°C for 5 min to block Fc receptors for IgG. Directly labeled antibodies were added to the cell suspensions at the final concentration of 2 µg/ml. Following incubation at 4°C for 30 min, the cells were washed twice and suspended with 200 µl FACS buffer. Antibodies used were against CD3, CD4, CD8, CD11c, CD19 (BD Bioscience), F4/80 (Caltag Laboratories, Burlingame, CA), or the isotype control Ig (BD Bioscience). FACSCalibur (BD Bioscience) was used to acquire 50,000 events in a live gate, and data were analyzed with FlowJo software (Tree Star, Inc. Ashland, OR).
Real-time PCR analysis
Total RNA was extracted from whole eyeball or lens using TRIZol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol and pooled within each group to obtain a sufficient amount as well as to decrease biological variability. Total RNA was incubated with DNase (Promega, Madison, WI) at 37°C for 10 min and reverse-transcribed into cDNA using oligo (dT)12–18 primers (Invitrogen) and Superscript II reverse transcriptase (RT; Invitrogen). A negative control sample was prepared without RT for each RNA sample. cDNA samples were used for real-time PCR, performed on an ABI Prism 7700 sequence detection system (Applied Biosystems, Foster City, CA) with Taqman primer/probe sets purchased from Applied Biosystems. The tested genes are listed in Table 1
. cDNA samples (5 µl) were mixed with 1.25 µl primer/probe mixture and 12.5 µl TaqMan Universal PCR master mix (Applied Biosystems) reagent in a final volume of 25 µl. Real-time PCR was performed by heating samples at 95°C for 10 min followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. cDNA synthesized from universal mouse total RNA (BD Bioscience) was used to generate standard curves for each reaction. Data obtained were normalized with values of GAPDH and calculated to obtain fold changes of values from each control mouse line, as indicated in Table 1
.
|
RESULTS
Morphological features of mouse eyes with EAU or with cytokine-induced inflammation
Figure 2
demonstrates the typical morphological changes in mouse eyes with inflammation induced by the different mechanisms. In line with previous reports [13
, 14
], the changes in eyes with EAU concentrated mainly in the posterior segment of the eye and typically included retinal detachment and folding, loss of photoreceptors, and inflammatory cell infiltration throughout the different retinal cell layers, subretinal space, choroids, and vitreous (Fig. 2A
2B
2C)
.
|
Eyes from IL-7 Tg mice exhibited inflammatory changes affecting the anterior and posterior segments, including optic nerve involvement (Fig. 2G) . A unique feature in eyes of these Tg mouse lines was the dense accumulation of lymphocytes in the limbus and the adjacent conjunctiva (Fig. 2H) , as well as at the optic nerve sheath (Fig. 2I) .
Severe morphological changes were seen in eyes of the IFN- Tg mice (Fig. 2J)
. These eyes showed total retinal detachment, formation of retinal folds, and rosettes. The lens was cataractous, and the optic nerve showed hypotrophy (Fig. 2L)
. It is important, however, that unlike in eyes of mice expressing IL-1 or IL-7, little or no cellular infiltration or exudate could be detected in eyes of the IFN- Tg mice.
Identification of the infiltrating cell populations in eyes of the different mouse groups
To identify the populations of infiltrating cells in inflamed mouse eyes of the different experimental groups, we performed flow cytometric analysis of cells harvested from these eyes following treatment with collagenase (Fig. 3
). This treatment yielded a population of resident ocular cells of 2–4 x 106 per eye, which were mixed with infiltrating cells in inflamed eyes. Essentially, no infiltrating cells could be found in eyes from IFN- Tg mice, but sufficient numbers for analysis of infiltrating cells were found in eyes of the other mouse lines.
|
Profiles of gene expression in eyes of mice of the different experimental groups: real-time PCR
To analyze for gene expression profiles in the mouse eyes with inflammation induced by TCR engagement (EAU) or by local cytokine production, we measured by real-time PCR the level of mRNA transcripts of 46 inflammation-related molecules. The data are summarized in Table 1
.
This table also records the levels of all tested transcripts, normalized against GAPDH, in normal (noninflamed) eyes of the mouse strains used in this study, i.e., B10.RIII for EAU, FVB/N for IL-1 and IL-7 Tg mice, and BALB/c for the IFN- Tg animals. With a few exceptions, levels of most tested transcripts were similar among the three strains of mice, despite the remarkable differences in their phenotypes. Differences were noted, particularly between FVB/N and the other two strains, B10.RIII and BALB/c. FVB/N mouse eyes demonstrated high expression of IL-6, MCP-1/CCL2, and VLA-4 transcripts but low expression of IL-5 and GATA-3 transcripts. FVB/N mice differ from the other two strains by their development of congenital retinal degeneration, a pathological process that could contribute to these differences in expression of immune- and inflammation-related molecules.
Transcripts of several molecules were found to be expressed constitutively at relatively high levels. These included two of the nine tested cytokines (IL-1Ra and TNF-), six of the 10 chemokines (MCP-1/CCL2, C-10/CCL6, eotaxin/CCL11, CTACK/CCL27, BRAK/CXCL14, and fractalkine/CX3CL1), and four of the nine adhesion molecules (VCAM-1, LFA-1, cadherin-1, and cadherin-4). The expression level of individual transcripts in the normal control eyes conversely affected the degree of up-regulation of the molecule in the inflamed eyes. Thus, for example, as a result of the exceptionally high expression of the IL-1Ra transcript in control eyes, the substantial increase in the absolute expression of this molecule in inflamed eyes is recorded in Table 1
as merely a moderate increase in the relative level of this transcript. In contrast, IL-4 and IFN- could not be detected by the assay in normal mouse eyes, and the high relative increases in their expression in inflamed eyes represent only moderate levels of absolute expression. The great majority of all these inflammation-related transcripts was found to be up-regulated in eyes with EAU and eyes of the IL-1 or IL-7 Tg mice. In contrast, minimal or no change was seen in the transcript levels in severely damaged eyes of IFN- Tg mice. Despite the same transcripts being up-regulated in eyes with EAU or with inflammation induced by Tg expression of IL-1 or IL-7, the profiles of up-regulation differed remarkably between these experimental eyes. Several observations are of particular interest:
In line with our previous study [15
], mouse eyes with EAU exhibited an expression profile with a strong bias toward T helper cell type 1 (Th1)-immune response, demonstrated here mainly by the high up-regulation of IFN-, RANTES/CCL5, MIG/CXCL9, and T-bet and low levels of IL-4, IL-5, and GATA-3 transcripts. No such bias was seen, however, in the profiles of the two Tg mouse lines; similar up-regulation levels of Th1- and Th2-specific transcripts were seen in these eyes.
Other transcripts, which were substantially higher in eyes of the Tg mice than in eyes with EAU, included CCR7 and L-selectin. Conversely, eyes with EAU showed higher expression levels of IL-6, LFA-1, VLA-4, P- and E-selectin, cadherin-1, and SOCS3.
The major differences between gene expression in eyes with EAU or with inflammation induced by Tg expression of IL-1 or IL-7 are listed in Table 2 , in which the differences are presented as the calculated ratios between the fold increase value of the indicated transcript in eyes of the Tg mice and eyes with EAU.
|
Unlike the up-regulation seen for most chemokines, minimal or no difference was noted among inflamed eyes of all tested mice in the transcript levels of three chemokines, CTACK/CCL27, BRAK/CXCL14, and fractalkine/CX3CL1. This observation is in line with data collected in our previous study [15 ]. As shown in Table 1 , genes of these molecules are highly, constitutively expressed in normal mouse eyes.
Transcripts of apoptosis-related molecules, Bcl-2, Bax, and Bad, were found at remarkably high levels in the eyes of all normal mice tested. The levels of these molecules were only slightly up-regulated or were the same in the different inflamed eyes.
Considerably high levels of transcripts for the regulators of cytokine signaling molecules SOCS1 and SOCS3 were found in normal mouse eyes. Their levels were elevated moderately in eyes with the three types of inflammation
iNOS transcript level increased markedly in EAU eyes and eyes of IL-1 Tg mice but not in IL-7 Tg eyes. The high levels of MHC-II transcript were elevated moderately in all inflamed eyes tested.
DISCUSSION
The present study provides new information about certain features of the inflammatory processes, which are induced by local expression of cytokines with no involvement of TCR engagement. The morphological changes, profiles of the infiltrating cell populations, and expression levels of individual genes in inflamed eyes of the Tg mice differed from those seen in eyes with EAU, in which TCR engagement initiated the disease.
Extremely intense inflammation was found in eyes of the IL-1 Tg mice; less severe inflammation was observed in eyes of the IL-7 Tg mice (Fig. 2) . Inflammatory processes were also reported in mice expressing IL-1 or IL-7 locally in other tissues [4 , 5 , 16 ]. It is noteworthy, however, that the inflammation type and profile of the cellular infiltration differed considerably among various organs in which the same cytokines were expressed. Thus, whereas macrophages comprised the major population in the infiltrate induced by IL-1 in the skin [4 ] or joints [16 ] or by IL-7 in the colon [5 ], these cells were a small minority in ocular inflammation induced by these two cytokines (Fig. 3) . These differences among the infiltrating cell populations in animals with the cytokines expressed locally in different organs could be attributed at least in part to differences in the biological environment in these tissues and to differences in the chemokines produced by resident cells in these tissues [17 ]. This notion is also in line with our unpublished observation that the infiltrate composition differed even between different tissues of the same affected eye, as seen between the infiltrates in the limbus (mostly lymphocytes) and anterior chamber (mainly polymorphonuclear leukocytes) of IL-7 Tg eyes.
Unlike our observations in IL-1 and IL-7 Tg mice, eyes of the IFN- Tg mice exhibited severe morphological changes but essentially no inflammation (Fig. 2J
2K
2L)
. The reason for lack of inflammation in eyes of IFN- Tg mice in our study is unknown, but the pathogenic process could be similar to the injury and apoptosis of resident cells and tissue remodeling, as observed in lungs in which this cytokine was Tg-expressed [18
]. Egan et al. [19
] reported a low level of inflammation, but only at young ages, in eyes of Tg mice expressing IFN- under control of the A-crystallin promoter. Conversely, local Tg expression of IFN- produced cellular infiltration in another eye tissue, i.e., the retina [20
], or in the pancreas [21
]. As mentioned above, these differences could be a result of different environments in these tissues as well as to different profiles of locally produced chemokines in various tissues.
Analysis of the infiltrating cell populations in eyes of the IL-1 and IL-7 Tg mice revealed that they resembled eyes with EAU, only by containing large proportions of CD4 cells. Conversely, the infiltration in these Tg mouse eyes differed from that in eyes with EAU in including substantial portions of B cells but only small numbers of macrophages; opposite observations with these cell populations were made in eyes with EAU (Fig. 3) . The relatively large numbers of B cells in eyes of the Tg mice could be attributed, at least in part, to the strong expression of Th2-immune molecules in these eyes but not in eyes with EAU (Table 1 and see below). In addition, this difference between the infiltrating cell populations could be a result of different expression levels of other molecules, mainly chemokines and adhesion molecules (Table 1) .
To learn about the inflammatory processes in eyes of the Tg mice and to compare them with the processes in eyes with EAU, we carried out a comprehensive study in which we measured the levels of 46 inflammation-related transcripts using the real-time PCR technology (Table 1) . The validity of our observations by this method is indicated by the close similarity between the data collected in the present study with normal and EAU eyes (Table 1) and the data of a previous study in which such eyes of another mouse strain (B10.A) were examined [15 ]. In addition to real-time PCR, the expression of major transcripts was verified by the RNase protection assay (RPA). A good correlation was found in general between the two assays in detecting expression of most transcripts, but quantitative differences were observed with a few transcripts between their levels as measured by real-time PCR and the intensity of their bands by RPA (data not shown). These quantitative differences are attributed to the basic differences between the mechanisms by which the two methods detect mRNA transcripts.
As detailed in Results, remarkable differences were noted between the transcript expression profiles of eyes with EAU and eyes with inflammation induced by local expression of IL-1 or IL-7. The transcripts showing the highest and/or noteworthy differences are listed in Table 2 .
The most conspicuous difference between eyes with EAU and eyes of the Tg mice is the bias toward a Th1 response in EAU eyes but not in Tg mouse eyes. This bias was noted in our previous study with EAU mouse eyes [15
] and is depicted here by the relatively high expression levels of IFN-, RANTES/CCL5, MIG/CXCL9, VLA-4, and T-bet but low levels of IL-4, IL-5, eotaxin/CCL11, and GATA-3 transcripts. Eyes of Tg mice, conversely, had similar expression levels of the Th1- and Th2-specific transcripts (Table 1)
. The bias toward Th1-immune response in mice with EAU could be attributed to the use of CFA in the induction of disease. A major component of this adjuvant is Mycobacterium tuberculosis bacteria, which contain high levels of CpG oligodeoxynucleotide, a ligand for the Th1-stimulating Toll-like receptor (TLR)-9 [22
]. Microbial-based adjuvants are used to induce the great majority of experimental autoimmune diseases, and therefore, the pathogenic process in all these disease models is likely TLR-mediated and Th1-biased [22
]. Cytokine-induced, local inflammation thus provides a model for TLR-independent and unbiased, immunopathogenic processes.
In addition to the Th1-associated transcripts mentioned above, several other transcripts related to inflammation were up-regulated in eyes with EAU, more than in eyes of the Tg mice. These included the cytokine IL-6 and the adhesion molecules LFA-1, E-selectin, and cadherin-1.
In contrast to the differences between the mice with EAU and the Tg mice, a noteworthy similarity was observed between the profiles of gene up-regulation in eyes of the two lines of Tg mice. The finding of higher levels of up-regulation of most tested genes in the IL-1 Tg mice could be attributed to the more severe, inflammatory changes in eyes of these mice than in eyes of the IL-7 Tg mice (Fig. 2) .
In spite of the quantitative differences mentioned above among the profiles of gene expression in eyes with the three models of ocular inflammation, it is remarkable that basically, the same inflammation-related molecules were used by inflammatory processes triggered by different mechanisms, i.e., TCR engagement or local expression of two different cytokines.
The majority of inflammatory diseases is triggered by TCR engagement, but excessive, local expression of IL-1 and related cytokines is assumed to initiate inflammatory processes in diseases such as Familial Mediterranean Fever [23 ]. Data collected with the Tg mice may help, therefore, in understanding the pathogenesis of these diseases.
The real-time PCR data also identified a number of inflammation-related molecules, which are expressed constitutively in normal eyes. These molecules belong to the families of cytokines, chemokines, adhesion molecules, and apoptosis-related molecules (Table 1) . The level of several of these molecules was increased in inflamed eyes (e.g., C-10/CCL6, VCAM-1), and it is conceivable that these molecules perform inflammation-related and noninflammation-related functions. Other molecules, in particular, those related to apoptosis and the chemokines CTACK/CCL27 and BRAK/CXCL14, showed minimal changes in expression level as a result of inflammation and are likely to have mostly noninflammation-related activities. Little is known about the noninflammation-related functions of these constitutive molecules. In a recent study [24 ], we found that the transcript for CTACK/CCL27 transcribes mostly the PESKY splice variant of this chemokine, a molecule thought to be involved in cytoskeletal rearrangements in cells [25 ]. Other constitutive chemokines are probably involved in cell movement in the normal eye. For example, our recent study suggests that fractalkine/CX3CL1 plays a role in physiological macrophage migration in the retina [26 ].
In summary, we present here for the first time data concerning the cellular composition of the infiltrate and the expression profiles of major inflammation-related genes in inflammatory processes induced in mouse eyes by local expression of two cytokines, IL-1 and IL-7. A comparison between these inflammatory processes and the one induced by TCR engagement, in eyes with EAU, demonstrated differences among these disease models in the participation of lymphoid cell populations, as well as in the expression levels of inflammation-related genes.
ACKNOWLEDGEMENTS
This work was supported by the Intramural Research Program of the NEI, NIH. We thank Robert S. Lee for tail DNA analysis, the staff of NEI Histology Core for tissue section preparations, and Ricardo Dreyfuss for digital microphotography.
Received December 8, 2005; revised April 25, 2006; accepted May 7, 2006.
REFERENCES
in basal epidermis Proc. Natl. Acad. Sci. USA 92,11874-11878 transgenic mice J. Immunol. 172,577-584 Interferon expression disrupts lens and retinal differentiation in transgenic mice Dev. Biol. 166,557-568[CrossRef][Medline] transgenic mice J. Clin. Invest. 107,1127-1135[Medline]}-induced alveolar remodeling and pulmonary emphysema J. Immunol. 174,8106-8115 in the murine lens results in multiple ocular abnormalities and an early but self-limited inflammatory response Curr. Eye Res. 14,1063-1071[Medline] in the retina develop inflammation of the eye and photoreceptor loss Invest. Ophthalmol. Vis. Sci. 35,2667-2681 Cell 52,773-782[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
