Extracellular Signal-Regulated Kinase Activation Predominantly in Müller Cells of Retina with Endotoxin-Induced Uveitis

Abstract

purpose. To evaluate the consequences of mitogen-activated protein kinase (MAPK)-mediated signaling in retinas with endotoxin-induced uveitis (EIU).

methods. EIU was induced with footpad inoculation of lipopolysaccharide (LPS). To identify the expression and activity of extracellular signal-regulated kinase (ERK), c-Jun NH2-terminal kinase (JNK), and p38, Western blot analysis and immunohistochemistry (IHC) were performed using antibodies against these kinases and phosphorylated forms. To evaluate the ERK mRNA expression level, semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) was performed. To identify cell species that express phosphorylated (p)-ERK, simultaneous demonstration of p-ERK and glial fibrillary acidic protein (GFAP) was performed with combined IHC and in situ hybridization. Dexamethasone (Dex) was used to reduce the LPS-induced inflammatory stimulus, and changes in p-ERK expression were evaluated by Western blot analysis after treatment.

results. Only p-ERK among the phosphorylated MAPKs increased after LPS stimulation, according to Western blot analysis. p-ERK increased after LPS injection, whereas both the Western blot and RT-PCR studies showed no apparent changes in ERK-1 and -2 expression. IHC revealed that strong p-ERK-positive staining initially appeared in the Müller cell bodies. Thereafter p-ERK immunostaining was also observed transiently in the radial processes of the Müller cells. The double-labeling study revealed that almost all Müller cells were positive for GFAP and p-ERK. Dex treatment substantially reduced expression of p-ERK, beginning 12 hours after treatment.

conclusions. The present study suggests that LPS stimulation activates ERK in Müller cells, whereas the total amount of ERK is unchanged. Because the LPS-induced p-ERK level was reduced by Dex treatment, its expression seems to be associated with ocular inflammatory stimulus. Because the inflammatory stimulus elicited in EIU upregulated ERK activity in Müller cells, activated Müller cells may play a crucial role in protecting retinal cells from such inflammation.

Lipopolysaccharide (LPS) stimulation causes endotoxin-induced uveitis (EIU) in a model of human ocular inflammation. ¹ EIU mainly leads to acute inflammation of the anterior ocular segment. ^{1 2 3 4} In the retina, EIU also elicits various expressions, such as inducible nitric oxide ^{5 6} ; interleukin (IL)-1α, -1β, and -6; tumor necrosis factor (TNF)-α; interferon (IFN)-γ ⁷ ; and β-amyloid precursor protein. ⁸ Most of these molecules are inflammatory cytokines or molecules associated with inflammation. Regarding the intracellular signaling elicited by these molecules, data are available primarily about macrophages, ^{9 10} monocytes, ^{11 12} and neutrophils. ¹³ STAT3, NF-κB, and mitogen-activated protein kinase (MAPK)-mediated pathways are associated with stimulation by LPS. In the case of macrophages, LPS induces extracellular signal-regulated kinase (ERK) and c-Jun NH2-terminal kinase (JNK) phosphorylation. ^{14 15 16 17 18} In the nervous system, a few studies have shown that ERK and p38 MAPK were activated by LPS stimulation in microglia or astrocytes. ^{19 20 21} However, in mammalian retina an alteration of the intracellular signaling activity in response to LPS stimulation is poorly understood. Therefore, in the present study, we attempted to determine which MAPK system is activated in specific retinal cells during ocular inflammation. 

Materials and Methods

Animal Procedures

All experiments were performed on male Wistar rats (8 weeks old). For all surgical procedures, the rats were anesthetized before surgery with pentobarbital administered intraperitoneally (0.3 mg/kg). EIU was induced according to a previously reported method. ⁶ The animals were injected with 150 μg LPS from Salmonella typhimurium (Sigma Chemical Co., St. Louis, MO) in the right footpad. The control was injected with saline in the same manner. For dexamethasone (Dex) treatment, the rats were injected with Dex (1 mg/kg) 12 hours after LPS injection. All animal procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Retinal Protein Extraction and Western Blot Analysis

The animals were killed with an overdose of pentobarbital at 6, 12, and 16 hours; 1, 2, 3, and 5 days; and 1 and 2 weeks after LPS injection, or 0, 3, 6, 12, and 24 hours after Dex injection. The total retinal protein was prepared according to a previously reported method. ²² The eyes were immediately enucleated and the retinas were dissected from the scleral wall. The extracted tissue was solubilized in 3% sodium dodecyl sulfate (SDS) buffer (1 mM o-vanadium, 0.19 μL/mL aprotinin, and 0.1 μg/mL phenylmethylsulfonyl fluoride) and boiled for 10 minutes. The lysates were added to the same volume of 0.3 M sucrose, homogenized, and centrifuged at 14,000 rpm for 15 minutes at 4°C. The lysates were stored at −80°C until use. 

For Western blot analysis, 50 μg total protein in SDS sample buffer was placed in each lane. The samples were gel electrophoresed onto 10% SDS-polyacrylamide gel. After transfer, polyvinylidene (PVDF) membranes were washed in Tris-buffered saline containing 0.1% Tween-20. For the MAPK examination, the membranes were incubated overnight at 4°C with primary antibodies (phosphorylated [p]-ERK, p-JNK, and p-p38, diluted 1:1000, New England Biolabs Inc., Beverly, MA; ERK-1 and ERK-2, diluted 1:3000; Santa Cruz Biotech, Santa Cruz, CA; and β-actin diluted 1:1000, Sigma Chemical Co.). Enhanced chemiluminescence (ECL) Western blot analysis (Amersham, Arlington Heights, IL) was used for detection. 

Section Preparation

For section preparation, the animals were killed as described previously and perfused with 4% paraformaldehyde and 0.1 M phosphate buffer. The eyes were enucleated and postfixed overnight with the same solution at 4°C. They were then dehydrated, embedded in paraffin wax, sectioned into specimens 7 μm thick, mounted onto 3-aminopropyltriethoxysilane-coated slides, and stored under dry conditions until histologic analysis. The histologic results shown are representative of six animals. 

Immunohistochemistry

The sections were deparaffinized, rinsed in PBS, and incubated in blocking solution containing 0.5% Triton X-100, 3% BSA, and 0.02% Na-azide in PBS for 30 minutes at room temperature. These pretreated sections were incubated with primary antibody (dilutions: ERK-1, 1:2000; p-ERK, 1:300) overnight at 4°C. For single staining, the sections were rinsed three times in PBS, incubated with secondary antibody (goat biotinylated anti-rabbit IgG or horse biotinylated anti-mouse IgG (diluted 1:500, Vector Laboratories, Inc., Burlingame, CA) for 2 hours at room temperature, rinsed three times in PBS, and incubated in avidin-biotin-peroxidase complex (Vector) in PBS for 1 hour at room temperature. They were rinsed in PBS and immersed in 0.05 M Tris-HCl (pH7.6). Colorization was performed in Tris-HCl containing diaminobenzidine and hydrogen peroxide. 

In Situ Hybridization

All in situ hybridization (ISH) procedures were performed according to a previously reported method, with minor changes. ²² For probe synthesis, rat cDNA fragments of glial fibrillary acidic protein (GFAP; nucleotides 2156-2693, 537 bp; GenBank L27219; provided in the public domain by the National Institutes of Health, Bethesda, MD and available at http://rsb.info.nih.gov/genbank) was isolated using reverse transcription-polymerase chain reaction (RT-PCR). These fragments were subcloned, and digoxigenin (DIG)-labeled cRNA probes were prepared by in vitro transcription. 

Prehybridization procedures were completed, and hybridization was performed for approximately 12 hours at 58°C. After hybridization, the slides were washed and equilibrated in buffer 1 (100 mM Tris-HCl [pH 7.5] and 150 mM NaCl) for 5 minutes. Blocking was performed with 20% sheep serum in buffer 2 (buffer 1 with 0.5% skim milk and 0.1% Tween 20) for 2 hours at room temperature. The specimens were incubated with alkaline phosphatase-conjugated Fab fragments against DIG (diluted 1:2000 in 5% sheep serum and buffer 2; Roche Molecular Biochemicals, Mannheim, Germany) overnight at 4°C. For colorization, the slides were washed three times for 30 minutes each time in buffer 1, equilibrated in buffer 3 (100 mM Tris-HCl [pH 9.5], 100 mM NaCl, and 50 mM MgCl₂) for 10 minutes and stained with nitroblue tetrazolium-5-bromo-4-chloro-3-indoyl phosphate (NBT-BCIP) stock solution (Roche) in buffer 3 at room temperature. The reaction was stopped with 10 mM Tris-HCl (pH 7.6) and 1 mM EDTA, and the slides were mounted or successively processed for IHC by double staining, as described previously. 

Semiquantitative RT-PCR

The retina was extracted from the control rats and 1 day after LPS injection, as described previously, and total RNA was isolated by the acid guanidinium-phenol-chloroform method. RNA was quantitated by absorbance at 260 and 280 nm, and 3 μg was used to synthesize cDNA by RT (SuperScript II; GibcoBRL, Rockville, MD). PCR amplification was performed with the following primers: AGGGGAACTGCTGGGTCGTCCCGGT (sense) and CGGGCCTTCATGTTAATGATACAATT (antisense) for ERK-1 (790 bp), and AAGTTCAATGGCACAGTCAAGGCT (sense) and GAGGGTGCAGCGAACTTTATTGAT (antisense) for glyceraldehyde 3-phosphate dehydrogenase (GAPDH; 946 bp). The reaction was performed for 23 to 29 cycles under the following conditions: denaturation, 95°C for 30 seconds; annealing, 60°C for 45 seconds; and extension, 72°C for 60 seconds. The PCR products were electrophoresed through 2% agarose gel and photographed. 

Results

Levels of p-ERK, ERK-1, and ERK-2 after LPS Treatment

We initially examined the expression of the activated form of the three MAPK family members, ERK, JNK, and p38, using antibodies that recognize the phosphorylated form of these kinases. Among those examined, only the p-ERK signal was detected by Western blot analysis (data not shown), and further investigation was thus performed for ERK responses. The p-ERK level increased slightly at 16 hours after LPS injection compared with control levels (Fig. 1A) , and the expression was greater at 1 day after injection and remained so for 2 weeks after the injection. In contrast, the expression levels of ERK-1 and -2, together with the internal control (β-actin), were not significantly changed at any time point (Figs. 1B 1C 1D) . 

Immunohistochemical Localization of p-ERK and ERK-1 after LPS Stimulation

Because the Western blot study revealed that the protein level of p-ERK increased after LPS injection (Fig. 1A) , we further examined p-ERK expression in the retina by IHC. No immunoreactivity was detected in control retinas (Fig. 2A) . In the LPS-treated retinas, positive staining initially appeared in the bodies of cells in the inner nuclear layer (INL) 6 hours after LPS injection (Fig. 2B) . Most positive cell bodies were in the narrow middle layer of the INL. At day 1 after injection, immunoreactivity markedly increased, not only in the INL, but also in the fibers, which often traversed the outer limiting membrane (OLM) to the inner limiting membrane (ILM) vertically (Fig. 2C) . Figure 2I is a higher magnification of the section derived 3 days after injection, in which very intense p-ERK immunoreactivity was observed in the radial processes extending from the ILM through the OLM. According to this staining pattern in shape and localization, the positive cells were presumed to be Müller cells. The positive signal observed in fiber decreased to the control level in a few days, but prolonged cell body staining was observed 2 weeks after LPS injection (Fig. 2D) . 

We then performed IHC using an ERK-1-specific antibody to evaluate the total level of ERK (containing the nonphosphorylated form). In the untreated retinas, the immunopositive reaction was faint in the cell bodies of the INL (Fig. 2E) . After LPS injection, the ERK-1 reaction was similar to that in control samples at an early time point (Fig. 2F) . The positive staining seen in the INL appeared to become slightly stronger 1 day after injection (Fig. 2G) . This slightly increased staining pattern was maintained 2 weeks after the LPS injection, the latest time point analyzed (Fig. 2H) . Although IHC demonstrated a very weak increase in ERK-1, Western blot analysis failed to show significantly upregulated ERK-1 after LPS injection. 

De Novo Synthesis of ERK

Whereas p-ERK expression substantially increased after LPS injection in the retina, the total amount of ERK proteins was unchanged. We therefore examined whether de novo synthesis of ERK mRNA transcript was upregulated. Figure 3 shows an evaluation of the ERK-1 mRNA level by semiquantitative RT-PCR. We performed PCR with various cycles. The ERK-1 mRNA level was unchanged 1 day after LPS injection in any condition compared with the control, indicating that de novo synthesis did not occur in the retina after LPS stimulation. This finding, together with the results from Western blot and IHC studies, suggests that the transcription and total protein level of ERKs were unchanged, whereas the phosphorylation of ERK was significantly induced after LPS injection. 

Simultaneous Localization of p-ERK and GFAP mRNA

To confirm that the p-ERK-positive cells were Müller cells, we performed simultaneous labeling of p-ERK and GFAP mRNA. GFAP is generally used as a marker for astrocytes and Müller cells in the retina, and its expression is also known to be upregulated after LPS injection. ⁶ p-ERK was visualized immunohistochemically, and GFAP mRNA was detected with nonradioactive ISH. In the non-LPS-treated retina, no p-ERK staining was observed, and GFAP mRNA was found only in the ILM. The GFAP mRNA-positive cells found in the ILM were presumably astrocytes (Fig. 4A) . In the LPS-treated retina, numerous double-labeled cells (both p-ERK- and GFAP mRNA-positive) were observed in the INL 6 hours after injection (Fig. 4B) . Almost all p-ERK cells expressed a GFAP mRNA signal. One day after injection, some p-ERK-positive and GFAP mRNA-negative cells appeared in the INL (Fig. 4C) . The numbers of these single-labeled cells (p-ERK only) increased thereafter, and most of the GFAP mRNA signal disappeared by 1 week after injection. However, p-ERK immunoreactivity was still observed 2 weeks after injection (Fig. 4D) . This double-labeling result confirmed that the p-ERK-positive cells in the INL were Müller cells. 

Dex Suppresses ERK Activity

As mentioned previously, because the data indicated that inflammation induced ERK activity in Müller cells, we then attempted to reduce the inflammatory stimulus by administering Dex 12 hours after LPS injection. The expression level of p-ERK was unchanged until 6 hours after the Dex treatment, and it substantially decreased from 12 hours after treatment with Dex (Fig. 5A) . As an internal control, β-actin remained unchanged at any time point (Fig. 5B) . 

Discussion

The present study revealed inflammation stimulated by LPS-induced ERK activation, predominantly in the Müller cells in the retina, whereas both the protein and transcription levels of ERK were unchanged. Similar activation was not observed in the other MAPK family members, JNK and p-38. The activation of ERK in Müller cells has been demonstrated in experimental models of retinal ischemic injury, ^{23 24} under diabetic conditions, ²⁵ high intraocular pressure, ²⁶ photoreceptor degeneration in the Fischer 344 rat, ²⁷ and bright-light exposure. ^{28 29} Our finding indicate that inflammatory stimulus also elicited ERK activation predominantly in Müller cells. The activation of ERK in Müller cells would thus be a rather general response seen in various types of retinal disease models including EIU. In general, JNK and p-38 are assumed to be implicated in stress responses and cell death, whereas ERK serves as an organizer of various events, such as proliferation, differentiation, and development. ³⁰ In a model of light damage, activation of ERK in Müller cells was suggested to be a crucial response to protect photoreceptor cells. ³¹ As a possible mechanism for this Müller cell-mediated protection, Harada et al. ³² demonstrated that exogenous neurotrophin (NT)-3 increases basic fibroblast growth factor (FGF) production in Müller cells, which can directly prevent photoreceptor apoptosis. ³² In addition to NT-3, brain-derived neurotrophic factor, ciliary neurotrophic factor, and FGF2 activate ERK in Müller cells. ³³ Because ERK is located downstream from those growth factor receptors, Müller cells possess the growth factor receptors to respond to the growth factors. This could be a characteristic of Müller cells in the retinal cell species. In this way, LPS may somehow induce release of growth factor from some cells and thereby lead to activation of ERK in Müller cells. 

Another possible mechanism underlying ERK activation in the inflammatory model is involvement of proinflammatory cytokines such as IL-6, IL-1β, and TNF. The proinflammatory cytokines were induced by the ocular inflammation produced by EIU, ⁷ and these factors are likely to affect the Müller cells directly or indirectly. To determine the implications of the inflammatory responses, we examined whether the p-ERK expression level after LPS injection responds to Dex treatment. In the EIU model, glucocorticoid dramatically inhibited cell infiltration into aqueous humor, ^{34 35} which indicated that ocular inflammatory responses were suppressed. When Dex was administered, the LPS-induced p-ERK level markedly decreased (Fig. 5) . Although we could not examine whether the decrease in the p-ERK level resulted from the weakened inflammatory responses or direct inhibition by Dex, it is likely that cytokine production and/or other inflammatory stimuli were suppressed by Dex and thereby caused reduced p-ERK expression. The implication of growth factors in this process is also unclear. However, because there are some intracellular signaling pathways from the cytokine receptors to the ERK activation, this pathway may be involved in the LPS model. Although further investigation is needed to clarify this mechanism, ERK activation in Müller cells may play a critical role in maintaining retinal order and protecting photoreceptors from LPS damage. 

Another intriguing finding in this study is that a significant amount of p-ERK was located in the processes of the Müller cells, as well as in the cell bodies, including the nucleus. Further, the expression profiles seen in these two regions appeared distinct. The expression of p-ERK in the soma, including the nucleus, persisted for a relatively long period after LPS injection, whereas the expression in the processes was observed for a short time in the early phase of the response. This distinct localization of p-ERK seen in the processes and soma may suggest a differential function of p-ERK in response to inflammatory stimulus. Generally, it is well known that activated ERK is translocated into the nucleus and plays a role in transcription. Thus, the p-ERK observed in the nucleus by LPS stimuli (Fig. 2) may be associated with ordinary transcription regulation, whereas the p-ERK in the processes may have another function besides transcription regulation. The functional significance of p-ERK localization in the processes of the Müller cells in response to LPS should be studied further. 

In conclusion, LPS activates ERK predominantly in the Müller cells, probably in an inflammatory-response-mediated manner. Therefore, ERK activation is almost entirely suppressed by treatment with Dex. Although the reason for the preponderance of ERK activation in Müller cells among the retinal neural cells is unclear, Müller cells may be primary cells that can sense inflammation and retinal damage and elicit various responses through ERK to protect from the retinal cells inflammation and repair them after injury. 

 Submitted for publication July 24, 2001; revised November 13, 2001; accepted November 21, 2001.

 Commercial relationships policy: N.

 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.

 

Figure 1.

 

Western blot analysis. The expression level of p-ERK was stronger at 1 day after LPS injection than the control and remained higher (A). There was no significant difference in ERK-1 (B) and -2 (C) expression between control and LPS stimuli throughout the study. (D) Expression of β-actin as an internal control. Results represent four independent experiments. Time points are shown above the lane. c, control; h, hours; d, days; w, weeks.

Figure 2.

 

The IHC expression profiles of p-ERK (A, B, C, D, I) and ERK-1 (E, F, G, H) are shown at each time point after LPS injection [(A, E) control retinas; (B, F) 6 hours; (C, G) 1 day; (D, H) 2 weeks]. (I) Higher magnification of p-ERK immunoreactivity 3 days after injection. p-ERK immunoreactivity was detected in processes extending from the ILM to the OLM (I). GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer. Original magnification, (A–H) ×200; (I) ×400.

Figure 3.

 

Semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) for ERK-1 mRNA. The mRNAs from the retinas treated with either LPS (+) or vehicle (−) 3 days before they were used for the RT-PCR study. The reactions were performed with various numbers of PCR cycles (23, 25, 27, and 29 cycles). For the internal control, GAPDH was used with 23 cycles of PCR. Results represent three independent experiments.

Figure 4.

 

The simultaneous demonstration of p-ERK immunoreactivity and GFAP mRNA. The expression profiles are shown in the control retina (A) and at 6 hours (B), 1 day (C), and 2 weeks (D) after LPS injection. Brown staining indicates p-ERK immunoreactivity by immunohistochemistry, and blue staining demonstrates GFAP mRNA by ISH. (A) In the control retina, p-ERK was not detected, and GFAP mRNA-positive cells were observed only in the nerve fiber layer where retinal astrocytes are present. (B) Six hours after LPS injection, there were numerous dark purple cells in the INL, indicating simultaneous staining of p-ERK immunoreactivity and a GFAP mRNA signal. These results confirm that the p-ERK-positive cells in the INL were Müller cells. (C) On day 1, only p-ERK-positive cells appeared in the INL (arrows), indicating that the GFAP mRNA signal was decreasing, and thereby the proportion of double-labeled Müller cells was lower. (D) At 2 weeks, p-ERK immunoreactivity was still positive, whereas the GFAP mRNA signal had disappeared. Abbreviations as in Figure 2 . Original magnification, ×200.

Figure 5.

 

Western blot analysis of p-ERK expression after Dex treatment. Dex was injected 12 hours after LPS. (A) The expression level of p-ERK was unchanged until 6 hours after Dex injection. However, it dramatically decreased to the untreated level at 12 hours after injection. β-actin was used as the internal control (B). Results represent four independent experiments.