The Presence of Macrophage Migration Inhibitory Factor in Human Trabecular Meshwork and its Upregulatory Effects on the T Helper 1 Cytokine

Abstract

purpose. To investigate the expression and secretion of macrophage migration inhibitory factor (MIF) in human trabecular meshwork (HTM) and evaluate its role in ocular inflammation.

methods. Tissue samples of HTM cells were isolated from donor human eyes or corneoscleral buttons, and the HTM cells were cultured. The expression of MIF on HTM cells was evaluated by RT-PCR, Western blot analysis, and ELISA. T-cell clones (TCCs) were established from ocular infiltrating cells of patients with uveitis. ELISA was used to evaluate the pathologic role of MIF, in relation to regulatory effects on cytokine production by T cells.

results. MIF was detected in the HTM by RT-PCR and Western blot analysis. MIF was also shown by ELISA to be secreted by the HTM cells in culture. The HTM supernatant enhanced IFN-γ production by TCCs, but not IL-10; and these effects were neutralized by anti-MIF antibodies. Similarly, recombinant MIF enhanced the IFN-γ production by the TCCs.

conclusions. MIF is expressed and secreted in the HTM, and MIF has the capacity to enhance T helper 1 cytokines and may play a role as an inflammatory cytokine in the eye.

Macrophage migration inhibitory factor (MIF) was originally described as a T-lymphocyte–derived factor that inhibits macrophage migration in capillary tubes. ¹ Subsequently, MIF has been shown to be a proinflammatory pituitary and macrophage cytokine and a critical mediator of septic shock. ² Furthermore, MIF is capable of overcoming the glucocorticoid-mediated inhibition of cytokines, such as interleukin (IL)-1β, IL-6, IL-8, tumor necrosis factor (TNF), and interferon (IFN)-γ. ³ Bacher et al. ⁴ found that anti-MIF antibodies inhibit T-cell proliferation and IL-2 production in vitro. Thus, MIF appears to have an important regulatory role in the activation of T cells. 

The eye has been known as an immune-privileged organ, in that immune response is generally limited. The exact mechanisms by which this is accomplished is not fully understood. Inflammatory responses in the eye, such as uveitis can severely impair intraocular tissues and can lead to loss of vision. MIF is constitutively expressed in the epithelium and endothelium of the cornea in humans ⁵ and in the epithelium of the iris-ciliary body and retina (astrocytes and Müller cells) in rats, ^{6 7} indicating the importance of this protein. The function of MIF in the eye has not been well studied, but Apte et al. ⁸ demonstrated that MIF shares more than 90% homology with a factor in the aqueous humor that inhibits natural killer (NK) cell–mediated lysis of corneal endothelial cells. The NK cell inhibitory effects of the aqueous humor can be neutralized by anti-MIF antibody. Moreover, recombinant (r)MIF, at high doses (1000–10,000 ng/mL), similarly inhibits NK cell activity. ⁸ These findings suggest that MIF is capable of inhibiting at least one immune effector element, the NK cell, and thereby contributes to immune privilege in the eye. Trabecular meshwork endothelial cells also have immune-like functions. They are actively phagocytic and often migrate to clear debris from the trabecular meshwork. ^{9 10 11} It has been hypothesized that this macrophage-like activity is very important in the normal functioning of the trabecular meshwork. ^{9 10 11 12 13} In this study, we demonstrated the expression of MIF by the human trabecular meshwork (HTM) cells and evaluated a possible role for MIF in ocular inflammation. 

Materials and Methods

This study was performed in accordance with the Declaration of Helsinki. Fourteen pairs of donor human eyes or corneoscleral buttons were used. Whole eyes from donors without known ocular disease were obtained from the National Disease Research Interchange (Philadelphia, PA) within 48 hours of death. Normal donor anterior segments containing the HTM were provided by the Lions Eye Bank of Delaware Valley (Philadelphia, PA). These segments had been maintained in preservative (Optisol-GS; Chiron Vision, Irvine, CA) medium. Tissue was collected from patients after informed consent had been obtained. 

Supernatant of HTM

The tissue samples of HTM were obtained from two normal donors (ages, 16 and 39 years). The dissection and explant techniques were similar to those previously described. ¹⁴ The globes were bifurcated at the equator with a scalpel blade. The ciliary body, iris, and lens were gently removed from the anterior segment. The HTM was isolated from the surrounding tissues with incisions anterior and posterior to the meshwork with a microscalpel, grasped with forceps, and removed. The HTM cells that grew out from the tissue were cultured in Dulbecco’s modified Eagle’s medium with 20% heat-inactivated fetal bovine serum (FBS), 2 mM l-glutamine, and 0.1 mg/mL gentamicin (Gibco-BRL Life Technologies, Gaithersburg, MD). The cells were cultured at 37°C in a 10% CO₂ atmosphere. Cell viability was checked with trypan blue staining. Cultured medium from the HTM cells before passage 7 was harvested every 3 to 4 days, and stored at −80°C until use. Approximately 1 mL culture medium was harvested for each 10⁶ cells. Culture medium that had not been incubated with HTM cells served as the control. 

Reverse Transcription–Polymerase Chain Reaction

Pairs of trabecular meshwork were dissected from five separate donors (one each aged 50, 63, and 87 and two aged 59 years). RNA was isolated with an extraction kit (RNeasy Mini Kit; Qiagen, Inc., Valencia, CA) in accordance with the manufacturer’s protocol. Reverse transcription–polymerase chain reaction (RT-PCR) was performed with a commercial system (SuperScript One-Step RT-PCR; Gibco-BRL Life Technologies) used in accordance with the manufacturer’s protocol. The sequence of the MIF-specific primers for the forward direction was 5′-TGCAGCCTGCACAGCATC-3′ and the reverse direction was 5′-ATTGGCCGCGTTCATGTC-3′. The primers were located in the second and third exons separated by a 96-bp intron to create a 200-bp product if the amplified product was from mRNA and a 296-bp product if the amplified product was from the DNA. Two tubes were run in parallel, except that one of the tubes was preheated at 94°C for 5 to 10 minutes to inactivate the reverse transcriptase enzyme sufficiently. RNA was obtained from primary cultured HTM endothelial cells from three donors of age 16, 35, and 39 years. RNA isolation and RT-PCR were performed in the same manner as described for the HTM tissue. PCR products were purified using a PCR purification kit (QIAquick; Qiagen, Inc.), in accordance with the manufacturer’s protocol. The DNA sequencing reaction was performed with a DNA sequencing kit (Big Dye Terminator Cycle Sequencing Ready Reaction; PE-Applied Biosystems, Foster City, CA). The reaction product was then sequenced (Prism 310; PE-Applied Biosystems). Results from the sequence were analyzed by a BLAST (provided in the public domain by theNational Center for Biotechnology Information, Bethesda, MD, and available at http://www.ncbi.nlm.nih.gov) search. 

Western Blot Analysis

HTMs were dissected from anterior segments of eyes of three donors (ages 19, 68, and 79 years). The tissue was homogenized in 30 microliters of buffer (M-PER; Pierce Chemical Co., Rockford, IL) and then 5 μL of the soluble fraction was placed in sample buffer and reducing agent (NuPage LDS; Invitrogen Corp., Carlsbad, CA) according to the manufacturer’s protocol. Homogenates (5 μg) from cultured HTM cells from donors aged 30 and 76 years and 1 μg rMIF were also placed in the sample buffer and reducing agent. The sample was run on 12% gels with a running buffer (NuPage Bis-Tris; Invitrogen Corp.). Western blot analysis and staining were performed in accordance with the manufacturer’s protocol. The blot was incubated in anti-MIF antibody (1:200) overnight. The blots were washed and incubated with goat peroxidase–labeled anti-mouse antibody (Kirkegaard and Perry Laboratories, Gaithersburg, MD). MIF was detected using a chemiluminescence kit (Renaissance NEN Life Sciences; Boston, MA). 

T-Cell Clones

T-cell clones (TCCs) were established from ocular infiltrating cells in four patients: one each with human T-cell lymphotropic virus (HTLV)-1 uveitis (age, 26 years), sarcoidosis (age, 24 years), Vogt-Koyanagi-Harada (VKH) disease (age, 75 years), and acute retinal necrosis syndrome (ARNS; age, 16). The method of establishing TCCs is described elsewhere. ¹⁵ Briefly, cells from ocular fluids were plated at a concentration of one cell/three wells in a 96-well U-bottomed tissue culture plate (Falcon, Lincoln Park, NJ) in RPMI-1640 medium (GibcoBRL, Grand Island, NY) containing 10% FBS, 10 μg/mL phytohemagglutinin-P (PHA-P; Difco Laboratories, Detroit, MI), 100 U/mL recombinant interleukin (rIL)-2, and antibiotics in the presence of x-ray–irradiated (50-Gy) allogeneic peripheral blood mononuclear cells obtained from healthy volunteers as feeder cells. 

Cytokine Assay

Cytokines in the supernatant of the HTM culture were measured by enzyme-linked immunosorbent assay (ELISA). Cytokines measured in the experiment were MIF, IL-2, IL-4, IL-6, IL-10, TNF-α, IFN-γ, and transforming growth factor (TGF)-β2; also measured were macrophage inflammatory protein (MIP)-1α, regulated on activation normal T-cell expressed and secreted (RANTES) protein, soluble (s)Fas, and sFas ligand (sFasL). ELISA was performed with commercial kits: IL-2, IL-4, IL-6, IL-10, TNF-α, IFN-γ, TGF-β2, MIP-1α, and RANTES from R&D Systems (Minneapolis, MN) and sFas from MBL (Nagoya, Japan). MIF was measured by a sandwich ELISA with an anti-human MIF antibody described elsewhere. ¹⁶ The amount of sFasL was determined by a sandwich ELISA with anti-human FasL monoclonal antibodies, NOK-1 (mouse IgG1, κ) and NOK-3 (mouse IgM, κ). The lowest detection limit of each molecule by ELISA assay was as follows: MIF, 1000 pg/mL; IL-2, 7 pg/mL; IL-4, 3 pg/mL; IL-6, 0.7 pg/mL; IL-10, 3.9 pg/mL; TNF-α, 4.4 pg/mL; IFN-γ, 8 pg/mL; TGF-β2, 7 pg/mL; MIP-1α, 10 pg/mL; RANTES, 8 pg/mL; sFas, 500 pg/mL; and sFasL, 100 pg/mL. 

The effects of HTM supernatant or human rMIF on cytokine production by TCCs from the patients with uveitis were also evaluated. The TCCs were washed twice with phosphate-buffered saline (PBS; GibcoBRL) and resuspended at a concentration of 1 × 10⁵ cells/mL in RPMI-1640 with 10% FBS in flat-bottomed 96-well tissue culture plates (Falcon), with HTM supernatant. TCCs stimulated by 5 μg/mL anti-CD3 mAb (clone UCHT1; BD PharMingen, San Diego, CA) and 2 μg/mL anti-CD28 mAb (clone CD28.2; BD PharMingen) were also cultured in the presence of rMIF. The concentrations of IFN-γ and IL-10 in the supernatant of the culture were measured by ELISA. 

Cell Proliferation Assay

Proliferation of TCCs was measured by [³H]-thymidine uptake. TCCs were cultured with HTM supernatant for 40 hours and pulsed with 1 μCi [³H]-thymidine for 16 hours. The incorporated [³H]-thymidine was measured by liquid scintillation counter. 

Neutralization of HTM Activity by Anti-MIF Antibodies

The supernatant of HTM was cocultured with mouse anti-MIF monoclonal antibodies (clone 3H2F or P3 3E12H), rabbit anti-MIF polyclonal antibodies, or mouse IgG1, on ice for 1 hour with shaking. ¹⁷ The supernatant of HTM after coculturing with anti-MIF antibodies was used to assay its effects on cytokine production by TCCs. 

Statistical Analysis

Statistical analysis was performed with Student’s t-test. The difference between the two groups compared was determined to be significant when at P < 0.05. 

Results

MIF mRNA and DNA in HTM Cells

Reverse transcription followed by PCR with MIF-specific primers on the RNA isolated from two HTM endothelial cell primary cultures and the trabecular meshwork from all five donors showed the presence of a band approximately 200 bp in length, as predicted by the primers selected. Representative samples from one primary HTM cell culture and one donor HTM are shown in Figure 1 . As expected, preheating of the reverse transcriptase prevented the PCR reaction from occurring. Two samples, one of the cultured HTM cells and one of the donor HTMs, were sequenced, and a BLAST search confirmed that the cDNA was consistent with the mRNA of MIF with EXPECT values of 7e⁻⁷⁰ and 2e⁻¹⁸, where the EXPECT value is the statistical probability that the match occurs by chance alone. 

MIF Protein in HTM Cells

Immunoblotting with antibody to MIF of the proteins isolated from the HTM tissues and cultured HTM cells showed a band at approximately 15 kDa (Fig. 2) that migrated at the same position as the rMIF. 

Cytokines in the HTM Supernatant

To investigate whether HTM secreted MIF into the culture medium, MIF and other cytokines in the supernatant of primary HTM culture were measured by ELISA. The supernatant of HTM cells contained a large amount of MIF (15.8 ± 1.1 ng/mL) and other cytokines, such as IL-6 and TGF-β2 (4.2 ± 2.5 and 3.0 ± 1.1 ng/mL, respectively; Table 1 ). However, IFN-γ, IL-2, IL-4, IL-10, MIP-1α, RANTES, sFas, sFasL, and TNF-α were not detected in the HTM supernatant (Table 1) . MIF and the cytokines were not detectable in control media. 

Effects of HTM Supernatant on Cytokine Production by TCCs

To study the immunomodulating activity of the HTM supernatant, the effects of the supernatant on cytokine (IFN-γ and IL-10) production by TCCs established from cells of the patient with HTLV-1 uveitis were analyzed. The HTM supernatant enhanced IFN-γ production by TCCs in a dose-dependent manner, but not IL-10 production (Fig. 3) . The [³H]-thymidine incorporation assay showed that the HTM supernatant did not cause cell proliferation of TCCs (Fig. 4) . The effects of the HTM supernatant on IFN-γ production by TCCs were neutralized by coculturing the supernatant with the mouse anti-MIF monoclonal antibodies or the mouse anti-MIF polyclonal antibodies (Fig. 5) . 

Effects of rMIF on IFN-γ Production by TCCs

The effects of rMIF on IFN-γ production by TCCs were evaluated with TCCs established from cells of patients with various types of uveitis (HTLV-1, sarcoidosis, VKH disease, and ARNS). All tested TCCs stimulated by anti-CD3 and anti-CD28 antibodies produced a great amount of IFN-γ. In addition, similar to the results with the HTM supernatant, human rMIF enhanced IFN-γ production by all tested antigen-stimulated TCCs in a dose-dependent manner (Fig. 6) . 

Discussion

This is the first report demonstrating expression and secretion of MIF by the human trabecular meshwork. In our study, MIF increased expression of IFN-γ and thus enhanced the T helper (Th)1 cytokines. The eye is a classic example of an immune-privileged site. It is characterized by the absence of draining lymph nodes, the presence of an anatomic barrier to the peripheral circulation, and the presence of immunosuppressive molecules. The immunosuppressive molecules in the aqueous humor of the eye include TGF-β, ^{18 19} vasoactive intestinal peptides, ²⁰ α-melanocyte–stimulating hormone, ²¹ and sFasL. ^{22 23} Our previous study ²⁴ demonstrated that the aqueous humor under normal conditions is capable of suppressing cytokine production by activated T cells of patients with uveitis, and sFasL and TGF-β were responsible for parts of this immunosuppressive activity of the aqueous humor. Our first hypothesis about the function of MIF in the aqueous humor was that it also maintains the immune privilege of the eye, similar to TGF-β and other immunosuppressive molecules. ^{18 19 20 21 22 23} MIF has been thought to be one of the immunosuppressive factors in the aqueous humor, because rMIF at high dosages (1000–10,000 ng/mL) suppresses the NK cell–mediated lysis of mouse corneal endothelial cells. ⁸ However, the results presented herein appear to be the opposite of what was expected. 

In the supernatant of HTM cells from normal donors, a high concentration of MIF (15.8 ± 1.1 ng/mL) was detected in an ELISA. This was similar to levels detected in the aqueous humor of patients with uveitis (15.8 vs. 1–38 ng/mL). ²⁵ The HTM supernatant was shown to enhance IFN-γ production, but not that of IL-10, in TCCs from cells of an HTLV-1 patient. This effect was not attributed to the cell proliferation of TCCs, because HTM supernatant did not induce TCC proliferation. The effect on IFN-γ production was neutralized by mouse anti-MIF antibodies, indicating that MIF has the function of enhancing the production of the Th1 cytokine but not the Th2 cytokine from T cells. As expected, rMIF enhanced IFN-γ production in TCCs from a variety of uveitis conditions, such as HTLV-1 uveitis, sarcoidosis, VKH disease, and ARNS. These data indicate that MIF enhances Th1 cytokines and may play a role as an inflammatory cytokine. However, it appears that this notion is limited to the disease state of patients with pathophysiological concentrations in the eye, because our experiments used TCCs of cells from patients with uveitis, and the concentrations used in the experiments were similar to the MIF concentrations detected in the aqueous humor of patients with uveitis. ²⁵  

The notion that MIF may play a role as an inflammatory cytokine is further supported by previous findings in animals and humans. The induction of experimental autoimmune uveitis in rats was significantly suppressed by treatment with monoclonal antibodies to MIF. ²⁶ In patients with uveitis the serum level of MIF was significantly higher than that in healthy donors, and the serum MIF level was in accord with disease activity. ¹⁶ A significant amount of MIF was detected in the ocular fluid of patients with uveitis, and it increased in parallel with the intensity of ocular inflammation. ²⁵ The present study and the previous findings indicate that MIF at high concentrations plays a role as an inflammatory cytokine by enhancing production of the Th1 cytokine and may compound problems within the eye once a disease process starts. However, it is still possible that MIF is normally produced in vivo in nondisease states at a very low concentration that may not promote T-cell IFN-γ production, but may protect the HTM cells from NK cell attack. The present study was unable to show MIF activity at such low concentrations, and so this remains for future studies to determine. 

The physiological function of MIF in HTM is unclear. Perhaps low but significant levels of MIF are needed regionally in the outflow pathway to be immunosuppressive and to maintain the immune privilege of the normal eye. There may be other physiological functions of this molecule in the HTM. The trabecular meshwork is the site of the highest resistance to aqueous humor outflow and is thought to be critical to the regulation of intraocular pressure. ^{27 28 29 30 31 32} One of the important components in the resistance to outflow through the TM is the extracellular matrix (ECM). Outflow facility studies with anterior segment perfusion systems with glycosaminoglycan degradation enzymes ^{33 34 35 36} and matrix metalloproteinases (MMPs) ³⁷ strongly indicate that ECM degradation and synthesis (i.e., turnover) are important in the regulation of intraocular pressure. In synovial fibroblasts, MIF induces an upregulation of MMP-1, MMP-3, and TIMP-1 through an AP-1–, tyrosine kinase–, and protein kinase c– dependent pathway, in a dose-dependent fashion. ³⁸ In addition, MIF has been shown to interact directly with and inhibit the transcriptional activator Jab1, ¹⁷ an activator of c-Jun amino terminal kinase activity and an enhancer of endogenous phospho-c-Jun levels. 

Previous studies have demonstrated that MIF is secreted by the lens ³⁹ and cornea of the eye, ⁵ and with this finding in the trabecular meshwork, the importance of this molecule in the eye can be appreciated. The present study demonstrated that MIF, expressed and secreted by HTM cells, enhanced Th1 cytokine produced by activated T cells of patients with uveitis. Thus, MIF could play a role as an inflammatory cytokine in intraocular inflammation. 

 Supported in part by Grants-in-Aid for Scientific Research 12771013 and 13771015 from the Ministry of Education, Science, Sports, and Culture of Japan.

 Submitted for publication November 6, 2001; revised March 18, 2002; accepted March 29, 2002.

 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.

 

Ethidium bromide-stained agarose gel containing the products of RT-PCR reaction with MIF-specific primers. Lane S: ΦX174 DNA/Hae III fragment standard. Lanes 1 and 2: RT-PCR reaction products of mRNA from a primary HTM endothelial cell culture from a 35-year-old donor. Lane 1: reverse transcriptase preheated before RT-PCR. Lane 2: ∼200-bp band of RNA. Lanes 3 (preheated) and 4: RT-PCR reaction products using pooled mRNA from the HTMs of both eyes of an 87-year-old donor.

Figure 2.

 

(A) SDS-PAGE of samples from cultures of HTM cells of (lane 1) a 30- and (lane 2) a 76-year-old donor; (lane 3) rMIF, and homogenates of HTM tissue from donors aged (lane 4) 19, (lane 5) 68, and (lane 6) 79 years. (B) Western blot from the gel in (A). All lanes show chemiluminescent antibody labeling of approximately 15 kDa in mass.

Table 1.

 

Cytokine Concentrations in HTM Supernatant

Cytokine	Concentration (ng/mL)
MIF	15.8 ± 1.1
IL-6	4.2 ± 2.5
TGF-β2	3.0 ± 1.1
IFN-γ	Below detectable level
IL-2	Below detectable level
IL-4	Below detectable level
IL-10	Below detectable level
MIP-1α	Below detectable level
RANTES	Below detectable level
sFas	Below detectable level
sFasL	Below detectable level
TNF-α	Below detectable level

 HTM supernatant contained large concentrations of IL-6, TGF-β2, and MIF, suggesting that HTM produces these cytokines. The concentrations of these cytokines were measured using ELISA, except for MIF, which was measured at Hokkaido University by sandwith ELISA, with anti-human MIF antibody. Data are the mean ± SD.

Figure 3.

 

(A) Effects of HTM supernatant at various concentrations on IFN-γ production by TCCs established from ocular infiltrating cells of a 48-year-old patient with HTLV-1 uveitis. *P < 0.05, **P < 0.005. (B) Effects of various concentrations of HTM supernatant on IL-10 production by TCCs. Error bar, SD.

Figure 4.

 

Proliferation of TCCs in the presence of various concentrations of HTM supernatant. Proliferation of TCCs was measured by [³H]-thymidine uptake. Error bar, SD.

Figure 5.

 

Effects of HTM supernatant cocultured with anti-human MIF monoclonal (clone 3H2F and P3 3E12H) and polyclonal antibodies on IFN-γ production by TCCs. *P < 0.05. Error bar, SD.

Figure 6.

 

Effects of various concentrations of human rMIF on IFN-γ production by TCCs established from cells of patients with (A) HTLV-1 uveitis, (B) sarcoidosis, (C) Vogt-Koyanagi-Harada disease, or (D) acute retinal necrosis syndrome. *P < 0.05, **P < 0.005 versus medium only. Error bar, SD.

The authors thank David L. Epstein, Duke University, for assistance in initiating these studies and Yuka Mizue, Sapporo Immunodiagnostic Laboratory, for preparation of the anti-MIF monoclonal antibodies.