Expression of Alternatively Spliced Growth Factor Receptor ... - IOVS

1 downloads 0 Views 2MB Size Report
monkey. Invest Ophthalmol Vis Sci. 1981;2():644-65I. Expression of Alternatively. Spliced Growth Factor Receptor. Isoforms in the Human. Trabecular Meshwork.
242

Reports

References 1. Volberg T. Gciger B, Citi S. Bcrshadsky AD. Effect of protein kinase inhibitor 11-7 on the comraciility, integrity, and membrane anchorage of the microfilamcnt system. Cell Motil Cytoskel. 1994:29:321-338. 2. Tian B. Kaufman PL. Volberg T. Gabelt BT. Cieiger B. H-7 disrupts the actin cytoskelcton and increases outflow facility. Arch Ophthalmol. 1998:116:633-643. 3. Kaufman PL. Aqueous humor dynamics. In: ID Duane. ed. Clinical Ophthalmology. Vol. 3. Philadelphia: Harper and Row; 1985:1-24. 4. Bill A. Conventional and uveoseler.il drainage of aqueous humor in the cynomolgus monkey (Macaca irus) at normal and high intraocular pressures. Exp Eye Res. 1966:5:45-54. 5. Gabelt BT, Kaufman PL. The effect of prostaglandin ¥2 on trabecular outflow facility in cynomolgus monkeys. Ex/> Eye Res. 1990: 51:87-91. 6. Kaufman PL. Bill. A, Barany 1:11. Effect of cytochalasin B on conventional drainage of aqueous humor in the cynomolgus monkey. Exp Eye Res. 1977:25(suppl):411-414.

Expression of Alternatively Spliced Growth Factor Receptor Isoforms in the Human Trabecular Meshwork RobertJ. Wordinger,x Abbot F. Clark,12 Rajnee Agarwal,' Wencli Lambert,' and Steven E. Wilson* PURIHXSI-. Growth factors act through high-affinity cell sur-

face receptors expressed by target cells and are critical modulators of cell function. Because aqueous humor is known to contain growth factors, these molecules may play a key role in maintaining the normal function of the human trabecular meshwork (HTM). Alternate mRNA splicing is an important mechanism used by cells to generate diverse isoforms of growth factor receptors. Although previous investigators have suggested that HTM cells may express alternative isoforms of several growth factor receptors, there have been no studies to verify these preliminary findings. The objective of this study was to determine whether cultured and ex vivo HTM cells express alternate isoforms of hepaiocyte, kcratinocytc, and transforming growth factor beta (TGF/3)-II receptors and to characterize the isoform molecular sequences.

IOVS, January 1999, Vol. 40, No. 1 7. Kaufman PL, Barany EH. Loss of acme pilocarpine effect on outflow facility following surgical disinsertion and retrodisplacement of the ciliary muscle from the scleral spur in the cynomolgus monkey. Invest Ophthalmol. 1976; 15:793- 807. 8. Perkins TW, Alvarado JA, PolanskyJR, Stilwell L, Maglio M, Luster R. Trabecular meshwork cells grown on fillers: conductivity and cytochalasin effects. Invest Ophthalmol Vis Sci. I988;29:I836-I846. 9. Kaufman PL. Aqueous humor dynamics following total iridectomy in the cynomolgus monkey. Invest Ophthalmol Vis Sci. 1979; 18:870-874. 10. Kaufman PL, Barany HI I. Cytoehalasin B reversibly increases outflow facility in the eye of the cynomolgus monkey. Invest Ophthalmol Vis Sci. 1977; 16:47-53. 11. Kaufman PL, Rent/hog L. Effect of total iridectomy on outflow facility responses to adrencrgic drugs in cynomolgus monkeys. Exp Eye Res. 1981:33:65-74. 12. Kaufman PL, Barany EH. Adrenergic dnig effects on aqueous outflow facility following ciliary muscle retrodisplacement in the cynomolgus monkey. Invest Ophthalmol Vis Sci. 1981;2():644-65I.

METHODS. To determine whether cells within the HTM

express mRNA for alternate isoforms of growth factor receptors, total RNA was isolated from several well-characterized HTM cell lines that were established from donors of various ages and from fresh ex vivo HTM tissues from healthy donors. After cDNA synthesis, polymeni.se chain reaction was initiated using specific primers for alternate forms of the following receptors: hepatocyte growth factor (HGFR), keratinocyte growth factor (KGFR), and transforming growth factor beta receptor II (TGF/3R-H). Specificity and characterization of the polymerase chain reaction amplification products were determined by nucleic acid sequencing. RESULTS. Amplification products of the expected size for

the growth factor isoforms were expressed in cell lines and in ex vivo tissues. Nucleic acid sequencing showed that cultured HTM cells and fresh ex vivo trabecular meshwork tissues expressed specific mRNA for alternatively spliced isoforms of HGFR, KGFR, and TGF/3IMI The HGFR alternate isoform contained a 96-bp insert in the C-terminal coding region of the cytoplasmic tyrosine kinase domain. The KGFR alternate isoform is a soluble, truncated form, because it has no transmembrane or cytoplasmic domain as does the normal membrane-associated form. The TGFjSR-lI alternate isoform contained a 75-bp insert in the N-terminal coding region of the extracellular domain. CONCLUSIONS. In vitro and ex vivo HIM cells express

horn the 'Department of Anatomy and Cell Biology and The North Texas Eye Research Institute. University of North Texas Health Science Center at Tort Worth: ^Glaucoma Research. Alcon Laboratories. Eort Worth. Texas; and ^Department of Ophthalmology, The Cleveland Clinic Foundation, Ohio. Supported in part by funds provided by the National Glaucoma Research Program of the American Health Assistance Foundation, Rockville, Mankind: The Glaucoma Research Foundation. San Francisco, California; Grant EY10056 (SEW) from the National Eye Institute. Bethesda. Maryland: and Alcon Laboratories, Eon Worth, Texas. Submitted for publication March 30. 1998; revised July 29, 1998; accepted August 19. 1998. Proprietary interest category: E. C5. Reprint requests: Robert J. Wordinger, Department of Anatomy and Cell Biology, University of North "Texas Health Science Center at Fort Worth. 3500 Camp Bowie Boulevard. Eort Worth, TX 76107-2699.

mRNA for alternatively spliced isoforms of HGFR, KGFR, and TGFJ3R-II. These alternatively spliced receptor isoforms may be functional within the HTM and may play a critical role in maintaining the normal microenvironment of this important tissue. {Invest Ophtbahnol Vis Sci. 1999; 40:242-247) efects in the structure, function, or number of human trabecular meshwork (HTM) cells may influence the pathogenesis of primary open-angle glaucoma. Thus, an understanding of the factors that control the normal microenvironment within the HTM is critical to our understanding of the disease. Polypeptide growth factors are important in the nor-

D

Reports

10VS, January 1999, Vol. 40, No. 1 mal physiology of cell populations. Growth factors are known to control a diverse number of cellular functions, including proliferation, motility, differentiation, phagocytosis, and extracellular matrix synthesis and degradation, and, therefore, may play key roles in the pathophysiology of the trabecular meshwork. The potential for growth factors to participate in controlling the microenvironment within the HTM is significant because aqueous humor, which bathes the trabecular meshwork, is known to contain growth factors,1 and cultured HTM cells express mRNA for several growth factors.2 Therefore, the potential for paracrine and autocrine signaling exists within the trabecular meshwork. Growth factors mediate their actions by binding to and then activating specific cell surface receptors. We have shown the mRNA expression of several growth factor receptors by cells within the HTM and additionally have shown that many of these receptors were functional.5 Alternate mltNA splicing is an important mechanism used by cells to generate diverse isoforms of protein molecules/' The use of mRNA splicing allows a single gene to generate more than one mRNA and therefore more than one protein. Increasing numbers of mammalian genes have been reported to use alternative mRNA splicing as a means of increasing molecular diversity/' Alternatively spliced isoforms of hepatocyte growth factor receptor (HGFR), keratinocyte growth factor receptor (KGFR), mid transforming growth factor beta receptor II (TGFjSR-11) have been reported, and we hypothesized that these isoforms may also be expressed in the HTM. The objectives of this study were to determine whether cultured HTM cells and dissected HTM tissues express alternative isoforms for these growth factor receptors and to determine the molecular sequence of these isoforms.

METHODS

Trabecular Meshwork Dissection Human donor eyes were obtained from regional eye banks within 24 hours of death. The eyes were equatorially bisected, and the lens, iris, and ciliary body were removed from the anterior segment. The trabecular meshwork from each eye was obtained by making parallel cuts anterior to the scleral spur and posterior to Schwalbe's line with the aid of a surgical microscope. Normal HTM samples were obtained from 75-, 79-, and 80-year old donors. Total RNA was obtained from the trabecular meshwork samples from each pair of donated eyes using a kit (Micro RNA Isolation Kit: Stratagene. La Jolla, CA). Subsequently, first-strand cDNA was synthesized as described below.

Trabecular Meshwork Cell Culture Early passaged, previously characterized, normal HTM cell lines from donors aged 6 days, 48 days, 6 months, 2 years, 18 years, 54 years, and 80 years were used. The cultured cells used in this stud)' displayed the same characteristics as HTM cells reported previously/' 5 6 The HTM cells appeared broad and flat, were partially elongated, and formed a monolayer when confluent. These cells were grown until confluent in Ham's F-IO medium (JRH Biosciences, Lenexa, KS) supplemented with 10% fetal bovine serum (HyClone Laboratories, Logan, UT), 2 mM i.-glutamine (0.292 mg/ml), 100 units/ml penicillin, and 0.1 mg/ml streptomycin (Life Technologies, Grand Island, NY). Cells were incubated at 37°C in 7% CO 2 and 93% air. The

243

medium was changed every 3 days. HTM cell lines were propagated as described by Steely et al.r> Briefly, number 3 beads (Cytodex; Sigma, St. Louis, MO) in a 2% suspension in sterile phosphate-buffered saline (PBS) were added to confluent monolayers. Seven days later, the monolayer was gently washed in a stream of culture medium, and dislodged cellcovered beads were transferred to new plates. The cells that remained on the original plate were allowed to regrow into an additional monolayer.

Total Cellular RNA Extraction and cDNA Synthesis Total cellular RNA from cultured HTM cells was prepared using a commercial kit (RNAzol B; Biotex, Houston, IX). After ethanol precipitation, the RNA was resuspended in 20 /xl water and stored at —80°C. First-strand cDNA synthesis was prepared from total cellular RNA. Initially, to reduce secondary structure, 20 jug RNA and 0.75 jag random primers (Promega, Madison, Wl) were combined and incubated at 85°C for 3 minutes. For each trabecular meshwork tissue sample, the entire 20 JU.1 of total RNA was used for cDNA synthesis. The following were then added to the reaction tube: 80 units RNasin (Promega), 40 units avian myeloblastosis virus (AMV) reverse transcriptase (Promega), 0.625 mM each deoxyribonucleotide, 50 mM Tris-HCI, 75 mM potassium chloride, 10 mM clithiothreitol, and 3 mM magnesium chloride. The reaction tube was incubated at 42°C for 30 minutes followed by an incubation at 94°C for 2 minutes. The cDNA was stored at -20°C until used for polymerase chain reaction (PCR).

Primer Design A software program (Oligos 4.0; National Biosciences, Plymouth, MN) was used to design PCR primers that had optimal annealing temperatures and that would amplify at similar temperatures and magnesium concentrations. All designed primers were submitted through BLAST (National Center for Biotechnology Information, Bethesda, MD) to ensure that they would not hybridize to any other known nucleic acid sequences under the conditions used. All primer pairs were designed so that amplification of potentially contaminating genomic DNA sequences would produce PCR products that were substantially larger than the expected PCR products. Larger than expected PCR products would result because intron sequences that were excised during RNA processing would be included in genomic DNA. The annealing temperature for all the primer pairs used in this experiment was 55°C. HGFR A specific 3' HGFR downstream primer (CTCATCAGCGTTATCITC) was designed as was a 5' upstream primer (TGGTCCnTGGCGTCGTCCTC). This primer design yields an expected product of 336 bp. An alternate HGFR-speciiic downstream 3' primer (CCriTGAAGGCAGGCATrrCT) was designed from the inverse complement of 21 bases from the 96-bp insert reported previously by Wilson et al.7 This downstream primer was used with the 5' upstream primer to yield an expected product of 199 bp that was specific for the previously reported alternate HGFR isoform. KGFR A specific 3' KGFR downstream primer (GCCCTATATAATTGGAGACCT) was designed as was a 5' upstream primer (GGAT-

244

Reports

IOVS, January 1999, Vol. 40, No. 1

CAAGCACGTGGAAAAGA). This primer design yields an expected product of 177 bp. A 21-nucleotide downstream 3' primer (CACCGGCGGCCTTGCTGTTTl') that was specific for the alternative nbroblast growth factor (FGF) receptor-2 primer-derived amplification product, previously reported by Wilson et al., was also designed. The design included the inverse complement to 11 nucleotides at the 3' end of the K exon and 10 nucleotides at the 5' end of the B exon. This downstream primer was used with the FGF rcceptor-2/KGF 5' upstream primer to yield an expected PCR product of 229 bp that was specific for the previously reported alternate KGFR isoform.

336 bp

HGFR L 1

2

3 C

TGF0R-II l'he primer pairs used for TGF/3R-II consisted of a 5' upstream primer (CCACCGCACGTTCAGAAGTCG) and a 3' downstream primer (C H GCCGGTH CCCAGGTTGA) that yields an expected PCR product of 543 bp. The PCR reaction yields an alternative band at approximately 600 bp. The band was sequenced to verify that it was an alternatively spliced form of TGF0R-I1 and contains 618 bp

PCR, Horizontal Gel Electrophoresis, and Sequencing of PCR Products All samples were amplified with a primer pair specific for each growth factor receptor using a master mix containing all the components in the PCR reaction except the target template cDNA or water negative control. All PCR reactions were prepared using a commercial method (Taq Start Antibody Hot Start; Clontech, Palo Alto. CA). The antibody is used to prevent nonspecific amplification of primer-dimer formation, thus enhancing the specificity and sensitivity of the PCR reaction. Control reactions without template were included with each

336 bp

HGFR

199 bp

HGFR-alt. L 1

2 3 4 5 6 7 C

177 bp

KGFR L 1 2 3 4 S 6 7 C

229 bp-

KGFR-alt. L 1 2 3

4

5 6 7 C

618 bp

TGFp-RII L 1

2 3 4 5 6 7 C

350 bp-

p-actin L 1 2 3 4 5 6 7 C

1 K.I KI 1. tithidium bromide-stained agarosc gel that compares ex prcssion of alternatively .spliced isolorms (alt) of hepatocyte growth factor receptor (HCiFK; met), keratinocytc growth factor receptor tk(iHR) and transforming growth factor beta receptor II (T(iF0R-lI) polynicra.se chain reaction (PCR) products with expected forms in cultured human trabecular meshwork (HTM) cells The cDNA samples were generated from human cultured HTM cells of donors of various ages (limes I through ~). Actin was included as an internal standard l.iini' I. DNA ladder; linw I, (i days; lune _'. 18 days; Uuw .5, (•> months. Imii' -v. 1 years; Uine 5. 18 years; time 6. Si years: Imw ~. 80 years; lime i , control

KGFR-alt. TGFp-RII L

1

2

3 C

E 2. Kthidium bromide-stained agarose gel that compares expression of alternatively spliced isoforms (alt) of hepatocyte growth factor receptor (HCilR; met), kcralinocyte growtli factor receptor (KliFR), and transforming growth factor beta receptor U (TGF/SR-II) polymerasc chain reaction (PCR) products with expected forms in ex vivo trabecular meshwork tissues The cDNA samples were generated from ex vivo trabecular meshwork tissues from donors of various ages (Umes I. 2, J). Actin included as an internal standard. Lane L, DNA ladder; lane I. 75 years; him' J. 79 years; lane .i, 80 years, tune C. control.

amplification for each pair of primers. Programmable temperature cycling (FTC-100; MJ Research, Watertown, MA) was performed with the following cycle profile: denaturation for 2 minutes at 94°C, followed by 92°C for 2 minutes, 40 cycles of annealing for 30 seconds at S5°C, extension for 90 seconds at 72°C, and denaturation for 45 seconds at 92°C. Horizontal 1.5% agarose (Life Technologies. Gaithersburg, MD) gel electro phoresis was performed using 20 /xl of each PCR reaction product and 4 pil of 10 X loading buffer per lane with 150 ml gel run in Tris-acetic acid-EDTA running buffer using a commercially available wide electrophoresis unit (Mini-Sub Cell Electrophoresis Unit; Hio-Rad, Richmond, CA). A 100-bp DNA ladder (Life Technologies) was used as a molecular size standard. Ten microliter of 10 mg/ml ethidium bromide was added to the Tris-acetic acid-EDTA running buffer, and 100 V was applied until the loading dye had traveled two thirds of the distance to the end of the gel (approximately 60 minutes). To ensure specificity of the PCR reaction products, nucleic acid sequencing was performed by cloning PCR products into a cloning vector (TA Cloning Vector; Invitrogen, San Diego, CA) and then using a DNA sequencing kit (Scquenase 2.0; United States Biochemical. Cleveland, OH). RESULTS

Control PCR Reactions Each of the cDNA samples from cultured HTM cell lines (Fig. 1) and normal ex vivo HTM tissues (Fig. 2) exhibited actin PCR amplification products of the expected size (350 bp). There was no evidence of a 790-bp genomic amplification product (data not shown), indicating the absence of contaminating genomic DNA These results show the quality of the cDNA

Reports

IOVS, January 1999, Vol. 40, No. 1 used for PCR detection of mRNA expression. During PCR amplification reactions, control reactions were negative for amplification products, showing that the PCR method and reagents used yielded specific amplified products only when a cDNA source was included.

Expression of mRNA in HTM Cell Lines for Alternate Spliced Isoforms of Growth Factor Receptors Amplification products of the expected size for the alternate spliced isoform of HGFR (bp 199), KGFR (229 bp), and TGFJ3R-U (618 bp) were expressed in cell lines originating from the HTM (Fig. 1). The PCR results for the expected isoforms of HGFR and KGFR are also presented in the same gel to compare relative expression levels of each isoform. Even though reverse transcription-PCR is known to be only semiquantitative under ideal conditions, this allows some comparison of expression levels of the expected isoforms with the alternate isoforms. Because the primers used for TGFjSR-II were not designed specifically for the alternate isoform, a prominent amplification product of 543 bp, representing the normally expressed TGF/3R-II receptor, was also present in all cell lines and in the 6l8-bp alternate isoform. There were no significant differences between ages in the expression of alternate isoforms of HGFR and TGFj3R-II (Fig. 1). However, the alternative and expected isoforms of KGFR were variably expressed. The cell lines derived from 2-, 18-, and 54-year-old donors expressed the alternate KGFR isoform, whereas the cell lines derived from younger (6-day-, 48-day-, and 6-month-oId) and older (80-year-old) donors did not. Nucleic acid sequencing of PCR-generatecl products showed that the amplified products were derived from mRNA for the alternate spliced isoform of the respective growth factor receptor. Nucleic acid sequencing of the alternate TGF/3R-II product (618 bp) showed that the amplified product was derived from mRNA for the previously reported alternate spliced isoform of TGF/3R-II. Sequencing of the TGFjBR-II alternative isoform showed that it contained a 75-bp insert in the N-terminal coding region of the extracellular domain between nucleotide 429 and 430.

Expression of mRNA for Alternate Spliced Isoforms of Growth Factor Receptors in Ex Vivo HTM Tissues Amplification products of the expected size for the alternate spliced isoform of HGFR (bp 199), KGFR (229 bp), and TGF/3R-1I (618 bp) are expressed by ex vivo isolated trabecular meshwork tissues originating from human donors (Fig. 2). Similar to the results presented in Figure I, the PCR results for the expected isoforms of HGFR and KGFR are also presented in the same gel to compare relative expression levels of each isoform. Because the primers used for TGF/3R-II were not designed specifically for the alternate isoform, a prominent amplification product of 543 bp, representing the normally expressed TGF/3R-II receptor, is also present in all tissue samples and the 6L8-bp alternate isoform. Because of the limited number of tissue samples, it would be somewhat speculative to compare expression patterns. However, there seems to be some variability of expression in the expected isoform of HGFR and KGFR (Fig. 2) In addition, the expression of the alternate isoform of KGFR is not as variable as that in cultured cells. Nucleic acid sequencing of the PCR-generated products

245

showed that the amplified products were derived from mRNA for the alternate spliced isoform of the growth factor receptors. Sequencing of the alternate TGF/3R-1I isoform revealed that it was identical with that expressed by trabecular meshwork cell lines, as described earlier. Expression of mRNA for alternate growth factor receptor isoforms by trabecular meshwork cell lines (Fig. 1) seemed to be similar to expression of mRNA in dissected HTM tissue (Fig. 2).

DISCUSSION

We have previously reported that cultured HTM cells express mRNA for several growth factor receptors.2'^ The results of the present study show, for the first time, that alternatively spliced isoforms of growth factor receptors are also expressed within the HTM. Specifically, we have identified alternate isoforms of HGFR, KGFR, and TGF/3R-M. Because it: is possible that cell culture conditions cause an alteration in the expression of growth factor receptors, our results are of further significance, because we show similar PCR expression patterns in cultured HTM cells and ex vivo HTM tissues. Thus, these results show that cells within the HTM produce mRNA for specific alternatively spliced growth factor receptors and that the expression is not an artifact of the in vitro cell culture system. Alternative mRNA splicing is an important mechanism that is used by cells to generate diverse protein molecules.'4 Use of this process provides a cell the unique potential of controlling the quantitative and qualitative expression of diverse proteins. In addition, it allows a single gene to generate more than one mRNA and thus more than one protein. Functional differences among alternatively spliced growth factor receptor isoforms are generally unknown; however, they could influence ligand specificity, intracellular and extracellular localization patterns, phosphorylation, kinase activity, and intracellular substrate interactions.8 Alternative splicing of growth factor receptor mRNA by trabecular meshwork cells provides the cell additional versatility in regulating the microenvironment of the trabecular meshwork. Most target cells for TGFjBs express two high-affinity cell surface receptors. The receptors are identified as TGF/3R-1 (55 kDa) and TGFjSR-II (80 IcDa). Recent information indicates a unique relation between these receptors. Although TGFjSR-l and TGF/3R-I1 are both required for TGFjS ligand signaling to occur, TGF/3R-II cannot generate a response independently of TGF/3R-l.y Both receptors are transmembrane serine-threonine kinases, and they form a heteromeric receptor signaling complex when activated.10 The TGF/3R-II receptor is the "primary" receptor that binds the ligand and then forms a complex with TGFj3R-l, which then acts as the "transducer" of the message. Nikawa" first reported the isolation of an alternate isoform of the human TGF/3 type II receptor. Recently Hirai and Fujita12 have reported the isolation of the same isoform. The alternative isoform includes a 75-bp insert between nucleotides 429 and 430. This results in an alteration of the amino acid sequence in the N-terminal coding region and the addition of 25 amino acids to the extracellular domain of the receptor. Our studies indicate that HTM tissue and cultured HTM cells expressed the identical alternate isoform of the TGF/3 type II receptor.

246

Reports

What might be the function of the alternate TGFjSR-II isoform in the trabecular meshwork? Massague9 has suggested that TGF/3 signaling through two sequential transmembrane kinases may accord the ligand direct control over signal diversity. For example, each TGF/3 ligand may recognize a different TGFjSR-II and TGF/3R-I receptor that are paired in a specific combination. The nature of the signal generated may thus depend on the composition of the receptor complex. A clear function for the alternative form of the TGF/3R-II receptor is unknown at the present time. I-Iirai and Fujita12 have showed that when the isoform is expressed in cells deficient in TGFjSR-II receptors, it can act as a functional receptor by transducing signals from TGF/3 ligands. The expression of a novel alternate isoform for TGF/3R-N by HTM cells increases the diversity of the TGF/3 receptor complex within the tissue and thus may allow trabecular meshwork cells to generate diverse signal patterns in response to different TGF/3 ligands. The diverse signaling pattern may then be reflected in altered cellular functions such as cell proliferation, cell motility, or alteration in the extracellular matrix. The biologic activity of HGF is transduced through the HGF (c-met) receptor.13 HGFR is a transmembrane 190-kDa heterodimer consisting of disulfide-linked chains of 50 kDa (a chain) and 140 kDa (j3 chain)."1 The receptor is synthesized as a single-chain precursor and then proteolytically cleaved to yield the mature two-chain protein. The j3 chain spans the membrane and contains the tyrosine kinase catalytic domain, whereas the a chain remains extracellular.M Wilson et al.' have reported the isolation of an alternative isoform of human HGFR in the cornea that consists of a 96-bp insert between nucleotide 4129 and 4130. This represents an intron that is retained during mRNA splicing and results in an alteration in the C-terminal coding region of the cytoplasmic tyrosine kinase domain. Translation of the alternate HGFR would be terminated near the 5' end of the 96-bp insert, which is distal to the major autophosphorylation site. Termination at this point produces a truncated receptor that would likely have kinase activity but with altered signaling function.7 Our laboratory was the first to report expression of HGF and HGFR mRNA in HTM cells.2•* Recently we have shown that the exogenous administration of HGF significantly stimulates trabecular meshwork cell proliferation.3 Because it is currently not known whether human aqueous humor contains HGF, it is possible that the HGF-HGFR system may function in an autocrine manner within the HTM. In our current studies, HTM tissue and cultured HTM cells expressed the identical alternate isoform of the HGFR. It has been shown that this receptor is also expressed by human corneal epithelial cells7 and human breast epithelial cells.7 The function of truncated growth factor receptors has not been well documented. Some truncated receptors are thought to participate in regulating the responses of cells to growth factors by competing for available ligand or by forming inactive heterodimers with the full-length receptors. It is currently not known whether this is the mechanism of action of the alternate HGFR isoform in the HTM. The specific function of this alternate HGF receptor continues to be investigated. Normally, the KGFR and the FGFR-2 receptor are derived by alternate mRNA splicing from the same gene (bek). In this unique situation, two growth factor receptors with different ligand binding specificity and expression patterns

IOVS, January 1999, Vol. 40, No. 1 are normally encoded by alternative transcripts of the same gene. The alternative KGFR showed in HTM tissue and cultured trabecular meshwork cells results in a termination at bp 13 in the B exon, because a frame shift is produced when the K and B exons are spliced.7 The alternative receptor is likely to have KGF ligand-binding specificity. In addition, the alternative KGFR would be soluble because it is truncated just before the transmembrane domain and would not have the transmembrane and cytoplasmic domains of the membrane-associated receptor. The identical alternative KGFR has also been shown to be expressed by human mammary epithelial cells and corneal epithelial cells,7 indicating that the receptor may occur widely in cells. There is currently little information concerning the role that KGF plays in the HTM. We first reported expression of KGF and KGFR mRNA in HTM cells.2* Recently, we have shown that the exogenous administration of KGF does not stimulate trabecular meshwork cell proliferation.* Because it is currently not known whether human aqueous humor contains KGF, it is possible that, similar to the HGF-HGFR system, the KGF-KGFR system may function in an autocrine manner within the HTM. In conclusion, we have shown that ex vivo HTM tissue and cultured HTM cell lines express mRNA for alternatively spliced growth factor receptors. Specifically, we have identified alternative forms for HGFR, KGFR, and TGF/3R-II. At present, we have an incomplete understanding of the significance of and the role played by these alternatively spliced isoforms. However, it is likely that if translated these alternate growth factor receptors would have distinct biologic activities within the HTM. We have shown that functional growth factor receptors are normally expressed by HTM cells.3 Several questions remain to be answered. Detection of alternative isoforms of growth factor receptor mRNA by PCR suggests that the corresponding proteins are produced within the HTM. However, the translation of a specific mRNA that is synthesized in a cell may be repressed by posttranscriptional regulatory mechanisms. Thus, it is important in future research to detect individual alternate isoform proteins in trabecular meshwork tissue and cultured trabecular meshwork cells. Currently, antibodies against these proteins are not available for these studies. In the future we will be challenged to ascertain the functional consequences resulting from the differential expression of these splice variants. It is also important to determine whether the binding affinity of these alternative growth factor receptors is different in trabecular meshwork cells from patients with glaucoma. In addition, we will be challenged to ascertain which extracellular regulators or combination of regulators can affect alternative splicing of growth factor receptors within the HTM. The microenvironment within the HTM is undoubtedly controlled by multiple factors. Our current results underscore the complexity of this important tissue.

Acknowledgments The authors thank Sherry English-Wright for assistance in cell culture and dissection of the human trabecular meshwork; The Central Florida Lions Eye and Tissue Bank, Tampa, FL. for ocular tissue; and Lawrence Oakford for help in figure presentation.

Reports

10VS, January 1999, Vol. 40, No. 1

References 1. Triputhi RC, Borisuth NSC, l.i J, Tripathi BJ. Growth factors in the aqueous humor and their clinical implications. J Glaucoma. 199-1; 3:248-258. 2. Wordingcr RJ, Clark Al;, Wilson SI!. 1-xpression of F.GF, HGI;, basic FGF, TGITj3l and their receptor mRNA in cultured human tr.ibccular meshwork cells (HTM). Invest Opbthahnol Vis Sci. I996;37:S895. 3. Wordingcr RJ, Clark AF, Agarwal R. McNatt L, Wilson SI£, Qu I;, Fung, BK-K. Cultured trabecular meshwork cells express functional growth factor receptors. Invest Opbthahnol Vis Sci. 1998; 39:1575-1589. 4. Andreadis A, Gallego Mil, Nadal-Ginard B. Generation of protein isoform diversity by alternative splicing. Annu Rev Cell liiol. 1997; 3:207-242. 5. Clark AF, Wilson K, McCartney Ml), Miggans SI', Kunkle M, Howe W. Glucocorticoid-induced formation of cross-linked actin networks in cultured human trabecular meshwork cells. Invest Opbthalmol Vis Sci. 1994; 3 5:281 - 294. 6. Steely HT, Browder SL, Julian MB, Miggens ST, Wilson KL, Clark AF. The effects of dexamethasone on libronectin expressed in cultured human trabecular meshwork cells. Invest Ophtbahnol Vis Sci. I992;33:2242-225O. 7. Wilson S1-, Weng J, Chwang E. Gollahon 1., Leitch AM, Shay JW. llepatocyte growth factor (HGF). keratinocyte growth fac-

Prominent Increase of Macrophage Migration Inhibitory Factor in the Sera of Patients with Uveitis Nobuyoshi Kitaichi,' Satoshi Kotake,' Yoichi Sasamoto,' Kenichi Narnba,x Akira Matsuda,1 Kazumasa Ogasawara,2 Kazunori Onoe2 Hidehiko Matsuda,' and fun Nishihira*

8.

9. 10.

11.

12.

13-

14.

15.

247

tor (KGF), and their receptors in human breast cells and tissues:alternative receptors. Cell Mol liiol Res. 1994:40:337-350. HouJ, Kan M, McKeehan K, McBride G, Adams P, McKechan WL. Fibroblast growth factor receptors from the liver vary in three structural domains. Science. 1991;251 665-668. Massague J. TGF/3 signaling. Receptors, transducers, and Mad proteins. Cell. 1996:85:947-950. Wrana JL, Attisano J. Carcamo A. el al. TGF-/3 signals through a heteromeric protein kinase receptor complex. Cell. 1992:71: IOO3-1OI4. Nikawa Y. A cDNA encoding the human transforming growth factor beta receptor suppresses the growtli defect of a yeast mutant. Gene. 1994; 149:367-372. Hirai R, Fujita T. A human transforming growth factor-/3 type II receptor that contains an insertion in the extracellular domain. lixp Cell Res. 1996:223:135-141. Weidner KM. Hartmann G, Sachs M, Birchmeier \V. Properties and functions of scatter factor/hepatocyte growth factor and its receptor c-met. AmJ Respir Cell Mol liiol. 1993;8:229-237. Giordanio S, Ponzetto C, l)i Ren/o MF, Cooper CS, Comoglio PM. Tyrosine kinase receptor indistinguishable from the c-met protein. Nature. 1989;339:155-156. Thach RI-. Cap recap: the involvement of elF-4F in regulating gene expression. Cell. 1992:68:177-180.

RESULTS. The mean MIF levels in the sera of the patients with IJehccts disease, VKH disease, and sarcoidosis and of healthy control subjects were 60.4 ± 9.0 (mean ± SE) ng/ml, 16.5 ± 2.9 ng/ml, 27.1 ± 5.6 ng/ml, and 5.4 ± 0.04 ng/ml, respectively. The average levels of MIF in the sera of uveitis patients were significantly higher (/-* < 0.0001) than those of healthy control subjects. The high levels of MIF were especially noted in patients with Behcet's disease at the ocular exacerbation stage and patients with sarcoidosis at the severe uveitis stage. CONCLUSIONS. Significant increase of MIF in sera was characteristic of uveitis, and MIF may be a useful laboratory parameter to use to comprehend the clinical course of uveitis. (Inuest Opbthalmol Vis Sci. 1999:40:247-250)

PURPOSE. TO investigate paihogcnesis underlying endogenous uveitis, macrophage

migration

inhibitory

factor

(MIF) was quantified in sera of patients. METHODS. Sera were obtained from the 55 patients with uveitis (24 with Hehcxt's disease; 9 with Vogt-KoyanagiHarada's |VKH| disease; 22 with sarcoiclosis) and 58 healthy control subjects. MIF levels were determined by a human MIF en/vme-linked immunosorbent assav.

From the 'Department of Ophthalmology, School of Medicine, the "Section of Pathology, Institute of Immunological Science, and the ^Central Research Institute. School of Medicine, Hokkaido University, Sapporo, Japan. Partially supported by a Grant-in-Aid for Scientific Research from The Ministry of Education, Science, Sports and Culture, Tokyo, Japan; by a Research Grant for Behcel's Disease from The Ministry of Health and Welfare, Tokyo. Japan; and by The Japan National Society for the Prevention of Blindness, Tokyo, Japan. Submitted for publication January 12, 1998: revised June 4 and August 19, 1998: accepted September 8, 1998. Proprietary interest category: N. Reprint requests: Kazunori Onoe, Section of Pathology, Institute of Immunological Science, Hokkaido University, Sapporo ()6(), Japan.

M

acrophage migration inhibitor)' factor (MIF) was reported as the first lymphokine produced by activated T cells in guinea pigs.1 The MIF was determined by an inhibitory capacity of macrophage migration. Although it was suggested that MIF production was associated with the generation of cellmediated immune responses, the biological role of MIF in vivo has been unclear for a long time. Recent studies have established that MIF is a critical cytokine in the delayed type hypersensitivity response, an important mediator of endotoxic shock, and a counter regulator of glucocorticoid action, liacher et al.2 showed that anti-MlF antibodies inhibited T-cell proliferation and interleukin-2 (IL.-2) production in vitro and suggested that MIF played an important regulator}' role in the activation of T cells induced by mitogenic or antigenic stimuli. Moreover, it was reported that anti-MIF antibodies inhibited immunologically induced crescentic anti-glomcrular basement membrane glomerulonephritis in rats and collagen type 11-induced arthritis in mice. v '' In human diseases, it was demonstrated that the serum MIF levels of the patients with atopic dermatitis were significantly liigher than those of the healthy control subjects. 5