From the ' Department of Ophthalmology, Duke University Medi- ..... in disc areas; TRD, traction retinal detachment; NVD, neovascularization of the disc; NVE,.
1256
Reports
10. Danforth JM, Strieter RM, Kunkel SL, Arenberg DA, VanOtteren GM, Standiford TJ. Macrophage inflammatory protein-la expression in vivo and in vitro: the role of lipoteichoic acid. Clin Immu-
nol Immunopathol. 1995;74:77-83.
Matrix Metalloproteinases and Their Inhibitors in Human Vitreous Monica A. De La Paz,l Yoshifumi Itoh,2 Cynthia A. Toth,l and Hideaki Nagase2 conduct zymographic analysis to study the matrix metalloproteinases (MMPs) and tissue inhibitor of metalloproteinases (TIMPs) in vitreous samples of patients undergoing pars plana vitrectomy as part of the treatment of vitreoretinal disease.
PURPOSE. TO
Forty-two vitreous samples were collected at the time of pars plana vitrectomy. Diagnoses included severe (exudative) age-related macular degeneration (AMD) (12), macular hole (10), presumed ocular histoplasmosis syndrome (6), proliferative diabetic retinopathy (PDR) (5), epiretinal membrane (4), vitreomacular traction syndrome (2), macroaneurysm with subretinal hemorrhage (1), central retinal vein occlusion with vitreous hemorrhage (1), and proliferative vitreoretinopathy (1). Gelatin zymography, reverse gelatin-zymography, carboxymethylated transferrin zymography, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis were performed on the liquid vitreous samples to assess for MMP and TIMP activity.
METHODS.
Progelatinase A occurred in all vitrectomy samples. In addition, a band consistent with TIMP-2 occurred in all samples on reverse zymography. An inhibitor of MMP of a lower molecular weight than TIMP-1 was found in all the samples. A serine proteinase with a broad band around 180 kDa was found in 2 of the 11 AMD vitreous samples. A 75-kDa metalloproteinase was found in several AMD samples, but it was much more abundant in the PDR samples. RESULTS.
Metalloproteinases and their endogenous inhibitors are present in human vitreous and may be involved in the pathogenesis of PDR and other vitreoretinal diseases. (Invest Ophthalmol Vis Sci. 1998;39: 1256-1260)
CONCLUSIONS.
From the ' Department of Ophthalmology, Duke University Medical Center, Durham, North Carolina; 2 Department of Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City. Supported by Research to Prevent Blindness, the Robert Wood Johnson Foundation (MADLP), and National Institutes of Health grants AR39189 (HN) and AR40994 (HN)Submitted for publication September 10, 1997; revised February 18, 1998; accepted March 9, 1998. Proprietary interest category: N. Reprint requests: Monica A. De La Paz, Duke University Eye Center, Box 3802 DUMC, Durham, NC 27710.
IOVS, June 1998, Vol. 39, No. 7 11. De Kimpe SJ, Kengatharan M, Thiemermann C, Vane JR. The cell wall components peptidoglycan and lipoteichoic acid from Staphylococcus aureus act in synergy to cause shock and multiple organ failure. Proc Natl Acad Sci USA. 1995:92:10359-10363.
T
he role of the vitreous in disorders of the retina is a subject of great interest that remains incompletely understood. The vitreous, which is structurally complex, is composed of a specialized extracellular matrix with a gel-like consistency. The water content is very high (98%-997% of the wet weight) and hyaluronan is the predominant glycosaminoglycan. Collagen types II, V, and IX are present in a 69:24:7 ratio and provide structural support to the gel. Various poorly characterized soluble proteins and glycoproteins are present, some of which are serum components.1 Age-related changes can occur in the vitreous that result in the loss of the gel-like consistency and liquefaction. The exact biologic mechanism of these changes is unknown. Changes in the vitreous are also known to occur in association with retinal diseases. For example, vitreous pathology in proliferative diabetic retinopathy (PDR) has been extensively described.2 It has been proposed that vitreous changes are involved in the pathogenesis of age-related macular degeneration (AMD).3 The matrix metalloproteinases (MMPs) are a group of enzymes involved in physiological and pathologic processes associated with extracellular matrix remodeling.4 These zincdependent endopeptidases function to degrade the extracellular matrix. The activity of this group of enzymes is in part regulated by a group of natural inhibitors of MMPs known as tissue inhibitors of metalloproteinases (TIMPs).4 The MMPs are secreted as inactive proenzymes, and removal of the amino terminal propeptide generates the mature, active form of the enzyme. The TIMPs act by restricting the conversion of the proenzyme to the active form and also directly inhibit the active form of the enzymes. Little is known about the role of MMPs and TIMPs in the normal vitreous and in disease states involving the retina and vitreous. Mutations in TIMP3 have been found in affected family members in Sorsby's fundus dystrophy pedigrees.5 In this hereditary macular dystrophy, which phenotypically resembles AMD, changes in Bruch's membrane as a result of impaired regulation of extracellular matrix turnover have been proposed to lead to the degenerative changes in the fundus. Progelatinase A (proMMP-2) has been identified in human vitreous, and progelatinase B (proMMP-9) has been described in vitreous of patients with PDR.6 In this report, we have studied proteinases, including MMP and its inhibitor, TIMP, in the vitreous of patients with severe exudative AMD, PDR, macular hole, and other disorders of the retina and vitreous. Zymographic analysis was conducted as an initial step in the evaluation of the level of enzymes and TIMPs and their comparison among different vitreoretinal disorders. METHODS
Vitrectomy Specimens All pars plana vitrectomies were performed at Duke University Medical Center by one of the authors (CAT) as part of the
Reports
IOVS, June 1998, Vol. 39, No. 7 therapeutic regimen of the disease process. The study was conducted in accordance with the Institutional Review Board of Duke University Medical Center, and all subjects were treated in accordance with the tenets of the Declaration of Helsinki. Informed consent was obtained, and a vitreous biopsy was performed, with removal of approximately 200 JLL! of a sample of undiluted vitreous at the start of the vitrectomy. Characteristics of patients are shown in Table 1. Vitreous samples were obtained from 12 patients with a diagnosis of subfoveal neovascular membrane secondary to AMD, 10 patients with macular hole, 6 with subfoveal neovascular membrane secondary to presumed ocular histoplasmosis syndrome, 5 patients with severe proliferative diabetic retinopathy, 4 patients with epiretinal membranes, 2 patients with vitreomacular traction syndrome, 1 patient with proliferative vitreoretinopathy, 1 patient with central retinal vein occlusion, who was to have a vitrectomy performed for the treatment of vitreous hemorrhage, and 1 patient with a macroaneurysm, who was to undergo surgical removal of a submacular hemorrhage. Polyacrylamide Gel Electrophoresis Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed as described by Bury,7 and proteins were stained with Coomassie Brilliant Blue R-250. Ten microliters of the liquid vitreous specimens was combined with 10 fx\ of the loading buffer without reducing agents. A 5-/LL1 vitreous sample was loaded in each lane. SDS-PAGE was conducted at room temperature. A similar processing of the vitreous samples was used for all zymographic analyses. Gelatin Zymography and Carboxymethylated Transferrin Zymography Gelatin zymography was conducted with SDS-polyacrylamide gel containing 0.8 mg/ml gelatin, as described by Hibbs et al.8 Samples were mixed with the SDS-PAGE sample buffer without a reducing agent and were subjected to electrophoretic analysis at room temperature. Enzymic activity was visualized through negative staining with Coomassie Brilliant Blue R-250 (Bio-Rad, Richmond, CA). Carboxymethylated transferrin (Cm-Tf) was prepared as described previously.9 Cm-Tf zymography was conducted using the SDS-polyacrylamide gel containing 0.25 mg/ml Cm-Tf. Reverse Zymography Reverse zymography was conducted with SDS-polyacrylamide gel containing gelatin and 1.2 mg/ml isolated proMMP-2, as described previously.8 Reverse zymography was performed on the control samples and on half of the vitrectomy samples. After electrophoresis the gels were washed with the buffer containing 2.5% Triton X-100 to renature proMMP-2 in the gel by the removal of SDS. The activity of TIMPs was visualized as an inhibition of MMP-2 activity on the gelatin in the gel, and, thus, positive staining with Coomassie Brilliant Blue R-250 could be observed.
RESULTS Results of the zymographic analyses are summarized in Table 1 for the cadaver donor eyes. Representative results of gelatin zymography, Cm-Tf zymography, reverse zymography, and
1257
SDS-PAGE are shown in Figure 1. All specimens contained a 68-kDa gelatinase band on gelatin zymography, which migrated identical with the proMMP-2 band of the HT-1080 cell culture medium. A band consistent with the migration pattern of TIMP-2 occurred in all samples evaluated by reverse zymography. In addition, reversed zymography demonstrated a TIMP activity of lower molecular weight than TIMP-1, which was detected in all samples, which is suggestive of TIMP-3 or an unknown TIMP. Vitreous samples of patients with AMD demonstrated an abundant proteinase activity with a broad band around 180 kDa that was identified by Cm-Tf zymography in 2 of the 12 samples. This proteinase occurred in relatively high amounts based on the relative intensity of the bands, and further characterization demonstrated an absence of the band in the presence of 2 mM phenylmethylsulfonyl fluoride, thus identifying it as a serine proteinase (Fig. 2). This serine proteinase was undetected in the other vitrectomy samples. There were no identified differences in severity of the AMD or in the clinical course of the two patients with serine proteinase in the vitreous in comparison with the other patients with AMD. An approximately 75-kDa proteinase was detected by Cm-Tf zymography infiveof the AMD vitreous samples, two of the macular hole vitreous samples, three of the presumed ocular histoplasmosis syndrome samples, and all the PDR samples (Fig. 1). Relative intensity of the band was highest in the PDR samples. The activity was inhibited by EDTA, indicating that this is a metalloproteinase. This band was of a lower molecular weight than proMMP-9, and it may represent the active form MMP-9. Further biochemical characterization of this metalloproteinase is in progress. Many of the samples contained an approximately 120-kDa gelatinase. It was noted in three of the AMD samples and in four of the PDR samples. Interestingly, it was not noted in any of the 10 macular hole vitrectomy samples. Further characterization of this gelatinase is in progress.
DISCUSSION In this report, we evaluated the expression of MMPs and TIMPs in the vitreous of patients undergoing pars plana vitrectomy for the treatment of severe, exudative AMD, macular hole, PDR, and other vitreoretinal disorders. Little is known about die biochemical basis of changes in the vitreous in vitreoretinal disease, including aging disorders of die retina such as AMD. The vitreous can be considered to be a specialized form of extracellular matrix. In view of the known importance of the metalloproteinase system in extracellular matrix remodeling and turnover, we herein have sought to evaluate diis system in die vitreous of eyes that were to undergo vitrectomy for diseases known to affect the retina alone or die vitreous and die retina. In AMD, the potential importance of the metalloproteinase system has been a subject of great interest, particularly in view of the phenotypic similarity to Sorsby's fundus dystrophy, in which mutations in TIMP-3 in family members have been described.5 However, De La Paz et al.10 recently found no evidence for a strong association between the AMD phenotype and mutations in TIMP-3. Recently, it has been proposed that the posterior hyaloid isfirmlyattached in some cases of severe exudative AMD and that thisfirmattachment may contribute to the disease process.3 A role for pars plana vitrectomy to relieve
1258
Reports
IOVS, June 1998, Vol. 39, No. 7
TABLE 1. Patient Characteristics and Results, Grouped by Diagnosis Summary of Findings Sample* Age-related macular degeneration 3 9 10 12 26 52 59 74 93 106 111 113 Macular hole 2 6 11 31 71 75 78 85 92 112 Presumed ocular histoplasmosis syndrome 1 4 7 33 109 110 Proliferative diabetic retinopathy 5 45 79 80 99 Epiretinal membrane 8 17 35 42 Other diagnoses 13 70 34 102 105
4gef
Clinical Characteristics
Eye
Sex
l A DA subfoveal CNVM 10 DA subfoveal CNVM 4 DA subfoveal CNVM 1 DA subfoveal CNVM 12 DA subfoveal CNVM and extensive submacular hemorrhage 1 DA subfoveal CNVM 4 DA subfoveal CNVM 10 DA subfoveal CNVM 1 DA subfoveal CNVM 4 DA subfoveal CNVM 2 DA subfoveal CNVM 4 DA subfoveal CNVM
OD OD OS OD OD
M F F F M
77 84 65 76 73
OS OD OD OD OD OD OS
M F M M M F F
80 74 69 63 62
(years)
79
84
Vision before Vision before Vision before hyaloid Vision before Vision before Vision before Vision before Vision before Vision before Vision before hyaloid
surgery 20/60, attached posterior hyaloid surgery 20/200 attached posterior hyaloid surgery 12/200 E attached posterior
OD OS OD
F F F
55 70 68
surgery surgery surgery surgery surgery surgery surgery
OS OS OD OD OS OD OD
F F F M F M F
71 69 55 81 65 63 68
2 4 1 2 1 1
subfoveal subfoveal subfoveal subfoveal subfoveal subfoveal
OD OS OS OS OS OD
F F F F M M
29 49 65 43 55 75
NVE, TRD Severe NVD Vitreous hemorrhage, TRD Vitreous hemorrhage, severe NVE NVE, TRD
OD OS OD OD OS
F M F F F
61 46 33 29 53
Vision Vision Vision Vision
OS OD OS OD
F M F M
69 75 79 72
OS OD OD OD
F F F M
73 50 91 63
OD
M
77
DA CNVM, DA CNVM, DA CNVM, DA CNVM, DA CNVM, DA CNVM,
before before before before
surgery surgery surgery surgery
20/400 attached posterior hyaloid 10/200 attached posterior hyaloid 20/100 attached posterior hyaloid 20/400 attached posterior hyaloid 20/160 attached posterior hyaloid 20/70 detached posterior hyaloid 20/400 detached posterior
20/50 20/50 20/200 20/70
Vitreomacular traction Vitreomacular traction Macroaneurysm with submacular hemorrhage Proliferative vitreoretinopathy with full thickness retinal folds CRVO, vitreous hemorrhage
DA, size of choroidal neovascularization (CNVM) in disc areas; TRD, traction retinal detachment; NVD, neovascularization of the disc; NVE, neovascularization elsewhere; CRVO, central retinal vein occlusion; M, male; F, female; +, present; —, absent; A, approximately 180-kDa serine proteinase; B, 120-kDa metalloproteinase; C, approximately 75-kDa metalloproteinase noted on carboxymethyl transferrin zymography, of lower molecular weight than matrix metalloproteinase (MMP>2; 1, 68-kDa metalloproteinase, possibly progelatinase A (proMMP-2); 2, tissue inhibitor of metalloproteinase (TIMP)-2; and 3, unknown TIMP of lower molecular weight than TIMP-1. * Sample number of reference also refers to lane number used in figures. t Age of donor at time of death.
Reports
10VS, June 1998, Vol. 39, No. 7 Normal Donor
PDR Donor
1259
A M D Donor
Sample * Cm-Tf Zymography
kDa ProMMP-9 (88 kDa) C
Gelatin Zymography
kDa
B ProMMP-9 (88 kDa) ProMMP-2 (68 kDa) 62 kDa MMP-2
kDa Reverse Zymography
ProMMP-9 (88 kDa)
TIMP-1 Unidentified TIMP TIMP-2
SDS/PAGE
kDa 92 61 46 34 25
—- M »
ff
FIGURE 1. Zymographic and reverse zymographic analyses of human vitreous samples. Human vitreous samples from normal, proliferative diabetic retinopathy (PDR), and age-related macular degeneration (AMD) donors were applied to carboxymethylated transferrin (Cm-Tf) zymography, gelatin zymography, reverse zymography, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under nonreducing conditions. Each lane contained an equal volume (2.5 jul) of the sample. SDS-PAGE was performed to distinguish the tissue inhibitor of metalloproteinase (TIMP) bands that appeared in reverse zymography from the protein bands. The conditioned medium (HT) of the phorbolmyristate acetate-treated HT-1080 cells was applied to show progelatinase B (proMMP-9), progelatinase A (proMMP-2), 62-kDa MMP-2, TIMP-1, and TIMP-2. Proteins were visualized by staining with Coomassie Brilliant Blue R-250. (A) A proteolytic activity around 180 kDa; (B) 120-kDa metalloproteinase.
this firm attachment has, therefore, been proposed. Further studies are needed to more fully understand the potential importance of the vitreous in AMD. In this report, we found evidence for the presence of a serine proteinase with a broad band around 180 kDa in 2 of 11 vitreous samples in the exudative AMD cases. The significance of this finding is unclear. A review of these two cases in comparison with the other AMD cases did not reveal any major difference in the severity of examination findings or in the
duration of symptoms. Whether they are involved in the pathogenesis of AMD remains to be more fully elucidated. Metalloproteinases previously have been identified in the human vitreous by Brown et al., 6 who identified proMMP-2 in the normal and diabetic vitreous and proMMP-9 activity in diabetic vitreous samples. In this report, we have identified a metalloproteinase that migrated in an identical manner to the latent form of MMP-2 on gelatin zymography, supporting the results of the previous report.
1260
Reports
IOVS, June 1998, Vol. 39, No. 7
Control Sample No. 5 10 HT
2mMPMSF 5 10 HT
50 mM EDTA 5 10 HT
10LIME-64
5 10 HT
kDc -ProMMP-9
-C
2. Sensitivity of the 75- and 180-kDa proteinases to proteinase inhibitors. Vitreous samples (no. 5 and no. 10) were applied to carboxymethylated transferrin (Cm-Tf) zymography. After washing, the gels were incubated at 37°C in the presence or absence of 2 mM phenylmethylsulfonyl fluoride, 50 mM EDTA, or 10 JLLM E-64 [L-£raHS-epoxysuccmyl-leucylamide-(4-guanidino)-butane], a cysteine proteinase inhibitor. The conditioned medium (HT) of the phorbolmyristate acetate-stimulated HT-1080 cells was applied to indicate the position where progelatinase B (proMMP-9) migrates. (A) A proteolytic activity around 180 kDa; (B) 75-kDa metalloproteinase. FIGURE
In addition, we have described the presence of TIMP activity with zymographic characteristics analogous to TIMP-2 in all samples evaluated with reverse zymography, which suggests the presence of an endogenous inhibitor of metalloproteinase activity in human vitreous. The presence of MMP-2 and a TIMP activity in the vitrectomy samples provides evidence suggesting that matrix remodeling may be a normal biochemical process in the vitreous. A 75-kDa metalloproteinase was noted on Cm-Tf zymography of many of the vitrectomy specimens, but it was particularly prominent in the PDR samples. It was noted in only one of the eight cadaver vitreous specimens. The tendency for this metalloproteinase activity to be elevated in PDR suggests that it may be involved in the disease process. Alternatively, it may represents a manifestation of increased vascular permeability in neovascular fronds or a diffusion from local tissue sites such as the ciliary body. This MMP may enter the vitreous from the serum, a mechanism proposed for die entry of proteins into the vitreous.' Additional studies would be beneficial in studying the role of this metalloproteinase in PDR and proliferative vitreoretinopadiy. In summary, we have described the presence of metalloproteinase activity and endogenous inhibitor activity in vitrectomy samples of patients with various diseases of the retina alone or of the vitreovis and the retina. Additional characterization of the identified gelatinase activities and inhibitors identified in this report are needed to more fully understand the significance to normal vitreous physiology and disease states. The sample size was small, and additional studies to elucidate the importance of the metalloproteinase system, particularly in PDR, are clearly indicated to more fully understand the importance of the pathophysiology of the vitreous in common diseases involving the retina and the vitreous.
Acknowledgments The authors acknowledge the assistance of Alison Cobb (Duke University Medical Center, Durham, NC) in the processing of the vitreous samples.
References 1. Sebag J. Biochemistry of the vitreous. In: The Vitreous. New York: Springer-Verlag; 1989:24-25. 2. Sebag J, Buckingham B, Charles A, Reiser K. Biochemical abnormalities in vitreous of humans with proliferative diabetic retinopathy. Arch Opbthalmol. 1992; 110:1472-1476. 3. Lambert HM, Capone A, Aaberg TM, Stemberg P, Mandell BA, Lopez PF. Surgical excision of subfoveal neovascular membranes in age-related tnacular degeneration. Am] Opbtbalmol. 1992;113: 257-262. 4. Nagase H. Matrix metalloproteinases. In: Hooper NM, ed. Zinc Metalloproteinases in Health and Disease. London: Taylor and Francis; 1986:153-204. 5- Weber BHF, Vogt G, Pruett RC, Stohr H, Felbor U. Mutations in the tissue inhibitor of metalloproteinases-3 (T1MP3) in patients with Sorsby s fundus dystrophy. Nat Genet. 1994;8:352-356. 6. Brown D, Hamdi H, Bahri S, Kenny MC. Characterization of an endogenous metalloproteinase in human vitreous. Curr Eye Res. 1994;13:639-647. 7. Bury AF. Evaluation of three sodium dodecyl sulfate-polyacrylamide gel electrophoresis buffer systems./ Chromatogr. 1981;213: 491-500. 8. Nagase H, Human strometysins 1 and 2. Methods Enzymol. 1995; 248:449-470. 9- Rankin CA, Suzuki K, Itoh Y, et al. Matrix metalloproteinases and TIMP3 in cultured C57BL/6j-cpk kidney tubules. Kidney Int. 1996; 50:835-844. 10. De La Paz MA, Pericak-Vance MA, Lennon F, Haines JL, Seddon JM. Exclusion of TIMP3 as a candidate locus in age-related macular degeneration. Invest Opbthalmol Vis Set. 1997;38: 1060-1065.
