- Email: [email protected]
Immunohistochemical study of matrix metalloproteinases (MMPs) and their tissue inhibitors (TIMPs) in pulmonary lymphangioleiomyomatosis (LAM)
Immunohistochemical Study of Matrix Metalloproteinases (MMPs) and Their Tissue Inhibitors (TIMPs) in Pulmonary Lymphangioleiomyomatosis (LAM) TOMAYOSHI HAYASHI, MD, PHD, MARIAN V. FLEMING, MD, WILLIAM G, STETLER-STEVENSON, MD, PHD, LANCE A. LIOTTA, MD, PHD, JOEL MOSS, MD, PHD, VICTOR J. FERRANS, MD, PHD, AND WILLIAM D. TRAVIS, MD To evaluate the role of matrix metalloproteinases (MMPs) and their specific tissue inhibitors (TIMPs) in the pathogenesis of the structural damage and cystic lesions found in pulmonary lymphangioleiomyomatosis (LAM), immunohistochemical studies were made of the localization of MMP-1, MMP-2, MMP-3, MMP-9, TIMP-1, TIMP2, HMB-45, and type IV collagen in sections of lung biopsy specimens from four patients with this disorder. These studies showed increased immunoreactivity compared with that in normal bronchiolar and vascular smooth muscle cells, of MMP-2 and, to a lesser extent, MMP9 and MMP-1 in the LAM cells. MMP-2 was also localized in some elastic fibers and in the basement membranes of LAM cells and overlying epithelial cells. The basement membranes in both of these sites often showed colocalization of MMP-2 and type IV collagen. Some epithelial basement membranes showing this colocalization were dis-
rupted. These changes were not accompanied by increased immunoreactivity for TIMPs. Taken together with previous observations showing structural damage to elastic fibers and collagen fibrils, and with the absence of demonstrable neutrophil or pancreatic types of elastase, these findings suggest that MMP-2 and MMP-9 (both of which can degrade elastin as well as collagens) are responsible for the connective tissue destruction and cyst formation in LAM. HUM PATHOL 28:1071--1078. Copyright © 1997 by W.B. Saunders Company Key words: malaix metalloproteinases, tissue inhibitors of metalloproteinases, lymphangiole'mmyomatosis, immanohistochemistry; confocal microscopy. Abbreviations: LAM, lymphangioleiomyomatosis; MMP, matrix metalloproteinase; TIMP, tissue inhibitor of metalloproteinase; H& E, hematoxylin and eosin; Ig, immunoglobulin.
Pulmonary lymphangioleiomyomatosis (LAM) occurs in women of reproductive age and is characterized by proliferation of smooth muscle ceils, which are derived from lymphatic vessels, and by formation of multiple, air-filled cysts. The proliferating smooth muscle cells, herein referred to as LAM ceils, show an immunophenotype that differs from that of normal smooth muscle cells in that they react with a n t i - H M t M 5 antibody. >5 Most studies of LAM have focused on the causes of the proliferation of I A M cells, with special emphasis on h o r m o n a l effects, and little is known about the mechanisms of formation of the cystic lesions. These lesions are associated with degeneration of collagen and elastic fibers] Lysis of the extracellular matrix components, including the various types of collagen, proteoglycans, and elastic fibers, is regulated by the balance of the activities of the matrix metaUoproteinases (MMPs) and their tissue inhibitors (TIMPs).8'9 High levels of MMP-2 and MMP-9 have been associated with lysis of basement membranes and with inva-
sive growth of tumor cells, s'9 These enzymes degrade native type IV collagen, denatured collagens (gelatin), and a n u m b e r of types of native collagens that contain regions in which the helical structure of the protein is disrupted. 8'9 They also degrade elastic fibers, t° Recent work in our laboratory has shown the importance of MMPs and TIMPs in the pathogenesis of destructive and fibrotic pulmonary lesions. In idiopathic pulmonary fibrosis, diffuse alveolar damage, and pulmonary Langerhans cell granulomatosis, increased activity of MMPs was found to occur in destructive lesions, whereas increased activity of TIMPs, particularly of TIMP-2, coexisted with fibrotic changes. 1~ In view of these findings, the current study was u n d e r t a k e n to determine whether abnormalities in the activities of MMPs and TIMPs are associated with the proliferative and cystic lesions of pulmonary LAM.
MATERIALS AND METHODS
Patients From the Pathology Section and Pulmonary-Critical Care Medicine Branch, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD; Department of Pulmonary & Mediastinal Pathology, Armed Forces Institute of Pathology, Washington, DC; and the Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, MD. Accepted for publication January 14, 1997. Address correspondence and reprint requests to Victor J. Ferrans, MD, PhD, Section Chief, Pathology Section; National Heart, Lung and Blood Institute, National Institutes of Health, Bldg 10, Rm 2N240, 10 Center Dr, MSC-1518, Bethesda, MD 20892-1518. Copyright © 1997 by W.B. Saunders Company 0046-8177/97/2809-001255.00/0
The study group consisted of four women (ranging in age from 24 to 38 years; mean age, 32 years) in whom the diagnosis of pulmonary LAM was established on the basis of clinical and radiological findings and the results of histological and immunohistochemical studies of open lung biopsy specimens.l-6
Histological Studies The open lung biopsy specimens were fixed with buffered 10% formalin, dehydrated, embedded in paraffin, and sectioned at a thickness of 5 #m. For histopathological study,
1071
HUMAN PATHOLOGY Volume28, No. 9 (September 1997) the sections were stained with hematoxylin and eosin (H&E), Masson trichrome, and Movat pentachrome methods.
Immunohistochemical Staining Paraffin sections were used to show the localization and degree of activity of HMB-45, type IV collagen, MMP-1, MMP2, MMP-3, MMP-9, TIMP-1 and TIMP-2. The antibodies against MMP-1, MMP-2, MMP-9, TIMP-1, and TIMP-2 (dilutions, 1:1,000; 1:2,000; 1:1,000; 1:4,000, and 1:1,000, respectively) were rabbit polyclonal antibodies that were prepared in the laboratory of one of us (W.S.S.). The characteristics and specificity of these antibodies have been described previously. 12-14 The antibodies against HMB-45 (DAKO Corp., Carpinteria, CA; dilution, 1:1,000), type IV collagen (DAKO, dilution, 1:50) and MMP-3 (Oncogene Science, Uniondale, NY; dilution, 1:50) were mouse monoclonal antibodies. The tissue sections were treated with 0.4% pepsin in 0.01 mol/L HCI for 15 to 30 minutes at 37°C before immunohistochemical staining. The reactions for TIMP-1 and TIMP-2 were also performed on sections not treated with pepsin, because such treatment often resulted in a decrease in the immunoreactivity for these components. Comparison of the results obtained in untreated sections and in sections treated with pepsin showed that normal basement membranes were not nonspecifically stained by the antibodies to the various MMPs and TIMPs after pepsinization, thus excluding the possibility of nonspecific adsorption of the antibodies to partially degraded type IV collagen. For single labeling, the sections were stained using the avidin-biotinylated peroxidase complex peroxidase method in conjunction with a Vector Elite Kit (Vector Laboratories, Burlingame, CA) and a Vector VIP kit to produce a violet color at the sites of reactivity. M1 sections were counterstained with Carazzi's hematoxylin. For double labeling for confocal laser scanning microscopy, the sections were stained by the double indirect immunofluorescence method, using combinations of a mouse monoclonal antibody against type IV collagen (dilution, 1:50) and a rabbit polyclonal antibody against MMP-2 (dilution, 1:400). The primary antibodies against type IV collagen and HMB-45 were reacted with a secondary antibody (horse antimouse immunoglobulin [Ig] G; Vector Laboratories; dilutions 1:100) labeled with fluorescein isothiocyanate. The primary antibody against MMP-2 and actin was reacted with a secondary antibody (goat anti-rabbit IgG; Vector Laboratories; dilutions 1:100) labeled with Texas red, followed by nuclear counterstaining with 4',6-diamidino-2-phenylindole (DAPI; Sigma Chemical, St Louis, MO). In these preparations, a yellow fluorescence indicated the colocalization of the red and the green fluorescence. Idennfication of the histopathological orientation was facilitated by nuclear staining with DAPI for confocal microscopy. All preparations were examined with a confocal microscope, Leica (Heidelberg, Germany) model TCS4D/DMIRBE, equipped with argon and argon-krypton lasers.
Immunohistochemical Control Procedures Negative control immunohistochemical procedures included (1) omission of the primary antibody from the staining protocol and (2) replacement of the primary antibody by normal rabbit or mouse IgG in appropriate concentrations.
RESULTS Histopathological Findings The lung biopsy specimens showed a combination of cysts and nodules of proliferating LAM cells (Fig
1A). T h e cysts varied in size a m o n g the four biopsy specimens. The larger cysts were lined mainly by thin, flat epithelial cells, whereas the smaller cysts were lined by mixtures of bronchiolar ciliated epithelial cells, hyperplastic cuboidal epithelial cells, and flattened epithelial cells. Similar types of epithelial cells f o r m e d the lining of large nodules of proliferating LAM cells. T h e smaller nodules were lined mostly by ciliated epithelial cells. T h e LAM cells were mainly distributed a r o u n d bronchioles and blood vessels and in the walls of the cysts. Small amounts of dense collagen were present at the periphery of the nodules. The smooth muscle cells in the nodules were haphazardly arranged, unlike the normal muscle layers of bronchioles and blood vessels. T h r e e types of LAM cells were identified in the LAM lesions in each of the four patients: relatively large, spindle-shaped cells (Fig 1B); small, slightly elongated cells that contained only small amounts of cytoplasm; and large, epithelioid cells (Fig 1C). T h e large, spindleshaped cells had r o u n d or oval nuclei with finely granular chromatin and occasional infoldings of the nuclear m e m b r a n e s . These cells h a d moderately a b u n d a n t amounts of cytoplasm and indistinct cell borders. T h e smaller LAM cells had only small amounts of cytoplasm and contained r o u n d to oval nuclei with dense, h o m o geneous chromatin and indistinct nucleoli. T h e epithelioid LAM cells were r o u n d to polygonal in shape, had r o u n d to oval nuclei with occasional nucleoli, and abundant, often vacuolated cytoplasm. In sections stained with H&E, all three types of cells a p p e a r e d paler than normal smooth muscle cells, although the epithelioid cells were m o r e eosinophilic than the o t h e r two types of LAM cells. The biopsy specimen f r o m patient 1 showed multiple cysts and nodules containing mixtures of all three types of smooth muscle cells. T h e biopsy specimen f r o m patient 2 showed only mild cystic changes and multiple nodules c o m p o s e d of a p r e d o m i n a n c e of small LAM cells. T h e biopsy specimen f r o m patient 3 was characterized by p r o m i n e n t cysts, which were s u r r o u n d e d by areas of cellular proliferation with n u m e r o u s epithelioid cells. The biopsy specimen f r o m patient 4 showed small cysts and only small nodules, which contained mixtures of large and small LAM cells. All biopsy specimens had m a c r o p h a g e s in the alveolar spaces and lymphocytes in the interstitium, hut both of these types of cells were inconspicuous; neutrophils were not found. T h e Movat stain highlighted areas where LAM cell infiltrates were associated with destruction of elastic fibers in the pleura or bronchiolar walls (Fig 1D).
Immunoperoxidase Staining T h e i m m u n o h i s t o c h e m i c a l findings were similar in all four cases. Data on the immunoreactivity of the different cellular and extracellular c o m p o n e n t s of the lungs for MMPs and TIMPs are summarized in Table 1. HMB-45. Large L~_M cells showed a strong granular reaction for HMB-45 (Fig 2A). T h e staining reaction for HMB-45 was focally distributed in the cytoplasm of
1072
MMPs AND TIMPs IN LAM (Hayashi et al)
FIGURE I. Histopathologica[ findings in pulmonary LAM (A) Low-magnification view showing cystic dilatation and nodular proliferation of LAM cells in pulmonan/tissue. Cleftlike spaces in the nodular lesions represent lymphatic vessels (arrowheads). (H & Estain; original magnification ×100.) (B) Large, spindle-like proliferating LAM cells form an interlacing paffern, These cells have indistinct borders, moderately abundant amounts of cytoplasm, and oval to elongated nuclei with finely granular chromatin and distinct nucleoli, (H & E stain; original magnification x400). (C) LAM cells of the smaller type have scanty cytoplasm. Their nuclei have dense, homogeneous chromatin and indistinct nucleoli. These cells are arranged in poorly defined fascicular patterns. A small cluster of epithelioid I_AM cells with round nuclei and vacuoiated cytoplasm is also present (arrowhead). (H & E stain; original magnification x400). (D) Bronchiolar elastic fibers adjacent to proliferating LAM cells are disrupted (arrowhead), (Movat pentachrome stain; original magnification x400).
these cells. Small LAM cells showed a negative to weak reaction. Epithelioid LAM cells were moderately to strongly reactive. Bronchial and vascular smooth muscle cells were negative, as were all other types of cells f o u n d in the biopsy specimens. MMP-1. The reaction of MMP-1 was weak to moderate in large LAM cells (Fig 2B) and negative to weak in small and epithelioid LAM cells. Vascular and bronchiolar smooth muscle cells and n o r m a l interstitial fibroblasts/myofibroblasts were weakly reactive. Endothelial cells in areas of LAM showed a negative to weak reaction; those in other areas, a weak reaction. Cuboidal epithelial cells covering proliferative lesions showed a weak reaction; o t h e r types of epithelial cells were negative or weakly reactive. Macrophages varied f r o m negative to weakly reactive. T h e matrix in LAM lesions showed a negative to weak reaction, whereas in o t h e r areas it was weakly reactive. MMP-2. Large LAM cells showed a m o d e r a t e re1073
action for MMP-2 (Fig 2C). The other two types of LAM cells generally were less intensely reactive than the large LAM cells (Fig 2D). N o r m a l bronchiolar and vascular smooth muscle cells were weakly reactive; however, fibroblasts/myofibroblasts were strongly reactive. Endothelial cells in b o t h LAM lesions and other areas showed a m o d e r a t e to strong reaction. T h e reaction was moderate in bronchiolar ciliated cells and type I pneumocytes, weak in type II pneumocytes, and m o d e r a t e to strong in cuboidal lining cells of proliferative lesions. Tissue a n d alveolar m a c r o p h a g e s showed a weak to m o d e r a t e reaction. The matrix was unreactive. Strong staining of elastic fibers was observed in the vicinity of some areas of LAM cell proliferation (Fig 2E, F) and, focally, on the internal and external elastic laminae of small arteries. MMP-3. No reactivity for MMP-3 was observed in LAM cells or other tissue components. MMP-9. Large LAM cells were weakly reactive (Fig 3A), whereas the other two types of LAM cells
HUMAN PATHOLOGY TABLE 1.
Volume 28, No. 9 (September 1997)
I m m u n o r e a c t M t y of C o m p o n e n t s of P u l m o n a r y LAM Lesions for MMPs a n d TIMPs
Large, spindle-shaped LAM cells Smaller LAM cells Epithelioid LAM cells Vascular smooth muscle cells Bronchiolar smooth muscle cells Normal fibroblasts/myofibroblasts Endothelial cells in areas of LAM Endothelial cells in other areas Bronchiolar epithelial cells Type I pneumocytes Type II pneumocytes Cuboidal epithelial cells in areas of proliferation of LAM cells Tissue macrophages Alveolar macrophages Matrix in areas of LAM Matrix in other areas
MMP-1
MMP-2
MMP-3
MMP-9
TIMP-1
TIMP-2
1-2 0-1 0-1 1 1 1 0-1 1 0-1 0-1 0-1
2 1 1-2 l 1 3 2-3 2-3 2 2 1
0 0 0 0 0 0 0 0 0 0 0
1 0 0 1 1 0 0-1 0-2 0-1 0-1 0-1
1 0 0-1 1 1 0 0-1 1 1 0-1 l
1-2 1 1-2 1-2 1-2 0-1 0-1 2-3 0-1 0-1 O-I
1 0-1 1 0-1 1
2-3 1-2 1-2 0 0
0 0 0 0 0
1 &l 0-1 0-1 1-2
0-1 &l 0-1 0-1 1
0-1 0-1 0-1 0-1 1-2
NOTE. The intensity of the immunohistochemical staining was graded as follows: 0 = no reaction; 1 = mild or weak staining; 2 = moderate staining; 3 = intense or strong staining.
FIGURE 2. lmmunohistochemical reactivity of LAM cells for HMB-45 and MMPs. (A) HMB-ZL5antibody. Proliferating, spindle-shaped LAM cells in peribronchiolar area show a strong granular reaction for HMB-45. (Original magnification x400) (B) MMP1. Large, spindle-shaped LAM cells show a weak to moderate cytoplasmic reaction for MMP-1. The cytoplasm of endothelial ceils in this lesion in very thin and shows a weak reaction (arrowheads). (Original magnification x800). (C) MMP-2. Large, spindleshaped and small LAM cells show a moderate reaction for MMP-2. Flat and cuboidal epithelial cells covering the lesion also show a moderate reaction. (Original magnification x400) (D) MMP-2. Epithelioid LAM cells with vacuolated or nonvacuolated cytoplasm show a moderate reaction for MMP-2. (Original magnification x800). (E) MMP-2. Interstitial elastic fibers show a strong reactivity for MMP-2. A moderate reaction is present in cytoplasm of adjacent large spindle-like LAM cells. (Original magnification x700). (F) Nomarski differential interference contrast view of the same area shown in E. Area of reactivity for MMP-2 corresponds to that of wavy elastic fibers (arrowheads). (Original magnification ×700).
1074
MMPs AND TIMPs IN LAM (Hayashi et at)
FIGURE 3. Immunohistochemical reactivity of LAM cells for MMPs and TIMPs. (A) MMP-9. Large, spindleshaped LAM cells are only weakly reactive. (Original magnification x800). (B) TIMP1. Large, spindle-shaped !_AM cells and their surrounding matrix show a weak reaction, (Original magnification x800). (C) TIMP-2. Large, spindle-shaped LAM cells show a weak to moderate cytoplasmic reaction for TIMP-2. (Original magnification x700). (D) TIMP-2. Both the small and large spindle-shaped cells and their matrix in a proliferative lesion show a weak, diffuse reaction. (Original magnification ×800).
were negative. N o r m a l s m o o t h muscle cells were weakly reactive. N o r m a l interstitial fibroblasts/myofihroblasts were unreactive. Endothelial cells, b o t h in proliferative lesions and in other areas, varied in reactivity f r o m negative to moderately reactive. Cuboidal epithelial lining cells in proliferative areas showed a weak reaction, whereas o t h e r epithelial cells and tissue and alveolar m a c r o p h a g e s varied f r o m negative to weakly reactive. The extracellular matrix in areas of LAM lesions showed a negative to weak reaction, and thick collagen bundles in o t h e r interstitial areas reacted weakly. TIMP-1. Large LAM cells showed a weak reaction (Fig 3B); epithelioid LAM cells, a negative to weak reaction, and immature LAM cells, a negative reaction. Normal smooth muscle cells showed a weak reaction. Normal interstitial fibroblasts and myofibroblasts were negative. Endothelial cells showed a negative to weak reaction in areas of LAM cells and a weak reaction in other sites. Bronchiolar ciliated cells and type II pneumocytes were
weakly reactive, whereas type I pneumocytes and cuboidal lining cells were negative to weakly reactive. Tissue and alveolar macrophages varied from negative to weakly reactive. The matrix in areas of LAM and in foci of dense collagen showed a negative to weak reaction. TIMP-2. Large and epithelioid LAM cells showed a weak to m o d e r a t e reaction (Fig 3C); small LAM cells exhibited a weak reaction. Normal smooth muscle cells were weakly to moderately reactive. N o r m a l interstitial fibroblasts and myofibroblasts were negative to weakly reactive. T h e cytoplasm of endothelial cells showed a negative to weak reaction in areas of LAM a n d moderate to strong reaction in other areas; in addition, the b a s e m e n t m e m b r a n e s of endothelial cells in larger vessels showed a m o d e r a t e reaction for TIMP-2. Epithelial cells and b o t h tissue and alveolar m a c r o p h a g e s were negative to weakly reactive. T h e matrix showed a negative to weak reaction in areas of I_AM (Fig 3D) and a weak to m o d e r a t e reaction in uninvolved areas of lung.
1075
HUMAN PATHOLOGY
Volume 28, No, 9 (September 1997)
T, r
v"
D
C
FIGURE 4, Confocal microscopy images of sections stained form the immunohistochemical demonstration of type IV collagen (green fluorescence) and MMP-2 (red fluorescence), Areas of colocalization of these two components show yellow to orange fluorescence, which results from the superimposition of the red and green fluorescence. (Each, original magnification ×1,200), (A) In this area, both the LAM cells and the overlying epithelial cells show a positive reaction for MMP-2. The basement membranes of the LAM cells are poorly developed and show only a very weak reaction for type IV collagen. The basement membranes of the epithelial cells lining the area of LAM cell proliferation are well defined and continuous with those of the subjacent capillary endothelial cells. Focal areas of colocalization (yellow) of MMP-2 and type iV collagen are present in the epithelial and endothelial basement membranes. (B) In comparison with the area shown in A, the basement membranes of the LAM cells are beffer developed and show more intense reactivity for type IV collagen, as well as more extensive colocalization with MMP-2. Similar colocalization is evident in the basement membranes of epithelial and endothelial cells. (C) The colocalization of type IV collagen and MMP-2 in this area is similar to that shown in B; however, the epithelial basement membranes (arrowhead) in this area are focally disrupted. (D) Areas of colocalization of MMP-2 and type IV collagen appear orange (because of the predominance of the red fluorescence), Disruption of the epithelial basement membranes is evident (arrowheads). The basement membranes of the LAM cells are poorly developed (as in A) and show very liffle colocalization of type IV collagen and MMP-2,
D o u b l e L a b e l i n g f o r M M P - 2 a n d T y p e IV Collagen
LAM cells showed a positive cytoplasmic reaction for MMP-2 (Fig 4A-D). The reaction for type IV collagen was positive only in certain LAM cells. Variable propor-
tions of each of the three types of LAM cells gave a positive reaction for type 1V collagen; in all instances the reactivity was discontinuous and focally distributed along the cell surfaces (Figs 4A-4D). Type IV collagen on the surfaces of the LAM cells was partially colocalized with MMP-2 (Figs 4B and 4C). The basement mem-
1076
MMPs AND TIMPs IN LAM (Hayashi et al)
branes of epithelial cells covering proliferative foci also showed focal colocalization of MMP-2 and type IV collagen; however, evidence of disruption of these membranes was observed only infrequently (Fig 4C, D). Similar colocalization was observed in the basement membranes of capillaries subjacent to the epithelial lining of areas of proliferation of LAM cells.
DISCUSSION This study of MMPs and TIMPs in open lung biopsy specimens from patients with LAM provides new insights into the pathogenesis of the structural changes that occur in this unusual disorder. Our results suggest that increased activity of MMPs, particularly of MMP-2, may be responsible for the cystic destruction of the lung parenchyma in LAM. We also observed immunohistochemical heterogeneity of LAM cells, in accord with previous morphological observations, 5'15'16 with three major types of cells: large cells, small cells, and epithelioid cells. Interestingly, this morphological heterogeneity correlated with different patterns of immunohistochemical staining (Table 1). The large LAM cells stained more strongly than the other two types of LAM cells for MMP-2, MMP-9 and HMB-45. The finding of heterogeneity in the immunohistochemical staining of the LAM cells for HMB-45 was also reported previously. 5'16 The ontogenic relationships of these three types of cells remain unclear. Bonetti et al 5 have suggested that these cells represent different functional stages of the same type of cell. Previous ultrastructural studies 7 have shown that the cystic lesions of LAM are associated with damage to collagen and elastic fibers. In early lesions, the collagen fibrils adjacent to the proliferating LAM cells often had a spiraling structure, which is usually considered 7'is to indicate partial lysis or structural disorganization of the collagen. In contrast, the collagen fibrils in late lesions f o r m e d compact bundles, consistent with the developm e n t of fibrotic changes. 7 Elastic fibers in the areas of proliferation of LAM cells showed irregularly outlined amorphous components, with few or no associated micro fibrils. Some elastic fibers had many electron-dense granular deposits] Similar deposits were present on epithelial basement membranes and their associated micro fibrils, and in some collagen fibrils, r Energy-dispersive X-ray microanalysis of the LAM lesions showed the presence of calcium, iron, and phosphorus in the deposits. 7 Evidence has been presented in detail elsewhere ~8-2°to show that these alterations occur in a variety of conditions associated with damage to elastic fibers, including pulmonary emphysema, 2° pulmonary toxicity of paraquat, ~9 and floppy mitral valves, is Immunohistochemical staining 7 showed that a-l-antitrypsin, a well-known inhibitor of elastase, was localized on the elastic fibers in areas of smooth muscle cell proliferation. 7 Fukuda et al 7 proposed that the a-l-antitrypsin that they detected on elastic fibers was actually complexed with elastase and that the cystic destruction in LAM is due to an imbalance in the elastase/oe-l-antitrypsin system, similar to that involved in the pathogenesis
of emphysema. However, they did not identify the source of this elastolytic and collagenolytic activity, because they were unable to detect neutrophil elastase or pancreatic elastase-1 in the LAM lesions. Thus, the binding of o~-l-antitrypsin to damaged elastic fibers in LAM remains unexplained. The current study shows that the LAM cells have strong imlnunoreactivity for MMP-2, and less intense reactivity for MMP-1 and MMP-9. In each of the four patients, the staining for these MMPs in LAM cells was greater than that in normal bronchiolar and vascular smooth muscle. The staining patterns of bronchiolar and vascular smooth muscle cells in these four patients was similar to that which we have observed in normal lung tissue from control subjects, n To verify this in a direct manner, the staining of these control tissues was repeated, and this p r o c e d u r e was p e r f o r m e d concurrently with the staining of tissue sections from the patients with LAM. Certain elastic fibers in the LAM lesions, including some of those in bronchiolar and vascular structures as well as in the interstitium, showed immunoreactivity for MMP-2. Nevertheless, TIMP-1 and TIMP-2 were not increased in LAM cells. The immunoreactivity for MMPs and TIMPs in other types of pulmonary cells in the biopsy specimens from patients with LAM did not appear unusual in comparison to that which we have observed in control lung. 1~ Importantly, we did not observe infiltration of the lung by other cell types, such as neutrophils, eosinophils, or macrophages, which could serve as other sources of lytic enzyme activity directed against connective tissue components in the LAM lesions. Our data suggest that the cystic destruction of lung parenchyma in LAM may be attributable to an imbalance in the activities of MMPs and TIMPs, with the proliferating LAM cells representing a primary source of MMP-2. The observation in the current study concerning the localization of MMP-2 on elastic fibers is in accord with our observation on the localization of MMP2 on basement membranes. In addition to the staining of elastic fibers, we observed extracellular localization of MMP-2 in focal areas of the basement membranes of LAM cells and of some alveolar epithelial cells overlying nodules of LAM cells. Furthermore, using confocal microscopy, we detected colocalization of MMP-2 and type 1V collagen at these sites. The basement membranes of LAM cells, as shown by staining by type IV collagen, usually were thin and discontinuous. However, some of the basement membranes of alveolar epithelial cells showing colocalization of MMP-2 and type IV collagen were discontinuous, in contrast to their normal continuous appearance in other locations. These findings suggest that increased activity of MMP-2 p r o d u c e d by the LAM cells may lead to alterations in basement membranes as well as in other components of the extracellular matrix, such as elastic fibers. The greater reactivity for MMP-9 and TIMPs in the collagen matrix of the E&M lesions correlates with the small a m o u n t of collagen deposition seen in the fibrous remodeling that occurs in this disorder. It is important to correlate our findings of in-
1077
HUMAN PATHOLOGY
Volume 28, No. 9 (September 1997)
creased immunoreactivity for MMPs in LAM with currently available data on the biochemistry of these enzymes. In contrast to neutrophil elastase and pancreatic elastase, which are serine proteases, MMP-2 and MMP9 are zinc-containing metalloenzymes that do not react with o~-l-antitrypsin but are inhibited by TIMPs, with which they form complexes. Both MMP-2 and MMP-9 can degrade various components of the extracellular matrix, including elastic fibers, denatured collagens (gelatins), and a n u m b e r of types of native collagens that contain regions in which the helical structure is disrupted. Our finding of very low levels of TIMPs in areas of increased immunoreactivity for MMP-2 and MMP-9 provides support for the conclusion that the proteolytic action of these enzymes is u n o p p o s e d and could result in significant destruction of connective tissue components. In conclusion, the results of the current study suggest that elevated, u n o p p o s e d activity of MMPs can induce the tissue destruction that occurs in LAM.
REFERENCES 1. Tortes VE, BjornssonJ, King BF, et al: Extrapulmonary lymphangioleiomyomatosis and lynlphangiomatous cysts in tuberous sclerosis complex. Mayo Clin Proc 70:641-648, 1995 2. NakajimaJ, Kotsuka Y, Yagyuu K, et al: Pulmonary hamartoangiomyomatosis (lymphangiomyomatosis) in tuberous sclerosis: A case in which clinical course had been modified by previous surgical intervention. Nippon Kyobu Shikkan Gakkai Zasshi 33:80-84, 1995 3. Guinee DGJR, Feuerstein I, Koss MN, et al: Pulmonary lmyphangioleiomyomatosis: Diagnosis based on results of transbronchial biopsy and immunohistochemical studies and correlation with highresolution computed tomography findings. Arch Pathol Lab Med 118:846-849, 1994 4. Hoon V, Thung SN, Kaneko M, et al: HMB-45 reactivity in renal angiomyolipoma and lymphangioleiomyomatosis. Arch Pathol Lab Med 118:732-734, 1994 5. Bonetti F, Chiodera PL, Pea M, et al: Transbronchial biopsy in lymphangiomyomatosis of the lung: HMB45 for diagnosis. Am J Surg Pathol 17:1092-1102, 1993 6. Chan JK, Tsang WY, Pau MY, et ah Lymphangiomyomatosis
and angiomyolipoma: Closely related entities characterized by hamartomatous proliferation of HMB-45-positive smooth muscle. Histopathology 22:445-455, 1993 7. Fukuda Y, Kawamoto M, Yamamoto A, et al: Role of elastic fiber degradation in emphysema-like lesions of pulmonary lymphangiomyomatosis. HUM PATHOL 21:1252-1261, 1990 8. Stetler-Stevenson WG, Aznavoorian S, Liotta LA: Tumor cell Interactions with the extracellular matrix during invasion and metastasis. Annu Rev Cell Biol 9:541-573, 1993 9. RayJM, Stetler-Stevenson WG: The role of matrix metalloproteases and their inhibitors in tumour invasion, metastasis and angiogenesis. Eur RespirJ 7:2062-2072, 1994 10. Senoir RM, Griffin GL, Fliszar CJ, et ah Human 92-and 72kilodalton type IV collagenases are elastases. J Biol Chem 266:78707875, 1991 11. Hayashi T, Stetler-Stevenson WG, Fleming MV, et al: Immunohistochemical study of metalloproteinases and their tissue inhibitors in the lungs of patients with diffuse alveolar damage and idiopathic pulmonary fibrosis. Am J Pathol 149; 1241-1256, 1996 12. Brown PD, Levy AT, Margulies IM, et al: Independent expression and cellular processing of Mr 72,000 type IV collagenase and interstitial collagenase in human tumorigenic cell lines. Cancer Res 50:6184-6191, 1990 13. Corcoran ML, Stetler-Stevenson WG, Brown PD, et al: Interleukin 4 inhibition of prostaglandin E2 synthesis blocks interstitial collagenase and 92-kDa type IV collagenase/gelatinase production by human monocytes. J Biol Chem 267:515-519, 1992 14. Stetler-Stevenson WG, Brown PD, Onisto M, et al: Tissue inhibitor of metalloproteinases-2 (TIMP-2) mRNA expression in tumor cell lines and human tumor tissues. J Biol Chem 265:1393313938, 1990 15. Dail DH: Lymphangioleiomyomatosis, in Dail DH, Hammar SP (ed) : Pulmonary Pathology (ed 2). New York, NY, Springer-Verlag, 1994, p 1414 16. Bonetti F, Pea M, Martignoni G, et ah Cellular heterogeneity in lyanphangiomyomatosis of the lung. HUM PATHOL 22:727-278, 1991 17. Fukuda Y, Ferrans ~(], Schoenberger CI, et ah Patterns of pulmonary structural remodeling after experimental paraquat toxicity: The morphogenesis of intraalveolar fibrosis. Am J Pathol 118:452475, 1985 18. Tamura K, Fukuda Y, Ishizaki M, et ah Abnormalities in elastic fibers and other connective-tissue components of floppy mitral valve. Am HeartJ 129:1149-1159, 1995 19. Fukuda ~; Ferrans VJ: Pulmonary elastic fiber degradation in paraquat toxicity: An electron microscopic immunohistochemical study. J Submicrosc Cytol Pathol 20:15-23, 1988 20. Fukuda Y, Masuda Y, Ishizaki M, et al: Morphogenesis of abnormal elastic fibers in lungs of patients with panacinar and centriacinar emphysema. HUN PATI4OL 20:652-659, 1989
1078
rupted. These changes were not accompanied by increased immunoreactivity for TIMPs. Taken together with previous observations showing structural damage to elastic fibers and collagen fibrils, and with the absence of demonstrable neutrophil or pancreatic types of elastase, these findings suggest that MMP-2 and MMP-9 (both of which can degrade elastin as well as collagens) are responsible for the connective tissue destruction and cyst formation in LAM. HUM PATHOL 28:1071--1078. Copyright © 1997 by W.B. Saunders Company Key words: malaix metalloproteinases, tissue inhibitors of metalloproteinases, lymphangiole'mmyomatosis, immanohistochemistry; confocal microscopy. Abbreviations: LAM, lymphangioleiomyomatosis; MMP, matrix metalloproteinase; TIMP, tissue inhibitor of metalloproteinase; H& E, hematoxylin and eosin; Ig, immunoglobulin.
Pulmonary lymphangioleiomyomatosis (LAM) occurs in women of reproductive age and is characterized by proliferation of smooth muscle ceils, which are derived from lymphatic vessels, and by formation of multiple, air-filled cysts. The proliferating smooth muscle cells, herein referred to as LAM ceils, show an immunophenotype that differs from that of normal smooth muscle cells in that they react with a n t i - H M t M 5 antibody. >5 Most studies of LAM have focused on the causes of the proliferation of I A M cells, with special emphasis on h o r m o n a l effects, and little is known about the mechanisms of formation of the cystic lesions. These lesions are associated with degeneration of collagen and elastic fibers] Lysis of the extracellular matrix components, including the various types of collagen, proteoglycans, and elastic fibers, is regulated by the balance of the activities of the matrix metaUoproteinases (MMPs) and their tissue inhibitors (TIMPs).8'9 High levels of MMP-2 and MMP-9 have been associated with lysis of basement membranes and with inva-
sive growth of tumor cells, s'9 These enzymes degrade native type IV collagen, denatured collagens (gelatin), and a n u m b e r of types of native collagens that contain regions in which the helical structure of the protein is disrupted. 8'9 They also degrade elastic fibers, t° Recent work in our laboratory has shown the importance of MMPs and TIMPs in the pathogenesis of destructive and fibrotic pulmonary lesions. In idiopathic pulmonary fibrosis, diffuse alveolar damage, and pulmonary Langerhans cell granulomatosis, increased activity of MMPs was found to occur in destructive lesions, whereas increased activity of TIMPs, particularly of TIMP-2, coexisted with fibrotic changes. 1~ In view of these findings, the current study was u n d e r t a k e n to determine whether abnormalities in the activities of MMPs and TIMPs are associated with the proliferative and cystic lesions of pulmonary LAM.
MATERIALS AND METHODS
Patients From the Pathology Section and Pulmonary-Critical Care Medicine Branch, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, MD; Department of Pulmonary & Mediastinal Pathology, Armed Forces Institute of Pathology, Washington, DC; and the Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, MD. Accepted for publication January 14, 1997. Address correspondence and reprint requests to Victor J. Ferrans, MD, PhD, Section Chief, Pathology Section; National Heart, Lung and Blood Institute, National Institutes of Health, Bldg 10, Rm 2N240, 10 Center Dr, MSC-1518, Bethesda, MD 20892-1518. Copyright © 1997 by W.B. Saunders Company 0046-8177/97/2809-001255.00/0
The study group consisted of four women (ranging in age from 24 to 38 years; mean age, 32 years) in whom the diagnosis of pulmonary LAM was established on the basis of clinical and radiological findings and the results of histological and immunohistochemical studies of open lung biopsy specimens.l-6
Histological Studies The open lung biopsy specimens were fixed with buffered 10% formalin, dehydrated, embedded in paraffin, and sectioned at a thickness of 5 #m. For histopathological study,
1071
HUMAN PATHOLOGY Volume28, No. 9 (September 1997) the sections were stained with hematoxylin and eosin (H&E), Masson trichrome, and Movat pentachrome methods.
Immunohistochemical Staining Paraffin sections were used to show the localization and degree of activity of HMB-45, type IV collagen, MMP-1, MMP2, MMP-3, MMP-9, TIMP-1 and TIMP-2. The antibodies against MMP-1, MMP-2, MMP-9, TIMP-1, and TIMP-2 (dilutions, 1:1,000; 1:2,000; 1:1,000; 1:4,000, and 1:1,000, respectively) were rabbit polyclonal antibodies that were prepared in the laboratory of one of us (W.S.S.). The characteristics and specificity of these antibodies have been described previously. 12-14 The antibodies against HMB-45 (DAKO Corp., Carpinteria, CA; dilution, 1:1,000), type IV collagen (DAKO, dilution, 1:50) and MMP-3 (Oncogene Science, Uniondale, NY; dilution, 1:50) were mouse monoclonal antibodies. The tissue sections were treated with 0.4% pepsin in 0.01 mol/L HCI for 15 to 30 minutes at 37°C before immunohistochemical staining. The reactions for TIMP-1 and TIMP-2 were also performed on sections not treated with pepsin, because such treatment often resulted in a decrease in the immunoreactivity for these components. Comparison of the results obtained in untreated sections and in sections treated with pepsin showed that normal basement membranes were not nonspecifically stained by the antibodies to the various MMPs and TIMPs after pepsinization, thus excluding the possibility of nonspecific adsorption of the antibodies to partially degraded type IV collagen. For single labeling, the sections were stained using the avidin-biotinylated peroxidase complex peroxidase method in conjunction with a Vector Elite Kit (Vector Laboratories, Burlingame, CA) and a Vector VIP kit to produce a violet color at the sites of reactivity. M1 sections were counterstained with Carazzi's hematoxylin. For double labeling for confocal laser scanning microscopy, the sections were stained by the double indirect immunofluorescence method, using combinations of a mouse monoclonal antibody against type IV collagen (dilution, 1:50) and a rabbit polyclonal antibody against MMP-2 (dilution, 1:400). The primary antibodies against type IV collagen and HMB-45 were reacted with a secondary antibody (horse antimouse immunoglobulin [Ig] G; Vector Laboratories; dilutions 1:100) labeled with fluorescein isothiocyanate. The primary antibody against MMP-2 and actin was reacted with a secondary antibody (goat anti-rabbit IgG; Vector Laboratories; dilutions 1:100) labeled with Texas red, followed by nuclear counterstaining with 4',6-diamidino-2-phenylindole (DAPI; Sigma Chemical, St Louis, MO). In these preparations, a yellow fluorescence indicated the colocalization of the red and the green fluorescence. Idennfication of the histopathological orientation was facilitated by nuclear staining with DAPI for confocal microscopy. All preparations were examined with a confocal microscope, Leica (Heidelberg, Germany) model TCS4D/DMIRBE, equipped with argon and argon-krypton lasers.
Immunohistochemical Control Procedures Negative control immunohistochemical procedures included (1) omission of the primary antibody from the staining protocol and (2) replacement of the primary antibody by normal rabbit or mouse IgG in appropriate concentrations.
RESULTS Histopathological Findings The lung biopsy specimens showed a combination of cysts and nodules of proliferating LAM cells (Fig
1A). T h e cysts varied in size a m o n g the four biopsy specimens. The larger cysts were lined mainly by thin, flat epithelial cells, whereas the smaller cysts were lined by mixtures of bronchiolar ciliated epithelial cells, hyperplastic cuboidal epithelial cells, and flattened epithelial cells. Similar types of epithelial cells f o r m e d the lining of large nodules of proliferating LAM cells. T h e smaller nodules were lined mostly by ciliated epithelial cells. T h e LAM cells were mainly distributed a r o u n d bronchioles and blood vessels and in the walls of the cysts. Small amounts of dense collagen were present at the periphery of the nodules. The smooth muscle cells in the nodules were haphazardly arranged, unlike the normal muscle layers of bronchioles and blood vessels. T h r e e types of LAM cells were identified in the LAM lesions in each of the four patients: relatively large, spindle-shaped cells (Fig 1B); small, slightly elongated cells that contained only small amounts of cytoplasm; and large, epithelioid cells (Fig 1C). T h e large, spindleshaped cells had r o u n d or oval nuclei with finely granular chromatin and occasional infoldings of the nuclear m e m b r a n e s . These cells h a d moderately a b u n d a n t amounts of cytoplasm and indistinct cell borders. T h e smaller LAM cells had only small amounts of cytoplasm and contained r o u n d to oval nuclei with dense, h o m o geneous chromatin and indistinct nucleoli. T h e epithelioid LAM cells were r o u n d to polygonal in shape, had r o u n d to oval nuclei with occasional nucleoli, and abundant, often vacuolated cytoplasm. In sections stained with H&E, all three types of cells a p p e a r e d paler than normal smooth muscle cells, although the epithelioid cells were m o r e eosinophilic than the o t h e r two types of LAM cells. The biopsy specimen f r o m patient 1 showed multiple cysts and nodules containing mixtures of all three types of smooth muscle cells. T h e biopsy specimen f r o m patient 2 showed only mild cystic changes and multiple nodules c o m p o s e d of a p r e d o m i n a n c e of small LAM cells. T h e biopsy specimen f r o m patient 3 was characterized by p r o m i n e n t cysts, which were s u r r o u n d e d by areas of cellular proliferation with n u m e r o u s epithelioid cells. The biopsy specimen f r o m patient 4 showed small cysts and only small nodules, which contained mixtures of large and small LAM cells. All biopsy specimens had m a c r o p h a g e s in the alveolar spaces and lymphocytes in the interstitium, hut both of these types of cells were inconspicuous; neutrophils were not found. T h e Movat stain highlighted areas where LAM cell infiltrates were associated with destruction of elastic fibers in the pleura or bronchiolar walls (Fig 1D).
Immunoperoxidase Staining T h e i m m u n o h i s t o c h e m i c a l findings were similar in all four cases. Data on the immunoreactivity of the different cellular and extracellular c o m p o n e n t s of the lungs for MMPs and TIMPs are summarized in Table 1. HMB-45. Large L~_M cells showed a strong granular reaction for HMB-45 (Fig 2A). T h e staining reaction for HMB-45 was focally distributed in the cytoplasm of
1072
MMPs AND TIMPs IN LAM (Hayashi et al)
FIGURE I. Histopathologica[ findings in pulmonary LAM (A) Low-magnification view showing cystic dilatation and nodular proliferation of LAM cells in pulmonan/tissue. Cleftlike spaces in the nodular lesions represent lymphatic vessels (arrowheads). (H & Estain; original magnification ×100.) (B) Large, spindle-like proliferating LAM cells form an interlacing paffern, These cells have indistinct borders, moderately abundant amounts of cytoplasm, and oval to elongated nuclei with finely granular chromatin and distinct nucleoli, (H & E stain; original magnification x400). (C) LAM cells of the smaller type have scanty cytoplasm. Their nuclei have dense, homogeneous chromatin and indistinct nucleoli. These cells are arranged in poorly defined fascicular patterns. A small cluster of epithelioid I_AM cells with round nuclei and vacuoiated cytoplasm is also present (arrowhead). (H & E stain; original magnification x400). (D) Bronchiolar elastic fibers adjacent to proliferating LAM cells are disrupted (arrowhead), (Movat pentachrome stain; original magnification x400).
these cells. Small LAM cells showed a negative to weak reaction. Epithelioid LAM cells were moderately to strongly reactive. Bronchial and vascular smooth muscle cells were negative, as were all other types of cells f o u n d in the biopsy specimens. MMP-1. The reaction of MMP-1 was weak to moderate in large LAM cells (Fig 2B) and negative to weak in small and epithelioid LAM cells. Vascular and bronchiolar smooth muscle cells and n o r m a l interstitial fibroblasts/myofibroblasts were weakly reactive. Endothelial cells in areas of LAM showed a negative to weak reaction; those in other areas, a weak reaction. Cuboidal epithelial cells covering proliferative lesions showed a weak reaction; o t h e r types of epithelial cells were negative or weakly reactive. Macrophages varied f r o m negative to weakly reactive. T h e matrix in LAM lesions showed a negative to weak reaction, whereas in o t h e r areas it was weakly reactive. MMP-2. Large LAM cells showed a m o d e r a t e re1073
action for MMP-2 (Fig 2C). The other two types of LAM cells generally were less intensely reactive than the large LAM cells (Fig 2D). N o r m a l bronchiolar and vascular smooth muscle cells were weakly reactive; however, fibroblasts/myofibroblasts were strongly reactive. Endothelial cells in b o t h LAM lesions and other areas showed a m o d e r a t e to strong reaction. T h e reaction was moderate in bronchiolar ciliated cells and type I pneumocytes, weak in type II pneumocytes, and m o d e r a t e to strong in cuboidal lining cells of proliferative lesions. Tissue a n d alveolar m a c r o p h a g e s showed a weak to m o d e r a t e reaction. The matrix was unreactive. Strong staining of elastic fibers was observed in the vicinity of some areas of LAM cell proliferation (Fig 2E, F) and, focally, on the internal and external elastic laminae of small arteries. MMP-3. No reactivity for MMP-3 was observed in LAM cells or other tissue components. MMP-9. Large LAM cells were weakly reactive (Fig 3A), whereas the other two types of LAM cells
HUMAN PATHOLOGY TABLE 1.
Volume 28, No. 9 (September 1997)
I m m u n o r e a c t M t y of C o m p o n e n t s of P u l m o n a r y LAM Lesions for MMPs a n d TIMPs
Large, spindle-shaped LAM cells Smaller LAM cells Epithelioid LAM cells Vascular smooth muscle cells Bronchiolar smooth muscle cells Normal fibroblasts/myofibroblasts Endothelial cells in areas of LAM Endothelial cells in other areas Bronchiolar epithelial cells Type I pneumocytes Type II pneumocytes Cuboidal epithelial cells in areas of proliferation of LAM cells Tissue macrophages Alveolar macrophages Matrix in areas of LAM Matrix in other areas
MMP-1
MMP-2
MMP-3
MMP-9
TIMP-1
TIMP-2
1-2 0-1 0-1 1 1 1 0-1 1 0-1 0-1 0-1
2 1 1-2 l 1 3 2-3 2-3 2 2 1
0 0 0 0 0 0 0 0 0 0 0
1 0 0 1 1 0 0-1 0-2 0-1 0-1 0-1
1 0 0-1 1 1 0 0-1 1 1 0-1 l
1-2 1 1-2 1-2 1-2 0-1 0-1 2-3 0-1 0-1 O-I
1 0-1 1 0-1 1
2-3 1-2 1-2 0 0
0 0 0 0 0
1 &l 0-1 0-1 1-2
0-1 &l 0-1 0-1 1
0-1 0-1 0-1 0-1 1-2
NOTE. The intensity of the immunohistochemical staining was graded as follows: 0 = no reaction; 1 = mild or weak staining; 2 = moderate staining; 3 = intense or strong staining.
FIGURE 2. lmmunohistochemical reactivity of LAM cells for HMB-45 and MMPs. (A) HMB-ZL5antibody. Proliferating, spindle-shaped LAM cells in peribronchiolar area show a strong granular reaction for HMB-45. (Original magnification x400) (B) MMP1. Large, spindle-shaped LAM cells show a weak to moderate cytoplasmic reaction for MMP-1. The cytoplasm of endothelial ceils in this lesion in very thin and shows a weak reaction (arrowheads). (Original magnification x800). (C) MMP-2. Large, spindleshaped and small LAM cells show a moderate reaction for MMP-2. Flat and cuboidal epithelial cells covering the lesion also show a moderate reaction. (Original magnification x400) (D) MMP-2. Epithelioid LAM cells with vacuolated or nonvacuolated cytoplasm show a moderate reaction for MMP-2. (Original magnification x800). (E) MMP-2. Interstitial elastic fibers show a strong reactivity for MMP-2. A moderate reaction is present in cytoplasm of adjacent large spindle-like LAM cells. (Original magnification x700). (F) Nomarski differential interference contrast view of the same area shown in E. Area of reactivity for MMP-2 corresponds to that of wavy elastic fibers (arrowheads). (Original magnification ×700).
1074
MMPs AND TIMPs IN LAM (Hayashi et at)
FIGURE 3. Immunohistochemical reactivity of LAM cells for MMPs and TIMPs. (A) MMP-9. Large, spindleshaped LAM cells are only weakly reactive. (Original magnification x800). (B) TIMP1. Large, spindle-shaped !_AM cells and their surrounding matrix show a weak reaction, (Original magnification x800). (C) TIMP-2. Large, spindle-shaped LAM cells show a weak to moderate cytoplasmic reaction for TIMP-2. (Original magnification x700). (D) TIMP-2. Both the small and large spindle-shaped cells and their matrix in a proliferative lesion show a weak, diffuse reaction. (Original magnification ×800).
were negative. N o r m a l s m o o t h muscle cells were weakly reactive. N o r m a l interstitial fibroblasts/myofihroblasts were unreactive. Endothelial cells, b o t h in proliferative lesions and in other areas, varied in reactivity f r o m negative to moderately reactive. Cuboidal epithelial lining cells in proliferative areas showed a weak reaction, whereas o t h e r epithelial cells and tissue and alveolar m a c r o p h a g e s varied f r o m negative to weakly reactive. The extracellular matrix in areas of LAM lesions showed a negative to weak reaction, and thick collagen bundles in o t h e r interstitial areas reacted weakly. TIMP-1. Large LAM cells showed a weak reaction (Fig 3B); epithelioid LAM cells, a negative to weak reaction, and immature LAM cells, a negative reaction. Normal smooth muscle cells showed a weak reaction. Normal interstitial fibroblasts and myofibroblasts were negative. Endothelial cells showed a negative to weak reaction in areas of LAM cells and a weak reaction in other sites. Bronchiolar ciliated cells and type II pneumocytes were
weakly reactive, whereas type I pneumocytes and cuboidal lining cells were negative to weakly reactive. Tissue and alveolar macrophages varied from negative to weakly reactive. The matrix in areas of LAM and in foci of dense collagen showed a negative to weak reaction. TIMP-2. Large and epithelioid LAM cells showed a weak to m o d e r a t e reaction (Fig 3C); small LAM cells exhibited a weak reaction. Normal smooth muscle cells were weakly to moderately reactive. N o r m a l interstitial fibroblasts and myofibroblasts were negative to weakly reactive. T h e cytoplasm of endothelial cells showed a negative to weak reaction in areas of LAM a n d moderate to strong reaction in other areas; in addition, the b a s e m e n t m e m b r a n e s of endothelial cells in larger vessels showed a m o d e r a t e reaction for TIMP-2. Epithelial cells and b o t h tissue and alveolar m a c r o p h a g e s were negative to weakly reactive. T h e matrix showed a negative to weak reaction in areas of I_AM (Fig 3D) and a weak to m o d e r a t e reaction in uninvolved areas of lung.
1075
HUMAN PATHOLOGY
Volume 28, No, 9 (September 1997)
T, r
v"
D
C
FIGURE 4, Confocal microscopy images of sections stained form the immunohistochemical demonstration of type IV collagen (green fluorescence) and MMP-2 (red fluorescence), Areas of colocalization of these two components show yellow to orange fluorescence, which results from the superimposition of the red and green fluorescence. (Each, original magnification ×1,200), (A) In this area, both the LAM cells and the overlying epithelial cells show a positive reaction for MMP-2. The basement membranes of the LAM cells are poorly developed and show only a very weak reaction for type IV collagen. The basement membranes of the epithelial cells lining the area of LAM cell proliferation are well defined and continuous with those of the subjacent capillary endothelial cells. Focal areas of colocalization (yellow) of MMP-2 and type iV collagen are present in the epithelial and endothelial basement membranes. (B) In comparison with the area shown in A, the basement membranes of the LAM cells are beffer developed and show more intense reactivity for type IV collagen, as well as more extensive colocalization with MMP-2. Similar colocalization is evident in the basement membranes of epithelial and endothelial cells. (C) The colocalization of type IV collagen and MMP-2 in this area is similar to that shown in B; however, the epithelial basement membranes (arrowhead) in this area are focally disrupted. (D) Areas of colocalization of MMP-2 and type IV collagen appear orange (because of the predominance of the red fluorescence), Disruption of the epithelial basement membranes is evident (arrowheads). The basement membranes of the LAM cells are poorly developed (as in A) and show very liffle colocalization of type IV collagen and MMP-2,
D o u b l e L a b e l i n g f o r M M P - 2 a n d T y p e IV Collagen
LAM cells showed a positive cytoplasmic reaction for MMP-2 (Fig 4A-D). The reaction for type IV collagen was positive only in certain LAM cells. Variable propor-
tions of each of the three types of LAM cells gave a positive reaction for type 1V collagen; in all instances the reactivity was discontinuous and focally distributed along the cell surfaces (Figs 4A-4D). Type IV collagen on the surfaces of the LAM cells was partially colocalized with MMP-2 (Figs 4B and 4C). The basement mem-
1076
MMPs AND TIMPs IN LAM (Hayashi et al)
branes of epithelial cells covering proliferative foci also showed focal colocalization of MMP-2 and type IV collagen; however, evidence of disruption of these membranes was observed only infrequently (Fig 4C, D). Similar colocalization was observed in the basement membranes of capillaries subjacent to the epithelial lining of areas of proliferation of LAM cells.
DISCUSSION This study of MMPs and TIMPs in open lung biopsy specimens from patients with LAM provides new insights into the pathogenesis of the structural changes that occur in this unusual disorder. Our results suggest that increased activity of MMPs, particularly of MMP-2, may be responsible for the cystic destruction of the lung parenchyma in LAM. We also observed immunohistochemical heterogeneity of LAM cells, in accord with previous morphological observations, 5'15'16 with three major types of cells: large cells, small cells, and epithelioid cells. Interestingly, this morphological heterogeneity correlated with different patterns of immunohistochemical staining (Table 1). The large LAM cells stained more strongly than the other two types of LAM cells for MMP-2, MMP-9 and HMB-45. The finding of heterogeneity in the immunohistochemical staining of the LAM cells for HMB-45 was also reported previously. 5'16 The ontogenic relationships of these three types of cells remain unclear. Bonetti et al 5 have suggested that these cells represent different functional stages of the same type of cell. Previous ultrastructural studies 7 have shown that the cystic lesions of LAM are associated with damage to collagen and elastic fibers. In early lesions, the collagen fibrils adjacent to the proliferating LAM cells often had a spiraling structure, which is usually considered 7'is to indicate partial lysis or structural disorganization of the collagen. In contrast, the collagen fibrils in late lesions f o r m e d compact bundles, consistent with the developm e n t of fibrotic changes. 7 Elastic fibers in the areas of proliferation of LAM cells showed irregularly outlined amorphous components, with few or no associated micro fibrils. Some elastic fibers had many electron-dense granular deposits] Similar deposits were present on epithelial basement membranes and their associated micro fibrils, and in some collagen fibrils, r Energy-dispersive X-ray microanalysis of the LAM lesions showed the presence of calcium, iron, and phosphorus in the deposits. 7 Evidence has been presented in detail elsewhere ~8-2°to show that these alterations occur in a variety of conditions associated with damage to elastic fibers, including pulmonary emphysema, 2° pulmonary toxicity of paraquat, ~9 and floppy mitral valves, is Immunohistochemical staining 7 showed that a-l-antitrypsin, a well-known inhibitor of elastase, was localized on the elastic fibers in areas of smooth muscle cell proliferation. 7 Fukuda et al 7 proposed that the a-l-antitrypsin that they detected on elastic fibers was actually complexed with elastase and that the cystic destruction in LAM is due to an imbalance in the elastase/oe-l-antitrypsin system, similar to that involved in the pathogenesis
of emphysema. However, they did not identify the source of this elastolytic and collagenolytic activity, because they were unable to detect neutrophil elastase or pancreatic elastase-1 in the LAM lesions. Thus, the binding of o~-l-antitrypsin to damaged elastic fibers in LAM remains unexplained. The current study shows that the LAM cells have strong imlnunoreactivity for MMP-2, and less intense reactivity for MMP-1 and MMP-9. In each of the four patients, the staining for these MMPs in LAM cells was greater than that in normal bronchiolar and vascular smooth muscle. The staining patterns of bronchiolar and vascular smooth muscle cells in these four patients was similar to that which we have observed in normal lung tissue from control subjects, n To verify this in a direct manner, the staining of these control tissues was repeated, and this p r o c e d u r e was p e r f o r m e d concurrently with the staining of tissue sections from the patients with LAM. Certain elastic fibers in the LAM lesions, including some of those in bronchiolar and vascular structures as well as in the interstitium, showed immunoreactivity for MMP-2. Nevertheless, TIMP-1 and TIMP-2 were not increased in LAM cells. The immunoreactivity for MMPs and TIMPs in other types of pulmonary cells in the biopsy specimens from patients with LAM did not appear unusual in comparison to that which we have observed in control lung. 1~ Importantly, we did not observe infiltration of the lung by other cell types, such as neutrophils, eosinophils, or macrophages, which could serve as other sources of lytic enzyme activity directed against connective tissue components in the LAM lesions. Our data suggest that the cystic destruction of lung parenchyma in LAM may be attributable to an imbalance in the activities of MMPs and TIMPs, with the proliferating LAM cells representing a primary source of MMP-2. The observation in the current study concerning the localization of MMP-2 on elastic fibers is in accord with our observation on the localization of MMP2 on basement membranes. In addition to the staining of elastic fibers, we observed extracellular localization of MMP-2 in focal areas of the basement membranes of LAM cells and of some alveolar epithelial cells overlying nodules of LAM cells. Furthermore, using confocal microscopy, we detected colocalization of MMP-2 and type 1V collagen at these sites. The basement membranes of LAM cells, as shown by staining by type IV collagen, usually were thin and discontinuous. However, some of the basement membranes of alveolar epithelial cells showing colocalization of MMP-2 and type IV collagen were discontinuous, in contrast to their normal continuous appearance in other locations. These findings suggest that increased activity of MMP-2 p r o d u c e d by the LAM cells may lead to alterations in basement membranes as well as in other components of the extracellular matrix, such as elastic fibers. The greater reactivity for MMP-9 and TIMPs in the collagen matrix of the E&M lesions correlates with the small a m o u n t of collagen deposition seen in the fibrous remodeling that occurs in this disorder. It is important to correlate our findings of in-
1077
HUMAN PATHOLOGY
Volume 28, No. 9 (September 1997)
creased immunoreactivity for MMPs in LAM with currently available data on the biochemistry of these enzymes. In contrast to neutrophil elastase and pancreatic elastase, which are serine proteases, MMP-2 and MMP9 are zinc-containing metalloenzymes that do not react with o~-l-antitrypsin but are inhibited by TIMPs, with which they form complexes. Both MMP-2 and MMP-9 can degrade various components of the extracellular matrix, including elastic fibers, denatured collagens (gelatins), and a n u m b e r of types of native collagens that contain regions in which the helical structure is disrupted. Our finding of very low levels of TIMPs in areas of increased immunoreactivity for MMP-2 and MMP-9 provides support for the conclusion that the proteolytic action of these enzymes is u n o p p o s e d and could result in significant destruction of connective tissue components. In conclusion, the results of the current study suggest that elevated, u n o p p o s e d activity of MMPs can induce the tissue destruction that occurs in LAM.
REFERENCES 1. Tortes VE, BjornssonJ, King BF, et al: Extrapulmonary lymphangioleiomyomatosis and lynlphangiomatous cysts in tuberous sclerosis complex. Mayo Clin Proc 70:641-648, 1995 2. NakajimaJ, Kotsuka Y, Yagyuu K, et al: Pulmonary hamartoangiomyomatosis (lymphangiomyomatosis) in tuberous sclerosis: A case in which clinical course had been modified by previous surgical intervention. Nippon Kyobu Shikkan Gakkai Zasshi 33:80-84, 1995 3. Guinee DGJR, Feuerstein I, Koss MN, et al: Pulmonary lmyphangioleiomyomatosis: Diagnosis based on results of transbronchial biopsy and immunohistochemical studies and correlation with highresolution computed tomography findings. Arch Pathol Lab Med 118:846-849, 1994 4. Hoon V, Thung SN, Kaneko M, et al: HMB-45 reactivity in renal angiomyolipoma and lymphangioleiomyomatosis. Arch Pathol Lab Med 118:732-734, 1994 5. Bonetti F, Chiodera PL, Pea M, et al: Transbronchial biopsy in lymphangiomyomatosis of the lung: HMB45 for diagnosis. Am J Surg Pathol 17:1092-1102, 1993 6. Chan JK, Tsang WY, Pau MY, et ah Lymphangiomyomatosis
and angiomyolipoma: Closely related entities characterized by hamartomatous proliferation of HMB-45-positive smooth muscle. Histopathology 22:445-455, 1993 7. Fukuda Y, Kawamoto M, Yamamoto A, et al: Role of elastic fiber degradation in emphysema-like lesions of pulmonary lymphangiomyomatosis. HUM PATHOL 21:1252-1261, 1990 8. Stetler-Stevenson WG, Aznavoorian S, Liotta LA: Tumor cell Interactions with the extracellular matrix during invasion and metastasis. Annu Rev Cell Biol 9:541-573, 1993 9. RayJM, Stetler-Stevenson WG: The role of matrix metalloproteases and their inhibitors in tumour invasion, metastasis and angiogenesis. Eur RespirJ 7:2062-2072, 1994 10. Senoir RM, Griffin GL, Fliszar CJ, et ah Human 92-and 72kilodalton type IV collagenases are elastases. J Biol Chem 266:78707875, 1991 11. Hayashi T, Stetler-Stevenson WG, Fleming MV, et al: Immunohistochemical study of metalloproteinases and their tissue inhibitors in the lungs of patients with diffuse alveolar damage and idiopathic pulmonary fibrosis. Am J Pathol 149; 1241-1256, 1996 12. Brown PD, Levy AT, Margulies IM, et al: Independent expression and cellular processing of Mr 72,000 type IV collagenase and interstitial collagenase in human tumorigenic cell lines. Cancer Res 50:6184-6191, 1990 13. Corcoran ML, Stetler-Stevenson WG, Brown PD, et al: Interleukin 4 inhibition of prostaglandin E2 synthesis blocks interstitial collagenase and 92-kDa type IV collagenase/gelatinase production by human monocytes. J Biol Chem 267:515-519, 1992 14. Stetler-Stevenson WG, Brown PD, Onisto M, et al: Tissue inhibitor of metalloproteinases-2 (TIMP-2) mRNA expression in tumor cell lines and human tumor tissues. J Biol Chem 265:1393313938, 1990 15. Dail DH: Lymphangioleiomyomatosis, in Dail DH, Hammar SP (ed) : Pulmonary Pathology (ed 2). New York, NY, Springer-Verlag, 1994, p 1414 16. Bonetti F, Pea M, Martignoni G, et ah Cellular heterogeneity in lyanphangiomyomatosis of the lung. HUM PATHOL 22:727-278, 1991 17. Fukuda Y, Ferrans ~(], Schoenberger CI, et ah Patterns of pulmonary structural remodeling after experimental paraquat toxicity: The morphogenesis of intraalveolar fibrosis. Am J Pathol 118:452475, 1985 18. Tamura K, Fukuda Y, Ishizaki M, et ah Abnormalities in elastic fibers and other connective-tissue components of floppy mitral valve. Am HeartJ 129:1149-1159, 1995 19. Fukuda ~; Ferrans VJ: Pulmonary elastic fiber degradation in paraquat toxicity: An electron microscopic immunohistochemical study. J Submicrosc Cytol Pathol 20:15-23, 1988 20. Fukuda Y, Masuda Y, Ishizaki M, et al: Morphogenesis of abnormal elastic fibers in lungs of patients with panacinar and centriacinar emphysema. HUN PATI4OL 20:652-659, 1989
1078