Current Eye Research
Volurnc 8 number 10 1989
Ultrastructural immunocytochemical localization of elastin in normal human trabecular meshwork Haiyan
Vickery Trinkaus-Randall* and Thomas F.Freddo2
Departments of 'Anatomy and 20phthalmology, Boston University School of Medicine, Boston, MA 021 18, USA Received o n Junc 5 , 1989; acccptcd on August 21, 1989
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
ABSTRACT
Previous studies have suggested that hydrophobic moieties within the aqueous outflow channels might interact with certain aqueous components to retard outflow. While elastin is among the most hydrophobic proteins in the trabecular meshwork, it reacts poorly with conventional ultrastructural staining methods, so its potential role in regulating outflow could not be assessed. It was our goal to specifically localize elastin ultrastructurally using polyclonal antibodies against alpha elastin and its soluble precursor, tropoelastin. Human aorta served as a positive control. Preadsorption of the primary antibodies or their substitution with either normal rabbit serum or Tris buffer resulted in negligible labelling. With either antibody, only the electron-lucent elements in the center of elastic fibers of the trabecular meshwork were labelled, indicating that only these elements truly represent elastin. The pattern of elastin distribution within these fibers is most consistent with that found in tendons elsewhere in the body. INTRODUCTION
Previous studies have shown that bovine aqueous humor obstructs flow through micro-porous filters with pore sizes similar to those found in the juxtacanalicular connective tissue ( J C T ) of the trabecular meshwork. The obstruction of these filters was eliminated using proteases but not hyaluronidase, suggesting that proteins or glycoproteins were implicated in the process(1). Ethier et a1 ( 2 ) investigated this phenomenon further and found that hydrophobic filters were more prone to obstruction than hydrophilic filters. They also found that the ionic detergent Triton-X eliminated
0 IRL Press
the filter obstruction phenomenon. From these studies, the authors concluded that the process of filter obstruction likely involved hydrophobic interactions between proteins in the aqueous humor and the filter surfaces. If this work on nonphysiologic filters is at all predictive of the behavior of aqueous humor in-vivo, it suggests that specific identification and localization of hydrophobic moieties in the extracellular matrix of the trabecular meshwork will advance our understanding of the factors which affect trabecular flow in normal and glaucomatous eyes. One of the most hydrophobic proteins in the trabecular meshwork is elastin. Elastic-like fibers were originally demonstrated in the human trabecular meshwork by Salzmann (1912)(3), but the question of whether elastin is truly present in these fibers has been controversial (4-9) Ultrastructural studies of these fibers have demonstrated that they contain two morphologically distinct zones, a central amorphous component and a peripheral fibrillar component ( 7 - 8 ) . Pursuing the matter further, Lutjen-Drecoll , et a1 (10) demonstrated that only a small amount of elastase-digestible material exists in these fibers and thus concluded that they should be called llelastic-likell. Most recently, Murphy, et a1 (11), using an indirect immunofluorescent antibody protocol, localized elastin in the
.
1071
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research trabecular meshwork at the light microscopic level. No fluorescent staining could be identified in any portion of the JCT. Fluorescent staining was only found within the osmiophilic deposits which occupy the cores of the trabecular beams. Unfortunately, no specific correspondence could be made between the known ultrastructural subcomponents of these deposits and the distribution of elastin within them. In the present study, polyclonal antibodies raised against human aortic alphaelastin and bovine tropoelastin were used in a post-embedding colloidal gold protocol to more specifically localize the distribution of elastin in normal human trabecular meshwork at the electron microscopic level. MATERIALS AND METHODS Eight normal, eye bank eyes, age 1 day-83 years, were used. The anterior segments of the eyes were cut radially into small pieces. A piece of human aorta, derived from autopsy was used as a positive control. The specimens were fixed in 2% paraformaldehyde-2.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.3, containing 0.2% CaCl,, for 3 hr at room temperature. They were rinsed in the same buffer and postfixed in 2% osmium tetroxide and 1.5% potassium ferrocyanide for 2 hr. The specimens were dehydrated in a graded series of ethanol and embedded in Epon-Araldite. Semi-thin sections for light microscopy were made to examine the orientation of the specimens. Two different planes of thin sections, radial and coronal, were cut for electron microscopy and mounted on nickel grids. Post-embedding immuno-histochemistry was performed using a modification of the protein-A, colloidal gold technique (12). Grids were floated, specimen-side down,
1072
onto drops of the following series of reagents at room temperature. A l l sections were incubated in 4% nonfat dry milk for 30 min in order to block non-specific binding (13), and then transferred onto drops of antiserum against human aorta elastin or bovine tropoelastin (Elastin Products Company, Pacific, MI), diluted 1:lOO with 0.1 M Tris-1% BSA for 2 hr. After rinsing with 0.05 M Tris, pH 7.4 and 0.05 Tris-0.2% BSA, pH 7.4, grids were incubated for 1 hr in a solution containing colloidal gold-labelled IgG (Janssen Pharmaceutica, Piscataway , NJ) , diluted 1:12 with 0.05 M Tris-1% BSA, pH 8.3. The sections were then rinsed in 0.05 M Tris-0.2% BSA, 0.05 M Tris and then in distilled water. All specimens were counter-stained with uranyl acetate and lead citrate and examined in a Philips-300 Electron Microscope (Eindhoven, The Netherlands) Control studies included: (1) substitution of the primary antibodies with either normal rabbit serum or Tris buffer, (2) preincubation of antibodies with tropoelastin or alpha-elastin and (3) use of human aorta as a positive control. Studies with elastase were also conducted. Sections were incubated with pancreatic, porcine elastase (600 U/ml, Type 111, Sigma, St. Louis, MO) in 0.5 M Tris, p~ 8.8 for 1 hr. at room temperature.
.
RESULTS The distributions of both the alphaelastin and tropoelastin antibodies were virtually identical in the human trabecular meshwork. Although morphometric studies were not completed, there was no apparent difference in the amount of colloidal gold present with either antibody. Adult eyes Trabecular beams. In radial sections, both antibodies were localized within
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 1: The antibody to tropoelastin is evident within the central electron-lucent area of the elastic fibers in a radial
section of the adult trabecular beams. E = endothelium; ITS = Intertrabecular space. (X41,OOO)
Figure 2: The tropoelastin antibody is evident almost entirely within the central, electron-lucent core of the
elastic fibers when they are cut longitudinally. (x41,OOO)
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research areas which appeared as electron-lucent dots dispersed within the electron-dense areas of the trabecular beams (Fig 1). In coronal sections, the elastic fibers were cut longitudinally or obliquely. If the sections included the centers of the elastic fibers, the electron-lucent areas appeared as a central core enclosed by electron-dense microfibrillar components, which were, in turn, surrounded by sheath material. The antibodies only labelled the electron-lucent central core. Neither the microfibrillar components nor sheath
material was labelled (Fig 2). JCT area. Both antibodies labelled the electron-lucent areas in the cribriform plexus (Fig 3 ) . A s in the beams, when these fibers were cut longitudinally, both antibodies were localized principally within the central core of the elastic fiber. The microfibrils and sheath materials were not labelled either. Infant eves Trabecular beams. Both antibodies were again localized to the amorphous, electron-lucent, discrete dots within each
Figure 3 : The antibody to tropoelastin is evident within the electron-lucent area of elastic fibers in the region of
juxtacanalicular connective tissue (JCT). SC = Schlemm's canal. (X41,OOO)
1074
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research microfibrillar bundle. The individual microfibrillar elements were more clearly discernable in infant eyes than in the adult (Fig 4 ) . Interestingly, the sheaths which are known to surround the elastic fibers of the meshwork in the adult eye were not observed in either radial or coronal sections of two, one day-old infant eyes (Fig 4 ) . JCT area. Both antibodies labelled the amorphous electron-lucent dots within the microfibrillar bundles. The microfibrils appeared as described above but were more
irregularly distributed. These fibers, like those of the trabecular beams, lacked sheaths as well (Fig 5). The amorphous elastin component was not completely enclosed by microfibrils in some areas, characteristic of immature elastic fibers. Control studies 1. Substitution of the primary antibody with either normal rabbit serum or Tris buffer resulted in negligible colloidal gold labelling (Fig 6A). 2 . Preincubation of antibody with either alpha elastin or tropoelastin gave
Figure 4 : The antibody to tropoelastin is evident within the central, amorphous, electron-lucent areas of elastic fibers in cross-section of a one day-old infant eye.
The surrounding sheath material, found in the adult eye, is not present in the infant eye. (x41,600)
1075
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 5 : The antibody to alpha elastin is evident principally within the electronlucent areas of the elastic fibers in the
JCT region of one day-old infant eye. SC Schlemm's canal. (X41,600)
similar negative results (Fig 6B). 3 . Because it is known that elastic arteries contain elastin, we used human aorta as a positive control. Both antibodies were localized to the electronlucent elastin laminae within the wall of this vessel. The thin microfibrillar envelope was not labelled (Fig 7 ) . Elastase disestion Elastase treatment only affected the central electron-lucent area of the elastic fibers in the trabecular meshwork. The density of the colloidal gold labelling was reduced but not eliminated
after elastase digestion (Fig 8). Similar results were obtained in human aorta.
1076
=
DISCUSSION Elastin is the principal extracellular matrix constituent responsible for the elasticity of connective tissue. Due to its largely hydrophobic nature, the protein reacts poorly with conventional ultrastructural staining methods (14). This problem becomes more apparent in studying elastic fibers in the trabecular meshwork, since, unlike the elastic tissue in the aorta where the amorphous component
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 6 : Oblique sections of adult trabecular beams. A. Substitution of the primary antibody with normal rabbit serum yielded negligible colloidal gold
labelling (X26,240). B. Preincubation of antibody with alpha elastin gave similar negative results. (X26,240)
represents greater than 90% of the mature
amount of elastase-digestible material
elastic fiber (15), elastic fibers in trabecular meshwork only contain a small
(10)
Several immunomicroscopic techniques 1077
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 7: The tropoelastin antibody is clearly localized to the electron-lucent elastin laminae in human aorta.(X41,600)
Figure 8: Elastase only affects the electron-lucent area of the elastic fibers in tie trabecular beams. The density of
the colloidal gold particles appeared to have been reduced after elastase treatment. (X41,600)
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research have been developed to more effectively localize mature elastin fibers in tissues from different species (16-19). Comparing various methods for demonstration of elastin, Daga-Gordini (1987) concluded that immuno-gold labelling, employed in a post-embedding protocol, was the most satisfactory (20). Most recently, polyclonal rabbit antihuman aortic alpha elastin and polyclonal rabbit antifetal bovine tropoelastin antibodies have been employed to examine the distribution of elastin in pseudoexfoliative material. This material, distributed in various ocular tissues including lens capsule, ciliary epithelium, iris epithelium, iris stroma and conjunctiva, was found to contain elastin (21). In the present study, similar antibodies were used in a post-embedding immunogold protocol to ultrastructurally localize the distribution of alpha elastin in human trabecular meshwork. The elastic fibers in the trabecular meshwork, particularly in coronal sections, definitely show the central amorphous component and peripheral microfibrillar component characteristic of elastic fibers. Neither the microfibrillar, nor the sheath material, was labelled by antibodies against alpha elastin or its precursor, tropoelastin. These results, confirm that only the amorphous component in the elastic fibers of the trabecular meshwork represents elastin (7-8,lO). The specificity of this reaction is demonstrated by the results of positive and negative controls. The observation that the antibody against tropo-elastin, the soluble precursor of elastin, showed the same labelling pattern as the antibody against alpha elastin in both human trabecular meshwork and human aorta is reassuring. This is because the amino acid composition
of tropoelastin is identical to that of elastin with the exception that the desmosines are absent and the lysine content is increased (15). It has been demonstrated that the antibody to whole elastin does not bind to desmosine (22). Both tropoelastin and insoluble elastin are multideterminant antigens and share enough common determinants so that they cross-react with each other (23-25). Crossreactivity of the anti-bovine antibodies with human tissue can be explained by the multiplicity of elastin antigenic determinants and the polyclonal character of the antibodies used. The individual microfibrillar elements of the elastic fibers were more clearly discernable in the infant eyes than in the adult eyes. In the JCT region, the microfibrils were loosely bound together and more irregularly oriented. The amorphous elastin component was not completely enclosed by microfibrils in some areas, suggesting that they were not fully mature elastic fibers. During elastogenesis, the microfibrils of the elastic fiber are known to play a central role in the initial accumulation of elastin, which appears to aggregate between and around the microfibril and take the shape already assumed by the fibrils. In the subsequent stages of fiber development,the majority of the micro-fibrils appear to be progressively displaced to the outer aspect of the fiber (26-27). This later stage continues during the early post-natal period. The microfibrils in the young adult and older eyes were more tightly bound together and appeared embedded in ground substance. Furthermore, the surrounding sheath material not observed in two infant eyes was present in adult eyes and appeared to increase with age: this is consistent with the findings of LutjenDrecoll et a1 ( 2 8 ) .
1079
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research Although previous immunofluorescent studies with anti-elastin antibodies showed no labelling for elastin in the JCT region (ll), we found elastic fibers which were labelled in a similar pattern to that seen in the trabecular beams. The negative results of immuno-fluorescent studies might be due to the small amount of elastin in this region and the lower resolution obtainable at the light microscopic level. Our findings demonstrate that, just as in the trabecular beams, both antibodies labelled only the electron-lucent areas within the center of the fibers in the JCT area. The localization of the elastin in the fibers of the cribriform plexus of the JCT region was particularly interesting. These fibers have been shown to represent the anterior extensions of the tendons of muscle cells from the longitudinal bundle of the ciliary muscle. This observation is critical as the ciliary muscle and the cribriform plexus have been suggested to be involved in the modulation of aqueous outflow resistance ( 2 9 ) . Two different types of elastic fibers have been described within various tissues. Elastic fibers in elastic arteries such as the aorta consist of a thick core of amorphous elastin component and a thin microfibrillar envelope. In contrast, elastic fibers in tendons consist of a thin core of amorphous elastin component within a thick microfibrillar envelope (30-33). Consistent with the findings of Lutjen-Drecoll et all the structural organization and iistribution of elastin within the elastic fibers is as one would predict for tendons. The relatively thin amorphous zomponent in these fibers also explains their relative insensitivity to elastase ligestion. The suggestion that the relatively small number of elastic fibers
I080
in tendon may play a role in the maintenance of a rest relationship between collagen fibrils following vigorous muscle contraction (31) emphasizes the importance of understanding the functional relationships between the ciliary muscle and the elastic fibers in the cribriform plexus. The fact that the amorphous component of the elastic fibers in the trabecular meshwork demonstrated an electron-lucent or an IIemptyIl appearance using conventional electron microscopy serves to remind us that other extracellular proteins likely fill the other Ilernpty"appearing 99spaces19 in the JCT contributing in a significant way to outflow resistance (34)* The question of whether elastin changes in amount or distribution in the trabecular meshwork of glaucomatous eyes and whether its hydrophobicity is of physiological importance remains to be established. ACKNOWLEDGEMENTS The expert technical assistance of Ms. Rozanne Richman is gratefully acknowledged. This work was supported in part by National Eye Institute Grant # EY-05503 to Dr. Roger Kamm. The eyes examined in these studies were provided by the National Disease Research Interchange (NDRI) and by the New England Eye Bank, a part of Tissue Banks International. CORRESPONDING AUTHOR Dr. Thomas Freddo, Eye Pathology Lab L905, Boston University School of Medicine, 80 E. Concord Street, Boston, MA., U.S.A. 2118
REFERENCES 1. Johnson, M. , Ethier, C.R., Kamm, R.D. , Grant, W.M. , Epstein, D.L. and
Current Eye Research
2.
3.
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
4.
5.
6.
Gaasterland, D. (1986) The flow of aqueous humor through micro-porous filters. Invest. Ophthalmol. Vis. Sci., 2 7 , 92-97. Ethier, C.R. (1986) The flow of aqueous humor in microporous materials. Doctoral thesis, Massachusetts Institute of Technology. Salzmann, M. (1921) "The Anatomy and Histology of Human Eyeballs1,(Trans. Brown, E.V.L.). Pp. 41-48. University of Chicago Press, Chicago. Garron, L.K. and Feeney, A.B. (1959) Electron microscopic studies of human eye 11. Study of the trabeculae by light and electron microscopy. Arch. Ophthalmol., 62, 966-973. Leeson, T.S. and Speakman, J.S. (1961) The fine structure of extracellular material in pectinate ligament (trabecular meshwork) of the human iris. Acta Anat., 46, 363-379. Spelsberg, W.W. and Chapman, G.B. (1962) Fine structure of human trabeculae. Arch. Ophthalmol., 67,
Tissue Matrix", (Eds. Ruggeri, A and Motta, P.M.). Pp. 126-139. Martinus Nijhoff Publishers, Boston. 15. ROSS, R. (1973) The elastic fiber, A review. J. Histochem. Cytochem., 2 l , 199-208. 16. Damiano, V.V., Tsang, A., Christer,
17.
18.
19.
773-784.
Iwamoto, T. (1964) Light and electron microscopy of the presumed elastic components of the trabeculae and scleral spur of the human eye. Invest. Ophthalmol., 3 , 144-156. 8 . Tripathi, R.C. (1969) Ultrastructure of the exit pathway of the aqueous. Ph.D Thesis, Univ. of London. 9. Hogan, M.J., Alvarado, J.A. and Weddel, J.E. (1971)" Histology of the human eye". Pp. 112-182. W.B. Saunders Company, Philadelphia. 10. Lutjen-Drecoll, E., Futa, R. and Rohen, J.W. (1981) Ultrahistochemical studies on tangential sections of the trabecular meshwork in normal and glaucomatous eyes. Invest. Ophthalmol. Vis. Sci., 21, 563-573. 11. Murphy, C.G., Yun, A.J., Newsome, D.A. and Alvarado, J.A. (1987) Localization of extracellular proteins of the human trabecular meshwork by indirect immunofluorescence. Am. J. Ophthalmol., 7.
104, 33-43. 12. Roth, J., Bendayan, M. and Orci, L. (1978) Ultrastructural location of intracellular antigens by the use of protein A-gold complex. J. Histochem. Cytochem., 26, 1074-1081. 13. Duhamel, R.C. and Johnson, D.A. (1985) Use of non fat dry milk to block nonspecific nuclear and membrane staining by avidin conjugates. J. Histochem. Cytochem.,
33, 14.
711-714.
Paaquall-Ronchetti, I. and Fornieri, C. (1984) The ultrastructural organization of the elastin fibre. In Wltrastructure of the Connective
20.
P. , Rosenbloom, J. and Weinbaum, G. (1979) Immunologic localization of elastin by electron microscopy. Am. J. Pathol., 96, 439-456. Fukuda, Y. and Ferrans, V.J. (1984) The electron microscopic immunohistochemistry of elastase-treated aorta and nuchal ligament of fetal and postnatal sheep. J. Histochem. Cytochem., 2 , 747-756. Fukuda, Y. , Ferrans, V. J. and Crystal, R.G. (1984) Development of elastic fibers of nuchal ligament, aorta and lung of fetal and postnatal sheep: An ultrastructural and electron microscopic study. Am. J. Anat. , 170, 653-665. Lethias, C., Hartmann, D.J., Masmejean, M. , Ravazzola, M., Sabbagh, I., Ville, G., Herbade, D. and Eloy, R. (1987) Ultrastructural immunolocalization of elastic fibers in rat blood vessel Using the protein A-gold technique. J. Histochem. Cytochem., 15-21. Dage-Gordini, D., Bressan, B.G., Castellani, I. and Volpin,D. (1987) Detection of elastin by immunoelectronmicroscopy: A comparison of different procedures. Histochem., @7,
=,
573-578.
Li, Z.Y., Streeten, B.W. and Wallace R.N. (1988) Association of elastin with pseudoexfoliative material: An immunoelectron microscopic study. Curr. Eye Res., 7 , 1163-1172. 22. King, G.S., Mohan, V.S. and Starcher, B.C. (1980) Radioimmunoassay for desmosine. Connect. Tiss. Res., 21.
z,
263-267.
Daynes, R.A., Thomas, M., Aivarez, V.L. and Sandberg, L.B. (1977) The antigenicity of soluble porcine elastins: I. Measurement of antibody by a radioinmunoassay. Connect. Tiss. Res. , 3, 75-82. 24. Keeley, F.W. (1981) Identification of elastin in developing aortic tissue by immuno biological methods. Connect. Tiss. Res., 8 , 193-198. 25. Ross, R. and Bornstein, P. (1969) The elastic fiber I. The separation and partial characterization of its macromolecular components. J. Cell Biol., 40, 366-381. 26. Daga-Gordini, D., Bressan, G.M., Castellani, I. and Volpin, D. (1987) Fine mapping of tropoelastin-derived components in the aorta of developing chick embryo.Histochem. J . , U , 623-32. 23.
1081
Volurnc 8 number 10 1989
Ultrastructural immunocytochemical localization of elastin in normal human trabecular meshwork Haiyan
Vickery Trinkaus-Randall* and Thomas F.Freddo2
Departments of 'Anatomy and 20phthalmology, Boston University School of Medicine, Boston, MA 021 18, USA Received o n Junc 5 , 1989; acccptcd on August 21, 1989
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
ABSTRACT
Previous studies have suggested that hydrophobic moieties within the aqueous outflow channels might interact with certain aqueous components to retard outflow. While elastin is among the most hydrophobic proteins in the trabecular meshwork, it reacts poorly with conventional ultrastructural staining methods, so its potential role in regulating outflow could not be assessed. It was our goal to specifically localize elastin ultrastructurally using polyclonal antibodies against alpha elastin and its soluble precursor, tropoelastin. Human aorta served as a positive control. Preadsorption of the primary antibodies or their substitution with either normal rabbit serum or Tris buffer resulted in negligible labelling. With either antibody, only the electron-lucent elements in the center of elastic fibers of the trabecular meshwork were labelled, indicating that only these elements truly represent elastin. The pattern of elastin distribution within these fibers is most consistent with that found in tendons elsewhere in the body. INTRODUCTION
Previous studies have shown that bovine aqueous humor obstructs flow through micro-porous filters with pore sizes similar to those found in the juxtacanalicular connective tissue ( J C T ) of the trabecular meshwork. The obstruction of these filters was eliminated using proteases but not hyaluronidase, suggesting that proteins or glycoproteins were implicated in the process(1). Ethier et a1 ( 2 ) investigated this phenomenon further and found that hydrophobic filters were more prone to obstruction than hydrophilic filters. They also found that the ionic detergent Triton-X eliminated
0 IRL Press
the filter obstruction phenomenon. From these studies, the authors concluded that the process of filter obstruction likely involved hydrophobic interactions between proteins in the aqueous humor and the filter surfaces. If this work on nonphysiologic filters is at all predictive of the behavior of aqueous humor in-vivo, it suggests that specific identification and localization of hydrophobic moieties in the extracellular matrix of the trabecular meshwork will advance our understanding of the factors which affect trabecular flow in normal and glaucomatous eyes. One of the most hydrophobic proteins in the trabecular meshwork is elastin. Elastic-like fibers were originally demonstrated in the human trabecular meshwork by Salzmann (1912)(3), but the question of whether elastin is truly present in these fibers has been controversial (4-9) Ultrastructural studies of these fibers have demonstrated that they contain two morphologically distinct zones, a central amorphous component and a peripheral fibrillar component ( 7 - 8 ) . Pursuing the matter further, Lutjen-Drecoll , et a1 (10) demonstrated that only a small amount of elastase-digestible material exists in these fibers and thus concluded that they should be called llelastic-likell. Most recently, Murphy, et a1 (11), using an indirect immunofluorescent antibody protocol, localized elastin in the
.
1071
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research trabecular meshwork at the light microscopic level. No fluorescent staining could be identified in any portion of the JCT. Fluorescent staining was only found within the osmiophilic deposits which occupy the cores of the trabecular beams. Unfortunately, no specific correspondence could be made between the known ultrastructural subcomponents of these deposits and the distribution of elastin within them. In the present study, polyclonal antibodies raised against human aortic alphaelastin and bovine tropoelastin were used in a post-embedding colloidal gold protocol to more specifically localize the distribution of elastin in normal human trabecular meshwork at the electron microscopic level. MATERIALS AND METHODS Eight normal, eye bank eyes, age 1 day-83 years, were used. The anterior segments of the eyes were cut radially into small pieces. A piece of human aorta, derived from autopsy was used as a positive control. The specimens were fixed in 2% paraformaldehyde-2.5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.3, containing 0.2% CaCl,, for 3 hr at room temperature. They were rinsed in the same buffer and postfixed in 2% osmium tetroxide and 1.5% potassium ferrocyanide for 2 hr. The specimens were dehydrated in a graded series of ethanol and embedded in Epon-Araldite. Semi-thin sections for light microscopy were made to examine the orientation of the specimens. Two different planes of thin sections, radial and coronal, were cut for electron microscopy and mounted on nickel grids. Post-embedding immuno-histochemistry was performed using a modification of the protein-A, colloidal gold technique (12). Grids were floated, specimen-side down,
1072
onto drops of the following series of reagents at room temperature. A l l sections were incubated in 4% nonfat dry milk for 30 min in order to block non-specific binding (13), and then transferred onto drops of antiserum against human aorta elastin or bovine tropoelastin (Elastin Products Company, Pacific, MI), diluted 1:lOO with 0.1 M Tris-1% BSA for 2 hr. After rinsing with 0.05 M Tris, pH 7.4 and 0.05 Tris-0.2% BSA, pH 7.4, grids were incubated for 1 hr in a solution containing colloidal gold-labelled IgG (Janssen Pharmaceutica, Piscataway , NJ) , diluted 1:12 with 0.05 M Tris-1% BSA, pH 8.3. The sections were then rinsed in 0.05 M Tris-0.2% BSA, 0.05 M Tris and then in distilled water. All specimens were counter-stained with uranyl acetate and lead citrate and examined in a Philips-300 Electron Microscope (Eindhoven, The Netherlands) Control studies included: (1) substitution of the primary antibodies with either normal rabbit serum or Tris buffer, (2) preincubation of antibodies with tropoelastin or alpha-elastin and (3) use of human aorta as a positive control. Studies with elastase were also conducted. Sections were incubated with pancreatic, porcine elastase (600 U/ml, Type 111, Sigma, St. Louis, MO) in 0.5 M Tris, p~ 8.8 for 1 hr. at room temperature.
.
RESULTS The distributions of both the alphaelastin and tropoelastin antibodies were virtually identical in the human trabecular meshwork. Although morphometric studies were not completed, there was no apparent difference in the amount of colloidal gold present with either antibody. Adult eyes Trabecular beams. In radial sections, both antibodies were localized within
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 1: The antibody to tropoelastin is evident within the central electron-lucent area of the elastic fibers in a radial
section of the adult trabecular beams. E = endothelium; ITS = Intertrabecular space. (X41,OOO)
Figure 2: The tropoelastin antibody is evident almost entirely within the central, electron-lucent core of the
elastic fibers when they are cut longitudinally. (x41,OOO)
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research areas which appeared as electron-lucent dots dispersed within the electron-dense areas of the trabecular beams (Fig 1). In coronal sections, the elastic fibers were cut longitudinally or obliquely. If the sections included the centers of the elastic fibers, the electron-lucent areas appeared as a central core enclosed by electron-dense microfibrillar components, which were, in turn, surrounded by sheath material. The antibodies only labelled the electron-lucent central core. Neither the microfibrillar components nor sheath
material was labelled (Fig 2). JCT area. Both antibodies labelled the electron-lucent areas in the cribriform plexus (Fig 3 ) . A s in the beams, when these fibers were cut longitudinally, both antibodies were localized principally within the central core of the elastic fiber. The microfibrils and sheath materials were not labelled either. Infant eves Trabecular beams. Both antibodies were again localized to the amorphous, electron-lucent, discrete dots within each
Figure 3 : The antibody to tropoelastin is evident within the electron-lucent area of elastic fibers in the region of
juxtacanalicular connective tissue (JCT). SC = Schlemm's canal. (X41,OOO)
1074
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research microfibrillar bundle. The individual microfibrillar elements were more clearly discernable in infant eyes than in the adult (Fig 4 ) . Interestingly, the sheaths which are known to surround the elastic fibers of the meshwork in the adult eye were not observed in either radial or coronal sections of two, one day-old infant eyes (Fig 4 ) . JCT area. Both antibodies labelled the amorphous electron-lucent dots within the microfibrillar bundles. The microfibrils appeared as described above but were more
irregularly distributed. These fibers, like those of the trabecular beams, lacked sheaths as well (Fig 5). The amorphous elastin component was not completely enclosed by microfibrils in some areas, characteristic of immature elastic fibers. Control studies 1. Substitution of the primary antibody with either normal rabbit serum or Tris buffer resulted in negligible colloidal gold labelling (Fig 6A). 2 . Preincubation of antibody with either alpha elastin or tropoelastin gave
Figure 4 : The antibody to tropoelastin is evident within the central, amorphous, electron-lucent areas of elastic fibers in cross-section of a one day-old infant eye.
The surrounding sheath material, found in the adult eye, is not present in the infant eye. (x41,600)
1075
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 5 : The antibody to alpha elastin is evident principally within the electronlucent areas of the elastic fibers in the
JCT region of one day-old infant eye. SC Schlemm's canal. (X41,600)
similar negative results (Fig 6B). 3 . Because it is known that elastic arteries contain elastin, we used human aorta as a positive control. Both antibodies were localized to the electronlucent elastin laminae within the wall of this vessel. The thin microfibrillar envelope was not labelled (Fig 7 ) . Elastase disestion Elastase treatment only affected the central electron-lucent area of the elastic fibers in the trabecular meshwork. The density of the colloidal gold labelling was reduced but not eliminated
after elastase digestion (Fig 8). Similar results were obtained in human aorta.
1076
=
DISCUSSION Elastin is the principal extracellular matrix constituent responsible for the elasticity of connective tissue. Due to its largely hydrophobic nature, the protein reacts poorly with conventional ultrastructural staining methods (14). This problem becomes more apparent in studying elastic fibers in the trabecular meshwork, since, unlike the elastic tissue in the aorta where the amorphous component
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 6 : Oblique sections of adult trabecular beams. A. Substitution of the primary antibody with normal rabbit serum yielded negligible colloidal gold
labelling (X26,240). B. Preincubation of antibody with alpha elastin gave similar negative results. (X26,240)
represents greater than 90% of the mature
amount of elastase-digestible material
elastic fiber (15), elastic fibers in trabecular meshwork only contain a small
(10)
Several immunomicroscopic techniques 1077
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research
Figure 7: The tropoelastin antibody is clearly localized to the electron-lucent elastin laminae in human aorta.(X41,600)
Figure 8: Elastase only affects the electron-lucent area of the elastic fibers in tie trabecular beams. The density of
the colloidal gold particles appeared to have been reduced after elastase treatment. (X41,600)
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research have been developed to more effectively localize mature elastin fibers in tissues from different species (16-19). Comparing various methods for demonstration of elastin, Daga-Gordini (1987) concluded that immuno-gold labelling, employed in a post-embedding protocol, was the most satisfactory (20). Most recently, polyclonal rabbit antihuman aortic alpha elastin and polyclonal rabbit antifetal bovine tropoelastin antibodies have been employed to examine the distribution of elastin in pseudoexfoliative material. This material, distributed in various ocular tissues including lens capsule, ciliary epithelium, iris epithelium, iris stroma and conjunctiva, was found to contain elastin (21). In the present study, similar antibodies were used in a post-embedding immunogold protocol to ultrastructurally localize the distribution of alpha elastin in human trabecular meshwork. The elastic fibers in the trabecular meshwork, particularly in coronal sections, definitely show the central amorphous component and peripheral microfibrillar component characteristic of elastic fibers. Neither the microfibrillar, nor the sheath material, was labelled by antibodies against alpha elastin or its precursor, tropoelastin. These results, confirm that only the amorphous component in the elastic fibers of the trabecular meshwork represents elastin (7-8,lO). The specificity of this reaction is demonstrated by the results of positive and negative controls. The observation that the antibody against tropo-elastin, the soluble precursor of elastin, showed the same labelling pattern as the antibody against alpha elastin in both human trabecular meshwork and human aorta is reassuring. This is because the amino acid composition
of tropoelastin is identical to that of elastin with the exception that the desmosines are absent and the lysine content is increased (15). It has been demonstrated that the antibody to whole elastin does not bind to desmosine (22). Both tropoelastin and insoluble elastin are multideterminant antigens and share enough common determinants so that they cross-react with each other (23-25). Crossreactivity of the anti-bovine antibodies with human tissue can be explained by the multiplicity of elastin antigenic determinants and the polyclonal character of the antibodies used. The individual microfibrillar elements of the elastic fibers were more clearly discernable in the infant eyes than in the adult eyes. In the JCT region, the microfibrils were loosely bound together and more irregularly oriented. The amorphous elastin component was not completely enclosed by microfibrils in some areas, suggesting that they were not fully mature elastic fibers. During elastogenesis, the microfibrils of the elastic fiber are known to play a central role in the initial accumulation of elastin, which appears to aggregate between and around the microfibril and take the shape already assumed by the fibrils. In the subsequent stages of fiber development,the majority of the micro-fibrils appear to be progressively displaced to the outer aspect of the fiber (26-27). This later stage continues during the early post-natal period. The microfibrils in the young adult and older eyes were more tightly bound together and appeared embedded in ground substance. Furthermore, the surrounding sheath material not observed in two infant eyes was present in adult eyes and appeared to increase with age: this is consistent with the findings of LutjenDrecoll et a1 ( 2 8 ) .
1079
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
Current Eye Research Although previous immunofluorescent studies with anti-elastin antibodies showed no labelling for elastin in the JCT region (ll), we found elastic fibers which were labelled in a similar pattern to that seen in the trabecular beams. The negative results of immuno-fluorescent studies might be due to the small amount of elastin in this region and the lower resolution obtainable at the light microscopic level. Our findings demonstrate that, just as in the trabecular beams, both antibodies labelled only the electron-lucent areas within the center of the fibers in the JCT area. The localization of the elastin in the fibers of the cribriform plexus of the JCT region was particularly interesting. These fibers have been shown to represent the anterior extensions of the tendons of muscle cells from the longitudinal bundle of the ciliary muscle. This observation is critical as the ciliary muscle and the cribriform plexus have been suggested to be involved in the modulation of aqueous outflow resistance ( 2 9 ) . Two different types of elastic fibers have been described within various tissues. Elastic fibers in elastic arteries such as the aorta consist of a thick core of amorphous elastin component and a thin microfibrillar envelope. In contrast, elastic fibers in tendons consist of a thin core of amorphous elastin component within a thick microfibrillar envelope (30-33). Consistent with the findings of Lutjen-Drecoll et all the structural organization and iistribution of elastin within the elastic fibers is as one would predict for tendons. The relatively thin amorphous zomponent in these fibers also explains their relative insensitivity to elastase ligestion. The suggestion that the relatively small number of elastic fibers
I080
in tendon may play a role in the maintenance of a rest relationship between collagen fibrils following vigorous muscle contraction (31) emphasizes the importance of understanding the functional relationships between the ciliary muscle and the elastic fibers in the cribriform plexus. The fact that the amorphous component of the elastic fibers in the trabecular meshwork demonstrated an electron-lucent or an IIemptyIl appearance using conventional electron microscopy serves to remind us that other extracellular proteins likely fill the other Ilernpty"appearing 99spaces19 in the JCT contributing in a significant way to outflow resistance (34)* The question of whether elastin changes in amount or distribution in the trabecular meshwork of glaucomatous eyes and whether its hydrophobicity is of physiological importance remains to be established. ACKNOWLEDGEMENTS The expert technical assistance of Ms. Rozanne Richman is gratefully acknowledged. This work was supported in part by National Eye Institute Grant # EY-05503 to Dr. Roger Kamm. The eyes examined in these studies were provided by the National Disease Research Interchange (NDRI) and by the New England Eye Bank, a part of Tissue Banks International. CORRESPONDING AUTHOR Dr. Thomas Freddo, Eye Pathology Lab L905, Boston University School of Medicine, 80 E. Concord Street, Boston, MA., U.S.A. 2118
REFERENCES 1. Johnson, M. , Ethier, C.R., Kamm, R.D. , Grant, W.M. , Epstein, D.L. and
Current Eye Research
2.
3.
Curr Eye Res Downloaded from informahealthcare.com by University of North Carolina on 12/05/14 For personal use only.
4.
5.
6.
Gaasterland, D. (1986) The flow of aqueous humor through micro-porous filters. Invest. Ophthalmol. Vis. Sci., 2 7 , 92-97. Ethier, C.R. (1986) The flow of aqueous humor in microporous materials. Doctoral thesis, Massachusetts Institute of Technology. Salzmann, M. (1921) "The Anatomy and Histology of Human Eyeballs1,(Trans. Brown, E.V.L.). Pp. 41-48. University of Chicago Press, Chicago. Garron, L.K. and Feeney, A.B. (1959) Electron microscopic studies of human eye 11. Study of the trabeculae by light and electron microscopy. Arch. Ophthalmol., 62, 966-973. Leeson, T.S. and Speakman, J.S. (1961) The fine structure of extracellular material in pectinate ligament (trabecular meshwork) of the human iris. Acta Anat., 46, 363-379. Spelsberg, W.W. and Chapman, G.B. (1962) Fine structure of human trabeculae. Arch. Ophthalmol., 67,
Tissue Matrix", (Eds. Ruggeri, A and Motta, P.M.). Pp. 126-139. Martinus Nijhoff Publishers, Boston. 15. ROSS, R. (1973) The elastic fiber, A review. J. Histochem. Cytochem., 2 l , 199-208. 16. Damiano, V.V., Tsang, A., Christer,
17.
18.
19.
773-784.
Iwamoto, T. (1964) Light and electron microscopy of the presumed elastic components of the trabeculae and scleral spur of the human eye. Invest. Ophthalmol., 3 , 144-156. 8 . Tripathi, R.C. (1969) Ultrastructure of the exit pathway of the aqueous. Ph.D Thesis, Univ. of London. 9. Hogan, M.J., Alvarado, J.A. and Weddel, J.E. (1971)" Histology of the human eye". Pp. 112-182. W.B. Saunders Company, Philadelphia. 10. Lutjen-Drecoll, E., Futa, R. and Rohen, J.W. (1981) Ultrahistochemical studies on tangential sections of the trabecular meshwork in normal and glaucomatous eyes. Invest. Ophthalmol. Vis. Sci., 21, 563-573. 11. Murphy, C.G., Yun, A.J., Newsome, D.A. and Alvarado, J.A. (1987) Localization of extracellular proteins of the human trabecular meshwork by indirect immunofluorescence. Am. J. Ophthalmol., 7.
104, 33-43. 12. Roth, J., Bendayan, M. and Orci, L. (1978) Ultrastructural location of intracellular antigens by the use of protein A-gold complex. J. Histochem. Cytochem., 26, 1074-1081. 13. Duhamel, R.C. and Johnson, D.A. (1985) Use of non fat dry milk to block nonspecific nuclear and membrane staining by avidin conjugates. J. Histochem. Cytochem.,
33, 14.
711-714.
Paaquall-Ronchetti, I. and Fornieri, C. (1984) The ultrastructural organization of the elastin fibre. In Wltrastructure of the Connective
20.
P. , Rosenbloom, J. and Weinbaum, G. (1979) Immunologic localization of elastin by electron microscopy. Am. J. Pathol., 96, 439-456. Fukuda, Y. and Ferrans, V.J. (1984) The electron microscopic immunohistochemistry of elastase-treated aorta and nuchal ligament of fetal and postnatal sheep. J. Histochem. Cytochem., 2 , 747-756. Fukuda, Y. , Ferrans, V. J. and Crystal, R.G. (1984) Development of elastic fibers of nuchal ligament, aorta and lung of fetal and postnatal sheep: An ultrastructural and electron microscopic study. Am. J. Anat. , 170, 653-665. Lethias, C., Hartmann, D.J., Masmejean, M. , Ravazzola, M., Sabbagh, I., Ville, G., Herbade, D. and Eloy, R. (1987) Ultrastructural immunolocalization of elastic fibers in rat blood vessel Using the protein A-gold technique. J. Histochem. Cytochem., 15-21. Dage-Gordini, D., Bressan, B.G., Castellani, I. and Volpin,D. (1987) Detection of elastin by immunoelectronmicroscopy: A comparison of different procedures. Histochem., @7,
=,
573-578.
Li, Z.Y., Streeten, B.W. and Wallace R.N. (1988) Association of elastin with pseudoexfoliative material: An immunoelectron microscopic study. Curr. Eye Res., 7 , 1163-1172. 22. King, G.S., Mohan, V.S. and Starcher, B.C. (1980) Radioimmunoassay for desmosine. Connect. Tiss. Res., 21.
z,
263-267.
Daynes, R.A., Thomas, M., Aivarez, V.L. and Sandberg, L.B. (1977) The antigenicity of soluble porcine elastins: I. Measurement of antibody by a radioinmunoassay. Connect. Tiss. Res. , 3, 75-82. 24. Keeley, F.W. (1981) Identification of elastin in developing aortic tissue by immuno biological methods. Connect. Tiss. Res., 8 , 193-198. 25. Ross, R. and Bornstein, P. (1969) The elastic fiber I. The separation and partial characterization of its macromolecular components. J. Cell Biol., 40, 366-381. 26. Daga-Gordini, D., Bressan, G.M., Castellani, I. and Volpin, D. (1987) Fine mapping of tropoelastin-derived components in the aorta of developing chick embryo.Histochem. J . , U , 623-32. 23.
1081
