Cytochemical Detection of Disulfide and Sulfhydryl Groups in Lamprey Aortic Microfibrils Abstract a Cytochemical Study Was Perfo

J. Nihon Univ. Sch. Dent., Vol. 32, 81-92, 1990 81

Cytochemical Detection of Disulfide and Sulfhydryl Groups in Lamprey Aortic Microfibrils

Keitaro ISOKAWA †

(Received 26 January and accepted 29 January 1990)

Key words: cytochemistry, lamprey, aorta, microfibrils, elastic fibers Abstract A cytochemical study was performed on the lamprey ventral aorta with special reference to disulfide and sulfhydryl groups of microfibrils, using the high-iron diamine-thiocarbohydrazide-silver proteinate (HID-TCH-SP) method combined with several other types of treatment. The HID-TCH-SP staining observed was classified into three categories: 1) weak staining in the periphery of collagen fibrils, 2) moderate staining in the boundaries of collagen fibrils, microfibrils and smooth muscle cells, and 3) intense staining of microfibrils. The first and second categories of staining were considered to represent chondroitin and/or heparan sulfate because of sensitivity of the staining to chondroitinase ABC (ChABC) and its specific localization. By contrast, the third category of staining was considered to represent disulfide and sulfhydryl groups of microfibrillar glycoprotein, because it was disclosed only after Oxone oxidation or thiosulfation and was not removed by ChABC digestion. Although this staining reactivity was not apparently altered by SH blockade prior to oxidation or thiosulfation, it was markedly diminished or completely inhibited by S-S reduction followed by SH blockade. These results indicate that lamprey aortic microfibrils contain more S-S groups than SH groups. Introduction Numerous previous studies on elastic fibers have revealed that elastin and elastin-associated microfibrils are structural entities of mammalian elastic fibers[1,2].A wide phylogenetic distribution of elastin is suggested by biochemical analyses of aortae from a wide variety of vertebrates[3,4]and the results have been confirmed morphologically and cytochemically[5-9]. On the other hand, no system- atic examinations of elastin-associated microfibrils have been carried out in lower vertebrates. However, microfibrils associated with elastin have been demonstrated morphologically in reptilian lung[9], amphibian notochord[8], and teleostean dermis[10].In addition, a recent cytochemical study by IsoKAwAet al.[7]has shown that microfibrils associated with teleostean elastin are very similar to mammalian 磯川桂太郎: Department of Anatomy, Nihon University School of Dentistry, 1-8-13 Kanda-Surugadai, Chiyoda-ku, Tokyo 101, Japan. † To whom all correspondence should be addressed: Kellam ISOKA WA, Department of Anatomy and Cellular Biology, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, Wisconsin 53226, U.S.A. 82 elastin-associated microfibrils. Thus, elastin and elastin-associated microfibrils seem to be two important components of elastic fibers not only in mammals but also lower vertebrates except for the lowest ones. In the aortic wall of the lowest vertebrates, such as the lamprey, desmosine and isodesmosine (amino acids peculiar to elastin) are not detected [3,4], and microfibrils without elastin have been demonstrated as elastic components[11,12]. Lamprey microfibrils are interesting in two aspects. First, in mammalian elastogenesis, microfibrils appear first and are followed by deposition of elastin[1,2], so that lamprey microfibrils may be a primitive form of elastic fibers. Second, lamprey microfibrils are similar to oxytalan microfibrils, which are distributed in mammalian mature tissues such as periodontal ligaments[13,14]. This study was designed to clarify the cytochemical properties of lamprey aortic microfibrils with special reference to their disulfide (S-S) and sulfhydryl

(SH) groups. For this purpose, SPICER'S[15,16] high-iron diamine (HID) method specific for sulfate groups was combined with several cytochemical modifications, and the sites of the HID reaction were localized accurately by the thiocarbohydrazide-silver proteinate (TCH-SP) method[17,18].

Materials and Methods

Fifteen adult lampreys (Lampetra japonica), each weighing about 500 g, were anesthetized with 3-aminobenzoic acid ethyl ester (MS-222; Sigma Chemical Co., USA). After intracardiac perfusion of 50 ml of 4% paraformaldehyde-0.5% glutar- aldehyde in 0.1 M cacodylate buffer (pH 7.35), the ventral aorta was removed and cut into pieces, each measuring about 0.2 X 1.0 •~ 2.0 mm. Subsequently the specimens were immersed in the same fixative for 2 h at 4•Ž and then rinsed thoroughly in 0.1 M cacodylate buffer (pH 7.35). Specimens for morphological observation were postfixed in 1% osmic acid in 0.1 M cacodylate buffer (pH 7.35) for 1 h at room temperature, routinely dehydrated in graded alcohols and propylene oxide, and embedded in SPURR'S [19] low- viscosity resin. Thin sections (75 nm thick) cut on a Sorvall MT-5 000 ultra- microtome were contrasted with uranyl acetate and lead (UA-Pb) double staining and viewed with a Hitachi H-800 electron microscope.

Specimens for cytochemical examination were not postfixed with osmic acid , because such postfixation is known to cause non-specific binding of TCH- sp[6,16,18] After each cytochemical modification described below , all the specimens were immersed in SPICER'S HID solutiont[5] for 18 h at 22•Ž, dehydrated routinely, embedded, and sectioned. For accurate localization of HID, ultrathin sections were stained by the TCH-SP method as described previously[17,18] and observed by electron microscopy. Oxone oxidation The aldehyde-fixed, rinsed specimens were oxidized in 10% Oxone (monoper- sulfate compound, 2KHS05.KHSO4-1(2SO4, Du Pont, USA) for 30 min at room temperature and then rinsed thoroughly in distilled water . Some specimens proc- essed without Oxone oxidation are designated "Oxone control" in this article. 83

Thiosulfation The aldehyde-fixed, rinsed specimens were thiosulfated for 3 h at 22•Ž in cupric sulfite reagent[20]. The reagent consisted of 40 ml of 0.5 M ammonia solution, 40 ml of 0.025 M cupric sulfate brought to pH 9 by adding a few drops of concentrated ammonia, and 10 ml of 0.2 M sodium metabisulfite (Sigma Chemical Co., USA). Control specimens were incubated for 3 h at 22•Ž in the reagent without sodium metabisulfite[20]. Chondroitinase ABC (ChABC) digestion Some specimens prior to or after either Oxone oxidation or thiosulfation were incubated in a ChABC (protease free; Seikagaku-Kogyo, Tokyo) solution for 18 h at 37°C. Specimens without Oxone oxidation or thiosulfation were incubated similarly. The ChABC solution contained 0.2 units per ml of ChABC in 0.1 M sodium acetate-0.1 M Tris-HC1 buffer (pH 7.3)[21]. Control specimens were incubated for 18 h at 37•Ž in the same buffer without ChABC. S-S reduction and SH blockade S-S reduction and SH blockade were performed by a slight modification of the previous methods for light microscopic histochemistry[22,23] Briefly, S-S groups in the aldehyde-fixed, rinsed specimens were reduced to SH groups by incubation for 4 h at 37•Ž in 0.05 M dithioerythritol (DTE; Sigma Chemical Co., USA) in 0.1 M Tris-HC1 buffer (pH 8.5) or in 0.5 M thioglycolic acid (TG; Sigma Chemical Co., USA) in distilled water brought to pH 8 with 0.1 N NaOH. Subsequently, SH groups including native ones and reductively generated ones were blocked by incubation for 4 or 18 h at 37•Ž in 0.2 M N-ethylmaleimide (NEM; Sigma Chemical Co., USA) in 0.1 M phosphate buffer (pH 7.4). Some aldehyde-fixed, rinsed specimens were treated with either the S-S reduction or the SH blockade procedure. Following these treatments, all specimens were oxidized with Oxone or thiosulfated.

Results

Three tunics were distinguishable in the wall of the lamprey ventral aorta; the tunica intima, tunica media and tunica adventitia. The major cell population in the wall was smooth muscle cells, and the most abundant extracellular matrix components were microfibrils, which showed a typical tubular appearance with a diameter of 11.6•} 1.1 nm (mean •} S. D.; n = 50). Also present extracellularly were collagen fibrils and filamentous structures. The latter were distributed in the periphery of collagen fibrils (Fig. 1). Cytochemical results, described in the following, did not differ throughout the wall.

Oxone oxidation After Oxone oxidation, lamprey aortic microfibrils were intensely stained with HID-TCH-SP, and weak HID-TCH-SP staining was demonstrated in the periphery of collagen fibrils (Fig. 2A, Table 1). In the control specimens without oxidation, weak staining in the periphery of collagen fibrils and moderate staining in the boundaries between microfibrils and collagen fibrils were found, but no

HID-TCH-SP was disclosed within microfibrils (Fig. 2B, Table 1). 84

Fig. 1 Ultrastructure of the tunica media in lamprey ventral aorta. Note numerous microfibrils (mf)

and the bundle of collagen fibrils (co) in the extracellular space . Microfibrils show a typical tubular appearance (arrows in inset). Filamentous structures (arrowheads in inset) are demonstrated in the periphery of the collagen fibrils. sm: smooth muscle cells. Osmicated specimens , contrasted with UA-Pb. Bar =0.5,ƒÊm

Thiosulfation The HID-TCH-SP staining pattern of thiosulfated specimens was similar to that of Oxone-oxidized specimens. That is, microfibrils were intensely stained with HID-TCH-SP, and the periphery of collagen fibrils was weakly stained (Fig . 3A, Table 1). In thiosulfation control specimens, no HID-TCH-SP staining was observable in microfibrils, whereas weak HID-TCH-SP staining in the periphery of collagen fibrils and moderate staining at the boundaries between microfibrils

(1) 85

( 2A) ( 2B)

Fig. 2A Oxone-oxidized specimen, stained with HID-TCH-SP. Note intense staining of microfibrils (mf) and weak staining in the periphery of collagen fibrils (co). Unosmicated. Bar=0.5,um Fig. 2B Unoxidized (Oxone control) specimen, stained with HID-TCH-SP. Microfibrils (mf) lack the staining. Weak staining in the periphery of collagen fibrils (co) and moderate staining in the boundary (arrows) are demonstrated. Unosmicated. Bar=0.5gm

Table 1 Effect of oxidization, thiosulfation and ChABC digestion on HID-TCH- SP staining of lamprey aortic microfibrilsa)

and collagen fibrils were demonstrated (Fig. 3B, Table 1). ChABC digestion Digestion with ChABC completely removed HID-TCH-SP staining from the periphery of collagen fibrils (Figs. 4A, B, Table 1). However, ChABC digestion prior to or after either Oxone oxidation or thiosulfation did not affect the HID-TCH-SP staining of microfibrils, and the staining remained intense (Fig. 4A, Table 1). The HID-TCH-SP staining also persisted in the boundaries of microfibrils, collagen fibrils and smooth muscle cells, although it was markedly reduced (arrows in Figs. 4A, B). In control specimens incubated without ChABC, the 86

( 3A) (3B)

Fig. 3A Thiosulfated specimen, stained with HID-TCH-SP. Note intense staining of microfibrils (mf) and weak staining in the periphery of collagen fibrils (co). Unosmicated. Bar=0.5pm Fig. 3B Thiosulfation control specimen, stained with HID-TCH-SP. Microfibrils (mf) lack the staining, although weak staining is evident in the periphery of collagen fibrils (co). The staining in the boundary (arrows) is slightly stronger than that in the periphery of collagen fibrils. Unosmicated. Bar =0.5ƒÊm

Table 2 Cytochemical detection of S-S and SH groups in lamprey aortic microfibrilsa)

a) Evaluation of staining intensity, O; negative, 1; very weak, 2; weak, 3; moderate, 4; intense, 5; very intense. b) Net S-S detected in the reduced form, SH. c) Reducing and blocking treatents were omitted. 87

HID-TCH-SP staining pattern was similar to that of the undigested specimens. S-S reduction and SH blockade DTE-NEM or TG-NEM pretreatment apparently diminished or completely inhibited Oxone- or thiosulfation-induced HID-TCH-SP staining of microfibrils (Table 2). The 18-h MEM treatment was more effective than the 4-h treatment, and DTE-NEM treatment was more effective than TG-NEM treatment (Table 2) .

( 4A) (4B)

Fig. 4A ChABC-digested, Oxone-oxidized specimen, stained with HID-TCH-SP. The staining in the periphery of collagen fibrils (co) has been removed, but intense staining of microfibrils (mf) persists. Weak staining also remains in the boundary (arrows) between collagen fibrils and smooth muscle cells (sm). Unosmicated. Bar= 0.5,um Fig. 4B ChABC-digested, unoxidized specimen, stained with HID-TCH-SP. The staining in the periphery of collagen fibrils (co) has been removed, but the staining in the boundary (arrows) between collagen fibrils and microfibrils (mf) persists. No staining of microfibrils is observed because specimen is unoxidized. Unosmicated. Bar = 0.5 pm

Thus, S-S reduction with DTE followed by SH blockade with NEM completely inhibited Oxone-induced HID-TCH-SP staining of microfibrils (Fig. 5A, Table 2), and S-S reduction with DTE or TG followed by SH blockade with NEM for 18 h almost completely inhibited Oxone- or thiosulfation-induced HID-TCH-SP staining of microfibrils (Fig. 5B, Table 2). After DTE or TG pretreatment, Oxone- or thiosulfation-induced HID-TCH- SP staining of microfibrils remained intense (Fig. 5C, Table 2), and the staining intensity did not differ from that of microfibrils without these reduction treatments (Figs. 2A, 3A). Only NEM pretreatment did not apparently diminish Oxone- or thiosulfation-induced HID-TCH-SP staining of microfibrils, the staining being moderate to intense (Fig. 5D, Table 2). These reducing and/or blocking pretreatments did not alter the HID-TCH-SP staining in the periphery of collagen fibrils or that in the boundaries between microfibrils and collagen fibrils (Figs. 5A-D). 88

(5A) ( 5B)

( 5D)

Fig. 5A DTE-MEM (4 h)-treated, Oxone-oxidized specimen, stained with HID-TCH-SP. Note complete inhibition of the staining in microfibrils (mf). Weak staining is demonstrated in the

periphery of collagen fibrils (co) and in the boundary (arrows). Unosmicated. Bar =0.5,ƒÊm Fig. 5B DTE-NEM (18 h)-treated, thiosulfated specimen, stained with HID-TCH-SP. Note extreme- ly weak staining of microfibrils (mf). Weak staining in the periphery of collagen fibrils (co) and moderate staining in the boundary (arrows) are observed. Unosmicated. Bar = 0.5pm Fig. 5C DTE-treated, Oxone-oxidized specimen, stained with HID-TCH-SP. Intense staining of microfibrils (mf) is not affected by DTE treatment. The periphery of collagen fibrils (co) shows weak staining. Unosmicated. Bar = 0.5 ,ƒÊm Fig. 5D NEM (18 h)-treated, Oxone-oxidized specimen, stained with HID-TCH-SP. The staining of microfibrils (mf) is only minimally decreased. Weak staining is evident in the periphery of collagen fibrils (co). Unosmicated. Bar = 0.5ƒÊm

Discussion Major sulfated glycosaminoglycans in vascular walls include chondroitin sulfate A, dermatan sulfate (chondroitin sulfate B), chondroitin sulfate C and heparan sulfate1241.Most chondroitin sulfates have a close relationship with collagen fibrils[25],and heparan sulfate has an important role in cell-matrix and/ or matrix-matrix association, similar to fibronectin[26]. Thus, the weak HID-TCH- SP staining observed in the periphery of collagen fibrils is considered to be chondroitin sulfates because of its sensitivity to ChABC and its specific localiza-

(5C) 89

tion. HID-TCH-SP staining in the boundaries of collagen fibrils, microfibrils and smooth muscle cells was partially resistant to ChABC, and since these boundaries are the sites of cell-matrix or matrix-matrix interaction, this staining would appear to represent heparan sulfate as well as chondroitin sulfate. On the other hand, the results obtained in this study support the contention that microfibrils do not contain sulfated glycosaminoglycans; i.e., HID-TCH-SP staining of microfibrils was disclosed only after Oxone oxidation or thiosulfation, and the staining was resistant to ChABC digestion. Thiosulfation is a cytochemical method known to produce S-sulfocysteine from cystine and cysteine in tissues, and the sulfonic acid groups in S-sulfocystein are derivatives of the S-S groups in cystine and the SH groups in cysteine[20,27].Sulfonic acid groups (-SO31 have steric similarity to sulfate groups (-O-SO3-) and are believed to be detected by HID-TCH-SP staining specific to sulfate groupsr28'291.Although the mechanism is not fully understood, Oxone is also considered to oxidatively generate sulfonic acid groups[28-30].This interpretation is supported by the present finding that the HID-TCH-SP staining pattern of Oxone-oxidized sepcimens closely resembled that of thiosulfated specimens. In addition, the number and size of the HID-TCH-SP staining deposits are reported to be proportional to the amount of reactive groups at the staining Siter31'321.Therefore, the intense HID-TCH-SP staining of microfibrils after Oxone oxidation or thiosulfation would appear to indicate an abundance of S-S and/or SH groups in microfibrils. At the light microscopic level, azo-dyes are utilized for histochemical differen- tiation of S-S and SH groups[22,23],but unfortunately these dyes cannot be visual- ized by electron microscopy. In the present study, S-S-reducing and/or SH- blocking methods were combined with Oxone oxidation or thiosulfation followed by the HID-TCH-SP method. Although SH blockade did not alter the intense HID-TCH-SP staining of microfibrils, S-S reduction prior to SH blockade markedly diminished or completely inhibited the staining. These findings indicate that lamprey aortic microfibrils contain many more S-S groups than SH groups. Mammalian elastin-associated microfibrils are known to be rich in cystine residues[33], so that the present results suggest a similarity between the proteinaceous components of lamprey aortic microfibrils and mammalian elastin- associated microfibrils. Previous ultrastructural observations and results of carbo- hydrate cytochemistry suggesting homology between the two kinds of microfibrils have led to the assumption that lamprey aortic microfibrils without amorphous elastin constitute a primitive form of elastic fiber[11,12].This idea is also supported by the present study. In addition, the present study demonstrated that lamprey microfibrils are HID-TCH-SP-positive when oxidized with Oxone. This tinctorial property is essential for identification of mammalian oxytalan microfibrils, which are stained with aldehyde-fuchsin or HID when oxidized with Oxone or peracetic aCidr13'34'351. The extracellular fibers in lamprey ventral aorta are not stained with conventional elastic fiber stains[36-38]but are stained with them after particular types of fixation or pretreatment[36,37].This discrepancy could be due to the fact that lamprey aortic microfibrils might share common cytochemical properties with mammalian 90 oxytalan microfibrils, as demonstrated in the present study. In mammals, similarity between elastin-associated microfibrils and oxytalan microfibrils has been repeatedly emphasized[28,29,39,40],and,COTTA-PEREIRA et al.[41,42]have proposed that these microfibrils in mammals should be classified into a family of "elastic system fibers." The present results obtained for lamprey aortic microfibrils support this assumption from a phylogenetic standpoint. Acknowledgements I am grateful to Drs. Yoshihisa Toda and Minoru Takagi for useful discussion, and to Hideo Nagai for valuable advice on operation of the electron microscope. This work was supported in part by a grant from the Sato Memorial Foundation. References [1] Ross, R.: The elastic fiber. A review,J. Histochem.Cytochem., 21, 199-208,1973 [2] CLEARY,E. G. and GIBSON,M. A.: Elastin-associatedmicrofibrils and microfibrillarproteins, Int. Rev. Connective Tissue Res., 10, 97-209, 1983 [31 SAGE,E. H. and GRAY,W. R.: Evolution of elastin structure, Adv. Exp. Med. Biol., 79, 291-309, 1977 [4] SAGE,H. and GRAY,W. R.: Studies on the evolution of elastin. I. Phylogenetic distribution, Comp. Biochem. Physiol. [B], 64, 313-327, 1979 [5] SERAFINI-FRACASSINI,A., FIELD, J. M., SPINA,M., GARBISA,S. and STUART,R. J.: The morphological organization and ultrastructure of elastin in the arterial wall of trout (Salmo gairdneri) and salmon (Salmo salar), J. Ultrastruct. Res., 65, 1-12, 1978 [6] ISOKAWA,K., TAKAGI,M. and ToDA,Y.: Ultrastructural cytochemistry of trout arterial fibrils as elastic components, Anat. Rec., 220, 369-375, 1988 [7] IsoxAwA,K., TAKAGI,M. and ToDA,Y.: Ultrastructural and cytochemical study of elastic fibers in the ventral aorta of a teleost, Anguilla japonica, Anat. Rec., 226, 18-26, 1990 [8] NAKAO,T.: Elastic fibres in the notochord of Rana rugosa tadpoles, Cell Tissue Res., 153, 245-251, 1974 [9] SATO,S.: The lung evolution and its design. A comparative ultrastructural study, J. Jpn. Med. Soc. Biol. Interface, 18, 59-80, 1987 [10] FERRI,S.: Elastic system fibers (oxytalan, elaunin and elastic fibers) in the skin of a freshwater teleost. Optical and electron microscopy study, Arch. Anat. Microsc. Morphol. Exp., 69, 259-266, 1980 [11] WRIGHT,G. M.: Structure of the conus arteriosus and ventral aorta in the sea lamprey, Petromyzon marinus, and the Atlantic hagfish, Myxine glutinosa. Microfibrils, a major component, Can. J. Zool., 62, 2445-2456, 1984 [12] ISOKAWA,K., TAKAGI,M. and ToDA,Y.: Ultrastructural cytochemistry of aortic microfibrils in the Arctic lamprey, Lampetra japonica, Anat. Rec., 223, 158-164, 1989 [13] FULLMER,H. M.: Differential staining of connective tissue fibers in areas of stress, Science, 127, 1240, 1958 [14] CARMICHAEL,G. G. and FULLMER,H. M.: The fine structure of the oxytalan fiber, J. Cell Biol., 28, 33-36, 1966 [15] SPICER,S. S.: Diamine methods for differentiating mucosubstances histochemically, J. Histochem. Cytochem., 13, 211-234, 1965 [16] SPICER,S. S., SCHULTE,B. A., THOMOPOULOS,G. N., PARMLEY,R. T. and TAKAGI,M.: Cytochemistry of complex carbohydrates by light and electron microscopy. Available methods and their application. In International Academy of Pathology Monograph, Vol. 24, Connective Tissue Diseases, 163-211, WAGNER,B. M., FLEISCHMAJER,R. and KAUFMAN, N., Eds., Williams and Wilkins, Baltimore, U. S. A., 1983 91

[17] THIERY,J. P.: Mise en evidence des polysaccarides sur coupes fines en microscopie electro- nique, J. Microsc., 6, 987-1018, 1967 [18] SANNES,P. L., SPICER,S. S. and KATSUYAMA,T.: Ultrastructural localization of sulfated complex carbohydrates with a modified iron diamine procedure, J. Histochem. Cytochem.. 27, 1108-1111, 1978 [19] SPURR,A. R.: A low-viscosity epoxy resin embedding medium for electron microscopy, J. Ultrastruct. Res., 26, 31-43, 1969 [20] CASTINO,F. and BUSSOLATI,G.: Thiosulphation for the histochemical demonstration of protein-bound sulfhydryl and disulphide groups, Histochemistry, 39, 93-96, 1974 [21] COUCHMAN,J. R., CATERSON,B., CHRISTNER,J. E. and BAKER,J. R.: Mapping by monoclonal antibody detection of glycosaminoglycans in connective tissues, Nature, 307. 650-652. 1984 [22] BARRNETT, R. J. and SELIGMAN, A. M.: Histochemical demonstration of sulfhydryl and disulfide groups of protein, J. Nat. Cancer Inst., 14, 769-803, 1954 [23] SIPPEL,T. O.: The histochemistry of thiols and disulphides. II. Methodology of differential staining, Histochem. J., 10, 585-595, 1978 [24] MULLINGER,R. N. and MANLEY,G.: Glycosaminoglycans and atherosclerosis in animal aortas, J. Atheroscler. Res., 9, 108-111, 1969 [25] RUGGERI,A. and BENAllO,R.: Collagen-proteoglycan interaction. In Ultrastructure of the connective tissue matrix, 113-125, RUGGERI,A. and MOTTA,P. M., Eds., Martinus Nijhoff Publishers, Boston, U.S.A., 1984 [26] WOODS,A. and CoucHmAN,J. R.: Focal adhesions and cell-matrix interactions, Collagen Rel. Res., 8, 155-182, 1988 [27] BocK, P.: Staining of elastin and pseudo-elastica ("elastic fiber microfibrils", type III and type IV collagen) with paraldehyde-fuchsin, Mikroskopie, 33, 332-341, 1977 [28] TAKAGI, M., BABA, T., BABA, H. and TODA, Y.: Ultrastructural cytochemistry of oxytalan fibres in monkey periodontal ligaments with the high iron diamine method, Histochem. J., 19, 75-84, 1987 [29] BABA,T., TAKAGI,M., KAGAMI,A., HISHIKAWA,H. and HOSOKAWA,Y.: Ultrastructural cytochemical properties of elastin-associated microfibrils and their relation to fibronectin, Histochem. J., 20, 688-696, 1988 [30] HIRAYAMA,H., TAKAGI,M. and TODA,Y.: Histochemical studies of oxytalan fibers in monkey periodontal ligaments. Jan. J. Oral Biol._ 27_931-941 1985 (in Tananecel [31] TAKAGI, M., PARMLEY, R. T., TODA, Y. and AUSTIN, R. L.: Ultrastructural cytochemistry and immunocytochemistry of sulfated glycosaminoglycans in epiphyseal cartilage, J. Histochem. Cytochem., 30, 1179-1185, 1982 [32] TAKAGI,M., PARMLEY,R. T., TODA,Y. and DENYS,F. R.: Extracellular and intracellular digestion of complex carbohydrates by osteoclasts, Lab. Invest., 46, 288-297, 1982 [33] Ross, R. and BORNSTEIN,P.: The elastic fiber. I. The separation and partial characterization of its macromolecular components, J. Cell Biol., 40, 366-381, 1969 [34] RANNIE,I.: Observations on the oxytalan fibre of the periodontal membrane, Trans. European Orthodontic Soc., 39, 127-136, 1963 [35] FULLMER,H. M., SHEETS,J. H. and NARKATES,A.: Oxytalan connective tissue fibers. A review, J. Oral Pathol., 3, 291-316, 1974 [36] KEIBEL,F.: Die Bulbus- and Arterienwillste bei Petromyzonten, Z. Mikrosk. Anat. Forsch., 5, 353-370, 1926 [37] WAGENVOORT,C. A.: Die Bedeutung der Arterienwillste far den Blutkreislauf, Acta Anat., 21, 70-99, 1954 [38] SAGE,H. and GRAY,W. R.: Studies on the evolution of elastin. II. Histology, Comp. Biochem. Physiol. [B], 66, 13-22, 1980 [39] TAKAGI,M., YAGASAKI,H., BABA,T. and BABA,H.: Ultrastructural visualization of carbohydrates in oxytalan fibers in monkey periodontal ligaments, J. Histochem. Cytochem., 33, 1007-1014, 1985 [40] BABA,T., TAKAGI,M. and YAGASAKI,H.: Ultrastructural cytochemistry of carbohydrates in microfibrils associated with the amorphous elastin in the monkey aorta, Anat. Rec., 213, 92

385-391, 1985 [41] COTTA-PEREIRA,G., GUERRA-RODRIGO,F. and BITTENCOURT-SAMPAIO,S.: Oxytalan, elaunin, and elastic fibers in the human skin, J. Invest. Dermatol., 66, 143-148, 1976 [42] COTTA-PEREIRA,G., GUERRA-RODRIGO, F. and DAVID-FERREIRA,J. F.: The elastic systemfibers. In Advances in Experimental Medicine and Biology, Vol. 79, Elastin and Elastic Tissue, 19-30, SANDBERG,L. B., GRAY,W. R. and FRANZBLAU,C., Eds., Pleunum Press, New York, U.S.A., 1977