immunohistochemical evidence fortheintracellular formation...
TRANSCRIPT
1267
0022-1554/80/12 126708$02.25The Journal of Histochemistry and Cytochemistrycopyright © 1980 by The Histochemical Society, Inc.
+
Vol. 28, No. 12, pp. 1267-1274, 1980Printed in U.S.A.
Immunohistochemical Evidence for the IntracellularFormation of Basement Membrane Collagen (Type IV) inDeveloping Tissues’
GORDON W. LAURIE, C.P. LEBLOND, ISABELLE COURNIL,
and GEORGE R. MARTIN
Department of Anatomy. McGill University, Montreal, Canada (G. W.L.; C.P.L.; IC.) and National institute of Dental Research.National institutes of Health. Bethesda. Maryland (G.R.M.
Received for publication February 4, 1980 and in revised form April 18, 1980 (OA 80-1 19)
Antibodies to type IV collagen obtained from the base-ment membrane of the mouse EHS tumor were incubatedwith sections of rat incisor teeth and other tissues forimmunostaining by direct or indirect methods. In alllocations, the immunostaining was pronounced in base-ment membranes in which it was restricted to the “basallamina” layer, from which “bridges” often extended tonearby basal laminae. Usually no immunostaining wasdetectable in the cells associated with the basement mem-branes. However, examination of the capillaries at theposterior extremity of the rat incisor tooth, where tissuesare at an early stage of development, showed immuno-staining not only of the basement membrane, but also ofthe endothelial cells. The staining was localized in roughendoplasmic reticulum cisternae, some Golgi saccules andtheir peripheral distensions, and structures believed to be
IntroductionBasement membranes are thin extracellular matrices separat-
ing connective tissue from epithelia, endothelia, muscle fibers,and the nervous system. Several components of basementmembrane have been identified, namely, type IV collagen
(22), the glycoproteins laminin (13, 41) and fibronectin (37),and a heparan sulfate proteoglycan (20).2 Physical chemicalstudies have indicated that type IV collagen is composed ofthree polypeptide chains comprising a triple helical domainresembling that in other collagens (22, 36, 40). Some studies
suggest that the polypeptide chains are of a single typereferred to as al(IV) ( 10, 22), while others indicate that two
‘Supported by grants from the Medical Research Council ofCanada.
2Hassell JR. Gehron Robey P, Barrack HJ, Wilizek J, Rennard SI,Martin GR: A basement membrane proteoglycan isolated from theEHS sarcoma. Manuscript in preparation.
secretory granules. These findings suggest that the syn-thesis of type IV collagen proceeds along the classicalsecretory pathway through rough endoplasmic reticulumand Golgi apparatus. At the same time, immunostainingwas usually lacking in the cells of the capillaries that hadmigrated about 2 mm away from the posterior end of theincisor tooth and also in the cells of most other tissuesexamined, even though the associated basal laminae werereactive. It is, therefore, presumed that the production oftype IV collagen may be high in cells at an early stage ofdevelopment and that any later production and turnoverofbasement membrane collagen can only be minimal.KEY WORDS: Type IV collagen; Anti-type IV collagen an-tibodies; Basal lamina; Basement membrane; Endothelial cell;Rough endoplasmic reticulum; Golgi apparatus.
different chains are present (4, 6, 42). It is not even knownwith certainty whether type IV collagen consists of one (22),two (12, 23, 42), or more proteins (35). However, mostinvestigators agree that the chains of type IV collagen are
larger than those of the other types and contain a lower
proportion ofglycine (9, 14, 18, 28), indicating that nontriplehelical sequences constitute a larger proportion of type IVthan ofother collagens.
Using radioautographic and immunohistochemical meth-ods, the biosynthetic pathway of the interstitial type I collagenin cells of bone and dentin has been traced from the roughendoplasmic reticulum (rER), through Golgi saccule disten-sions and secretory granules, to the extracellular matrix (21,
25, 44, 45). Similar studies have been attempted with base-ment membranes. Martinez-Hernandez et al. (27) immunized
rabbits with an aqueous extract of Reichert’s membrane-thebasement membrane from the parietal yolk sac. The Fab’fragment of the immunoglobulin obtained from these animals
was coupled to peroxidase and used to stain the cells of the
1268 LAURIE, LEBLOND, COURNIL, MARTIN
parietal yolk sac. Within these cells, the Fab’ was localized inthe rER cisternae but not in the Golgi saccules. The authorssuggested that basement membrane antigens were elaboratedin the rER and released directly through the cell membrane.However, it was difficult to interpret these results, since theantigens capable of reacting with the Fab’ had not been
identified. In other studies, antibodies to type IV collagen
were used for immunolocalization at the light microscopelevel in cultured aortic endothelial cells (19) and parietal yolksaccells (1).
Using the parietal yolk sac, Clark et al. (7) and Minor et al.(29) followed the synthesis, secretion, and deposition of
material that had incorporated 3H-proline. Radioactivity wasobserved over the cells at 30 mm and over the adjacentbasement membrane after 3 hr. Delorme and Grignon( 1 1 ) showed radioautographic reactions over the rER andGolgi saccules of endothelial cells after 3H-proline injection
and attributed them to the production of collagenous proteindestined for the capillary basement membrane.
In recent years, the mouse EHS sarcoma has been used as a
source of type IV collagen (31, 40). The collagenous material
extracted from the tumor consists of triple helical protein(s),from which chains of 160,000 and 140,000 daltons areobtained after denaturation and reduction of disulfide bonds(42). Larger forms of these two chains are also present in thetumor extract, but in very small amounts. Since studies with
‘4C-proline indicate that the larger forms are precursors of thesmaller ones (42), it is presumed that procollagen-likemolecule(s) are synthesized and later shortened to collagen-
like molecule(s). In common with other collagens, cross linksare then formed to give rise to dimeric and polymeric compo-
nents (22).Since present views are that the purified collagenous
material obtained from the tumor is composed of type IVcollagen with only traces ofprocollagen (42), it will no longer
be called type IV procollagen as was done in two previous
reports (8, 24) but will be referred to hereafter as type IVcollagen. This material served to prepare antibodies thatstrongly reacted with both the collagen-like and procollagen-
like fractions (42), but not significantly with laminin and
fibronectin (33). These antibodies were used to localize type
IV collagen to the basement membrane ofepidermis (47) andto those of enamel organ and blood vessels in the incisor tooth(8).
In this article, we have used antibodies to the tumor type
IV collagen to study the light and electron microscopicdistribution of type IV antigenicity in the basement mem-
branes and cells of incisor tooth and other tissues. Let us recall
that the basement membrane is considered by some authors tobe the fairly dense layer referred to by electron microscopists
as “basal lamina” (22). Others believe that the basement
membrane includes not only the basal lamina, but also a lucentlayer separating it from the adjacent cells, the “lamina lucida,”
and on the connective tissue side, a region of reticular fibers
(38, 47). Besides examining which of these structures are
immunostained, we have paid particular attention to the
adjacent cells. In most sites, these cells have been found to be
unreactive. However, immunostaining has been observed in
some cells, particularly the endothelial cells of the rapidly
growing capillaries at the posterior end of the incisor tooth.The stained sites within these cells have been identified andtaken to be the intracellular loci of the precursors of type IV
collagen.
Materials and MethodsType IV collagen was extracted from the mouse EHS tumor with 0.5M acetic acid and purified by salt precipitation (1.71 M NaCl) and ionexchange chromatography (40). This material closely resembled thetype IV collagen obtained from other sources as judged by amino acidcomposition and cyanogen bromide cleavage (42). In addition, typeIV antibodies prepared from the tumor cross-reacted with those fromthe parietal yolk sac (39). They immunostained authentic basementmembranes in all investigated tissues (8, 40, 47). These observationssuggest that the tumor type IV collagen is very similar, if not identical,to the type IV collagen in normal basement membranes.
The type IV collagen from the tumor was used to immunize NewZealand white rabbits. Anti-type IV antibodies were obtained andpurified by immunoabsorption according to the method of Nowack etal. (30) as modified by Cournil et al. (8). First sepharose beads werereacted with glutaraldehyde to form a protein binding complex. 0.8mg ofantigen was bound to the complex and, after ensuring by glycineaddition that no further binding sites remained, anti-type IV serumwas added. The bound antibodies were removed by 3 M sodiumthiocyanate and 0.4-0.6 mg of antibodies/mi were obtained. Theactivity of these antibodies was assessed by an indirect immunostain-ing method (described later in this section) using unfixed frozensections of the tooth. The antibodies were then linked to horseradish
peroxidase by the two step procedure of Avrameas and Ternynck (2)as described by Karim et al. (2 1) for immunostaining by the directmethod. To assess nonspecific staining, immunoglobulin fromnonimmunized rabbits was similarly linked to peroxidase.
Sherman rats aged 20±2 days were perfused with cool 5%formaldehyde in 0.08 M phosphate buffer at pH 6. The posterior 2mm end of each incisor tooth was separated and chopped on a SorvalTC-2 sectioner at about a 60- .cm thickness. The slices were cut eitheralong the length of the tooth or, less often, in cross section. The sliceswere first kept for 20 mm in a 30% solution of normal goat serum inphosphate buffered saline. They were then incubated overnight withperoxidase-linked anti-type IV antibodies (0.4 mg of IgG per ml),postfixed in 2.5% glutaraldehyde, stained for the peroxidase moietyof the antibodies using diaminobenzidine-hydrogen peroxide, treatedwith osmium tetroxide, and processed for Epon embedding, all aspreviously described by Karim et al. (21).
Indirect immunostaining with unlabeled antibodies was also usedin experiments in which the peroxidase-antiperoxidase (PAP) proce-dure was applied to unfixed frozen sections or to chopped slices offixed teeth and other tissues. On unfixed frozen sections the methodofCournil et al. (8) was used. Ten-micron thick sections on glass slideswere incubated overnight at 4#{176}Cwith purified antibodies (0.05 mg/mI)or with anti-type IV serum (1:20,000), followed the next day at roomtemperature by the PAP sequence, diaminobenzidine-hydrogenperoxide, osmium tetroxide, and mounting. With chopped slices, thePAP procedure was modified as follows. The slices were placed in asmall vial, exposed for 20 mm at room temperature to a 30% solutionof normal goat serum in phosphate buffered saline, incubated with a1:20,000 solution of anti-type IV serum overnight at 4#{176}C,then atroom temperature with a 1:50 solution of goat anti-rabbit globulinserum for 20 mm and, finally, a 1:50 solution of PAP for 5 mm, withwashes between each step. The slices were postfixed in glutaraldehyde
4
COLLAGEN IV ANTIGENICITY IN GROWING CAPILLARIES 1269
and then processed through diaminobenzidine and succeeding steps,as after incubation with peroxidase-linked anti-type IV antibodies.
Because of difficulties in selecting the posterior end of the toothwhere capillaries were frequently immunostained, this region waslocalized under the 10 or 40 objective of the light microscope. TheEpon block was stood on a glass slide on the microscope stage, so thatthe tissue end of the block was near the objective and was illuminatedby light passing through the block. In this way, it was possible tochoose the area of interest and trim away the rest. Then semithin(-0.5 jsm thick) and thin (-0.07 tm thick) Epon sections were cutand examined, respectively, in the light and electron microscope.
ResultsWhen tissues of 20-day-old rats were immunostained withantibodies to type IV collagen, basement membranes in
kidney, tooth, lung, intestine, and muscle showed a strong
reaction, but, with rare exceptions, the cells associated with
these membranes remained unstained. For example, in kidney
sections, staining was noted in the basement membranes of
proximal and distal tubules (Figure 1), glomeruli (not shown),
and blood vessels, but not in the epithelial cells of the tubules,
nor in the endothelial cells of blood vessels (Figure 1).
Figure 1. Light micrograph of a portion of the kidney of a 20-day-oldrat. The tissue was fixed by formaldehyde perfusion; 60-Mm thickslices were treated with antiserum to type IV collagen followed by thePAP sequence; they were then embedded in Epon. Semithin (0.5-gmthick) sections were cut from the surface of the embedded slices.Immunostaining can be seen in the basement membrane of theproximal tubules (P), which occupy most of the field, but as well in thebasement membrane of distal tubules (D), loops of Henle (H), andarterioles (A). However, the cells associated with these membranes donot contain immunoreactive material. x400.
‘, ‘ .
1 ‘ .
1’’
‘a... ‘
.0
., ‘ aa..
#{149} 0 1 ’ k#{149}
/
Figure 2. Light micrograph of a cross section of the rat incisor toothabout 2 mm anterior from the posterior end. 10-gm thick frozensections cut from a fresh, unfixed tooth were exposed to antiserumagainst type IV collagen and then to the PAP sequence. The pulp (P)contains reactive structures corresponding to the basement mem-branes of blood vessels; these include capillaries, which are scatteredthroughout, and thick walled arterioles, which are located along thevertical axis of the pulp. No staining is seen in odontoblasts (0) andameloblasts (A), nor in the thin layer ofdentin which separates them.
The periodontal tissue (P0) contains two layers of reactivecapillaries, one along its outer surface, and a more discrete oneadjacent to the extensive basement membrane of the enamel organ.This membrane, which partially encloses ameloblasts and the rest ofthe enamel organ (arrows), extends lingually as far as the apicalforamen, centrally located next to the lowest arrow; there, themembrane folds to return in a labial direction (thus enclosingHertwig’s root sheath) and ends between odontoblasts and amelo-blasts. x48.
Similarly, cross sections of rat incisor tooth located about 2
mm forward of the posterior end usually displayed immuno-staining of only the basement membranes of capillaries, ar-terioles, and enamel organ (Figure 2). At higher magnification,
it was possible in immunostained sections (Figure 3), but not
in controls (Figure 4), to distinguish the basement membraneof the enamel organ from that around nearby capillaries,
although a region of apposition existed where the two ap-peared to be continuous and the staining was darker. How-
ever, when this region was examined in the electron mi-croscope (Figure 5), two bands of stained material were
distinguished, one along the outer epithelium of the enamel
organ and the other lining the capillary endothelium. The
bands corresponded to the respective basal laminae. The laminalucida adjacent to the enamel epithelium was generally un-
#{149} ‘<t’ <’’i
1270 LAURIE, LEBLOND, COURNIL, MARTIN
Figures 3 and 4. Semithin (0.5-gm thick) sections of capillariesassociated with the basement membrane of the enamel organ. Thesections are taken from the surface of 60-jim thick slices offormaldehyde-fixed incisor teeth that have been cut in cross sectionabout 2 mm anterior to the posterior end, exposed to eitherperoxidase-labeled anticoilagen IV antibodies or to peroxidase-labeled nonimmune IgG, stained with diaminobenzidine-hydrogenperoxide and Epon embedded. x684. Figure 3 depicts a sectiontreated with peroxidase-labeled anticollagen IV antibodies. Theenamel organ is composed of the ameloblasts (A), stratum inter-medium (SI), stellate reticulum (SR), and outer enamel epithelium(Ep); it is separated from the periodontal tissue (P0) by a linecorresponding to the stained basement membrane of the enamelorgan (slanted arrows). The periodontal tissue includes threecapillaries (C) whose stained basement membranes appear to fuse withthat of the enamel organ. Neither the cells of the enamel organ, northe endothelial cells show any stain. The reactive basement membraneof a pericyte (vertical arrow) is associated with one capillary. Figure 4depicts a control section treated with peroxidase-labeied nonimmuneIgG. No reaction is seen around capillaries or along the limit betweenenamel organ and periodontal tissue, although there is slight nonspe-cific staining.
stained, except for some scattered reactions, while that of theendothelium was not always clearly identified. The two basal
laminae were joined on their connective tissue side by irregu-lar bands of various thicknesses, stained with the same inten-
sity. These connecting bands, referred to as “bridges,” did not
seem to correspond to reticular fibers but to strands of basallamina-like material. (Bridges were also observed in smallintestine between the basal laminae of epithelium and adjacentcapillaries.) Briefly then, both the outer enamel epitheliumand the endothelium had distinctly stained basal laminae,
whereas the epithelial and endothelial cells in the depicted
areas (Figures 3, 5) showed no, or only infrequent, reaction. Inthe rare cases of reaction, this consisted of a weak staining of
the rER content.
In contrast, the capillaries located exactly at the posterior
end of the tooth showed not only stained basement mem-
branes, but also many strongly reactive endothelial cells. Thus,the capillary in Figure 6 is surrounded by a band of stained
material corresponding to the basal lamina, whereas the
endothelial cell shows an equivalent deposition of stain within
the cisternae of the rER. In general, the stained material of the
Figure 5. Electron micrograph showing the region of apposition of aperiodontal capillary with the outer enamel epithelium. The micro-graph is taken from a region located about 2 mm anterior to theposterior end of a tooth fixed by formaldehyde perfusion. The toothwas cut cross-sectionally in 60-gm thick slices and exposed toperoxidase-labeled anticollagen IV antibodies, stained withdiaminobenzidine-hydrogen peroxide, and Epon embedded. Thinsections were then made from the surface of the slices. The region ofapposition contains the stained basal lamina (BL,) of the outer enamelepithelium (Ep) and the less regular, stained basal lamina (BL ) of thecapillary endothelium (En). Stretching between basal laminae aresimilarly stained bands ofvariable thickness, which have been termed“bridges” (Br). In the basal laminae and ‘bridges, immunostainingcan be distinguished as fine dots (arrowheads). The lamina lucida (11)by the outer enamel epithelium contains a few reactive dots, but isgenerally unstained. The adjacent cells and capillary lumen (L) do notshow staining. Mitochondria (M) and filaments (F) can be distin-guished in the epithelial cell. Occasionally staining of the plas-malemma is considered to be an artifact of DAB diffusion. x45,000.
COLLAGEN IV ANTIGENICITY IN GROWING CAPILLARIES 1271
Figure 6. Electron micrograph of acapillary located in the pulp of the pos-tenor end of a tooth fixed by formalde-hyde perfusion. 60-gm thick slices werecut, exposed to peroxidase-labeled an-ticollagen IV antibodies, and processedas for Figure 5. The capillary is lined byan endothelial cell containing a nucleus(N), cisternae of rER, most of whichcontain immunoreactive material, and afew mitochondria which are not stained.Surrounding the capillary is the im-munostained basal lamina (BL), whichdemarcates the adjacent connective tis-sue. Within the lumen of the capillary, aplatelet and a blood cell are unstained.x 12,600.
cisternae lines the wall, as may be seen at higher magnification
(Figure 7). In the Golgi apparatus depicted in this figure,
immunostaining may be observed in some of the saccules and
their peripheral distensions, as well as in small spherical
structures that may be intermediate vesicles (Figure 7). Near
the lumen of the capillary, a stained structure is observed that
may be a secretory granule. The other organelles, particularlynucleus, mitochondria, and cell coat (glycocalyx), do not take
up the stain. Blood cells and platelets in the capillaries are not
stained (Figure 6). Control sections exposed to nonimmune
IgG are not stained to any significant degree.Taken together, the results indicate that the occurrence of
type IV antigenicity is readily demonstrated in the basal
lamina of basement membranes of endothelia and epithelia, as
well as in bridges connecting them. Type IV antigenicity may
also occur in the rER cisternae and Golgi apparatus of the cells
in contact with basement membranes, but only in exceptional
cases, such as observed in the endothelial cells at the posterior
end of the incisor tooth.
DiscussionThe biosynthesis of collagen types I, II, and III occurs inseveral distinct steps (16). Within producer cells, the polypep-
tide chains of these proteins are synthesized in precursorforms, which are known after removal of the signal peptides aspro a chains. Groups of three chains coil into a triple helix to
form the procollagen molecule. This is later released toextracellular spaces where specific proteases cleave peptides
from the N and C termini to give rise to the collagen molecule(17, 26).
The use of radioautography after 3H-proline injection incombination with the immunohistochemical localization of
type I procollagen, has demonstrated that the pathway of
biosynthesis is the same as that of secretory proteins, proceed-ing from the rER to the Golgi apparatus and from there to
secretory granules that release their content by exocytosis (21,
25, 44,45).
The goal of the present study was to identify the intracellu-
1272 LAURIE, LEBLOND, COURNIL, MARTIN
3Smith CE, private communication, Department of Anatomy,McGill University, Montreal, Canada.
Figure 7. Electron micrograph of aportion of an endothelial cell from aperiodontal capillary located at theposterior end of a formaldehyde-perfused tooth. The procedure usedwas identical to that of Figure 5, ex-cept that the original treatment wasanti-collagen IV serum followed bythe PAP sequence. Immunostainingmay be seen in the cisternae of rER,some of the saccules in several Golgistacks (G), a Golgi-associated struc-ture (Sd), which may be a sphericaldistension, and a secretory granule-
like structure (Sg) near the lumen (L)of the capillary. Arrows point to smallimmunostained structures that may beGolgi intermediate vesicles. The im-munoreactive basal lamina (BL) maybe distinguished along the border ofthe endothelial cell. Other compo-nents of the cell, including the nucleus(N), are not stained. x31,680.
lar steps followed by the precursors of type IV collagen duringthe elaboration of basement membranes. For this purpose,antibodies were prepared against type IV material isolatedfrom the EHS mouse sarcoma and purified by affinitychromatography. The specificity of these antibodies wasshown in several ways. They reacted with basement mem-branes in all tissues examined (8, 40). Part of the antibodiesrecognized collagenous sequences in the type IV protein(s),while others reacted with the noncollagenous sequences (39);as a result, they could bind to both the collagen-like and theprocollagen-like fractions of the tumor material (42). Hence,they would be, expected to react with intracellular precursorforms, presumably pro a chains and procollagen, as well aswith the extracellular definitive form of type IV collagen.Finally, the properties of type IV materials were the same
whether they were extracted from the EHS tumor or from theparietal yolk sac, as mentioned previously (42). In the presentstudy, the basement membranes of kidney tubules, blood
vessel endothelium, enamel organ, epithelia of intestine andlung, striated muscle, etc. were found to be immunoreactive.
In these cases, examination of the stained basement mem-brane in the electron microscope showed that the basallamina portion rather than the lamina lucida was stained.Moreover, neighboring basal laminae were connected by“bridges” of material with the same immunoreactivity. Similarstructures had been observed by others using ruthenium redstaining(3, 34) or even routine preparations (15).
Unexpectedly, few of the cells in apposition to basementmembrane were stained. Thus, in the case of the incisor tooth,the cells associated with the basement membranes of enamelorgan and blood vessels were usually unstained. However, itmay be recalled here that the immunostaining was done on60-j.tm chopped slices, in which only the organelles opened bythe chopper knife could show a reaction. Hence, a negative
result could be due not only to the absence ofantigen, but alsoto a lack of entry of the reagents. Nevertheless, intracellularreactions were common at the posterior extremity of thetooth, but rare or absent 2 mm anteriorly. It is known that, atthe posterior end, new tooth tissues are being formed rapidlyand, in particular, the endothelial cells of the growingcapillaries have a high mitotic index.3 Moreover, the local
increase in cell number, cell size, and interstitium results indisplacement of tooth tissue in an anterior direction. There-fore, the endothelial cells of capillaries are at an early stage ofdevelopment in the posterior end and at a more advancedstage just 2 mm anteriorly. The immunostaining in posteriorend cells is localized to the cisternae of the rER, some Golgisaccules and distensions, and structures believed to be se-cretory granules. These findings suggest that many of theendothelial cells in this region of the tooth are actively
synthesizing type IV collagen. Since rER, Golgi apparatus,and, probably, secretory granules are reactive in these cells, it
is likely that the synthesis of type IV material follows the
pathway demonstrated earlier for type I procollagen (21).Preliminary observations of intracellular reactions in young
ameloblasts and smooth muscle cells lead to the same conclu-
sion.The endothelial cells located 2 mm forward of the posterior
end ofthe incisor tooth are usually unstained. The rare stainedcells found in this region show less reactive material in theirrER cisternae and presumably produce less type IV collagenthan the most posterior endothelial cells. In fact, most of thecells found 2 mm anteriorly, as well as the cells in the othertissues examined, are unstained and, therefore, do not pro-
COLLAGEN IV ANTIGENICITY IN GROWING CAPILLARIES 1273
duce any type IV or produce it at a level too low for detection
by the method used.
Presumably, the active synthesis of type IV collagen re-
quired to build up the basal lamina is a transient activity of the
producer cells that occurs at an early stage of development.
Thereafter, the immunostaining of the basement membrane
remains, but the cells are no longer stained. It is likely that the
collagenous part of the basement membrane persists in situ
with minimal addition from the cells. There is evidence thatsmall amounts of type IV collagen are being added continually
to the basement membrane of bovine retinal vessels (46) and
kidney glomerulus (5). Nevertheless, the turnover is minimal,
since rat basement membranes have been assigned a turnovertime ofat least 100 days (32) or even a year (43). Briefly then,
any addition of new collagen to basement membrane can only
occur at a very slow rate.
Literature Cited1. Adamson ED, Ayers SE: The localization and synthesis of some
collagen types in developing mouse embryos. Cell 16:953, 1979
2. Avrameas 5, Ternynck T: Peroxidase labeled antibody and Fabconjugates with enhanced intracellular penetration. Im-munocytochemistry 8:1175, 1971
3. Bunge MB, Williams AK, Wood PM, Uitto J, Jeffrey JJ: Com-parison of nerve cell and nerve cell plus Schwann cell cultures,with particular emphasis on basal lamina and collagen formation. JCell Biol 84:184, 1980
4. Burgeson RE, El Adli FA, Kaitila II, Hollister DW: Fetalmembrane collagens: identification of two new collagen alphachains. Proc Natl Acad Sci USA 73:2579, 1976
5. Chow AYK, Drummond KN: Incorporation and hydroxylationof proline-3-4-H3 as an index ofglomerular basement membranesynthesis in normal and nephrotoxic nephritic rats. Lab Invest20:213, 1969
6. Chung E, Rhodes RK, Miller EJ: Isolation of three collagenouscomponents of probable basement membrane origin from severaltissues. Biochem Biophys Res Commun 71: 1 167, 1976
7. Clark CC, Tomichek EA, Koszalka TR, Minor KR, KefalidesNA: The embryonic rat parietal yolk sac. The role of the parietalendoderm in the biosynthesis of basement membrane collagenand glycoprotein in vitro.J Biol Chem 250: 5259, 1975
8. Cournil I, Leblond CP, Pomponio J, Hand AR, Sederlof L,Martin GR: Immunohistochemical localization ofprocollagens. I.Light microscopic distribution of procoliagen I, III and IVantigenicity in the rat incisor tooth by the indirect peroxidase-antiperoxidase method.J Histochem Cytochem 27:1059, 1979
9. Dechenne C, Foidart-WillemsJ, Mahieu P: Collagen biosynthesisin cultures ofepithelial cells isolated from malignant hypertensivekidneys.J Submicrosc Cytol 8:101, 1976
10. Dehm P, Kefalides NA: The collagenous component of lensbasement membrane. The isolation and characterization of an achain size collagenous peptide and its relationship to newlysynthesized lens components. J Biol Chem 253:6680, 1978
1 1. Delorme P. Grignon G: An autoradiographic study on thebiosynthesis of the capillary basal lamina in the chick embryotelencephalon. Cell Tissue Res 192:299, 1978
12. Dixit SN: Isolation and characterization of two collagenouscomponents from anterior lens capsule. FEBS Lett 85:153, 1978
13. FoidartJM, Bere EWJr, Yaar M, Rennard SI, Gullino M, MartinGR, Katz SI: Distribution and immunoelectron microscope
localization of laminin, a noncollagenous basement membraneglycoprotein. Lab Invest 42:336, 1980
14. Foidart-Willems J, Dechenne C, Mahieu P: Biosynthesis ofbasement membrane collagen in cultures of renal glomerular andtubular epithelial cells. Diabetes Metab 1:227, 1975
1 5. Garant PR, Gillespie R: The presence of fenestrated capillaries inthe papillary layer ofthe enamel organ. Anat Rec 163:71, 1969
16. Gay 5, Miller EJ: Collagen in the Physiology and Pathology ofConnective Tissue. Gustav Fischer Verlag, New York, 1978, p1-110
17. Goldberg B, Taubman MB, Radin B: Procollagen peptidase: itsmode ofaction on the native substrate. Cell 4:45, 1975
18. Grant ME, Schofield JD, Kefalides NA, Prockop DJ: Thebiosynthesis of basement membrane collagen in embryonic chicklens. III. Intracellular formation of the triple helix and theformation of aggregates through disulfide bonds. J Biol Chem248:7432, 1973
19. Howard BV, Macarak EJ, Gunson D, Kefalides NA: Characteri-zation of the collagen synthesized by endothelial cells in culture.Proc Natl Acad Sci USA 73:2361, 1976
20. Kanwar YS, Farquhar MG: Presence of heparan sulfate in theglomerular basement membrane. Proc Natl Acad Sci USA76:1303, 1979
21. Karim A, Cournil I, Leblond CP: Immunohistochemical localiza-tion of procollagens. II. Electron microscopic distribution ofprocollagen I antigenicity in the odontoblasts and predentin of ratincisor teeth by a direct method using peroxidase linked an-tibodies.J Histochem Cytochem 27:1070, 1979
22. Kefalides NA, Alper R, Clark CC: Biochemistry and metabolismofbasement membranes. mt Rev Cytol 61:167, 1979
23. Kresina TF, Miller EJ: Isolation and characterization of basementmembrane collagen from human placental tissue. Evidence forthe presence of two genetically distinct collagen chains.Biochemistry 18:3089, 1979
24. Laurie G, Cournil I, Martin GR: Ultrastructural localization oftype IV procollagen antigenicity in the basal laminae (basementmembranes) of enamel organ and blood vessels of the rat incisortooth. Anat Rec (Proc) 193:600, 1979
25. Leblond CP, Wright GM: Steps in the elaboration ofcollagen byodontoblasts and osteoblasts. In Basic Mechanisms of CellularSecretion. Edited by AR Hand, C Oliver. Academic Press, NewYork, in press
26. Leung MKK, Fessler LI, Greenberg DB, Fessler JH: Separateamino and carboxy procollagen peptidases in chick embryotendon.J Biol Chem 254:223, 1979
27. Martinez-Hernandez A, Nakane PK, Pierce GB: Intracellularlocalization of basement membrane antigen in parietal yolk saccells. AmJ Pathol 76:549, 1974
28. Minor RB., Clark CC, Strause EL, Koszalka TR, Brent RI,Kefalides NA: Basement membrane procollagen is not convertedto collagen in organ cultures of parietal yolk sac endoderm. J BiolChem2Sl:1789, 1976
29. Minor RR, Hoch PS, Koszalka TR, Brent RI, Kefalides NA:Organ cultures of the embryonic rat parietal yolk sac. I. Mor-phologic and autoradiographic studies of the deposition of thecollagen and noncollagen glycoprotein components of basementmembrane. Dev Biol 48:344, 1976
30. Nowack H, Gay 5, Wick G, Becker U, Timpl R: Preparation anduse in immunohistology ofantibodies specific for type I and typeIII collagen and procollagen.J Immunol Methods 13:117, 1976
31. Orkin RW, Gehron P, McGoodwin EB, Martin GR, Valentine T,Swarm R: A murine tumor producing a matrix of basementmembrane.J Exp Med 145:205, 1977
1274 LAURIE, LEBLOND, COURNIL, MARTIN
32. Price RB, Spiro RG: Studies on the metabolism of the renalgI omerular basement membrane. Turnover measurements in therat with the use of radiolabeled amino acids. J Biol Chem252:8597, 1977
33. Rennard SI, Berg R, Martin GR, Foidart JM, Gehron Robey P:Enzyme linked immunoassay (ELISA) for connective tissue com-ponents. Anal Biochem 104:205, 1980
34. Sanders EJ: Development of the basal lamina and extracellularmaterials in the early chick embryo. Cell Tissue Res 198:527,1979
35. Sato T, Spiro RG: Studies on the subunit composition of the renalglomerular basement membrane. J Biol Chem 251:4062, 1976
36. Schwartz D, Veis A: Structure of bovine anterior lens capsulebasement membrane collagen molecules from electron micro-scopy. Biopolymers 18:2363, 1979
37. Stenman 5, Vaheri A: Distribution of a major connective tissueprotein, fibronectin, in normal human tissues. J Exp Med147:1054, 1978
38. Threadgold LT: The Ultrastructure of the Animal Cell. PergamonPress, Oxford, l976,p 150-151
39. Timpl R, Glanville RW, Wick G, Martin GR: Immunochemicalstudies on basement membrane (type IV) collagens. Immunology38:109, 1979
40. Timpl R, Martin GR, Bruckner P, Wick G, Wiedemann H:Nature of the collagenous protein in a tumor basement mem-brane. EurJ Biochem 84:43, 1978
41. Timpl R, Rhode H, Gehron Robey P, Rennard SI, Foidart JM,Martin GR: Laminin-A glycoprotein from basement mem-branes. J Biol Chem 254:9933, 1979
42. Tryggvason K, Gehron Robey P. Martin GR: Biosynthesis oftype IV procoilagens by tumor and parietal yolk sac tissue inmouse. Biochemistry 19:1284, 1980
43. Walker F: The origin, turnover and removal of glomerularbasement membrane. J Pathol 1 10:233, 1973
44. Weinstock M, Leblond CP: Synthesis, migration and release ofprecursor collagen by odontoblasts as visualized by radioautog-raphy after 3H-proline administration. J Cell Biol 60:92, 1974
45. Weinstock M: Elaboration of precursor collagen by osteoblasts asvisualized by radioautography after 3H-proline administration. InExtracellular Matrix Influences on Gene Expression. Edited byHC Slavkin, RC Greulich. Academic Press, New York, 1975, chl3,p 119-128
46. White R, Carlson EC, Brendel K, Meezan E: Basement mem-brane biosynthesis by isolated bovine retinal vessels: incorpora-tion of precursors into extracellular matrix. Microvasc Res 18:185, 1976
47. Yaoita H, Foidart JM, Katz SJ: Localization of the collagenouscomponent in skin basement membrane. J Invest Dermatol70:191, 1978