Surface ultrastructure of the isolated avian notochord in vitroThe effect of the perinotochordal sheath.код для вставкиСкачать
THE ANATOMICAL RECORD 197:257-276 11980) Surface Ultrastructure of the Isolated Avian Notochord In Vitro: The Effect of the Perinotochordal Sheath EDWARD C. CARLSON ANU M. CRISTINA KENNEY Department of Human Anatomy, University of California, Davis, California 9561 6 ABSTRACT The perinotochordal sheath (PNS) is a "tube" of extracellular matrix (ECM) that surrounds the avian notochord beginning in the second day of development. Somites, like the notochord, derive from chordamesoblast but are encased by a less substantial perisomitic matrix (PSM). Initially both tissue types exhibit epithelioid characteristics. Somitic cells subsequently disperse, however, while notochordal histoarchitecture is maintained until much later. To test the possible shape-preserving role of the PNS, notochords were isolated from chick embryos by homogenization (which retains the sheath) or by trypsinization (which removes the sheath). Somites were similarly isolated. Tissues were cultured 12-72 hours and studied by LM, SEM and TEM. Mechanically isolated notochords are initially rigid with smooth surfaces. During the culture period a few cells grow outward from cut ends of the notochord, but its overall rod shape and intact PNS are maintained. In contrast, uncultured trypsinized notochords a r e flaccid, denuded cylinders with numerous cytoplasmic blebs. They adhere to the substratum within 12 hours of culture when a few cells break away from the central tissue rod, migrate laterally, and appear mesenchymal. This cellular dispersion is directional (perpendicular to the long notochordal axis) and continuous (up to 72 hours). At this time a flattened ovoid growth area is formed. Cultured somites form flat circular growth areas within 12 hours of culture irrespective of the isolation method. These data suggest that the maintenance of a n epithelial configuration by notochords in vivo may be due in part to physical restraints of the PNS. It seems possible that notochordal secretions (manifested by the formation of a PNS) could result in its compartmentation and axial confinement while its unrestrained somitic relatives are free to disperse. The early chick embryo is particularly well- Histochemical and enzyme digestion studies suited for studies of notochordal morphogene- of t h e sheath demonstrate t h e presence of sis (Jurand, '62; Bancroft and Bellairs, '76). acid mucopolysaccharides (O'Connel a n d This structure is of special interest because it Low, '70; Frederickson and Low, '71; Ruggeri, induces neural plate formation from overly- '70, '72) and collagen (Frederickson and Low, ing ectoderm (Holtfreter, '68) and chondro- '71; Kenney and Carlson, '78). This latter obgenesis in somitic mesoderm (Lash, '63; 68a, servation is consistent with biochemical inb, c). Therefore it has been the subject of nu- vestigations which show that a cartilage-like merous morphological and biochemical inves- collagen (type 11) is associated with the nototigations. chord in vivo (Miller and Mathews, '74). At the level of transmission electron microThe P N S was first described by Duncan scopy (TEM), extracellular matrix (ECM) - ('57), but its origin was at first controversial including basal lamina, microfibrils, a n d because of its association with a n epithelioid other amorphous materials - forms a perino- structure prior to the differentiation of sectochordal sheath (PNS) surrounding the noto- ondary mesenchyme from t h e sclerotome. chord beginning in the second day of development (Low, '68; Bancroft and Bellairs, '76). Received July 11, 1979; Accepted November 21, 1979. 0003-276X/80/1972-0257$03.50 0 1980 ALAN R. LISS, INC. 257 258 EDWARD C. CARLSON AND M. CRISTINA KENNEY More recently Carlson and co-workers ('74) showed that cultured mesenchyme-free, isolated notochords produced connective tissue materials that were ultrastructurally indistinguishable from those which comprised the PNS in vivo. This "reappearance" of ECM in vitro h a s been confirmed by other studies (Carlson a n d Upson, '74; L a u s c h e r a n d Carlson, '75; Kenney and Carlson, '78) and it is now generally believed that the notochord is the progenitor of its own sheath. Other in vitro studies (Hay and Meier, '74; Kosher and Lash, '75) have shown t h a t t h e notochord s y n t h e s i z e s s u l f a t e d glycosaminoglycans (GAGS) and collagen(s), t h e majority of which consists of three identical a1 chains (Linsenmayer et al, '73). These results are consistent with our recent studies which demonstrate the production of ruthenium red-positive, hyaluronidase-sensitive ECM, and striated collagen fibrils by notochordal organ cultures within three days of incubation (Kenney and Carlson, '78). It is interesting t h a t during the development of the PNS, the notochord becomes more rod-like and remains confined to its axial position until its eventual demise. This is particularly noteworthy in light of the concomit a n t disorganization and dispersion of adjacent somitic cells. Like the notochord, somites are derived from chordamesoblast and appear as epithelial structures early in development. They are surrounded by components of ECM similar to those which comprise the PNS. In contrast, however, somites do not remain compartmentalized i n segmental tissue blocks. Rather, they organize further to form sclerotome, myotome and dermatome subdivisions, each of which eventually disperses to give rise to various mesodermal derivatives. Somitic mesoderm is a t least a s active a s the notochord in the production of ECM both in vivo (Kvist and Finnegan, '70; O'Connel and Low, '70; Olson and Low, '71; O'Hare, '72a, b; Minor, '73; Lipton and Jacobson, '74; von der Mark et al, '76) and in vitro (Lash, '68b; OHare, '72a, b; Orkin e t al, '73; Gordon and Lash, '74; Lash and Vasan, '78). Furthermore, the ECM is composed of GAGS (Minor, '73; O'Hare, '73; Gordon and Lash. '74; Lash and Vasan,'78) as well as collagen (von der Mark et al, '76), which are similar to those produced by the developing notochord. Unlike the PNS, however, perisomitic matrix (PSM) does not continue to a c c u m u l a t e a t t h e somitic external surface, nor does it restrict subsequent dispersion of somitic cells. Indeed most current embryology textbooks consider somitic break-up and dispersion a necessary prerequisite to mesodermal differentiation. The present study examines the capability of the intact PNS to prevent or permit notochordal cell migration in a n in vitro system and compares the results with similar data derived from cultured somites. Light (LM), scanning electron (SEMI, and transmission electron (TEM) microscopic data are correlated to demonstrate t h e surface morphology and histoarchitecture of t h e PNS. The primary aim is to provide evidence for the maintenance of t h e notochordal epithelioid rod shape in terms of physical restriction and compartmentation by the PNS during early embryogenesis. MATERIALS AND METHODS Tissue preparation and organ culture Fertile White Leghorn hens' eggs were incubated 24-60 hours a t 37°C (60% relative humidity). Sterile filter paper rings aided removal of embryos from the egg as previously described (Lauscher and Carlson, '75). Embryos were than transferred to sterile Hank's Balanced Salt Solution (HBSS) and staged according to Hamburger and Hamilton ('51). Cross-sectional scissor cuts were made caudal to the heart and rostra1 to terminal somites on embryos of the same or similar stages. Intermediate t r u n k segments were placed in a) cold (4°C) Karnovsky's ('65) fixative, b) cold Earle's Balanced Salt Solution (EBSS) buffered with 0.028 M HEPES, or c) cold tryspin (1% in HBSS). Notochords and somites were mechanically isolated from trunk segments in "b" (above) by mild homogenization in a hand-held glass tissue homogenizer with specially machined, conical (small end directed upward) teflon pestle. The homogenate was passed over a 153-pm mesh nylon sieve and washed generously with buffered EBSS. The residue was rinsed into a 10-cm petri dish, and somites and fragments of isolated notochords were removed and collected with a Pasteur pipette. Trunk segments in "c" (above) were trypsinized for 25 minutes and then flushed gently through Pasteur pipettes to isolate somites and notochords.These were collected separately and washed three times in 15% fetal calf serum to inactivate the trypsin. All isolated tissues were incubated in Falcon plastic 35-mm tissue culture dishes in M-199 culture medium supplemented with 2% fetal bovine serum, 8% bovine serum, ascorbic acid (100 pg/ml), penicillin-streptomycin (100 units/ PERINOTOCHORDAL SHEATH IN VITRO ml), fungizone (2.5pgiml), 10% tryptose phosphate broth and NaHCO, (2.2 mg/ml). Light and transmission electron microscopy All t r u n k segments, notochords, and somites were fixed one hour in cold Karnovsky's ('65) fixative (0.2 M sodium cacodylate buffer, pH 7.4). Cultured tissues were rinsed three times in buffered EBSS prior to treatment with the same fixative. Phase-contrast micrography was carried out on cultured tissues during the aldehyde fixation using an Olympus IMT inverted microscope. All tissues were post-fixed one hour at room temperature with 2% OsO, (0.144 M cacodylate buffer, pH 7.4) prior to dehydration in a n ascending series of graded ethanols. Uncultured isolated notochords or somites were surrounded by several drops of warm 2% Nobel agar in order to facilitate their manipulation. These agar-embedded tissues were exposed t o propylene oxide for additional dehydration. Tissues were embedded in a n EponAraldite mixture (Anderson and Ellis, '65) and cured overnight a t 37°C and for an additional 48 hours a t 60°C. Sections 1 p thick were stained with toluidine blue (1% in 1% sodium borate) and observed and photographed with an Olympus F H photomicroscope. Thin sections (500-800 A) were placed on uncoated 200- or 300-mesh copper grids prior to staining with uranyl acetate (5% in absolute ethanol) and lead citrate (Venable and Coggeshall, '65). These were observed and recorded a t original magnifications of 4,600 to 35,000 diameters in a Philips EM 400 transmission electron microscope. Scanning electron microscopy Uncultured notochords were fixed one hour a t room temperature i n aldehyde fixative (Karnovsky, '651, post-fixed in OsO, (2% in 0.144 M sodium cacodylate buffer), and dehydrated in a series of graded ethanols. Notochords were critical point-dried in CO, and mounted on aluminum specimen stubs. They were coated with evaporated carbon and gold, and observed with a Philips 501 scanning electron microscope. OBSERVATIONS In vivo trunk segments Cross sections through t r u n k segments (seventh somite level) from chick embryos a t stage 1 0 show prominent centrally located neural tubes (fig. 1).The recently closed dorsal surface is closely associated with surface 259 ectodermal cells a n d a transitory surface groove is present in t h e dorsal embryonic midline. The neural tube is flanked laterally by block-shaped somites which are separated from surrounding tissues by a cell-free connective tissue space. Somites are composed of irregular columnar cells surrounding a n illdefined lumen in a rosette configuration. They have not yet organized into sclerotome, dermatome and myotome subdivisions. The notochord is unvacuolated and appears as a rod of cells, flattened dorsoventrally and located in the midline approximately equidistant between the overlying neural tube and ventral developing entoderm. A primitive but continuous endothelium forms the walls of paired aortae in these specimens. At t h i s stage and segmental level the embryo is composed entirely of epithelial structures separated from a single uncompartmented tissue space. Electron micrographs show t h a t the perinotochordal area is occupied by electrondense extracellular connective tissue materials including a discontinuous basal lamina (fig. 2). These matrix components, including microfibrils, i n t e r s t i t i a l bodies and basal lamina, constitute the PNS. By stage 15, the dorsal ectoderm is completely separated from the underlying neural tube, while neural crest cells stream ventrad along its lateral margins (fig. 3). Somitic cells reorganize to form compact dorsolateral dematomyotomes and more dispersed ventromedial sclerotomes. Identical magnification micrographs show that at this stage sclerotoma1 cells have not begun to disperse. The entodermal lining moves ventrad away from the developing notochord, resulting in a n appare n t enlargement in t h e connective tissue space. Cross-sectioned notochords show small vacuoles, and demonstrate a more circular profile t h a n a t earlier stages. At this stage notochordal basal lamina is relatively continuous, and numerous small (5-10 nm) microfibrils are present within the PNS (fig. 4). The total area occupied by the sclerotome is enlarged dramatically a t stage 17 (primarily by increased intercellular spaces), and sclerotomal cells are closer to the notochord than at earlier stages (fig. 5 ) . Such proximity is probably a function of increased notochordal diameter except in the area ventral to the notochord, where mesenchymal cells (possibly resulting from sclerotomal proliferation and migration) a r e associated with aortic endothelium. The notochord increases in diameter and vacuoles are enlarged, but its crisp circular profile and close proximity to t h e 260 EDWARD C. CARLSON AND M. CRISTINA KENNEY PERINOTOCHORDAL SHEATH IN VITRO overlying n e u r a l t u b e a r e relatively u n changed. The PNS, including the subjacent b a s a l l a m i n a , i s s i m i l a r t o t h a t seen i n stage-15 embryos (fig.4) but appears denser than those seen at stage 10 + (fig. 2). A remarkable increase in notochordal girth between stages 17 and 20 (compare figs. 5 and 7) probably results in its lateral association with sclerotomal cells. However, similar cells are also present in the embryonic midline immediately dorsal and ventral to the developing notochord (fig. 7). These late-arriving midline cells from the sclerotome occupy a space which has been previously cell-free (fig. 1 , 3 , 5). At stage 20 more t h a n half of t h e total cross-sectional a r e a of t h e notochord is occupied by spherical vacuoles (fig. 7). Its rodlike configuration is maintained and its remarkably circular shape in cross section sug- 261 gests confinement to a limited space by an external retainer. The PNS is more electrondense at this stage than a t shorter incubation times (fig. 8). Its basal lamina is composed of a lamina densa (3G40 nm thick) separated from subjacent notochordal cells by a clear 20-30 nm area. Microfibrils parallel the notochordal surface in the adjacent connective tissue space. Somite isolation and culture In an effort to demonstrate the possible effects of perisomitic matrix (PSM) on growth patterns of isolated somites in vitro, these tissues were isolated from stage-12-14 chick embryos either by mechanical methods (to retain the PSM) or by trypsinization (to remove the PSM) and cultured as described in Materials and Methods. Following isolation by mild homogeniza- Fig. 1.Light micrograph of cross section through seventh somite of chick embryo a t stage l o + . At this stage embryos are composed almost entirely of epithelial tissues including ectoderm (ect), entoderm (ent), neural tube (NT), notochord (N), somites (S), and primitive endothelium of the developing aortae (a).Toluidine blue. x 220. Fig. 2. Electron micrograph of notochord surface i n area shown by arrows in figure 1. Intermittent basal lamina (bl) material surrounds the notochord (N) and small ( 5 C 1 0 0 A) microfibrils embedded in electron lucent ground substance forms the early perinotochordal sheath (PNS). x 21,000. Fig. 3. Light micrograph of cross section through seventh somite of chick embryo a t stage 15. The notochord ( N ) appears more cylindrical than at earlier stages and the connective tissue space interposed between notochord, somites, and underlying entoderm is increased in volume (compare with figure 1).At this time the ventromedial portion of the somite or sclerotome (SC) has hegun to disperse. Somite remnant h a s moved dorsolaterally. Aortae, a; toluidine blue. x 220. Fig. 4. Electron micrograph of notochord (N) surface in area shown by arrows in figure 3. Perinotochordal sheath (PNS) consists of continuous basal lamina (bl) and numerous microfibrils (mfb). These completely surround the notochord. x 21,000, Fig. 5. Light micrograph of cross section through chick embryo at stage 17 dorsal and j u s t rostral to the heart. Notochord (N) is increased in diameter compared to earlier stages. Intercellular spaces in the sclerotome (SC) have enlarged, and some sclerotomal cells approach the midline between notochord and underlying aortae ( a ) . Large vacuoles i n notochordal cells a r e normal f e a t u r e s a t t h i s incubation age. Toluidine blue. x 220. Fig. 6. Electron micrograph of notochord (N) surface i n area shown by arrows in figure 5. By stage 17 the perinotochordal sheath (PNS) consists of numerous microfibrils arranged parallel to and associated with the perinotochordal basal lamina (bl). x 21,000. Fig. 7 . Light micrograph of cross section through embryo a t stage 20 dorsal and just rostral to the heart. Notochord (N) is highly vacuolated and isenlarged in diameter by nearly double compared with stage 17 (see figure 5). The notochord maintains its axial position and cylindrical configuration. The once-epithelial somite is now highly disorganized, and sclerotomal cells (SC) merge i n the midline (dorsal and ventral to the notochord) with cells from the corresponding contralateral segment. Aorta, a. Toluidine blue. x 220. Fig.8. Electron micrograph of notochord (N)surface in area shown by arrows in figure 7 . Perinotochordal basal lamina (bl) is continuous and exhibits clearly demarcated lamina densa and lamina lucida. Perinotochordal sheath (PNS) contains dense concentrations of 15&200 A microfibrils (arrows) and amorphous materials. x 21,000. 262 EDWARD C. CARLSON AND M. CRISTINA KENNEY tion and sieving, somites from stage-12-14 embryos appear as condensed tissue blocks with smoothly contoured external surfaces and average 230 x 330 pm (fig. 9). By TEM, basal surfaces of somitic cells are covered by a continuous basal lamina surrounded by an electron-lucent matrix t h a t contains faintly cross-banded, 15-nm microfibrils (fig. 10). Somitic cells are undifferentiated and show little smooth- or rough-surfaced endoplasmic reticulum. Their peripheral cytoplasm is composed primarily of polyribosomes. When these mechanically isolated somites are submerged in embryo extract-free culture medium and grown 12 hours on Falcon plastic, they do not remain compartmentalized by PSM. Rather, somitic cells become stellate, disaggregate rapidly, and migrate radially on the substrat u m forming a circle-shaped growth a r e a (fig. 11). When somites are isolated by a brief treatm e n t w i t h t r y p s i n followed by f l u s h i n g through Pasteur pipettes, they appear approximately equal in overall size and shape but more translucent than their mechanically isolated counterparts. Moreover, trypsinized somites show convex surfaces virtually covered with small blebs. TEM demonstrates t h a t cells with these cytoplasmic evaginations show no evidence of cytotoxicity and t h a t t h e enzyme treatment completely re- moves adjacent PSM (fig. 13). After 12 hours of culture these denuded somites exhibit growth patterns t h a t a r e indistinguishable from those of somites isolated by mechanical m e a n s (fig. 1 4 ) . I n both c u l t u r e t y p e s , fibroblast-like somitic cells migrate peripherally from the central cell cluster and exhibit approximately equal mobility in all directions (compare figs. 11 and 14). Isolation and culture of mechanically isolated notochords In this group of experiments we endeavored to determine whether notochords isolated and cultured with the PNS intact might retain this sheath and if the presence of the PNS could play a role in maintaining notochordal shape under in vitro conditions. Notochords isolated by a hand-held tissue homogenizer and passed over nylon sieves average 20&1,500 p m in length. Typically they express resiliency and resist mechanical distortion. I t is not possible to prepare completely clean cylindrical notochord segments by this isolation procedure, and those grossly contaminated with mesenchymal cells were discarded. Following preparation for SEM, isolated notochords appear as rigid rods with slight regional constrictions (fig. 15). In these preparations small, irregular surface projections Fig. 9. Bright field micrograph of unfixed somite (S) isolated from stage-12 chick embryo by mild homogenization and sieving. Somites maintain the rosette configuration and exhibit smooth external surfaces (arrows). x 160. Fig. 10. Electron micrograph of surface of mechanically isolated somite (S)similar to that shown in figure 9. Basal lamina material (bl) surrounds t h e somite . A few microfibrils which exhibit faint irregular striations (arrows) are found in the associated perisomitic matrix (PSM). x 22,000. Fig. 11.Phase-contrast micrograph of mechanically isolated somite (S)cultured 12 hours on Falcon plastic. Cells migrate radially (arrows) and are not restrained by somitic basal lamina and other components of the perisomitic matrix. Most cells exhibit stellate fibroblast-like shapes. x 70. Fig.12. Bright field micrograph of unfixed somite (S) enzymatically isolated from stage-I2 chick embryo. Trypsinized somites exhibit blebbed surfaces (arrows) and are more transparent than those isolated by mechanical methods (compare with figure 9). x 160. Fig. 13. Electron micrograph of surface of enzymatically isolated somite (S)similar to t h a t shown in figure 12. Cellular organelles are unaffected by this treatment but cytoplasmic blebbing (B) is common a t external surfaces. Perisomitic matrix including basal lamina and microfibrils is completely removed (arrows) by the enzyme. x 22,000. Fig. 14. Phase-contrast micrograph of enzymatically isolated somite (Sj cultured 12 hours on Falcon plastic. Somitic cell growth patterns are virtually identical to those seen in cultures of mechanically isolated somites (compare with figure 11). Cells migrate radially (arrows) from the original cell cluster and form circular growth areas on the culture dish. x 70. PERINOTOCHORDAL SHEATH IN VITRO 263 264 EDWARD C. CARLSON AND M. CRISTINA KENNEY a r e relatively common. They do not appear cell-like, however, and serial cross-sections show that cellular materials are rarely assoc i a t e d w i t h t h e notochordal surface. As pointed out above, however, mesenchymal “contamination” i s possible a n d , indeed, phase-contrast monitofing indicates t h a t a few sclerotomal cells are occasionally carried into culture along with the intact PNS. The number of such cells is not significant and they probably do not materially affect the outcome of the experiment. LM shows t h a t other features of uncultured notocfiords are within the range of normal variation (fig. 16) and, except for occasional increased vacuolation, they are virtually indistinguishable from their in vivo counterparts (fig. 3). Although ECM surrounding the notochord is poorly resolved by LM (fig. 161, the presence of an intact PNS up to 2.0 p m thick is clearly demonstrable by TEM (fig. 17). No electron-dense counterparts of t h e surface projections imaged by SEM are observed, but basal lamina, microfibrils, and other components of the PNS are ultrastructurally indistinguishable from those observed in vivo (fig. 4). After 12 hours of incubation on Falcon plastic, mechanically isolated notochords maintain their rod shape (fig. 18). Cells from the cut ends of the notochord attach and appear to “fan out” on the substratum. Cross sections through t h e middle portion of these notochords (fig. 19) show that their circular profile is preserved and that their central parts remain unattached to the surface of the culture dish. At 36 hours in vitro, notochords show dumb bell-shaped growth areas by phase-contrast microscopy (fig.20). Extensive cell migration from both the cephalic and caudal ends (indistinguishable in our preparations) of the notochord continues but the mid-section remains rod-shaped and the outline of the original tiss u e is easily discernable. Cross sections through the central region show remarkably circular and smoothly contoured tissues with no evidence of lateral migration (fig. 21). The continued presence of a well-established PNS is confirmed by TEM (fig.22). Basal lamina material, microfibrils, and associated ground substance combine to compartmentalize notochordal cells and preserve its rod shape. Surface ECM is not diminished up to 72 hours in culture and often appears more concentrated after several days in vitro. A simil a r a c c u m u l a t i o n of ECM m a t e r i a l s is observed in notochordal interstitial spaces (in the rod-shapped mid-section) and is associated with cell surfaces (in the fan-shaped growth areas). Unquestionably, however, the largest population of microfibrils, basal lamina materials, and interstitial bodies is located in the “tube” of matrix-the PNS surrounding the intact notochord. Isolation and culture of enzymatically isolated notochords Notochords isolated by trypsinization and microdissection average 1-3 mm in length. Following treatment with the enzyme, they lose their characteristic rigidity and are easily distorted by physical manipulation. They exhibit little elasticity and do not usually regain their straight rod shape following mechanical bending or coiling. In the following experiments mesenchymal contaminants and other non-notochordal debris (usually visible by stereo microscopy) were removed by gentle flushing with pipettes and only the cleanest notochords were selected for in vitro morphological studies. When isolated uncultured notochords are prepared for SEM, they appear as cylindrical structures with highly irregular surfaces (fig. 23). They show relatively uniform diameters (about 25 p m ) and occasionally coil upon themselves. Their most striking topographical feature is the bulb-shaped blebs which virtually cover the entire notochordal surface. Most blebs are approximately 3-4 p m in diameter and do not appear stressed or torn. Mesenchymal cells or extracellular surfaceassociated substances (PNS) are not evident in these preparations. Following embedding in epoxy, orientation for cross-sectional microtomy and preparation for LM, cross sections of uncultured isolated notochords appear as solid circular tissues with numerous cytoplasmic projections (fig. 24). These projections correspond directly with blebs observed by SEM and are not vacuolated. Trypsinized notochords usually demonstrate few intracellular vacuoles, and dilated interstitial spaces which are prominent in vivo at later stages of development (Carlson, ’73a) are not present. Sections of notochords show 10-15 contiguous cells with prominent nuclei and dark-staining nuceoli. At the level of TEM, basal surfaces of notochordal cells are clearly resolved (fig. 25). Cytoplasmic blebs a r e lined by a continuous plasmalemma and their internal structure shows no evidence of cytotoxicity. They PERINOTOCHORDAL SHEATH IN VITRO usually demonstrate a paucity of membranebound organelles and contain numerous polyribosomes and cytoplasmic microfilaments. When trypsinized notochords are cultured under the same conditions as those described for notochords isolated mechanically, their respective patterns of growth on the substratum are widely disparate. Phase-contrast microscopy shows t h a t by 12 hours of culture, cells from enzyme-treated notochords migrate randomly from the original tissue mass and attach to the substratum (fig. 26). At 36, 48, a n d 72 h o u r s of i n vitro incubation (figs. 27-29), notochordal cells expand across the substratum so that the final growth area exhibits a relatively circular shape (fig. 29). This is in contrast to ensheathed notochords (mechanically isolated) where peripheral migration emanates primarily (if not totally) from the ends of notochordal segments (figs. 18, 20). DISCUSSION Notochordal and somitic morphology in uiuo The early chick notochord develops from specialized mesoderm (chordamesoblast) near Henson’s node (at about stage 5 ) and appears early a s a bilaminar-flattened strip of cells (Bancroft and Bellairs, ’76). It gradually becomes more rod-shaped as cell-cell contacts between presumptive notochordal cells increase (Jurand, ’62). This occurs during the first two and one-half days of development when the notochord loses contact with surr o u n d i n g t i s s u e s a n d a s s u m e s a “frees t a n d i n g ” a x i a l position v e n t r a l to t h e developing neural tube. The connective tissue space surrounding t h e notochord appears “ e m p t y ” by o r d i n a r y l i g h t microscopic methods, possibly because of fixation shrinkage. Nevertheless several investigators (Low, ’68; Ruggeri, ’72; Lauscher and Carlson, ’75; Revel and Brown, ’76; Bancroft and Bellairs, ’76; Ebendal, ’77) have clearly demonstrated by SEM and TEM that this area contains numerous fibrils and other ECM components. The developing notochord is larger in diame t e r , s h o w s a more c i r c u l a r profile, a n d a p p e a r s more highly differentiated a t i t s cephalic end t h a n caudally. The more advanced cephalic end is also surrounded by a thicker, denser PNS than its caudal counterpart (Carlson, unpublished observations). In order t o compare cross-sectional morphological characteristics of the notochord and associated PNS at successively later stages of de- 265 velopment, it is a requirement that sections be taken from the same position relative to some reference point (such as the heart). In addition, magnifications of sections to be compared must be kept identical. With these guidelines in mind, the present study shows that successively older embryos demonstrate a progressively denser, more fibrillar PNS, including a better-developed basal lamina. The sheath encases sequent i a l l y larger, more vacuolated notochords with crisp circular profiles, which give the impression of highly compartmentalized tissues surrounded by an extracellular barrier that disallows cellular dispersion. This overa l l picture of notochordal development is consistent with the suggestion that the notochord becomes progressively more turgid and thus provides early longitudinal support for the embryo (Romanoff, ’60). It seems possible, therefore, that the developing notochord may be responsible for the production of a n extracellular “tube” of matrix that surrounds and physically compartmentalizes itself to the ext e n t t h a t its rod shape remains unaltered prior to involution. Furthermore, t h i s a r rangement may be a requirement for normal embryonic axial rigidity prior to the establishment of segmental chondrification centers. Like t h e notochord, somites a r e derived from chordamesoblast. Pre-somitic mesoderm (primary mesenchyme) (Hay, ’68) from t h e primitive streak apparently disperses beneath the epiblast and in the paraxial region is segmented by intersomitic clefts leaving bilaterally symmetrical blocks of tissue flanking the notochord (Hamilton, ’52). The present study shows that at stage 10 (10 somites) the seventh somite exhibits a n epithelioid, rosette-like configuration with a poorly defined lumen. Perisomitic m a t r i x (PSM) is sparse compared to t h e P N S although a relatively continuous basal lamina surrounds the basal plasmalemmae. Following formation of the dermatomyotome and disorganization of the sclerotome these latter ventro-medial cells temporarily m a i n t a i n their position with respect to the midsagittal plane. By stage 17, however, sclerotomal intercellular spaces further increase and t h e cells appear closer to the notochordal surface. This phenomenon apparently is exaggerated by a concomitant increase in notochordal diameter (Gasser, ’79). Nevertheless, by stage 20 the block-like somite is highly dispersed, and sclerotomal cells interposed between no- 266 EDWARD C. CARLSON AND M. CRISTINA KENNEY Fig. 15. Low power scanning electron micrograph of mechanically isolated notochord. These tissues are quite rigid and regain their straight rod shape following mechanical bending. Surface projections are non-cellular (see arrows in fig. 16) and are not stained by usual TEM techniques (see fig. 17). Double-headed arrow indicates position of section in fig. 16. x 650. Fig. 16. Light micrograph of cross section through mechanically isolated notochord (N) (see double-headed arrow in fig. 1 5 for section location). The smooth circular profile of in vivo notochords (see fig. 3) is maintained by this procedure and no blebbing is observed. Arrows indicate non-cellular amorphous materials (see projections in fig. 15). Square indicates approximate area shown in fig. 17. Toluidine blue. x 285. Fig. 17. Low power transmission electron micrograph of notochord iN) surface (see square in fig. 16 for approximate area shown). A clearly identifiable perinotochordal sheath (PNS) is retained and completely encases the isolated notochords except a t its cut ends. x 20,100. Fig. 18. Phase-contrast micrograph of mechanically isolated notochord (N) cultured 12 hours. The cent,ral portion of the notochord has not attached to the substratum. It maintains its crisp edges (small arrows) and rod shape. Cellular migration is evident a t the cut ends of not,ochord. These cells firmly adhere to culture dishes. Double-headed arrow indicates approximate position of section in fig. 19. x 150. Fig.19. Light micrograph of cross section through notochord (N) similar to that shown i n figure 18 (see double-headed arrow in figure 18 for approximate location of section). The cylindrical shape of the notochord is maintained. Cells adjacent to notochord probably migrated from cut end. Toluidine Blue. x 190. Fig. 20. Phase-contrast micrograph of mechanically isolated notochord (N) cultured 36 hours. Crisp surface contrast in mid-section (small arrows) indicates preservation of tissue cylinder in this region. Cell migration and tissue attachment occur almost exclusively a t cut ends of notochord segment producing a dumb bell-shaped growth pattern. Double-headed arrow indicates approximate position of section in figure 21. x 150. Fig. 21. Light micrograph of cross section through mechanically isolated notochord (N) cultured 36 hours (see double-headed arrow in fig. 20 for approximate location of section). A few small vacuoles are present. The maintenance of the notochord rod shape is clearly shown by its circular profile. Square indicates approximate area shown in figure 22. Tnluidine blue. x 190. Fig. 22. Transmission electron micrograph of surface of notochord (N) in area similar to that shown by square in figure 21. Perinotochordal sheath is intact and is composed of basal lamina ibl) microfibrils and other components of the extracellular matrix. x 23,400. PERINOTOCHORDAL SHEATH IN VITRO 267 268 EDWARD C. CARLSON AND M. CRISTINA KENNEY Fig. 23. Scanning electron micrograph of uncultured notochord isolated by trypsinization from stage15 chick embryo. Following enzymatic isolation rigidity is lost and the normally smooth cylindrical surface is profusely blebbed (B(.Isolated notochords average 1-3 mm in length, approximately 25 W m in diameter and are flaccid rods which may be bent or coiled by mechanical manipulation. Double-headed arrow, position of cross section shown by light microscopy in figure 24. x 960. Fig. 24. Light micrograph of cross section through notochord (N) in area similar to that indicated by double-headed arrow in figure 23. Numerous cytoplasmic projections are seen a t the notochordal surface. Cross sections usually show 10-15 cells. Nuclei and nucleoli are readily distinguished. Rectangle indicates approximate area shown in figure 25. Toluidine blue. x 1.700. Fig. 25. Transmission electron micrograph of surfce of contiguous notochordal cells in area similar to that seen in rectangle in figure 24. Trypsinization does not alter the morphological integrity of cellular organells. Note complete absence of extracellular matrix on basal surfaces of cells. B. cytoplasmic blebs; nuc, nucleus. x 17,300. PERINOTOCHORDAL SHEATH IN VITRO 269 270 EDWARD C. CARLSON AND M. CRISTINA KENNEY Fig. 26. Enzymatically isolated notochord cultured 12 hours on Falcon plastic. Rod shape profile of notochord segment is maintained. A few randomly scattered cells (arrows) are associated with the notochord. x 68. Fig. 27. Enzymatically isolated notochord cultured 36 hours on Falcon plastic. The original notochordal attachment site is recognizable but the cylindrical shape is lost. Cells migrate a t approximately equal rates i n all directions (arrows). x 68. Fig. 28. Enzymatically isolated notochord cultured 48 hours on Falcon plastic. Only a poorly defined “ghost” of the initial notochordal tissue adherence is evident. The growth area is oval, suggesting t h a t cells migrating laterally (horizontal arrows) may have greater mobility than those a t terminal ends (vertical arrows). x 68. Fig. 29. Enzymatically isolated notochord cultured 72 hours on Falcon plastic. By this culture time notochordal cells form a circular growth pattern and the original notochordal tissue is nearly indistinguishable. X 68. PERINOTOCHORDAL SHEATH IN VITRO 27 1 272 EDWARD C. CARLSON AND M. CRISTINA KENNEY tochord and neural tube dorsally, and between notochord and aorta ventrally, are located approximately 8@120 p m dorsomedial or ventromedial from their position a t stage 10+ (compare figs. 1 and 7). Recently Gasser ('79) provided good evidence that, in the rat, sclerotomal cells do not migrate medially and their final position can be explained on t h e basis of dorsolateral migration of the somite remnant and sclerotoma1 cell proliferation. Gasser ('79) also suggested t h a t gaps between tissues at early stages of development may be due to fixation shrinkage. Applying these arguments to the present study it may be inferred that the apparent distance traversed by sclerotomal cells (< 120 pm relative to the notochord) may result from random cellular redistribution due to a high mitotic rate and also from some preparation shrinkage. Although our d a t a tend to support Gasser's ('79) observations, the mechanisms by which sclerotomal cells attain their ultimate midline location remain unresolved. In the present study our general account of somite formation and dispersion is consistent with the summaries of Hay ('68) and Lipton and Jacobson ('74) as well a s with the more general overviews offered by most current textbooks of embryology. However, our study emphasizes the striking contrast between morphogenetic changes of t h e somite and those of the notochord. It seems possible that the lack of a dense fibrillar PSM prior to dispersion may be assistive in somitic disorganization while the relatively more advanced PNS may be more confining. Isolation and culture of somites In the past decade, numerous investigations of somite explants have been carried out primarily to study their production of cartilage in vitro (Lash, '67; O'Hare, '72b; Kosher et al, '73; Lash et al, '73; Gordon and Lash, '74; Minor, '73; Lash and Vasan, '78). Clusters of somites were cultured on a variety of substrata including various types of collagens, nucleopore filter, millipore filter, n u t r i e n t agar, and chorioallantoic membrane. With the exception of Minor's ('73) TEM study, somite explant investigations h a v e been aimed largely a t elucidating biosynthetic products. Lash and Vasan ('78) recorded the degree of somitic spreading on various substrata but made no attempts to relate this parameter to the presence or absence of PSM a t the beginning of culture. Moreover, the resolution of their photomicrographs did not per- mit determinations of growth patterns and cell migrations. No previous ultrastructural studies have ben carried out to describe surface characteristics of somite explants prior t o incubation, and investigations of the effects of PSM on growth patterns and tissue expansion in vitro have not been reported. Our studies show that somites isolated by mechanical means from t r u n k segments of stage-1P16 chick embryos exhibit clearly defined shapes with smooth contours. They appear as flattened blocks with approximately two-thirds of their periphery smooth and undisturbed. The remaining one-third appears rough and irregular, suggesting mechanical dissociation of adjacent tissues. This is consistent with Hamilton's ('52) description of six somitic surfaces, all of which are free except one (the lateral), which is continuous with intermediate mesoderm. Somites isolated by trypsinization compare favorably in size (about 260 x 360 pm) and overall shape to those isolated by homogenization. By TEM, however, enzyme-isolated somites a r e completely denuded, while mechanically isolated somites show a PSM which is indistinguishable from their in vivo counterparts. In addition, trypsinized somites are decorated with cytoplasmic blebs, an apparently innocuous feature which is also seen i n enzyme-isolated notochords and o t h e r embryonic epithelia from which basal lamina has been remove (see Hay and Dodson, '73, for review). When somites isolated either mechanically or by trypsinization (ie, with or without PSM) a r e cultured on a non-collagenous (Falcon plastic) substratum, they spread rapidly in radial growth patterns. Within 12 hours cells migrate in a monolayer in all directions and consistently form ovoid to circular growth patterns 0.8-1.3 mm in diameter. These diameters vary within groups and are unrelated to the method of isolation. The data indicate that PSM associated with somites from stage-14-16 chick embryo trunk s e g m e n t s i s not competent to m a i n t a i n somitic block shape nor to prevent cell migration in culture. Although it is possible that the substratum may stimulate somite spreading, the stimulus is equally effective regardless of the presence or absence of PSM. Isolation and culture of notochords Mechanically isolated notochords. Most in vitro investigations of notochordal tissues have begun with notochords isolated with the aid of enzymes. A notable exception is the PERINOTOCHORDAL SHEATH IN VITRO work of Kosher and Lash ('75) who dissected notochords from stage-17 chicks with or without enzymatic treatment prior to co-culture studies with somites. Their freely dissected (no enzyme treatment) notochords were surrounded by alcian blue-positive materials as d e m o n s t r a t e d by conventional LM histochemical techniques. No SEM or TEM observations of these notochords were carried out but, based on the presence of a perinotochordal halo of polyanionic materials, it was concluded that the PNS was intact. The present study shows t h a t notochords isolated from stage-1616 trunk segments by mild homogenization behave a s relatively rigid cylinders that regain their straight rod shape following mechanical bending or coiling. By conventional LM, they show sharply demarcated circular profiles similar to those described by Kosher and Lash ('75). By SEM they exhibit smooth surfaces marked by irregular protrusions. Since non-notochordal cells are infrequently observed by serial LM sections, it seems likely that the irregularities imaged by SEM may represent attached particles of proteoglycans or other extracellular macromolecules that are not clearly visualized by routine LM and TEM procedures. This interpretation is consistent with other investigations (Jacob et al, '75; Bancroft and B e l l a i r s , '76; Revel a n d Brown, '76; Schoenwolf, '79) t h a t show perinotochordal ECM manifested as surface irregularities and wispy fibrillar materials. Most importantly, however, our TEM studies confirm the presence of a n i n t a c t P N S t h a t is u l t r a s t r u c t u r a l l y i d e n t i c a l to t h a t s e e n i n vivo (Frederickson and Low, '71; Ruggeri, '72; Bancroft and Bellairs, '76). Following 12 hours in vitro a few cells begin to migrate from the cut ends of mechanically isolated notochords. These cells are not derived from non-notochordal (mesenchymal) cells, since our cross-sectional LM and TEM observations show that the PNS is cell-free. The body of t h e notochord remains solidly cylindrical with sharp, straight borders, indicating the continued presence of a PNS. No lateral cell migration is noted. At 24 hours, the PNS is densely fibrillar and the notochord mid-section remains rod-shaped while "end" cells continue to migrate outward. The fanlike growth pattern and cellular adhesion resulting from this peripheral expansion is further exaggerated a t 36 hours. Still the central portion of the tissue remains cylindrical, unattached to the substratum and invested with a dense "tube" of matrix (PNS). Com- 27 3 partmentalization of notochordal tissues is even more striking a t 72 hours of culture. Cells from the cut ends continue their expansion of the substratum while the PNS-encased mid-section retains its pre-incubation rod shape. Growth patterns of mechanically isolated notochords differ m a r k e d l y from t h o s e exhibited by cultured, mechanically isolated (PSM intact) somites. The PSM is ineffective i n preventing somitic cell spread (at least on Falcon plastic substrata), while notochordal cells are completely immobilized by the PNS. It is noteworthy that a similar phenomenon occurs in vivo where the dissociation of the sclerotome (and concomitant deposition of ECM in intercellular spaces) is contrasted by continued restriction of notochordal cells (and concomitant accumulation of matrix in the PNS). In an effort to compare the in vitro growth patterns of notochords stripped of their PNS with those cultured with PNS intact, the PNS was completely removed from the notochordal surface prior to culture. The results of these experiments are disscussed below. Enzymatically isolated notochords The present study shows that following isolation by trypsinization and microdissection, notochords from s t a g e - 1 6 1 6 chick embryos lose their smoothly contoured shapes and show highly irregular surfaces. Their diameters are comparable to in vivo equivalents but they do not exhibit the rigidity expressed by mechanically isolated notochords. This suggests that the enzyme-sensitive PNS may provide some longitudinal resilience which is transmitted to the entire embryonic axis. It seems likely that the sheath also offers some circumferential support since its removal results in cytoplasmic blebbing similar to that observed in somites following enzymatic removal of PSM. Our SEM studies indicate that the blebs virtually cover isolated notochordal surfaces. No patterns such as segmentations or other non-uniform distributions of blebs a r e recognizable along t h e i r lengths. It is noteworthy that although other investigators (Kosher and Lash, '75) have demonstrated reduced histochemical staining of acid mucopolysaccharides following isolation of notochords by trypsinization, cytoplasmic blebbing was not evident. This is inconsistent with our results and other studies that show that blebbing seems to be a common phenomenon following enzymatic removal of basal lamina (Cohen and Hay, '71; Hay and Dodson, 274 EDWARD C. CARLSON AND M. CRISTINA KENNEY ’73; Carlson et al, ’74; Lauscher and Carlson, ’75; Kenney and Carlson, ’78). Cross sections through trypsinized notochords show 7-10 contiguous cells, radially arranged and morphologically similar to the same structure in vivo. Intracellular vacuoles are rarely observed in freshly isolated notochords, suggesting that these tissues may be dehydrated to some extent during the isolation procedure. Surface-associated cytoplasmic blebbing also could be a result of such dehydration. TEM observations confirm the absence of non-notochordal cells or contaminating ECM i n enzymatically isolated notochords a n d shows that trypsinization does not interrupt or otherwise damage cell membranes. Following 12 hours of in vitro incubation, a few notochordal cells adjacent to the plastic substratum break away from the central cell cluster and begin to migrate laterally. These early in vitro cell movements are not merely transitory but they signal the initiation of a mass lateral migration which is directional (perpendicular to the long axis of the isolated notochord) and continuous (up to 72 hours in culture). A few cells migrate from the cephalic and caudal ends but by far the greatest tissue expansion occurs laterally. By 72 hours the growth area is nearly circular and has a diameter approximately t h e length of t h e original notochord segment. Most investigators consider the avian notochord a n epithelioid s t r u c t u r e (Hay, ’68; Carlson, ’73 a, b; Carlson et al., ’74; Lauscher and Carlson, ’75). As Bancroft and Bellairs (’76) correctly point out, however, a clearly resolvable lumen is not present and its cells do not demonstrate close intercellular junctions near their “apical” surfaces. Therefore, to designate the notochord a s a t r u e epithelium may be misleading. This is particularly true in light of the present study, which suggests that the notochordal “epithelial” histoarchitecture may be a t least partially dependent upon an intact PNS. Certainly one could not expect an embryonic epithelium to maintain all of its morphological characteristics when removed from its environment, stripped of its “exoskeleton,” and placed i n culture. The changes we describe, however, suggest a major departure from a more epithelial structure following culture with PNS intact (ie, radially arranged cells with close intercellular junctions and surrounded by basal lamina) to a more mesenchymal structure following cul- ture with PNS removed (ie, layers of spreading stellate cells). This is in striking contrast to other true embryonic epithelia from which surrounding ECM has been removed (Cohen and Hay, ’71; Hay and Dodson, ’73) and which maintain their epithelial morphology in organ culture. Data in the present study indicate that und e r identical culture conditions t h e PNS plays a role i n m a i n t a i n i n g tissue shape while PSM does not. It seems reasonable that the restriction of notochordal cells and the concomitant dispersion of somitic cells in vivo may likewise reflect differing functions of their respective extracellular encasements. For example, cellular dispersion may be prevented by physicochemical properties of the PNS while PSM is incompetent to act in this way. Alternatively, the PNS may prevent cells from attaching to a substratum that promotes migration while the PSM is not capable of such activity. It must also be pointed out, however, t h a t notochordal disintegration in vitro following PNS removal may represent a n artifact of culture conditions and therefore, the possibility t h a t the notochord may remain compacted in vivo if the PNS is removed cannot be ruled out. Nevertheless, observations in the present study support our general hypothesis that the avian notochord is held passively in an epithelioid configuration by a self-perpetuated accumulation of matrix (PNS) on its surface. This interpretation is consistent with the concept proposed by Hay (’68) t h a t suggests t h a t embryonic epithelia derived from primary mesenchyme (ie, notochord, somites, lateral mesoderm, etc) may subsequently disperse to form secondary mesenchyme. It seems possible that the dispersion and subsequent expansion of notochordal cells described in the present study may be a morphological manifestation of this principle in a tissue which, under in vivo conditions, is incapable of such expression. ACKNOWLEDGMENTS We gratefully acknowledge the expert technical assistance of Mr. Armando Romero and Mr. Thomas Demlow in this project. Thanks are offered to Dr. Allen C. Enders for his critical evaluation of the manuscript. LITERATURE CITED Anderson, W.A., and R.A. Ellis (1965) Ultrastructure of n’ypanosoma lewisi: Flagellum, microtubules and kinetoplast. J. Protozool., 12:48%499. PERINOTOCHORDAL SHEATH IN VITRO Bancroft, M., and Bellairs (1976) The development of the notochord in the chick embryo, studied by scanning and t r a n s m i s s i o n electron microscopy. J. Embryol. Exp. Morph., 35:38%401. Carlson, E.C. (1973a) Intercellular connective tissue fibrils in the notochordal epithelium of the early chick embryo. Am. J. Anat., 136:77-90. C a r l s o n , E . C . i1973b) Periodic f i b r i l l a r m a t e r i a l i n membrane-bound dense bodies in notochordal epithelium of t h e e a r l y chick e m b r y o . J. U l t r a s t r u c t . R e s . , 42:287-297. Carlson, E.C., and R.H. Upson (1974) “Native” striated collagen fibrils produced by chick notochordal epithelial cells i n vitro. Am. J. Anat., 141:441-445. Carlson, E.C., R.H. Upson, and D.K. Evans (1974) The production of extracellular connective tissue fibrils by chick notochordal epithelium in vitro. Anat. Rec., 179:361-374. Cohen, A.M., and E.D. Hay (1971) Secretion of collagen by embryonic neuroepithelium a t the time of spinal cordsomite interaction. Devel. Biol., 26:576605. Duncan, D. (1957) Electron microscope study of the embryonic neural tube and notochord. Tex. Rep. Biol. Med., I5:367-377. Ebendal, T (1977) Extracellular matrix fibrils and cell contacts in the chick embryo. Cell Tiss. Res., 175:439-458. Frederickson, R.G., and F.N. Low (1971) The fine structure of perinotochordal microfibrils in control and enzymetreated chick embryos. Am. J. Anat., 130:347-376. Gasser, R.F. (1979) Evidence that sclerotomal cells do not migrate medially during normal embryogenic development of the rat. Am. J. Anat., 154:50%524. Gordon, J.S., and J.W. Lash (1974) In vitro chondrogenesis and cell viability. Develop. Biol. 36:88-104. Hamilton, H.H. (1952) Lillie’s development of the chick, Ed, 3. New York Holt Rinehart and Winston. Hamburger, V., and H.L. Hamilton (1951) A series of normal stages in the development of the chick embryo. J. Morph., 88:4%92. Hay, E.D. (1968) Organization and fine structure of epithelium and mesenchyme in the developing chick embryo. In R. Fleischmajer and R.E. Billingham, eds: “EpithelialMesenchymal Interactions; 18th Hahnemann Symposium.” Baltimore: Williams and Wilkins Co., pp. 31-55. Hay, E.D., and J.W. Dodson (1973) Secretion of collagen by corneal epithelium. I. Morphology of the collagenous products produced by isolated epithelia grown on frozen killed lens. J. Cell Biol., 57:19@213. Hay, E.D., and S. Meier (1974) Glycosaminoglycan synthesis by embryonic inductors: Neural tube, notochord, and lens. J. Cell Biol., 62:88%898. Holtfreter, J. (1968) Mesenchyme and epithelia in inductive and morphogenetic processes. In R. Fleischmajer and R.E. Billingham, eds.: “Epithelial-Mesenchymal Interactions; 18th Hahnemann Symposium.” Baltimore: Williams and Wilkins Co., pp. 1-30. Jacob, M., H.J. Jacob, and B. Christ (1975) The early differentiation of the perinotochordal connective tissue. A scanning and transmission electron microscopic study on chick embryos. Experientia 31:108%1086. Jurand, A. (1962) The development of the notochord in chick embryos. J. Embryol. Exp. Morph., 10:60%621. Karnovsky, M.J. (1965) A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy J. Cell Biol., 27:137a-l38a. Kenney, M.C., and E.C. Carlson (1978) Ultrastructure identification of collagen and glycosaminoglycans i n notochordal extracellular matrix in vivo and in vitro. Anat. Rec., 190:827-850. Kosher, R. A,, and J.W. Lash (1975) Notochordal stimulation 275 of in vitro somite chondrogenesis before and after enzymatic removal of perinotochordal materials. Devel. Biol., 42:36%378. Kosher, R.A., J.W. Lash, and R.R. Minor (1973) Environmental enhancement of in vitro chondrogenesis. IV. Stimulation of somite chondrogenesis by exogenous chondromucoprotein. Devel. Biol., 35:21@220. Kvist, T.N., and C.V. Finnegan (1970) The distribution of glycosaminoglycans in the axial region of the developing chick embryo. I Histochemical analysis. J. Exp. Zool., 175:221-240. Lash, J.W. (1963) Tissue interaction and specific metabolic responses: chondrogenic induction and differentiation. In M. Locke, ed.: “Cytodifferentiation and Macromolecular Synthesis.” New York: Academic Press, pp. 235-260. Lash, J.W. (19671 Differential behavior of anterior and posterior embryonic chick somites i n vitro. J. Exp. Zool. 165:47-56. Lash, J.W. (1968a) Chondrogenesis: Genotypic and phenotypic expression. J. Cell Physiol., 72(Suppl. 1):3%46. Lash, J.W. (1968b) Phenotypic expression and differentiation: In vitro chondrogenesis. In H. Ursprung, ed.: “The Stability of the Differentiated State.” New York: SpringerVerlag, pp, 17-24. Lash, J.W. ( 1 9 6 8 ~Somitic ) mesenchyme and its response to cartilage induction. In R.F. Fleischmajer and R.E. Billingham, eds: “Epithelial-Mesenchymal Interactions; 18th Hahnemann Symposium.” Baltimore: Williams and Wilkins Co., pp. 165-172. Lash, J.W., K. Rosene, R.R. Minor, J.C. Daniel, and R.A. Kosher (1973) Environmental enhancement of in vitro chondrogenesis. 111. The influence of external potassium ions a n d c h o n d r o g e n i c d i f f e r e n t i a t i o n . Develop. B i o l . 35:37@375. Lash, J.W., and N.S. Vasan (19781 Somite chondrogenesis in vitro. Develop. Bio1.66:151- 171. Lauscher, C.K., and E.C. Carlson (1975) The development of proline-containing extracellular connective tissue fibrils by chick notochordal epithelium i n vitro. Anat. Rec., 182:151-167. Linsenmayer, T.F., R.L. Trelstad, and J. Gross (1973) The collagen of chick embryonic notochord. Biochem. Biophys. Res. Comm., 53:3%45. Lipton, B.H., and A.G. Jacobson (1974) Analysis of somite development. Develop. Biol. 38:75-90. Low, F.N. (1968) Extracellular connective tissue fibrils in the chick embryo. Anat. Rec., 160:95-108. Miller, E.J., and M.B. Mathews (1974) Characterization of notochord collagen a s a cartilage-type collagen. Biochem. Biophys. Res. Comm., 60:424-430. Minor, R.R. (1973) Somite chondrogenesis, a structural analysis. J. Cell Biol. 56:27-50. O’Connel, J.J.,and F.N. Low (1970) A histochemical and fine structural study of early extracellular connective tissue in the chick embryo. Anat. Rec., 267:425-438. O’Hare, M.J. (1972a) Differentiation of chick embryo somites in chorioallantoic culture. J. Embryol. Exp. Morph. 27:215-228. OHare, M.J. (197213) Chondrogenesis in chick embryo somites grafted with adjacent and heterologous tissues. J. Embryol. Exp. Morph. 27:22%234. O’Hare, M.J. (1973) A histochemical study of sulphated glycosaminoglycans associated with the somites of the chick embryo. J. Embryol. Exp. Morph. 29:197-208. Orkin, R.W.,T.D. Pollard, and E.D. Hay (1973) SDS gel analysis of muscle proteins in embryonic cells. Develop. Bid. 35:388-394. Olson, M.D., and F.N. Low (1971) The fine structure ofdeveloping c a r t i l a g e i n t h e chick embryo. Am. J. A n a t . , 131:197-216. 276 EDWARD C . CARLSON AND M. CRISTINA KENNEY Revel, J.P, and S.S. Brown (1976) Cell junctions in development, with particular reference to the neural tube. Cold S p r i n g H a r b o r Symposia o n Q u a n t i t a t i v e Biology XL:44%455. Romanoff, A. (1960) “The Avian Embryo.” New York: Macmillan. Ruggeri, A. (1970) Sulle possibili correalazioni fra le modificazioni morfologiche della notochorda e la sua azione induttiva. Bull. SOC.Med.-Chir. Pavia, 4-6:17%194. Ruggeri, A. (1972) Ultrastructural, histochemical a n d autoradiographic studies on the developing chick notochord. Z. Anat. Ent. Gesch., 138:2&33. Schoenwolf, G.C. (1979) Histological and ultrastructural observations of tail bud formation in t h e chick embryo. Anat. Rec. 193:131-148. Venable, J.H., and R. Coggeshall (1965) A simplified lead citrate stain for use in electron microscopy. J. Cell Biol., 25:407-408. van der Mark, H., K. yonder Mark, and S.. Gay (1976) Study of differential collagens synthesis during development of the chick embryo by immunofluorescence. I. Preparation of collagen type I and type I1 specific antibodies and their application to early stages of the chick embryo. Develop. Biol. 48:237-249.