ABSTRACT
The properties of the early chick embryonic heart cardiac jelly were studied. The cells of the heart were removed by sequential treatments with calcium magnesium-free medium; the same medium containing 5 mM EDTA; and aqueous 0·1 % deoxycholate. The transparent, naked cardiac jelly retained the original shape and size of the untreated original heart when immersed in physiological ionic strength medium. Its size and shape responded to changes in the ionic strength of the surrounding media. Alcian blue, cetylpyridinium chloride and testicular hyaluronidase abolished the ability of the jelly to respond to ionic strength changes. Electron microscope examination of the negatively stained spread cardiac jelly revealed an extensive network of collagenous fibrils and fine filaments with some amorphous adhering material. Treatment with testicular hyaluronidase removed much of the amorphous material and improved the details of the filaments. These results suggest that glycosaminoglycans play an important part in the hydration of the cardiac jelly and that the stability of the cardiac jelly shape is mainly due to the filamentous network and their possible interactions with macromolecules of the cardiac jelly matrix. It is suggested that the factors that control the deposition of the connective tissue macromolecules and the assembly of the filamentous network are significant factors which influence the morphogenesis of the early embryonic heart.
INTRODUCTION
During early developmental stages, the heart has an extensive connective tissue layer situated between the myocardium and endocardium. This extracellular material is called ‘cardiac jelly’ (Davis, 1924), and is particularly prominent at the time the heart rudiment bends to the right, a process called looping (for discussion see Manasek, 1976a).
The cardiac jelly contains glycosaminoglycans as shown histochemically (Barry, 1951 ; Markwald & Adams Smith, 1972; Ortiz, 1958). Later biochemical and ultrastructural studies revealed that the cardiac jelly contains three major components, glycosaminoglycans, glycoproteins and collagen (Gessner & Bostrom, 1965; Gessner, Lorincz & Boström, 1965; Johnson, Manasek, Vinson & Seyer, 1974; Manasek et al. 1973; Manasek, 1976b). The presence of these components suggested various possible physiological roles of the cardiac jelly such as the formation of a microenvironment controlling the passages of substances (Manasek, 1975). Earlier, Barry (1948) and Patten, Kramer & Barry (1948) had suggested a functional significance of cardiac jelly as a viscoelastic element effecting diastolic rebound.
Recent studies (Manasek, 1976b; Markwald, Fitzharris & Adams Smith, 1975 a) have suggested that cardiac jelly is not simply a solution of macromolecules. Different solute molecules such as glycosaminoglycans and fuco-sylated glycoproteins interact to form exceedingly large complexes (Manasek, 1977). This observation suggested that the large components of cardiac jelly might not be able to translocate freely relative to each other. If this is true then one would expect to be able to demonstrate that the cardiac jelly compartment has an intrinsic shape that is not disrupted readily.
The experiments we report in this paper explore these properties of cardiac jelly. We have succeeded in isolating the cardiac jelly from single isolated embryonic hearts and have shown that it retains its shape even in the absence of the myocardial cell layer. Further, we examined the effects of differing ionic strengths on the shape of the cardiac jelly and found that it can be made to swell reversibly, without losing its shape by alternating between a low and physiological ionic strength environment. These findings have enabled us to propose a structural morphogenetic role for cardiac jelly during early heart development.
MATERIALS AND METHODS
Isolation of embryonic heart
Hearts were collected from stages (Hamburger & Hamilton, 1951) stage 10 + to 17-chick embryos (11–30 pairs of somites). After carefully removing the pericardial splanchnopleure with fine glass needles, the heart was removed by cutting the cephalic end of the bulbus arteriosus and, caudally, the omphalomesenteric veins. Some of the early embryonic stage hearts often included a small part of the foregut. The isolated hearts were washed several times with Tyrode’s medium.
Experimental manipulation of isolated hearts
Isolated hearts were placed individually in glass troughs measuring 55 ×; 10 ×;1 mm containing Tyrode’s medium. A cover-glass (22 × 30 mm) was laid over the trough. There was ample space for the heart to float between the cover-glass and the bottom of the trough. Solutions were introduced into one open end of the trough and withdrawn from the other end with a piece of paper. All of the operations were carried out under a dark-field light microscope and recorded with Polaroid film. During the exchanges of solutions great care was necessary to prevent loss of the sample.
Experimental manipulation of intact hearts
The Tyrode’s medium in the trough was replaced with distilled water (DW) photographs were taken at various time intervals. Then DW was exchanged with 0·1 M-NaCl, 5 mM sodium phosphate buffer, pH 7·4, and pictures were taken. After repeating this sequence with 0·1 M-NaCl solution or with Tyrode’s, 1 % cetylpyridinium chloride (CPC) was added to the trough.
Removal of cell layers
(a) 0·1 % Deoxycholate (DOC)
0·1% DOC (Na-salt, Sigma Chemical Co., St Louis) was prepared from 10% stock aqueous solution. Although it is not necessary, the isolated heart was washed with 10 mM phosphate buffer, pH 8·0, before being placed in the trough. 0·1 % DOC was introduced into the trough and the effects were photographed. After 0·5–2 h of treatment, DOC was replaced sequentially with DW, 0·1–0·15 M-NaCl, DW, 0·1–0·15 M-NaCl. Some of the DOC treated samples were further treated with 0·8 % Alcian blue, 1 % CPC or testicular hyaluronidase (THyase; 1 mg/ml in 0·15 M-NaCl, 0·1 M sodium phosphate buffer, pH 5·3) for 1 h at 37 °C.
(b) Calcium magnesium-free medium (CMF) -CMF.EDTA
In this procedure the intention was to reduce the adhesion between cells and between cells and connective tissue. Isolated hearts were washed with CMF (NaCl 8 g, KCl 0·2 g, NaH2PO4-H2O 0·05 g, NaHCO3 1 g, glucose 2 g, per liter; Moscona, 1952) and incubated in the same solution for 25 min at 37 °C. This was followed by CMF.EDTA (5 mM) for 25 min at 37 °C. After a brief rinse with DW, the heart was treated with 0·1% DOC at room temperature while mechanically shaking the trough. This treatment was continued for 25 min to 1·5 h. Then the sample was bathed sequentially with DW and 0·1–0·15 M-NaCl, 5 mM sodium phosphate, pH 7·4. Some of the samples were treated with 0·8 % Alcian blue, or 1 % CPC, or THyase.
The initial CMF treatment can be omitted without any noticeable effect on the results. The concentration of EDTA in CMF. EDTA should be above 1 mM since the removal of cell layers is more difficult at or below 2 mM EDTA. The transparent, barely visible cardiac jelly was negatively contrasted by adding India ink to the medium and photographed with a bright-field light microscope.
Electron microscopy
DOC-treated (0·1 % DOC) samples and also the bare cardiac jelly were fixed with full strength Karnovsky’s (1965) fixative for 5 min at 4 °C and then with the same fixative containing 0·8% Alcian blue. The total fixation time was 20 min. Samples were washed with 0·1 M cacodylate buffer pH 7·4 for 30 min at 4 °C and were post-fixed with 1 % OsO4 in the same buffer for 1 h at 4 °C. Dehydration was carried out through graded ethyl alcohol and samples were embedded in Araldite 502. Sections were cut on a Sorval MT-2 microtome with a diamond knife. Sections were mounted on copper grids, double stained with alcoholic uranyl acetate and lead citrate.
The trough is not convenient for preparing a number of naked cardiac jelly preparations simultaneously. Multiwell tissue culture plates (Falcon, no. 3008) are more suitable for this purpose. The cell layers of the heart were removed by the same treatment as before, but in this case samples were transferred from one well to another with a Pasteur pipette while they were being observed with a dark-field dissecting microscope. The naked cardiac jelly in DW was picked up with a Pasteur pipette with a minimal amount of fluid and was mixed with a drop of 0·3–0·5% bovine serum albumin (BSA) in a Microtest Terasaki tissue culture plate (Falcon, no. 3034) to improve the hydrophilicity of supporting film. The sample was picked up with a small loop of copper wire and mounted on a carbon enforced formvar or collodion film coated copper grid. Some of the samples mounted without BSA were digested with THyase (1 mg/ml) for 1 h at 37 °C prior to negative staining. Negative staining was done with 2 % aqueous uranyl acetate. All samples were examined in RCA-EMU 4 electron microscope operated at 50 kV.
Planimetry
Images of cardiac jelly were carefully cut out of photographic enlargements, desiccated under vacuum and weighed.
RESULTS
Studies on the intact heart
The freshly dissected heart maintained its shape throughout the initial handling period, during which time it was removed from the embryo and placed in the trough. Under the dark-field microscope the untreated freshly isolated embryonic heart appeared to be covered with numerous brightly contrasting, tightly packed granules. When the Tyrode’s medium was exchanged with distilled water, the heart swelled rapidly (Fig. 1). Maximum size was reached within 5–10 min, after which the heart gradually became somewhat smaller again. The rapid swelling was accompanied by the loss from the heart of many of the granular surface particles, probably as a result of cell lysis. Replacement of the DW with 0·1–0·15 M-NaCl solution restored the heart to nearly its normal size. Alternation of DW and salt solution caused the heart to undergo swelling and shrinking cycles that could be repeated at least twice. The ability to swell and shrink in response to medium changes was abolished by the addition of a few drops of 1 % CPC. This was accompanied by marked permanent shrinkage of the specimen.
Effects of ionic strength on the isolated early chick embryonic heart (11 somites). (A) The dissected heart. Distilled water 0 min. (B) Distilled water 5 min. Heart swelled extensively with simultaneous removal of granular material. (C) Distilled water 30 min. Note the increased loss of granular material and concomitant increase in cardiac jelly transparency. (D) 0·1 M-NaCl, 5 mM sodium phosphate, pH 7·4. The heart shrank to its nearly normal size and shape, but some cellular materials remained. (E) Distilled water 30 min. The heart swelled again as salt was washed out. (F) Tyrode’s solution 5 min. The heart shrank again to approximately the normal size and shape. Dark-field light micrographs. Scale 1 mm. ×;35.
Effects of ionic strength on the isolated early chick embryonic heart (11 somites). (A) The dissected heart. Distilled water 0 min. (B) Distilled water 5 min. Heart swelled extensively with simultaneous removal of granular material. (C) Distilled water 30 min. Note the increased loss of granular material and concomitant increase in cardiac jelly transparency. (D) 0·1 M-NaCl, 5 mM sodium phosphate, pH 7·4. The heart shrank to its nearly normal size and shape, but some cellular materials remained. (E) Distilled water 30 min. The heart swelled again as salt was washed out. (F) Tyrode’s solution 5 min. The heart shrank again to approximately the normal size and shape. Dark-field light micrographs. Scale 1 mm. ×;35.
Studies on cardiac jelly
(a) Effects of 0·1 % deoxycholate (DOC)
Within a few minutes after replacement of the Tyrode’s solution with 0-1 % DOC, the myocardial layer became more transparent and started swelling (Fig. 2). The entire heart became very soft. Simultaneously, a number of particles began to fall off from the myocardial layer. The size of the heart increased, reaching about twice that of the original heart. After 30 min in 0·1 % DOC solution, a large number of particles had been shed by the myocardium into the medium. Prolonged 0·1 % DOC treatment (up to 1 h), and subsequent shaking, DW and salt solution treatment (0·1–0·15 M-NaCl) removed most of the granular material from the myocardial layer, although some material still remained.
Effects of the 0·1 % deoxycholate treatment on the isolated early chick embryonic heart (14 somites). (A) The dissected heart. Deoxycholate (DOC) 0 min. (B) DOC 60 min. The heart swelled and most of the granular materials were more effectively removed than they were by distilled water. (C) Washed with distilled water. Some cellular materials were released. (D) 0·15 M-NaCl, 5 mM sodium phosphate, pH 7·4, 5 min. The swollen heart quickly shrank and returned to its original shape and size. (E) Distilled water 10 min. The heart swelled again and more cellular materials were released. The degree of swelling decreased gradually with repeated cycles. (F) 1 % cetylpyridinium chloride. The heart shrank extremely and irreversibly. Dark field light micrographs. Scale 1 mm. ×; 35.
Effects of the 0·1 % deoxycholate treatment on the isolated early chick embryonic heart (14 somites). (A) The dissected heart. Deoxycholate (DOC) 0 min. (B) DOC 60 min. The heart swelled and most of the granular materials were more effectively removed than they were by distilled water. (C) Washed with distilled water. Some cellular materials were released. (D) 0·15 M-NaCl, 5 mM sodium phosphate, pH 7·4, 5 min. The swollen heart quickly shrank and returned to its original shape and size. (E) Distilled water 10 min. The heart swelled again and more cellular materials were released. The degree of swelling decreased gradually with repeated cycles. (F) 1 % cetylpyridinium chloride. The heart shrank extremely and irreversibly. Dark field light micrographs. Scale 1 mm. ×; 35.
Electron microscopic examination of the 0·1 % DOC-treated heart showed clearly that there were still many myofibrillar and nuclear remnants in the myocardial and endocardial layers (Fig. 3). The myofibrils (Fig. 3, M), although grossly disorganized, were still recognizable. The intercalated discs (Fig. 3) were relatively intact, despite the loss of large amounts of membranes.
Electron micrograph of the DOC-treated heart. 0·1 % DOC treatment followed by distilled water and salt solution caused the disruption of myocardial cells. Although a large part of membranous structures are removed, disorganized myofibrils (M) and intercalated discs (arrow) are still present. Scale 1 μm. ×; 28000.
Electron micrograph of the DOC-treated heart. 0·1 % DOC treatment followed by distilled water and salt solution caused the disruption of myocardial cells. Although a large part of membranous structures are removed, disorganized myofibrils (M) and intercalated discs (arrow) are still present. Scale 1 μm. ×; 28000.
Replacement of 0·1 % DOC with DW caused further swelling of the heart (Fig. 2C). Although the general shape of the original heart still could be detected, some parts of the swollen structure appeared deformed. Salt solution had the same effect on the DOC-treated heart as it did on the intact heart, and caused shrinkage. Cyclic swelling and shrinking could also be repeated at least twice by alternating DW and 0·1 M-NaCl (Fig. 2C-E). It is important to point out that the shape of the heart after shrinking was very similar to that of the freshly dissected heart (Fig. 2A, D).
If, following incubation with 0·1 % DOC, the specimens were exposed to Alcian blue (0·8 %), CPC (less than 1 %) or testicular hyaluronidase they shrank markedly and no longer changed their dimensions in response to changes in salt concentration (Fig. 2F).
(b) CMF-CMF.EDTA-0·1 % DOC treatment
An attempt was made to effect complete myocardial removal. Since it is common procedure to use CMF medium with or without EDTA in loosening up the intercellular junctions (Moscona, 1952; Muir, 1966), the dissected heart was incubated with CMF and CMF. EDTA (above 2 mM-EDTA) prior to DOC treatment.
The effects of CMF-CMF.EDTA treatment became apparent during the subsequent 0·1 % DOC treatment (Fig. 4). Small patches of flocculent material with granules came off the heart and left transparent areas on the surface. The removal of this material from the heart was facilitated by vigorous shaking of the trough or by flushing out the DOC with DW and then with 0·1–0·15 M-NaCl while shaking the trough. In the latter case the swollen heart shrank to almost physiological shape and size (Fig. 4E). This shrinkage forced out the endocardial cell debris also, which was otherwise trapped in the lumen of the heart. Even after overnight exposure to DW, the swollen heart was able to shrink when placed in salt solution. During the course of the removal of the myocardial layer, the cells along the dorsal mesocardium came off most easily (Fig. 4C, D).
Removal of the cell layers from the isolated chick embryonic heart (16 somites) by calcium magnesium-free medium (CMF)-CMF.EDTA-0·l % DOC treatment. (A) Dissected heart. (B) CMF.EDTA 10 min at 37 °C. No appreciable change is seen. (C) 0·1 % DOC 30 min. Heart swelled and cells along the dorsal mesocardium were removed. (D) 0·1 % DOC 1·5 h. Heart swelled further and patches of myocardial cells began to come cff. (E) India ink. Naked cardiac jelly is virtually invisible using either bright- or dark-field illumination and can be visualized best by means of negative contrast, in this case using dilute India ink. Note that the cardiac jelly has the shape and size of the original heart (A). A small remnant of cellular material is stained dark by india ink. Dark field light micrographs except (F), which is bright field. Scale 1 mm. ×;36.
Removal of the cell layers from the isolated chick embryonic heart (16 somites) by calcium magnesium-free medium (CMF)-CMF.EDTA-0·l % DOC treatment. (A) Dissected heart. (B) CMF.EDTA 10 min at 37 °C. No appreciable change is seen. (C) 0·1 % DOC 30 min. Heart swelled and cells along the dorsal mesocardium were removed. (D) 0·1 % DOC 1·5 h. Heart swelled further and patches of myocardial cells began to come cff. (E) India ink. Naked cardiac jelly is virtually invisible using either bright- or dark-field illumination and can be visualized best by means of negative contrast, in this case using dilute India ink. Note that the cardiac jelly has the shape and size of the original heart (A). A small remnant of cellular material is stained dark by india ink. Dark field light micrographs except (F), which is bright field. Scale 1 mm. ×;36.
Conventional electron microscopic examination of the resulting transparent material revealed that the myocardium, as well as endocardium, was almost completely removed from the cardiac jelly (Fig. 5). Thus, the transparent material remaining after the complete removal of the myocardial flocculent materials was undoubtedly the naked cardiacjelly itself (Fig. 5, inset). The cardiac jeíly contained electron-dense materials of various sizes (up to 111 nm thick). Most of this material appeared amorphous, but sometimes some thin filaments (6–10 nm) were seen (Fig. 5, arrowheads).
Electron micrograph of the naked cardiac jelly (27 somites). No cellular components are discernible. The jelly appeared to be a mixture of varied sized cross or obliquely cut electron-dense rods and some granular material. Fine filamentous materials often interconnect these rods (arrows). Frequently filamentous substructures can be seen in the rods (arrowheads). Scale 1 μm. ×; 37200. Inset: dark field light micrograph of the naked cardiac jelly. Scale 01 mm. ×;48. The naked cardiac jelly was prepared by the CMF-CMF.EDTA method, and was fixed with Karnovsky’s fixative for 5 min followed with Karnovsky-0·8 % Alcian blue fixative.
Electron micrograph of the naked cardiac jelly (27 somites). No cellular components are discernible. The jelly appeared to be a mixture of varied sized cross or obliquely cut electron-dense rods and some granular material. Fine filamentous materials often interconnect these rods (arrows). Frequently filamentous substructures can be seen in the rods (arrowheads). Scale 1 μm. ×; 37200. Inset: dark field light micrograph of the naked cardiac jelly. Scale 01 mm. ×;48. The naked cardiac jelly was prepared by the CMF-CMF.EDTA method, and was fixed with Karnovsky’s fixative for 5 min followed with Karnovsky-0·8 % Alcian blue fixative.
Spread and negatively stained cardiac jelly showed a completely different picture. The negatively stained cardiac jelly preparations were often too thick to reveal details. However, structure could be discerned in well spread peripheral regions of the preparations (Fig. 6, 7). There were numerous fibrils running in various directions. The fibrils were extremely long and relatively straight, but they did not appear to be entangled extensively in spite of their length (Figs. 6 A, 7 A). They could be roughly classified into two groups based on the presence or absence of cross-banding pattern. Cross-banded fibrils were uniformly about 30 nm thick throughout their length. They often showed some adhering materials which tended to obscure the details of the cross-banded pattern (Fig. 6B). Testicular hyaluronidase treatment of the mounted sample prior to negative staining removed most of the adhering materials (Fig. 7) and greatly enhanced detail. The periodicity of the cross-banding pattern was about 65 nm, and the alternating light and dark band pattern was characteristic of fibrillar collagen (Fig. 7B). The fibrils frequently appeared side by side in a parallel fashion, but their cross-banding pattern was not necessarily in register. The apparent number of these fibrils increased with embryo age. Filaments without cross striations were thinner and had varying diameters of up to 10 nm. Some of these thin filaments had globular clumps of material associated with them, spaced about 90-100 nm along the filament (Fig. 7B). The regularity of the periodic association of the globules with filaments suggests that this association is not an artifact.
Electron micrographs of a negatively stained spread cardiac jelly. Spread cardiac jelly shows numerous filaments of varying diameters. They are very long and show some adhering substances which tend to obscure their detail. Negatively stained with 2 % aqueous uranyl acetate. Scale 1 μm. (A) ×; 6160; (B) ×; 18100.
Electron micrographs of a negatively stained spread cardiac jelly. Spread cardiac jelly shows numerous filaments of varying diameters. They are very long and show some adhering substances which tend to obscure their detail. Negatively stained with 2 % aqueous uranyl acetate. Scale 1 μm. (A) ×; 6160; (B) ×; 18100.
Electron micrographs of testicular hyaluronidase treated spread cardiac jelly. Much of the adhering substance is removed, revealing filament detail (7 A). There are two major types of filaments (7B), a thin type (up to 10 nm) showing globular material at regular intervals (90–100 nm) and another which is typical fibrillar collagen with characteristic cross-banding (7B). Negatively stained with 2 % aqueous uranyl acetate. Scale 1 μm (A) and 0·1 μm (B). (A) ×; 8800; (B) ×; 66000.
Electron micrographs of testicular hyaluronidase treated spread cardiac jelly. Much of the adhering substance is removed, revealing filament detail (7 A). There are two major types of filaments (7B), a thin type (up to 10 nm) showing globular material at regular intervals (90–100 nm) and another which is typical fibrillar collagen with characteristic cross-banding (7B). Negatively stained with 2 % aqueous uranyl acetate. Scale 1 μm (A) and 0·1 μm (B). (A) ×; 8800; (B) ×; 66000.
After complete removal of the myocardial layer, the remaining material was transparent. It could not be seen by transmitted light microscopy and its outlines could only barely be perceived by dark field microscopy. The only way it could be studied, visually or photographically, was to outline it negatively by flooding the trough with dilute India ink and examining the negative image (Fig. 8). Despite removal of the cell layers, the shape of the remaining material was very similar to that of the originally dissected heart. The naked cardiac jelly retained the ability to swell and shrink in response to changes in salt concentration. Although this response was much less than that exhibited by specimens containing either complete (Fig. 1) or partial (Fig. 2) cell layers, it was still clearly demonstrable (Fig. 8). Because of the smaller response, the changes shown by the naked cardiac jelly were measured by planimetry. Distilled water (Fig. 8B) resulted in an increase in image area to 1·6 times that of the original heart (Fig. 8 A). The swollen size was restored to the original size when the distilled water was replaced by 0·1 M-NaCl (Fig. 8C). The ability to respond to changes in salt concentration was lost rapidly by the naked cardiac jelly preparations. However, even after no visible swelling was elicited by distilled water, the addition of 0·8 % Alcian blue or a few drops of 1 % CPC resulted in shrinkage of the cardiac jelly (Fig. 8 F).
Cyclic response of the naked cardiac jelly (20 somites) to changes in ionic strength. (A) Calcium magnesium-free medium 0 min. (B) Distilled water-India ink. Just after removal of the myocardium. Note the swollen size of the jelly. (C) 0·1 M-NaCl, 5 mM Na-phosphate, pH 7·4 India ink. The cardiac jelly shrank to its original size and shape. (D) Distilled water-India ink 20 min. The cardiac jelly is somewhat swollen. (E) Distilled water 15 min. The cardiac jelly is swollen further and the size is very similar to that of (B). (F) 0·8 % Alcian blue. 0·8 % Alcian blue was diffused into the trough. The cardiac jelly shrank irreversibly. Bright field light micrographs except (A) and (E), which are dark field. Scale 1 mm. ×; 27.
Cyclic response of the naked cardiac jelly (20 somites) to changes in ionic strength. (A) Calcium magnesium-free medium 0 min. (B) Distilled water-India ink. Just after removal of the myocardium. Note the swollen size of the jelly. (C) 0·1 M-NaCl, 5 mM Na-phosphate, pH 7·4 India ink. The cardiac jelly shrank to its original size and shape. (D) Distilled water-India ink 20 min. The cardiac jelly is somewhat swollen. (E) Distilled water 15 min. The cardiac jelly is swollen further and the size is very similar to that of (B). (F) 0·8 % Alcian blue. 0·8 % Alcian blue was diffused into the trough. The cardiac jelly shrank irreversibly. Bright field light micrographs except (A) and (E), which are dark field. Scale 1 mm. ×; 27.
DISCUSSION
The present study demonstrates a major property of the cardiac jelly. The cardiac jelly has a shape that reflects the shape of the heart itself, and this shape can be manipulated experimentally by altering the ionic environment of the cardiac jelly. This finding is important to furthering our understanding of the regulation of embryonic morphogenesis.
The cardiac jelly is not simply a viscous solution. Complete removal of the myocardial layer by a variety of procedures did not cause the jelly to lose the shape it had while it was still confined by the myocardial investment. There was no notable slumping, or time-dependent shape change of the bare cardiac jelly in 0·1–0·15 M-NaCl. Such intrinsic stability might, of course, simply reflect an extremely high viscosity. This possibility is not. consistent with the results from our experimental manipulation of cardiac jelly shape. Furthermore, it is unlikely that our procedures to remove the myocardium resulted in this shape stability since all the procedures were extractive. Hence, any putative action on the cardiac jelly itself would have tended to remove material, making the structure less viscous and less stable rather than increasing its structural stability.
The shape and volume of the cardiac jelly can be altered reversibly by altering the ionic strength of the surrounding solution. This ability to expand in a low ionic medium (distilled water) is lost if the jelly is treated with testicular, hyaluronidase, Alcian blue or CPC. All of these agents are known to act on the glycosaminoglycans which are known to be in the cardiac jelly (Gessner & Boström, 1965; Gessner et al. 1965; Manasek et al. 1973). Thus, THyase hydrolyses chondroitin sulfate and hyaluronic acid; CPC precipitates these molecules and Alcian blue will bind to them. The increase in jelly size (swelling) in low ionic strength therefore appears to result from hydration of the resident glycosaminoglycans, a situation analogous to that shown by Fessler (1960) in Wharton’s jelly.
The naked cardiac jelly does not swell in response to distilled water as dramatically as do those specimens that have either a complete or partial myocardial covering. This suggests that some of the increase in size shown by the latter specimens may represent swelling of the cells or their remnants in addition to swelling of the jelly. However, the response of the specimens devoid of myocardium clearly establishes that the naked cardiac jelly has an intrinsic ability to respond to changes in ionic strength. The rapidity with which this is lost in naked specimens, as a function of repeated cycling, suggests that extraction is occurring more rapidly in the absence of a myocardial investment.
The observation that the grossly deformed and swollen jelly returns to its normal shape with increasing ionic strength suggests that the jelly has intrinsic structure that serves to regulate or stabilize its overall shape. Without such intrinsic regulation it would not be expected to resume its morphology after being distorted grossly in distilled water.
There are two lines of evidence that provide clues to the nature of the intrinsic shape of the cardiac jelly. Careful light and electron microscope observation (Johnson et al. 1974; Markwald et al. 1975 a; Patten et al. 1948) have demonstrated the presence of fibrils within the cardiac jelly. The present electron microscope study of negatively stained preparations of cardiac jelly supports these earlier observations. The cardiac jelly is therefore a multiphasic system. It is interesting to note that the fibrous components, particularly fibrillar collagen, appear to increase during the time that the heart undergoes looping. We therefore propose that presence of matrical fibrils increases the morphological stability of the cardiac jelly. The recent observation that matrical glycoproteins, which may also be filamentous (Markwald, Fitzharris & Bank, 1975b), interact with glycosaminoglycans (Manasek, 1977) is also noted as a mechanism whereby the molecular composition of the cardiac jelly is capable of stabilizing its morphology.
Up to now we have considered factors that stabilize the shape of the cardiac jelly. The cardiac jelly however is a dynamic structure that undergoes ontogénie remodelling as the organ changes its shape. For example, in a time period of about 5 h, the heart transforms itself from a midline to a ‘c’ shaped structure. This process is called ‘looping’. Obviously, the cardiac jelly must be able to change its shape too. Therefore, when we speak of ‘morphological stability’ of the jelly we do so in a restricted sense, involving a single developmental time point such as that utilized, of necessity, in our experiments. The factors that induce such stability must also be capable of physiological modification to permit normal organ morphogenesis. We therefore note another required property of cardiac jelly; it must be capable of undergoing time dependent shape change. Obviously, the most simple mechanism would be for the enveloping myocardium to simply deform the jelly, and hold it in its new shape. This, however, is not consistent with our demonstration that the shape of the jelly is maintained without myocardium. If morphogenetic movements exceed the elastic limit of the jelly one could imagine that such a bend could become permanent. However, one cannot experimentally ‘bend’ the jelly into a new or different shape suggesting that this is not a normal morphogenetic mechanism. Indeed, with each systole the jelly is deformed far more than it is during morphogenetic changes, yet such contractile deformations are also transient.
There are a number of possible mechanisms by which the shape of the cardiac jelly can be altered. There are three concomitant events that occur during looping: (i) continuous synthesis of matrix macromolecules (Manasek, 1973) which interact with each other to form supramolecular aggregates (Manasek, 1977); (ii) increased fibrillar structures within the cardiac jelly (Johnson et al. 1974); (iii) a coordinated change in shape of myocardial cells which results in localized changes in myocardial surface area (Manasek, Burnside & Waterman, 1972). Taken collectively, these observations as well as the present results permit us to propose the following model: myocardial cell shape changes, occurring in small continuous increments, result in small incremental deformational changes in organ shape. While the cells are changing shape they are also secreting matrix macromolecules which continuously and rapidly interact with each other in the extracellular compartment. These interactions (possibly covalent) serve to stabilize continuously the shape of the newly deposited matrix, which conforms to the limits of the myocardial mantle. Thus, the shape change of the myocardium does not have to deform the entire jelly continuously or at any given time; rather it needs to provide a mold only for those newly elaborated matrix components not yet fully stabilized within the jelly. Once they are incorporated into the matrix no additional work on the part of the myocardium is required to maintain their relative positions, hence shape of the jelly. In this model, we view the morphogenetic force as being myocardial in origin (cell shape change) and the jelly as a filler that assumes the shape dictated by the myocardium, and retains this shape by means of cross links. We therefore expect the jelly to demonstrate a continuum that ranges from older, ‘fixed’ molecules to those newly synthesized ones that are relatively mobile and conform to the new shape prior to being ‘fixed’.
There is an additional mechanism that could be operational. Since we have demonstrated that the cardiac jelly can swell and shrink in response to ionic environment it is possible that a similar mechanism operates in situ. Thus, by regulating the local ionic environment in different regions of cardiac jelly local volume changes could be effected. These local volume changes would be manifested as changes in the shape of the organ. Again, such changes would have to be stabilized (since the cardiac jelly does indeed retain its shape as shown by our experiments), most likely by matrix molecule interaction. This hydration model requires that the myocardium regulates the ionic strength of its substrate.
The role of the myocardium is predictably different in these two models. In the first, the myocardium itself exerts a deforming force directly by changing the shape of its cells. In the second, it produces local shape changes in the cardiac jelly by ionic means and, it would be assumed, conforms to the resulting shape. The shape changes detected in the myocardial cells (Manasek et al. 1972) would then not be primary determinative morphogenetic events, but secondary ones. We cannot decide which of these models is correct. Both give us predictions which are testable experimentally and which should clarify the mechanisms involved in morphogenesis.
ACKNOWLEDGEMENTS
This work was supported by grant HL 13831 from the National Heart, Lung and Blood Institute, National Institutes of Health.