ABSTRACT
Three lectins, wheat germ agglutinin (WGA), soybean agglutinin (SBA) and Ricinis communis agglutinin I (RCA), were used to study the basement membrane of developing chick lungs. Thinning of the basement membrane at the tips of newly formed bronchi was visualized with all three lectins, but was particularly evident using SBA. Control sections established the ability of the lectins to stain hyaluronic acid and chondroitin sulphate. Neuraminidase, bovine testes hyaluronidase and Streptomyces hyaluronidase removed some of the staining, but none were able to affect the staining of the basement membrane. Possible explanations for this are discussed in the text. Incorporation of [3H]thymidine is enhanced at the tips relative to the interbud area in stage-30 lungs, extending previous studies on stage-26 lungs. Evidence has been presented here which demonstrates that mechanisms of morphogenesis used in avian embryos are similar to those already elucidated in work on mammalian embryos.
INTRODUCTION
The development of many organs begins as an interaction between epithelium and mesenchyme (Grobstein, 1954; Cunha, 1976; Wessells, 1977; Goldin, 1980; Bernfield, Banerjee, Koda & Rapraeger, 1984). Greater complexity may characterize these events than that seen in earlier inductive interactions. One aspect of this complexity may be the specificity of signals exchanged between the tissues; therefore, the study of epithelial–mesenchymal interactions is likely to yield insight into other developmental processes (Grobstein, 1954).
Studies of salivary gland morphogenesis have produced a detailed picture of epithelial–mesenchymal interactions. The basement membrane, which lies at the interface between the two tissue types, has two main components: the trilayered basal lamina, directly adjacent to the basal surface of the epithelium, and the fibrillar mesh work, containing type I collagen (Kallman & Grobstein, 1964; Kallman, Evans & Wessells, 1967; Bloom & Fawcett, 1975). At the tips of growing glands, glycosaminoglycan (GAG) turnover and cell proliferation are enhanced. The clefts are sites of increased collagen deposition and of GAG accumulation due to stabilization by the collagen. GAG synthesis is reduced at the tips, leading to the appearance in the light microscope of a thinner basement membrane at the tips compared to that seen in the clefts (Bernfield & Banerjee, 1972; Bernfield, Banerjee & Cohn, 1972; Bernfield, Cohn & Banerjee, 1973; Banerjee, Cohn & Bernfield, 1977; David & Bernfield, 1979,1981; reviewed in Bernfield et al. 1984).
Previously, developing organs that have been examined for heterogeneities in the basement membrane have been derived from mammalian sources. It would be instructive to learn whether avian tissues use similar strategies in development. The chick lung has a monopodial pattern of branching, in which a single main stem continues to extend at the apex in its original line of growth while giving off lateral branches. This is different from the mammalian tissues in previous studies, which have new branches arising from the tips of previous branches.
To study these events, lectins have been used as probes of cell-surface chemistry. Lectins are glycoproteins or sometimes proteins that have the property of binding to specific sugars (Lis & Sharon, 1977). Wheat germ agglutinin (WGA) binds to N-acetylglucosamine (glcNAc) and sialic acid (SA). Because hyaluronic acid (HA) is made up of repeating disaccharides which contain glcNAc, WGA is likely to show affinity for this GAG. Likewise, chondroitin contains alternating units of N-acetylgalactosamine (galNAc), a sugar to which soy bean agglutinin (SBA) binds. Ricinis communis agglutinin I (RCA) binds to galactose and therefore will bind to keratan sulphate, which contains this sugar. Commercially available HA and CS were stained with WGA and SBA, respectively and confirm the expectations in these two cases, as shown in the Results section.
The purpose of the present study was to visualize changes in the basement membrane over a period of development when the branches are first emerging to a time when the older buds are beginning to branch. Fluorescein-conjugated WGA (Fl-WGA), SBA (Fl-SBA) and RCA (Fl-RCA) were used to stain the lungs. Incorporation of [3H]thymidine was examined in stage-30 lungs to determine any local differences in levels of cell proliferation that might correlate with lectin-staining patterns and, or, morphogenesis. Thinning of the basement membrane at the tips of newly formed bronchi was visualized with all three lectins. Cell proliferation was enhanced at the tips relative to the interbud area.
MATERIALS AND METHODS
Fl-WGA, Fl-SBA, and Fl-RCA were purchased from Vector Laboratories, Inc. Streptomyces hyaluronidase (Hs) and glcNAc were obtained from Calbiochem-Behring. Bovine testes hyaluronidase (Hbt), Type VI-S, Clostridium perfringens neuraminidase Type VI (<0 002 units mg−1 casein substrate, <0-003 units mg−1N-acetylneuraminic acid aldolase activity), synthetic N, N′, N″ triacetylchitotriose, galactose, grade III galNAc, grade III mixed isomers of whale and shark cartilage chondroitin sulphate and alcian blue 8-GX were purchased from Sigma. [3H]thymidine was bought from New England Nuclear. Nuclear Track Emulsion (NTB2) was from Eastman Kodak. Horse serum 4000, sterile and filtered, was from Irvine Scientific Co. Nutrient mixture F-12 (HAM), specially formulated to contain twice the amount of amino acid and pyruvate concentrations (special F-12) and penicillin–streptomycin solution were from Gibco Diagnostics Laboratories. Human umbilical cord hyaluronic acid was from Nutritional Biochemical Corp. Nuclepore filters, polycarbonate disc membranes, 0·1 μm pore size, and 10 urn thickness, were from Nuclepore. Bacto agar was obtained from Difco.
Chick lungs from stages 26-30 were dissected in a 1:1 mixture of horse serum and Ham’s F-12, then fixed and dehydrated in cold ethanol (Sainte-Marie, 1972), transferred to cold methyl benzoate and embedded in paraffin. WGA was applied to 7 gm-thick mounted sections at a concentration of 3 μg ml−1; SBA and RCA were applied at 50 μg ml−1. All staining was carried out in the dark at room temperature for h, followed by three washes of 10 min in phosphate-buffered saline (PBS). The sections were overlaid with a few drops of 80 % glycerol, 20 % PBS, and cover slips placed on top. Hyaluronic acid (HA) and chondroitin sulphate (CS) were added as dry powder to autoclaved Bacto agar (2 % in water) at a concentration of 1 mg ml−1. After the agar mixture had cooled to room temperature, it was fixed and embedded according to the schedule used for lung tissue. Agar without GAG was also fixed and embedded as a control.
To demonstrate the binding of a lectin to a particular sugar, the hapten sugars were added to solutions of the appropriate lectin at the same lectin concentration as control solutions. Control and hapten-containing solutions were allowed to stand for 1–2 h prior to staining. The concentrations of sugars used were: 0·2M-galNAc, 0·2M-gal, 0·2M-glcNAc and 0·002 M-triacetyl-chitotriose. Control solutions contained only lectin. Sections of lung tissue were then stained with control solutions, and adjacent sections were stained with the hapten-saturated lectin.
Some sections were preincubated with either Hs at 100 TRU ml−1, pH 5·3 in PBS; Hbt at 3000 NFUml−1, pH 5·3 in PBS, or neuraminidase at l unit ml−1, pH 5·3 in 0·05 M-acetate buffer at 37°C for 18 h. Adjacent control sections were incubated in buffer without the enzyme. Enzyme pre treatment was followed by three washes with stirring in PBS, before lectins were applied. Sections were photographed with a Nikon Optiphot microscope equipped with epifluorescent optics. Light with a major peak at 365 nm wavelength was used to illuminate the tissue. Kodak Tri-X film pushed to ASA 1600 by developing in Diafine was used to record the staining patterns.
For [3H]thymidine incorporation studies, lungs were dissected at stage 26, then placed on Nuclepore filters supported on steel mesh grids in organ culture dishes. The organ cultures were supplied with a medium containing 79% special F-12, 10% 9-day-old embryo extract, 10% horse serum, and 1 % penicillin-streptomycin. The lungs were incubated for 2 days at 37 °C in a 5 % CO2, humidified atmosphere in order to allow them to flatten. 20 μCi ml−1 of [3H]thymidine were added to each culture dish for 2h before removal of the tissue for fixation in Bouin’s solution, dehydration in ethanol and embedding in paraffin. 7 μm sections were cut, then hydrated before dipping in the NTB2 emulsion at 41 °C and allowed to air dry. The liquid photographic emulsion had been diluted 1:1 with a solution of 2 % glycerol in water. The slides were kept desiccated in light-tight boxes at 4°C for 10 days before developing in 1:1 Dektol, 2 min at 20°C. The sections were stained with 1 % alcian blue, pH 2·5, 0·1 mM-MgCl2 5 min and counterstained with neutral fast red for 1 min, dehydrated and mounted in Permount before being photographed in a Nikon Optiphot microscope with Kodak Panatomic-X film developed with HC110, dilution B. In all cases, several sections were examined for patterns of grain distribution.
RESULTS
Lung development
The embryonic chick lung first emerges as two ventral outgrowths from the floor of the oesophagus (Locy & Larsell, 1916). Each lung primordium consists of a single blind tube lined with a pseudostratified epithelium of endodermal origin surrounded by mesenchyme. This tube, called the mesobronchus, proceeds to elongate, and at stage 24 (Hamburger & Hamilton, 1951), the first of the primary buds appears from the dorsomedial side approximately midway down the length of the tube. The next primary buds appear in anterior-posterior sequence along the length of the mesobronchus, also from the dorsal side until the seventh bud appears at stage 28. At this time, branching is initiated from the ventral surface. At about the 10-bud stage (stage 30), the more anterior buds begin to branch, although the first bud branches earlier.
The gross morphology of the chick lung can be seen at stages 24–30 in 2-stage (1 day) intervals (Fig. 1). After stage 28, the posterior mesobronchus is displaced laterally so that the first through sixth primary buds, or bronchi, remain on the medial side of the lung, and the seventh through sixteenth bronchi are located laterally.
Development of the avian lung at five 1-day intervals. Line in corner of (D) represents 1 mm. (A) Stage 24,0–1 buds. (B) Stage 26,3 buds. (C) Stage 28,5–7 buds. (D) Stage 30,10–12 buds. (E) Stage 32,13 buds. There is much branching of the buds at this stage. At stage 30 (D) and stage 32 (E), trypan blue has been injected into the lumen to make the branching pattern visible. All photographs ×20.
Development of the avian lung at five 1-day intervals. Line in corner of (D) represents 1 mm. (A) Stage 24,0–1 buds. (B) Stage 26,3 buds. (C) Stage 28,5–7 buds. (D) Stage 30,10–12 buds. (E) Stage 32,13 buds. There is much branching of the buds at this stage. At stage 30 (D) and stage 32 (E), trypan blue has been injected into the lumen to make the branching pattern visible. All photographs ×20.
Lectin staining Fl-WGA
Fluorescence microscopy of stage-28 and -30 lungs stained with Fl-WGA reveals a preferential binding of the lectin to the basement membrane of both the lung epithelium per se and the epithelium covering the outer surface of the lung, the mesothelium (Fig. 2). In all of the tissue examined, what appears to be a basement membrane is staining, but the basal regions of the epithelial cells could also be the site of staining. In this paper, the term ‘basement membrane’ refers to staining in a linear pattern at the basal surface of epithelia. In any particular lung, a progression of development can be seen in successive buds – the anterior buds are older than the posterior buds.
The basement membranes of the buds are stained somewhat more intensely at the interbud area than at the tips. Buds ‘2’ and ‘3’ are in an early stage of outgrowth, and are thinner at their tips than bud 7’ at ‘nz’. Bud 7’ is beginning to branch. The new interbud area, ‘nz’, on its left is stained more heavily than the new tips (marked with arrows). In addition to the basement membranes, the mesothelium [‘mes’ on figure) stains in its apical and basal portions. The tips of the fourth and fifth buds do not lie in the plane of the section. The matrix stains diffusely. Figs 2–4. Patterns of staining with WGA. Mag. ×134. Bar, 0·1 mm.
The basement membranes of the buds are stained somewhat more intensely at the interbud area than at the tips. Buds ‘2’ and ‘3’ are in an early stage of outgrowth, and are thinner at their tips than bud 7’ at ‘nz’. Bud 7’ is beginning to branch. The new interbud area, ‘nz’, on its left is stained more heavily than the new tips (marked with arrows). In addition to the basement membranes, the mesothelium [‘mes’ on figure) stains in its apical and basal portions. The tips of the fourth and fifth buds do not lie in the plane of the section. The matrix stains diffusely. Figs 2–4. Patterns of staining with WGA. Mag. ×134. Bar, 0·1 mm.
In early stages of growth, the basement membrane at the tip of the bud will exhibit less staining than the interbud area. Later, when the bud has begun to branch, sparse amounts of lectin-binding material are present at the new tips, whereas staining is more evident at the new interbud area (Fig. 3).
Bud 2 (the second bud to appear) at a later stage, when it is branching. Most of the basement membrane stains intensely, but the new bud (arrow) has a very lightly staining basement membrane. The staining in the matrix seems to be associated with areas closer to the bronchi.
The basement membrane of the distal mesobronchus stains uniformly throughout the stages observed, although there is slightly less staining at the tip (Fig. 4). This portion of the mesobronchus contains sites of future bud initiation on its dorsal and ventral sides; however there appear to be no obvious discontinuities as judged by the light microscope in either area of the basement membrane. The apical surface of the mesothelium is another component of lung tissue which stains with Fl-WGA. Occasionally, the luminal surface of the inner epithelium will also stain, but not in a discernible pattern. Diffuse staining of the mesenchyme is also seen, preferentially associated in the area near the bronchi rather than near the mesothelium (Figs 2, 3).
Lectin staining Fl-SBA
SBA-binding sites are distributed in a pattern similar to that seen using WGA, but with a more pronounced difference between the tips and interbud areas. Initially, a bud shows little staining at the tip (Figs 5, 6); in a later bronchus the stain is distributed uniformly around the surface. Finally a branching bronchus will have sparse staining at the new tips, but the new interbud area, stalks and old interbud area will bind the lectin (Fig. 7).
In an older lung, buds that are just emerging follow the same pattern as buds that have arisen earlier: lightly stained at the tips (arrow), heavier staining at the interbud area.
A branching tip (arrow) is lightly stained compared to the old and new interbud area. The new interbud area is below and to the right of the arrow.
The mesothelium shows an affinity for Fl-SBA, as it does for Fl-WGA, and a faint staining of the mesenchyme is seen, in a pattern similar to that achieved with WGA. The only major difference in staining patterns using the two lectins is that the luminal surface of the bronchial epithelium is not stained with Fl-SBA, whereas it is stained by Fl-WGA. Finally, RCA also binds to sites that are codistributed with those which bind SBA (Fig. 8). Because an initial survey of the binding patterns displayed by RCA showed the pattern to be similar to SBA and WGA, further studies were not conducted on this lectin.
RCA stains the basement membranes of bronchi. The tip of the second bud from the left is branching (arrow), and stains less heavily than the interbud area, stalks and primary bronchus (mesobronchus). Mag. ×96. Bar, 0·1 mm.
Competition with hapten sugar
When the hapten sugar is combined with the lectin, 0-2M-galNAc and 0-2M-gal completely abolish the staining otherwise produced by Fl-SBA (Figs 9, 10) and Fl-RCA (data not shown). 0·2M-glcNAc abolishes all WGA staining except for that in the basement membrane (Figs 11, 12) where the residual staining is considerably diminished compared with that seen without the competing sugar. The complete removal of staining can only be achieved using 0-002 M-triacetyl-chitotriose as the hapten sugar (Figs 13,14).
An adjacent section stained with 50 μg ml−1 SBA which has been combined with 0·2 M-galNAc and allowed to stand 1 h at room temperature before application to sections. The staining is abolished.
An adjacent section stained with 3 μg ml−1 WGA after it has stood 1 h at room temperature with 0·2M-glcNAc before application to sections. All staining has been abolished except for the basement membranes.
(Control). Staining pattern with 3 μg ml−1 after it has stood 1 h at room temperature before staining.
An adjacent section stained with 3 μg ml−1 after it has stood in solution with 0·002 M-triacetylchitotriose for 1 h before staining is completely abolished.
Lectin staining of GAGs embedded in agar
Preparations of 1 mg ml−1 hyaluronic acid or chondroitin sulphate in agar were fixed and embedded in the same manner as the tissue. These were then sectioned and stained with the lectins to determine whether these GAGs are stained with WGA and SBA.
L mgml−1 HA stains with 3 μg ml−1 Fl-WGA, but no staining can be seen with 50 μg ml-1 Fl-SBA (Figs 15,16). The converse is true for 1 mg ml−1 CS: it does not stain with WGA, but does stain with SBA (Figs 17, 18).
1 mg ml−1 HA stained with 50 μg ml−1 SBA. No staining is seen with these concentrations of GAG and lectin.
Enzymic digestions
After enzymic digestions of sections with 1 unit ml−1 neuraminidase, the apical surfaces of the bronchial epithelium are markedly reduced in their ability to bind WGA. The mesenchyme is also only lightly stained in comparison with the control. The basement membrane appears to be unaffected by the neuraminidase digestion (Figs 19, 20). Streptomyces hyaluronidase (Hs) digestion of the sections reduces the staining of the mesenchyme with Fl-WGA, but the basement membrane appears unchanged in its staining characteristics (Figs 21, 22). Bovine testes hyaluronidase (Hbt) increases the overall Fl-SBA staining (data not shown). When the agar-embedded GAGs are predigested with the enzymes, HA is no longer stained with Fl-WGA after incubation in both types of hyaluronidase, whereas the control is still stained (Figs 23, 24). Hbt reduces the staining of CS by Fl-SBA (Figs 25, 26).
Adjacent section incubated with 1 units ml−1 neuraminidase 16 h at 37°C, then stained with WGA. Mesenchyme and apical surfaces of epithelium (arrows) are diminished in their staining.
(Control). Section incubated in buffer only in parallel with experimental sections, then stained with WGA.
Adjacent section incubated with 100 units ml−1 hyaluronidase 16 h at 37°C, then stained with WGA. Mesenchyme is reduced in its staining.
1 mgml−1 HA incubated with 3000 units ml−1 bovine testes hyaluronidase (A) and 100 TRU units ml−1Streptomyces hyaluronidase (B) for 16 h at 37°C. WGA does not stain the sections.
1 mgml−1 CS incubated with buffer in parallel with experimental sections before staining with SBA. Staining can be seen.
Autoradiography
3H-Tdr is incorporated into a higher percentage of cells at the tips of buds than the interbud areas (Fig. 27). This is characteristic of newly formed buds, but in older buds, the difference in DNA synthesis is not as marked. The distal mesobronchus has a high rate of incorporation which extends back to the most posterior bud. When an older bud branches, the correlation of new tip with higher incorporation than at the new interbud area is only vaguely apparent. At the junction of the two anterior mesobronchi that forms the trachea, DNA synthesis is reduced compared to the remainder of the epithelium, which shows heavy incorporation of the label (Fig. 28).
DISCUSSION
The distribution of glcNAc and, or, sialic acid (SA) as visualized by Fl-WGA staining is predominant in the basement membranes of both the lung epithelium itself and the mesothelium. Somewhat lesser amounts of these compounds can be seen at the tips of newly formed buds. After the buds enlarge, they reach a stage in which the basement membrane is uniformly covered with glcNAc- and SA-containing substances. Later, when the bronchi ramify, the new tips and interbud area repeat the sequence followed by the primary branches.
The presence of galNAc and gal, also primarily in the basement membrane, is seen by staining with Fl-SBA and Fl-RCA, respectively. The regional differences in the basement membrane during morphogenesis appear using these probes also and the same temporal changes in the pattern are seen as with Fl-WGA.
The competition experiment with the hapten sugars demonstrates the specificity of the lectins for these sugars. The fact that Fl-WGA staining of the basal lamina is not entirely prevented by the previous saturation of the lectin with 0-2M-glcNAc is unexpected. There are several alternative explanations for this phenomenon. The binding site for glcNAc might extend beyond the volume of the monomer, leaving room for additional binding to glcNAc. Perhaps the binding sites for sialic acid and glcNAc are spatially distinct, so that occupation of the site for glcNAc will not affect binding to SA, which could also be present in the basement membrane.
There is evidence that the two binding sites in WGA are indeed noncooperative and spatially distinct (Wright, 1984), but according to a recent study, SA and glcNAc bind to both subsites (Kronis & Carver, 1985). If 0·2M-glcNAc does not saturate both binding sites, the remaining site can remain free to bind either hapten, but there is, at present, no evidence in favour of an exclusive binding site for glcNAc or SA.
Alternatively, a polysaccharide may be present in the basement membrane in molecules not yet biochemically identified. The putative molecule might bind more tightly to WGA than glcNAc. For example, triacetylchitotriose, a trimer of glcNAc (Barker, Foster, Stacey & Webber, 1958), eliminates the staining at 1/100 the concentration of the monomer.
As anticipated, both HA and CS at a concentration of 1 mg ml−1 in agar stained with WGA and SBA respectively; thus it can be assumed that when these GAGs are present in the tissue, and are not prohibited sterically from binding to a lectin, their presence can be detected by this method. One cannot attribute all staining to these molecules, as other substances may contain the lectin-binding sugars. Glycoproteins such as laminin (Timpl & Martin, 1982), fibronectin (Yamada, 1981) and type IV collagen (Kefalides, 1970) contain the hapten sugars studied in this report. However, the density of these sugars in a microenvironment is low when compared to GAG’s such as HA and CS, which are composed of alternating units of the sugars. Proteoglycans are molecules that consist of a core protein with a large percent by weight (e.g. 80%) of GAG’s and therefore are likely to bind lectins.
Because the binding sites for the three lectins codistribute (except at the apical surface of the epithelium), the lectins may be staining the same substance. To explore this possibility, each agar-embedded GAG was stained with Fl-SBA and Fl-WGA, but significant staining was seen only with the appropriate lectin. Dermatan sulphate also contains galNAc and therefore could be stained by SBA as well as CS.
Alcian blue staining at different Mg2+ concentrations and digestion with different hyaluronidases are two methods which are often used to separate the presence of HA from CS (Quintarelli, Scott & Dellovo, 1964). In both cases, quantification is achieved by substraction of the first from the second amount. In addition, the alcian-blue-staining procedure is not specific for CS in the presence of embryonic HA (Derby & Pintar, 1978). Because SBA and WGA bind CS and HA, respectively, the exclusive distribution of each can be directly visualized.
The neuraminidase-treated sections showed that much of the apical surface of the bronchial epithelium is coated with a sialic acid-containing substance. This has also been demonstrated in the embryonic kidney with WGA (Ekblom, 1981) and neural tube using ruthenium red (Hay, 1978). The decrease in mesenchymal staining after neuraminidase treatment indicates that sialic acid is also present in the mesenchymal spaces.
Streptomyces hyaluronidase (Hs) specifically digests hyaluronic acid (Ohya & Kaneko, 1970; Derby & Pintar, 1978). The mesenchyme stained significantly less after digestion with this enzyme, indicating that hyaluronic acid is present in the mesenchyme in sufficient amounts to stain with the lectin.
The conclusions regarding neuraminidase- and Hs-sensitive material must be tempered with caution, as the enzymes may be contaminated. Also, in the present experiment, material may be stabilized by HA or SA, and thus could be removed by Hs or neuraminidase without being in itself a substrate for the enzymes.
The concentrations of control HA and CS were 1 mg ml−1, which correlates with a quantitative estimate of GAG levels in embryonic mouse interstitial matrix (Derby, 1978). Both positive controls ceased to stain after treatment with the appropriate enzyme. This property would extend to any HA and GAG in the tissue, except where steric conditions prevent the access of enzymes to the GAG’s. Hbt digests GAG’s into polymers of 4–14 sugar units (Weissman, Meyer, Sampson & Linker, 1954). In the tissue, GAG may be anchored to a protein in the basement membrane, and an oligosaccharide will remain bound to the protein after enzymic treatment. This may be one reason why Hbt did not reduce the staining of Fl-SBA in sections. On the other hand, the increase in staining seen after HBt may be due to an increase in general affinity of the tissue for the lectin or the digested polymers, perhaps mediated via endogenous lectins (Matsutani & Yamagata, 1982; Kobiler & Barondes, 1977; Pitts & Chang, 1981).
An oligosaccharide remaining bound to a protein core could also explain the effect of Hs on staining with Fl-WGA. Although staining with Fl-WGA was decreased by Hs in the mesenchyme, the basement membrane appeared unaffected.
The autoradiography experiments indicate greater mitotic activity at the tips than interbud areas. Several photomicrographs were scored for the number of cells per length along the perimeter in the interbud area and the tips. No evidence of different cell densities in either area was obtained; therefore the higher number of thymidine (Tdr)-incorporating cells does not reflect a higher cell density. A higher incorporation of 3H-Tdr in a single pulse at the tips could be due to a smaller proportion of non-cycling cells relative to the interbud area, as was found in the rodent lung (Lawson, 1983).
Greater incorporation of 3H-Tdr at the tips was also obtained in an experiment on chick lung at an earlier stage (Goldin & Opperman, 1980). The present investigation extended these results to cover the period of lectin staining studied.
At the tips, an inverse relationship is seen between the number of cells incorporating 3H-Tdr and a well-defined basement membrane. The distal mesobronchus shows an elevated incorporation rate, yet the basement membrane appears well stained with lectins (except for the tip), so that in areas of incipient bud formation, the above relationship does not obtain. Within the limits of resolution of this method, the distribution of DNA-synthesizing cells appears uniform beyond the most newly formed bud, therefore the site of new bud formation cannot be identified by 3H-Tdr labelling patterns.
CONCLUSIONS
Fluorescent WGA, SBA and RCAI were used to locate glcNAc, galNAc, and gal moieties in the embryonic chick lung, in hopes of staining HA, CS and perhaps KS. Positive control samples of HA and CS revealed that the GAG’s were stained by the lectins, and that the lectins showed a marked preference for specific GAG’s.
All three lectins stained the basement membrane in a developmentally significant manner: sparsely at the tips and densely in the interbud areas. The pattern was particularly evident when staining for galNAc.
The distribution of the hapten sugars correlates well with the model of branching events as proposed for salivary glands. Cell proliferation is enhanced at the tip relative to the interbud region, and therefore also agrees with the developmental programme seen in salivary glands.
Evidence has been presented here which demonstrates that similar events of morphogenesis are used in organs which differ in their pattern of branching. These similarities exist in epithelial-mesenchymal organs derived from avian as well as from mammalian embryos.
ACKNOWLEDGEMENTS
I wish to thank Norman K. Wessells, Geoffrey Goldin, Paul Green and Steve Klein for encouragement, support and thoughtful criticisms. I am grateful to Linda Jones for her expert help in typing this manuscript. This work was supported by Grant no. HD-04708 to N.K.W. and a predoctoral NSF minority fellowship.