I A Quantitative Study of the Segmental Distribution of Somitic Cells in the Developing Chick Limb Bud Using Laser-scanning Confocal Microscopy

Transcription

1 /96/$3.30 The Journal of Histochemistry and Cytochemistry Copyright by The Histochemical Society, nc Vol. 44, No. 4, pp , 1996 Printed in US.A. Original Article A Quantitative Study of the Segmental Distribution of Somitic Cells in the Developing Chick Limb Bud Using Laser-scanning Confocal Microscopy KENNETH B. R. EWAN and ALAN W. EVERETT2 Department of Physiology, The University of Western Australia, Nedlands, Australti Received for publication June 19, 1995 and in revised form October 23, 1995; accepted November (5A3704). Our aim was to map the segmental distribution of somitic cells in the limb before and after the fusion of these cells into myotubes. Somitic cells of the brachial somites were labeled by injection of D and DiO into the somitoceles of embryonic Day 2 (E2) embryos. The quantitative distribution of dye-labeled cells from injected somites was examned in whole mounts of E4 wing buds and in sections of E5 Wing buds using confocal microscopy. Dye derived from cells of anterior brachial somites 16 and 17 was highly concmtrated in the anterior half of E4 wings, whereas dye from posterior brachial somites 20 and 21 was concentrated in the posterior half of the wing. Not more than 14% (on average) of the dye in the wing was found outside the prefered half. Dye was equally dispersed in the anterior and posterior halves of the wing when the middle two somites contributing to the wing musculature, numbers 18 and 19, were injected. The total amount of dye in the whole limb at E4 was not significantly different after injection of similar amounts of dye into either pair of brachial somites 16/17, 18/19, or 20/21. n E5 embryos the distribution of labeled cells in the newly formed musde masses, identified with antibody 2H2 to myosin light chains, was examined in cryostat cross-sections of the wing. Dye was more widely distributed in the wing than at E4. As much as 32% (on average) of the dye was found in the musde mass outside the perkcred half after administration of dye into the anterior or posterior pair of somites. We condude that the branchial somites contribute similar numbers of cells to the wing musculature and that the segmental origin of these cells is rigidly maintained during their lateral migration into the limb. Cells from adjacent segments then mix when they fuse to form myotubes at E5. n addition, we found a significantly greater amount of dye per unit of musde mas in the proximal compared with distal parts of the limb. We argue that this fmding is due to division of myogenic cells during the course of their migration in the wing. (J Histochem Cytochem 4ik , 1%) KEY WORD% Myogenic cell fate; Migration of myogenic cells; Muscle development; Somites; Di; DiO; Chicken; Confocal microscopy. ntroduction Study of the fate of myogenic cells of quail somites transplanted into chick embryos has provided a detailed description of the segmental origins of the musculature of the fore (Beresford, 1983) and hind (Lance-Jones, 1988) limbs. The wing musculature is derived from cells of brachial somites 16 to 21 inclusive, whereas the leg musculature is derived from the subset of somites numbering 26 to 32. An important observation reported in both of these studies is that each somite contributes cells to only a subset of the limb muscles determined by the anteroposterior position of a somite within its regional pool. Furthermore, Lance-Jones (1988) found that there was a topographic ordering of the somitic contributions to the limb musculature that was independent of the muscle borders. t was suggested that a segmental ordering of somitic cells along Supported in part by an NHMRC Biomedical Postgraduate Scholarship (KBRE). Correspondence LO: Alan W. Everett, University of Western Australia, Nedlands, Western Australia, 6907 Australia. the anteroposterior axis of the limb most likely occurs at a very early stage of development before cleavage of the premuscle cell masses. The present study follows up these findings by examining the role of cell migration and cell fusion in the somitic contributions to the wing musculature. Our study utilizes vital labeling of the somitic cells with carbocyanine dyes and quantitative confocal microscopy to analyze the distribution of the dyes in the limb. Somites are formed in an anterior to posterior sequence from the segmental plate mesoderm during the second and third days of embryonic development. They initially consist of an epithelial ball surrounding a central cavity or somitocele containing cells of epithelial morphology. Several hours after their formation, cells of the ventral and medial portions of the somites differentiate into the sclerotome, while the remainder of the tissue gives rise to the dermomyotome and subsequently to the dermatome and myotome (see reviews by?g, and Trainor, 1994; Keynes and Sterns, 1988). Myogenic cells destined for the limb (Ordahl and Le Douarin, 1992) migrate laterally from the myotome and are present in the wing bud from HH (Hamburger and Hamilton, 1951) stage 15 onwards 347

2 348 EWAN, EVERETT (Kenny-Mobbs, 1985; Christ et al., 1977). The last wing cell leaves the somitic mesoderm by HH stage 18 (Chevallier, 1978) and reaches the distal wing by HH stage (Krenn and Wachtler, 1990). The cells aggregate into the dorsal and ventral premuscle masses from HH stage 23 (Hayashi and Ozawa, 1991; Schramm and Solursh, 1990; Newman et al., 1981) and express the myogenic regulatory factor MyoD a short time later (Williams and Ordahl, 1994). Fusion of myoblasts into myotubes and expression of muscle-specific proteins begins around HH stage 26 in the wing (Noakes et al., 1986). The findings from several studies suggest that a number of factors influence myogenic cell migration in the wing. An early report showed that migration of quail myogenic cells that had been implanted into chick limb buds was greatly impaired if the apical ectodermal ridge (AER) was removed at the time of transplantation (Gumpel-Pinot et al., 1984). Brand-Saberi et al. (1989) reported that the nature of the mesoderm of the wing was crucial to myogenic cell migration. They found that implanted quail myogenic cells migrated into the host tissue only if the host tissue was developmentally younger than the grafted quail tissue, even in the absence of the AER. The orientation of the wing mesoderm has also been shown to influence the direction of myogenic cell migration. Quail myogenic cells fail to migrate distally if grafted proximal to a block of wing mesoderm rotated 90" or 180" (Krenn and Wachtler, 1990). The authors postulated that the cells and/or the extracellular matrix of the limb bud mesoderm are structured in such a way as to ensure that the myogenic cells migrate in only one direction. We have attempted here to make quantitative estimates of the anteroposterior distribution of somitic cells during and after their migration phase in the wing. We find that anteroposterior movements are highly restricted during migration throughout the length of the limb bud but that "mixing" then occurs between cells from adjacent somites during the process of myotube formation. Materials and Methods njecting Somites. Fertile chicken eggs were incubated on their side at 37-38'C in a humidified incubator to HH (Hamburger and Hamilton, 1951) stages (50-56 hr). The uppermost part of the shell was windowed to expose the embryo. A small volume of sterile ndian ink (Hunt Speedball) diluted 1:4 in Ca++/Mg*+-free Hank's buffered saline was injected underneath the embryo to highlight the somites. Somites were counted according to the scheme of Beresford (1983). Di and DiO (Molecular Probes; Eugene, OR) were solubilized by sonication in Ca"/Mg"-free Hank's buffered saline (1 mg/ml) supplemented with 800 Wglml phosphatidyl choline (Sigma; St Louis, MO) according to the method of Hayashi and Ozawa (1991). This preparation was backfilled into a pulled glass microelectrode connected to a fluid-filled, gas-tight 1O-pl Hamilton syringe mounted on a micromanipulator. Somites were impaled through the dermomyotome and a micrometer screw mechanism was used to depress the plunger of the syringe and inject nl of the dye solution into the somitocele. An experiment was conducted in which a radioactive tracer was added to the dye to determine the volume being delivered in our experiments (data not presented). The size of the somites did not appear to change when they were injected. njections were always carried out on somites located on the right-hand side of the embryo. The window in the shell was sealed with tape and the egg was returned to the incubator. Confocal Microscopy. Digitized images of E4 whole-mounted wings and transverse sections of E5 wings were acquired with a BioRad MRC-1000 laser-scanning confocal imaging system fitted to a Nikon Diaphot inverted microscope with a Nikon PLAN-APO x 10 air objective. The system is equipped with the BioRad COMOS operating software version Di labeling was detected at the 568-nm excitation laser setting for Texas Red using the 605/32 emission filter (peak emission at 605 nm, 32-nm window at 50% of peak emission) attached to the red band epidetector. DiO and FTC labeling was detected at the 488-nm excitation laser setting using the emission filter attached to the green band epidetector. The gain and background settings were adjusted to ensure that the fluorescence intensity was within the working range of the epidetector. mages were checked for this criterion by using the SET PMT-1 look-up table macro which displays pixels in the image with intensities below the limit of detection as green and intensities above the saturation level as blue. Each image was Kalman-averaged over five scans to reduce the noise level. mages were displayed on the confocal microscope work station with the CorelPhoto-Paint program. They were either left black and white or were pseudo-coloured and captured on Kodak EPP-100 color transparency film or lford PAN-F black-and-white film using a Focus magecorder Plus camera system. A linear response of the photomultipliers to change in the amount of fluorescent dye in the specimen was demonstrated using standard solutions of DiO and Di (Figure 1). Both dyes were dissolved in 100% ethanol and serially diluted in the same solvent. A small volume of each solution was then sealed between a glass slide and a coverslip separated by a 1-mm spacer. Fifteen optical sections (2-Sections) 20 pm apart were captured and projected as a single image using the "Summary" option in COMOS, which determines the mean intensity of corresponding pixels in all the optical sections. The total pixel intensity of the projected images was plotted against the concentration of dye in the sample to obtain points shown in Figure 1. E4 Embryos. Embryos were fixed at HH stages 22 and 23 with 2% paraformaldehyde in 0.1 M phosphate buffer (ph 7.3) for 30 min at 4'C and washed in PBS overnight. Each limb bud was whole-mounted in an antifade medium consisting of 1 mglml p-phenylenediamine dihydrochloride in 90% glycerol/lo% PBS, ph 8. Using the 2-sectioning program of the COMOS software, optical Kalman-averaged sections were taken 20 pm apart. Projections of the 2-series of images ofthe whole limb were then made as described above. Examples of images so generated are shown in Figures 3A-3C. Embryos were rejected from analysis if dye was seen to extend beyond the boundaries of two labeled myotomes in the whole-mounted preparations. E5 Embryos. Embryos were fixed at HH stages as described for E4 embryos, ntact wings and trunks were immersed overnight in a 1:jOO dilution of ascites 2H2 monoclonal antibody (Noakes et al., 1986) to myosin light chains. After several washes in PBS, the wings were immersed in a 1:50 dilution of FTC-conjugated sheep anti-mouse gg (Silenus; Melbourne, Australia) and washed again in PBS. The wings were then left overnight in 15% sucrose in PBS. embedded in Bright CryoEmbed, and frozen. Transverse 20-pm sections the limbs were mounted with antifade medium (see E4 embryos above) and immediately stored at -20 C until examination by confocal microscopy. Antibody staining of limb whole mounts before sectioning minimized the spread of the Di by diffusion. Similarly, diffusion of Di was avoided after sectioning of the limb if the mounted sections were frozen and thawed only for a short observational period. Embryos were rejected from analysis if dye in the tissue sections was found to extend beyond the boundaries of two labeled myotomes. Results To trace the movement of somitic cells in the limb bud, somitic cells of adjacent pairs of brachial somites inclusive in HH stage embryos were labeled by microinjection into the somitoceles of nl of an aqueous solution of Di or DiO (Figure 2). The distribution of dye in the limb buds on the injected

3 349 DSTRBUTON OF SOMTC CELLS N THE LMB BUD o 1.5 Dye Concentration (mg/l) Figure 1. Relationship between the amount of DiO (filled circles) and Dil (open circles) and the pixel intensity (arbitrary units) due to fluorescence of the dyes. A gain setting was selected for the respective photomultipliers that gave images with pixel intensities ranging between zero (black) and 5 (white or saturation) over the range of dye concentrations shown. Three determinations were made for each dye at all concentrations tested. At any given concentration, DiO produces a slightly greater fluorescent signal than Dil. Data are background corrected by subtracting the fluorescent signal due to the solvent (ethanol). side was then analyzed in whole from ~4 embryos and in cryostat sections from Embryos were from analysis if dye was observed outside the boundaries of two labeled myotomes. Figure 2. Confinement of DiO to a single somite. An E2 embryo was injected with an aqueous solution of DiO into one somite and was fixed 1 hr later. (A) Micrograph of a projected confocal image of the whole mounted embryo optitally sectioned throuqh the dorsoventral axis. Only Part of the dorsal view is illustrated and includes the neural tube (nt) and several adjacent somites (arrows), one of which has been injected with the dye (*). (E) Phase micrograph of the same region of the embryo shown in A. Bar = 0 pm. E4 Embryos E2 embryos had DiO injected into either somite pair 16/17, 18/19, or 20121, and 2 days later (HH stage 22-23) whole mounts of the limb buds from these embryos were examined by confocal microscopy. Between 18 and 20 optical sections (1.7 pm thick) were taken at 20-pm intervals, extending from the dorsal to the ventral surface of the limb buds. These Z-series of images were used to obtain a single two-dimensional image in which each pixel (2.7 pm2)was assigned an intensity value equal to the mean intensity of the corresponding pixels in all the optical sections; this is referred to as a Z-projection. Although some loss of fluorescence signal because of light scattering would occur as the optical plane of focus penetrates deeper into the tissue during optical sectioning, such losses would be expected to be the same for all limbs. and therefore no correction has been made for it in the present work. Figures 3A-3C show representative projected images of limb buds from three embryos in which dye was injected into somites 16/17 (Figure 3A). 20/21 (Figure 3B), or 18/19 (Figure 3C). The images have been pseudo-colored so that the strongest staining appears white and the weakest green. t can be seen that the DiO has a very restricted distribution in the limb and is confined to a pathway that projects laterally from the injected somites (to the right of each image just out of the field of view). A quantitative assessment of these data was undertaken by determining the anteroposterior distribution of dye in the limb buds. This involved determining the mean pixel intensity within rectangles (163 x 24.7 pm) positioned at increments extending from the anterior edge to the posterior edge of the buds at a proximal and distal location using the COMOS image analysis software; i.e., along two pathways (see lines drawn across the limb bud in Figure 3A) parallel to the anteroposterior axis of the limb. The precise proximal and distal locations used for this analysiswere selected according to strict criteria (see legend to Figure 4). n this way we could evaluate if the anteroposterior distribution of dye changed as the cells migrated along the proximodistal axis of the limb. Mean pixel intensities were similarly obtained for the contralateral limb of all embryos and were subtracted from the data for the dye-labeled limb to correct for tissue autofluorescence. The results of this analysis are presented in Figure 4. Figure 4A shows that dye from somites 16 and 17 is concentrated in the anterior half of E4 wing buds, whereas dye from somites 20 and 21 is concentrated in the posterior half of the wing buds (Figure 4B). n fact, from inspection of these curves it is apparent that most of the dye is confined to either the anterior or

4 . -.,-. H Figure 3. Pseudo-colored fluorescence images of carbocyanine dye-labeled E4 limb buds (A-C) and sections of E5 limb buds (D,F,H) and anti-myosin staining of these sections (E,G,). (A-C) Whole mounts of HH stage 22 wings buds from embryos injected with DiO into somite pairs 16/17, 20121, and 18/19, respectively. These are projected images each derived from optical sections taken 20 pm apart through the dorsoventral axis of the limb buds. Anterior is toward the top of the figures. (D-) Micrographs of confocal images of 20-fim cross-sections of wings from HH stage embryos showing Dil labeling (D,F,H) and antimyosin light-chain labeling of the same sections (E,G,) after injection of Dil into somite pairs 16/17 (D,E), 20/21 (F,G), and 18/19 (H,) of E2 embryos. The sections are oriented with the dorsal surface of the wing toward the top of the figure. The strongest fluorescence in all images appears white. a, anterior; p. posterior. Bars = 0 pm.

5 DSTRBUTON OF SOMTC CELLS N THE LMB BUD Somites 16/17 A Somites 20/21 B Figure 4. Distribution of DiO in E4 wing buds. The anteroposterior distribution of DiO fluorescence in the limb buds was determined at two positions, one proximal (filled symbols) and one distal (open symbols). These positions were, respectively, at one-quarter (224 pm on average) and three-quarters (672 pm on average) of the distance from the lateral edge of the myotomes to the proximal boundary of the progress zone of the growing end of the limb bud. The proximal boundary of the progress zone was assumed to be 0 pm from the wing tip; this assumption was based on findings from earlier studies of the migration of somitic cells into the wing (Hayashi and Ozawa, 1991; Newman et al., 1981). The anterioposterior axis was constructed parallel to the lateral edge of posterior third of these wings. The anteroposterior distribution of dye was almost identical in the proximal and distal regions of the limbs (open and filled symbols, respectively, in the figures). From the area under these curves we calculated that 3-14% of the dye was found outside the preferred half-wing when the dye was injected into either the two anterior (16117) or posterior (20121) somites (Table l). Very similar amounts of the dye are found in the anterior and posterior halves of the limb when somites 18 and 19 are injected with DiO (Figure 4C; Eble 1). Batches of E4 embryos were also assessed for the total amount of dye in the whole limb bud in an effort to evaluate the contribution of cells made by each somite pair to the limb musculature. Wing buds were viewed on the confocal microscope only after all the embryos had been injected with similar amounts of dye into the somites. We found that the total pixel intensities of projected images were not significantly different after injection of DiO into somites 16/17, 18/19, or (Figure 5). E5 Embryos The objective with these embryos was to investigate the distribution of dye in the limb in relation to the newly forming muscle masses. Myotubes were identified immunocytochemically in sections of limb with monoclonal antibody 2H2 to myosin light chains (Noakes et al., 1986). Antibody binding was detected with a fluorescein-conjugated anti-mouse antibody. Fluorescein fluorescence could not be resolved from DiO fluorescence with our miccroscope, so the longer wavelength-emitting carbocyanine dye Di was used to label somitic cells. DiO fluorescence is compatible with rhodamine-conjugated second antibodies, but we avoided this combination because the rhodamine produced high tissue background fluorescence. E2 embryos were injected with Di and left for another 3 days. The limbs were then sectioned for analysis of the distribution of Di within the early developing dorsal and ventral muscle masses in both the proximal and distal parts of the limb. Dye was concentrated in the anterior parts of the muscle masses when somites 16 and 17 were injected with Di (Figure 3D). Newly forming myotubes in the dorsal and ventral muscle masses are clearly delineated by anti-myosin staining of the same section (Figure 3E). The converse is found when Di is injected into somites 20 and 21, i.e., most of the label is confined to the posterior part of the dorsal and ventral muscle masses (Figures 3F showing Di and 3G showing anti-myosin). Di is generally more widespread throughout the muscle masses when somites 18 and 19 are injected (Figures 3H showing Di and 31 showing anti-myosin). A quantitative analysis of the distribution of dye in the muscle bellies was made by determining the total pixel intensity due to fluorescence in the anterior and posterior portions of the area delineated by anti-myosin staining in the section. Determining what the myotomes. The data have been corrected for tissue autofluorescence. (A-C) Embryos injected with DiO into somite pairs 16/17, 20/21, and 18/19, respectively. Shown are the means f SEM for not less than four embryos in each group. Within each group the proximal and distal DiO intensity values (% maximum) were not significantly different at any given anteroposterior position on the limb except at the 80% value in B (p < 0.05; Student's t-test). We do not consider the difference between the intensity values at 80% to be physiologically significant and this is not considered further.

6 EWAN, EVERETT 352 Table 1. Distribution of DiO in the antenor and postenor halves of E4 wing buds Percentage of DiO Somites injected , Proximal Anterior 97 -c 2 14 f 5 59 f 7 Distal Posterior Anterior Posterior c 5 41 t 7 89 * 3 11 f c 4 1 -c 3 89 f c 4 a Thc antcropostcrior distribution of DiO in thc limb was dctcrmincd at a proximal and distal location as described in the text and thc lcgcnd to Figurc 4. Shown arc thc means * SEM for not lcss than four embryos in each group. A significant diffcrcncc (p < paircd f-tcst) was found bcrwccn antcrior and postcrior values (proximal and distal) whcn somites 16/17 and were injected with dye. were the true anterior and posterior halves of E5 limbs proved difficult because the limb after E4 undergoes a medial rotation. At issue was where to draw the dorsoventral axis through the section of wing to divide the tissue into its anterior and posterior parts. We opted to construct the axis through the midpoints of the dorsal and ventral muscle masses by the procedure illustrated in detail in Figure 6, and to determine the total pixel intensity due to Di fluorescence in the muscle masses on either side of the division. Vindication of this method for dividing the muscle masses into the anterior and posterior portions is provided by the results of the distribution of dye in the muscle after injection of somites 18 and 19 (Figure 7A). Because dye injected into these somites distributes evenly between the anterior and posterior halves of the limb at E4 (Figure 4C; Table ), equal amounts of dye would also be expected U) iz m C T 16/17 18/19 20/21 Somlte Number Figure 5. Total fluorescence intensity due to DiO in E 4 wing buds. Batches of E2 embryos (and all within the somite stages of development) were injected with DiO into somites 16/17,18/19.or 20/21. Projected images of optically sectioned wing buds from these embryos at E4 (HH stages for all embryos) were used to determine the total pixel intensity due to DiO fluorescence. The proximal boundary of the limb bud was set in line with the lateral edge of the myotomes. A total pixel intensity value due to intrinsic fluorescence of limb tissue was obtained by scanning the contralateral limb and subtracting this value from that obtained for the dye-containing limb. Fluorescence intensity measurements of all limbs were carried out using the same photomultiplier sensitivity settings. The same settings were also used to obtain the data on the relationship between dye and fluorescence intensity that are presented in Figure 1. Shown are the means f SEM for not less than four embryos in each group. Figure 6. Method for dividing muscle masses into anterior and posterior portions in sections of E5 wing. The micrograph is of a 2C-vm cryostat section from an HH stage 27 wing stained with antibody 2H2 to myosin light chains. The midpoint of the long axis of both muscle masses in cross-section was determined (dot in center of long thin lines), and then a line connecting these points (thick line) was drawn to divide the two muscle masses into the anterior and posterior portions. The dorsal muscle mass is toward the top of the figure. a, anterior; p. posterior. Bar = 0 vm. to be found in the anterior and posterior portions of the muscle masses at E5. n the proximal limb this is the case; the total amount of dye in the anterior and posterior muscle portions was not significantly different (see Figure 7A). n the distal limb there is a small, albeit significant, difference. Di was clearly confined to the anterior or posterior portions of the muscle masses after injection into somites or (Figures 7B and 7C, respectively). However, approximately 30% of the dye is found outside the preferred half in both the proximal and distal limb, indicating a broader distribution of dye in the limb compared with E4 embryos. Finally, it was apparent that in all embryos the pixel intensity due to Dil was higher in the proximal muscle masses compared with that in the distal masses. n the group of four embryos injected with dye into somites 18/19, for example, the mean pixel intensity (arbitrary units, corrected for background tissue fluorescence) was 38 k 4 (mean k SEM) in the proximal masses (dorsal and ventral) compared with 13 f 1 in the distal masses. The difference was highly < paired Student s t-test). This result suggests that less dye was delivered to the distal limb per unit of muscle mass. The significance of this observation is considered further in the Discussion to follow. Discussion This study makes use of the carbocyanine vital dyes DiO and Di to determine the fate of somitic cells in the developing wing. These highly lipid-soluble dyes dissolve in cell membranes and are not known to produce any toxic effects (see Honig and Hume, 1986). except perhaps when nonaqueous solvents are used to solubilize the dyes. n our experiments we bypassed this problem by solubiliz-

7 DSTRBUTON OP SoMFlC CELLS N THE LMB BUD 353 '"1 Somites 18 and 19 A ing the dyes in an aqueous vehicle containing phosphatidylcholine according to the method of Hayashi and Otawa (1991). The usefulness of the dyes to mark cells declines when cells grow or divide, and for this reason their use in the present study could not be extended beyond E5, when wing muscle growth is extensive and fluorescence due to the dyes becomes all but undetectable. n 8 v C. v) S 0) U r Q) X e. dentity of Dye-labeled Cells in Limb Buds The fate of cells of somitic origin is well characterized, and we ar- gue that the movement of the dyes into the limb after their injection into the somites is almost entirely due to the migration of myogenic precursors. Other cells of somitic origin contribute to the vertebral column and dermis of the back, and therefore are not Proxima Distal a complicating factor in our analysis. Staining was observed in the dorsal root ganglia in E5 embryos, suggesting that migrating neural crest cells also become labeled, most likely during their migration through the anterior halves of somites (Serbedzija et al., 1989). Somites 16 and 17 B 1001 L Ant Ant 75 ".& Post... T n -... Proximal Distal However, we did not observe staining in the ventral roots of E5 embryos (data not presented), making it unlikely that those crest cells destined to become Schwann cells would have contributed significantly to the dye in the limb. Moreover, although melanoblasts of neural crest origin migrate along the posterior border of the wing bud at HH stages (Grim and Christ, 1993), we found no evidence of unusually high intensity of dye along the posterior border of E4 or E5 wings after injection of dye into somites t is also unlikely that diffusion of the highly lipid-soluble dyes between membranes of adjacent limb mesenchymal cells contributed to the dye in the limb. Diffusion would be expected to lead to a broader distribution of dye in the distal limb and this did not happen; the anteroposterior distribution of DiO was identical in the proximal and distal regions of the limb buds (see Figure 4). This issue has been addressed by others (Honig and Hume, 1986), who reported no evidence for transfer of the dyes between cell membranes Somites 20 and 21 C Qauntitative Analysis of Cell Fate We showed that the intensity (between 0 and 5 arbitrary units) Post T - Post of pixels making up a confocal microscope-generated image was proportional to the amount of fluorescent dye in the specimen (see Figure 1); the specimen in this instance was dye dissolved in ethanol. We are uncertain to what extent the same relationship holds true when the dye is contained within the tissues of the limb. However, because the tissue being examined was identical for all embryos, relative fluorescence intensity measurements should reflect the relative amounts of dye in the limb buds after injection of the n - P roxi ma Distal Figure 7. Distribution of Dil in E5 wing buds. Batches of E2 embryos (all within the somite stages of development) were injected with Dil into somites 18/19 (A), 16/17 (E), or 20/21 (C). The embryos were sacrificed about 3 days later (at HH stages 26-27) for serial cryostat sectioning of the wing on the injected side. Distal and proximal 20-em sections were taken 845 f pm (mean -c SEM; n = 13 wings) and 1204 f em from the wing tips, respectively. The sections were then stained for myosin light chains to delineate the muscle 4 bellies. An optical section (2.2 am thick) was collected of both Dil and fluorescein (myosin) fluorescence from approximately the middle of the section of tissue. Pixel intensity values due to Dil fluorescence in the anterior (Ant) and posterior (Post) portions of the muscle bellies (see Figure 6) are expressed as a percent of the total. ntensity values are corrected for background fluorescence determined by scanning sections taken at random from the wing on the uninjected side of each embryo. A significant < 0.05; paired t-test) was found between anterior and posterior values in all cases except data from the proximal wing in A.

8 354 EWAN, EVERETT somites. Accordingly, the similar fluorescent intensities found in whole limb buds after injection of dye into either pair of somites suggested that each somite contributes a similar amount of dye, and therefore cells, to the wing. This analysis would depend on the amount of dye injected into the somites and on whether migrating cells were fully saturated with dye. Although we have no data about the level of saturation of somitic cells with the dye, judging from the very intense staining in the myotomes of all embryos it was likely that we were injecting a great excess. Dye became more widely dispersed in the limb buds between E4 and E5, when myotubes first begin to appear in the limb. Dye from the injected somites was largely confined to the developing muscle masses (see Figures 3D-31) at E5 and was mostly still within the anterior or posterior halves of the muscle masses when injected into the anterior and posterior pairs of somites, respectively. However, as much as 30% of the dye was outside the preferred halfwing compared with less than halfthis percentage at E4. This finding could be due to some anteriorposterior movement of postmigratory myogenic cells (after E4) and/or to fusion of labeled cells with those originating from adjacent unlabeled somites. n E5 embryos we showed that the total amount of dye per unit area of muscle was least in the distal limb. Therefore, less dye was being carried to distal sites in the limb by migrating myoblasts than was delivered proximally when normalized for the muscle mass. Although it may be that fewer myoblasts per unit of muscle mass do reach the distal limb, our findings can also be explained by myoblast division in the course of their movement through the limb, diluting the dye in those myoblasts that eventually make it to the distal musculature. We favor this interpretation in the light of an autoradiographic study of the migration of tritiated thymidinelabeled quail wing tissue grafted into host stage 20- wing buds (Lee and Ede, 1990). The study showed that quail cells become less intensely labeled at more distal locations from the graft site. Migration of Myogenic Ce1h into the Wing A variety of extracellular matrix molecules (see Brand-Saberi and Christ, 1993) and, more recently, agrowthfactor (Bladt et al., 1995) have been implicated in the migration of myogenic precursor cells into the limb. These cues are probably nonspecific in the sense that they appear to be followed by somitic cells from any segmental level. Early studies examined the issue of segmental identity of somite cell derivatives and showed that somites grafted to different segmental positions contributed to the musculature in accord with the new (host) segmental position (Lance-Jones, 1988; Chevalher et al., 1977). A model postulating the physical restriction of myogenic cell migration to the proximodistal axis was proposed based on the results of experiments in which quail myogenic cells were grafted into chick wings proximally to a block of wing mesoderm that had been rotated 90 or 180 (Krenn and Wachtler, 1990). The rotated block of mesoderm behaved like a barrier, preventing the distally directed migration of the grafted quail myogenic cells. The barrier effect was not seen if the graft was made into a wing bud younger than HH stage 24, which is beyond the stage at which our embryos were sacrificed (HH stages 22 and 23). However, the barrier model may still be relevant in younger embryos if the limbs of pre-stage 24 chicks are able to respecify the properties of a rotated block of limb mesenchyme. A salient finding from our work is that there is very little anteroposterior movement of the myogenic precursor cells during their migration along the entire proximodistal axis of the limb bud. This finding is consistent with the physical restriction model described above for ensuring the distal migration of these cells. n addition, our results suggest that when the migration phase is over, mixing of the cells between adjacent segmental levels occur and, as a result, only a fairly coarse representation of the segmental origins of the muscle fibers is preserved in the mature musculature. Acknowledgments Very special thanb to Dr Ulrich SeydeL, Centre for Celland Molecular Biology, University of Western Australair, for he& with use of the confocal microscope. Literature Cited Beresford B (1983): Brachial muscles in the chick embryo: the fate of individual somites. J Embryol Exp Morphol 7799 Bladt F, Riethmacher D, senmann S, Aguzzi A, Birchmeier C (1995): Essential role for the c-met receptor in the migration of myogenic precursor cells into the limb bud. Nature 376:768 Brand-Saberi B. Christ B (1993): nhibition of myogenic cell migration by the application of antibodies raised against limb bud mesenchyme. n Fal- on JF, Goetinck PF, Kelly RO, Stocum DL, eds. Limb development and regeneration. New York, Wiley-Liss, 541 Brand-Saberi B, Krenn V, Christ B (1989): The control of directed myogenic cell migration in the avian limb bud. Anat Embryol 180:555 Chevallier A (1978): Etude de la migration des cellules somitiques dans le mcsoderme somatopleural de Cbauche de aile. Rouxs Arch 184:57 Chevallier A, Kieny M, Mauger A (1977): Limb-somite relationship: origin of the limb musculature. J Embryol Exp Morphol 41:245 Christ B, Jacob M, Jacob H-J (1977): Experimental analysis of the origin of the wing musculature in avian embryos. Anat Embryol 150:171 Grim M, Christ B (1993): Neural crest migration into the limb bud of avian embryos. Prog Clin Biol Res 383A:391 Gumpel-Pinot M, Ede D, Flint 0 (1984): Myogenic cell movement in the developing avian limb bud in presence and absence of the apical ectodermal ridge (AER). J Embryol Exp Morphol 80:105 Hamburger V, Hamilton HL (1951): A series of normal stages in the development of the chick embryo. J Morphol 88:49 Hayashi K, Ozawa E (1991): Vital labeling of somite-derived myogenic cells in the chicken limb bud. Rouxs Arch 200:188 Honig MG, Hume R (1986): Fluorescent carbocyanir 2 dyes allow living neurons of identified origin to be studied in long-term cultures. J Cell Biol 103:171 Kenny-Mobbs T (1985): Myogenic differentiation in early chick wing mesenchyme in the absence of the brachial somites. J Embryol Exp Morphol 90:415 Keynes RJ, Sterns CD (1988): Mechanism of vertebrate segmentation. Development 103:415 Krenn V, Wachtler F (1990): THe orientation of the wing mesenchyme influences the direction of the migration of myoblasts in the avian embryonic wing bud. Anat Embryol 181:453 Lance-Jones C (1988): The somitic level of embryonic chick hindlimb muscles. Dev Biol 126:394 Lee KKH, Ede DA (1990): An autoradiography study of myogenic cell movement in avian limb buds following heterospecific and homospecific transplantation. Anat Embryol 182:547 Newman SA, Pautou M-P, Kieny M (1981): The distal boundary of myo-

9 DSTRBUTON OF SOMMlC CELLS N THE LMB BUD 355 genic primordia in chimaeric avian limb buds and its relation to an accessible population of cartilage progenitor cells. Dev Biol 84:440 Noakes PG, Everett AW, Bennett MR (1986): The growth of muscle nerves in relation to the formation of primary myotubes in the developing chick forelimb. J Comp Neurol 219:133 Ordahl CP, Lee Douarin NM (1992): Two myogenic lineages within the developing somite. Development Schramm C, Solursh M (1990): The formation of premuscle masses during chick wing development. Anat Embryol 182:235 Serbedzija GN, Bronner-Fraser M, Fraser SE (1989): A vital dye analysis of the timing and pathways of avian trunk neural crest cell migration. Development Tam PPC, Trainor PA (1994): Specification and segmentation of the paraxial mesoderm. Anat Embryol 189:275 Williams BA, Ordahl CP (1994): Pax-3 expression in segmental mesoderm marks early stages in myogenic cell specification. Development 120:785