AK and JPB conceived and designed the experiments. AK, CV, and YI performed the experiments. AK, CV, YI, and JPB analyzed the data. CV, YI, and JPB contributed reagents/materials/analysis tools. AK and JPB wrote the paper.
The authors have declared that no conflicts of interest exist.
The conversion of multinucleate postmitotic muscle fibers to dividing mononucleate progeny cells (cellularisation) occurs during limb regeneration in salamanders, but the cellular events and molecular regulation underlying this remarkable process are not understood. The homeobox gene
Amphibians such as the salamander can regenerate their limbs. This paper explores how multinucleate muscle cells transform into mononuclear cells and begin to proliferate during regeneration.
There is currently a significant focus on strategies to promote regeneration in adult mammals and therefore a renewed interest in the mechanisms that underlie regeneration in urodele amphibians. An adult salamander such as the newt or axolotl can regenerate its limbs and tail, jaws, ocular tissues such as the lens, and small sections of the heart (
Our understanding of this intriguing process has received considerable impetus from the recognition of two manipulations that induce mammalian myotubes to undergo fragmentation. The first is exposure to myoseverin, a trisubstituted purine derivative isolated from a combinatorial library (
The studies on cellularisation by myoseverin and
In order to obtain striated myofibers, tissue was isolated from the limbs of two species of larval salamander (
Photomicrograph of a live striated myofiber dissociated from the larval limb musculature and adhering to the culture dish in serum-free medium. This cell has the appearance of a normal quiescent fiber and was photographed with VAREL optics at 48 h after plating. Scale bar, 50 μm.
When cultures were labelled with tritiated thymidine, no labelled nuclei were observed in multinucleate cells after labelling for 24 h (540 myofibers, five cultures) or 48 h (263 myofibers, three cultures), while 16% of the mononucleate cells were labelled in the latter case. It is noteworthy that the absence of S-phase entry in nuclei within multinucleate cells includes the population of myofibers that undergoes the events of cellularisation or fragmentation described below.
The dissociation of viable myofibers has allowed us to evaluate their plasticity after implantation into a limb blastema, a procedure that has previously been performed only on myotubes formed in cell culture. Fibers were labelled with a cell tracker dye in suspension after dissociation, and single fibers were examined to verify the absence of any adherent mononucleate cells and were drawn into a glass micropipette (
(A) Schematic diagram of procedure. After dissociation of larval limb musculature, the cells were loaded with a cell tracker dye and single myofibers taken up into a suction micropipette, prior to injection into a larval limb blastema as detailed in the
(B) Section of a limb at 48 h after implantation of CellTracker Orange-labelled myofibers. The section has been counterstained with the nuclear stain Sytox green. Note the dye-labelled mononucleate cells (arrowed). Scale bar, 20 μm.
After 48 h in culture, some of the striated fibers remained viable, but showed no significant change in morphology and retained the appearance of the cell shown in
Approximately 10% of the total population of myofibers underwent changes in nuclear localisation within the cells such that a lobulated or ‘cauliflower’ structure formed in the middle or end of the cell (
(A) Phase-contrast micrograph of a live cell at 3 d after plating, showing a lobulated structure in the middle of the fiber.
(B) Micrograph of a live fiber at 2 d after plating, showing budding of nuclei at one end. The cell has been counterstained with Syto 13.
(C) Fluorescence micrograph of a myofiber at 24 h after microinjection with TR–dextran. The cell has been counterstained with Syto 13 dye to show the nuclei.
(D) Fluorescence micrograph of a colony formed from a single myofiber injected 24 h earlier with TR–dextran. The cell has flattened on the substrate and the nuclei are stained with Syto 13 dye.
(E) Fluorescence micrograph of a colony formed from the progeny of several myofibers in proximity that were injected 5 d earlier with TR–dextran.
(F) Analysis of the DNA content of cells derived from myofibers injected 5 d earlier with TR–dextran. The DNA content was determined by image analysis of the nuclei of mononucleate TR-positive cells that had been stained with Hoechst (see
(G) Photomicrograph of a live myofiber, 48 h after plating, showing a binucleate bud formed at the end. The cell was stained as for (B).
(H) Fluorescence micrograph of a bud containing three nuclei stained with Syto13 (yellow) derived from a myofiber that contained at least five nuclei and that was injected with TR–dextran (red).
Scale bars: (B), (C), and (G), 100 μm; (A), (E), and (H), 50 μm; and (D), 10 μm.
We have analysed the DNA content of single Hoechst-stained nuclei by normalised measurements of fluorescence intensity in TR-labelled cells within such colonies, and an example of a representative distribution for a single colony is shown in
In a second aspect of plasticity, which was shown by 40%–70% of the total population of myofibers, the initial stages also involved the migration of nuclei to form local aggregates, often of two or three nuclei, within the fiber. The migration of nuclei into a terminal aggregate is illustrated by selected images from a time-lapse video analysis (
(A) Single frames illustrating the migration of three nuclei (yellow arrows) along a myofiber, of which two are incorporated into a terminal aggregate by 11.4 h. One nucleus (green arrow) remained stationary during this period.
(B) Single frames illustrating the production of viable multinucleate fragments from a myofiber. Note the presence of a trinucleate aggregate (arrowed green) that separates after lateral breakage of the fiber (0 min, arrowed yellow). This fragment subsequently extends cytoplasmic processes (14.3 and 15.4 h) and migrates over the culture substratum.
Series (A) and (B) begin at 6 h after plating. Scale bars: (A) 50 μm; (B) 200 μm.
In view of the evidence that implicates microtubules as a target for myoseverin, we stained the cultures with antibody to β-tubulin. Although tubulin was polymerised in microtubules parallel to the axis of the fibers, the regions of nuclear aggregation were associated with depolymerised tubulin (
(A) The distribution of microtubules surrounding a multinucleate aggregate on a myofiber, as analysed by staining with anti-β-tubulin. Note the relatively disordered state of the tubulin (arrowed) in the vicinity of the nuclei. The fiber was stained at 48 h after plating. Scale bar, 50 μm.
(B) Taxol inhibits the activation of myofibers after dissociation. Myofibers were dissociated and cultured in the presence of taxol as described in the
The cultures were reacted with a digoxygenin-substituted antisense riboprobe to axolotl
(A) Expression of
(B) Absence of significant
(C) Expression of
(D) Expression of
In an initial investigation of this possibility, cultures were arrested as before by treatment with taxol, followed by reaction with the
In order to evaluate the uptake of morpholino-substituted oligonucleotides, larval myofibers were dissociated as usual in the presence of 10 μM biotinylated morpholinos or underivatised morpholinos. The cells were cultured for 48 h and then analysed by a detection procedure involving tyramide signal amplification (see
(A and B) Uptake of morpholino by myofibers. Myofibers were dissociated in the presence of biotinylated (A) or control (B) morpholinos and analysed by tyramide signal amplification at 24 h after plating. Note the positive signal in (A), dependent on the presence of biotin moiety. In three different experiments 70%–90% of the fibers were loaded as determined with this assay. Scale bar, 50 μm.
(C) Functional effect of loading various morpholinos. Note that loading
(D–G) Staining of myofibers with antibody to Msx1 protein. (D and E) Fluorescence micrograph of a nucleus in a quiescent myofiber stained with Hoechst for DNA (D) and Msx1 protein (E). (F and G) Fluorescence micrograph of a nucleus in an active myofiber stained for DNA (F) and Msx1 protein (G). These images (D–G) were taken from the same culture. Scale bar, 20 μm.
(H) Distribution of fluorescence intensity of nuclei in myofibers after staining with antibody to Msx1. The distributions for control active fibers and control quiescent fibers were determined for cells in the same culture and are significantly different (ANOVA,
Limb tissue was dissociated in the presence of control morpholinos or a morpholino antisense reagent directed at the translation initiation sequence of axolotl
The myofiber cultures were stained by indirect immunofluorescence with a rabbit antibody to Msx1 in order to evaluate the level of expression of the homeoprotein in the nuclei. There was a significant difference is staining of nuclei between active and quiescent fibers in the same culture (
The plasticity of isolated urodele myofibers as described here has not been observed in previous work on dissociated mouse myofibers (
One striking consequence of fiber activation is the appearance of the
The activity of the
The forelimbs and hind limbs of the larval spotted salamander (
Myofibers were dissociated as above, retained in suspension in a sterile bacteriological dish (Bibby Sterilin, Stone, United Kingdom), and incubated with 10 μM CellTracker Orange CMTMR (Molecular Probes, Eugene, Oregon, United States) for 30 min at 25 °C. The labelling was terminated by addition of 10% AMEM, and the cells were incubated for 45 min at 25 °C to permit enzymatic activation of the dye. The cell suspension was diluted several fold to allow observation of myofibers at low density. The forelimbs of axolotl larvae (7–10 cm size) were amputated at mid humerus level under tricaine (0.1%) anaesthesia 48 h before injection of labelled myofibers (see
Myofibers were incubated in AMEM containing 2,3-butanedionne monoxime (BDM) (4 mM) for 30 min to prevent contraction of the myofibers (
To record the coordinates of the myofibers in culture, the dish was scored underneath with a scalpel, and cells in each grid square were observed daily and images were acquired with a CCD camera (Sony, Tokyo, Japan). For time-lapse microscopy, myofiber cultures were placed under an Axiovert microscope fitted with an incubation chamber maintained at 26 °C and 3% CO2, and phase contrast or variable relief contrast (VAREL) (Zeiss) images were acquired using a digital camera controlled through Image-Pro Plus software (Media Cybernetics, Silver Spring, Maryland, United States). A sequence gallery was created using Image Pro-Plus and images of interest were selected, digitally enhanced, and processed in Adobe Photoshop 6.0 (Adobe, San Jose, California, United States).
Myofibers were incubated in 1 μCi/ml [3H] thymidine (Amersham Biosciences, Little Chalfont, United Kingdom) for 24 h, fixed in 1% glutaraldehyde, and processed for autoradiography (
DNA content in myofiber nuclei and TR–dextran-labelled mononucleate progeny was measured quantitatively after fixation and staining of the nuclei with Hoechst 33258 (2 μg/ml; Sigma). Baseline values for nuclear DNA content in cycling axolotl mononucleate cells were measured in parallel after incorporation of 5-deoxy-2′-bromouridine (BrdU) (1 μM; see
A viable nucleic acid stain such as Syto 13 or Hoechst 33342 (Molecular Probes) was routinely used in cultures to visualize and score the myofibers. The quiescent or inactive myofiber nuclei were aligned along the fiber (see
Dissociated myofibers were plated in medium containing taxol (2 μM; Sigma). Parallel control cultures were incubated in vehicle (DMSO) in a similar way. The cultures were fixed at 48 h after treatment and processed for tubulin immunofluorescence or in situ hybridisation.
Morpholino-based antisense oligonucleotides of 25 oligomere were prepared to target the translation initiation site of axolotl
Tyramide signal amplification (PerkinElmer Life Sciences Inc., Wellesley, Massachusetts, United States) coupled with enzyme-linked immunofluorescence (ELF97, Molecular Probes) was employed to localize the uptake of morpholino oligos in cultured myofibers. Myofiber cultures were fixed at 48 h in 0.5% PFA containing 0.05% glutaraldehyde for 15 min on ice. The fixative was replaced with freshly made 0.1% NaBH4 solution and incubated for 5 min. The manufacturer's protocol was employed for TSA amplification with the ELF97 modification. The samples were developed in ELF reaction buffer under fluorescence microscope for 10–20 s and images were acquired using a cooled digital camera.
The axolotl
Myofiber cultures were routinely fixed in ice-cold 0.5% PFA containing 0.05% glutaraldehyde for 10 min on ice. For β-tubulin staining, 5 μM Taxol (Sigma) was incorporated into the fixative. After fixation, the culture was treated with freshly prepared 0.1% NaBH4 for 5 min to reduce nonspecific fluorescence. The samples were post-fixed in ice-cold methanol at −20 °C for 10 min, washed three to four times in 0.3% PBT, and blocked in PBT containing 10% goat serum. The primary antibodies were to MHC and titin, and BrdU monoclonal antibody and rabbit polyclonal antibodies to phosphohistone H3, were all as described elsewhere (
The myofiber cultures stained for β-tubulin, MHC, or titin, or cultures injected with TR–dextran were observed under confocal laser scanning microscope (Leica). The images were acquired as
The myonuclei incorporate Syto13 live nuclear stain. The myofiber was observed with VAREL optics at 24 h after plating. Scale bar, 100 μm.
(4.1 MB TIF).
Nuclear localisation of Msx1 protein (green) was detected with a rabbit polyclonal antibody generated against the full-length mouse Msx1 homeoprotein.
(5.6 MB TIF).
Time-lapse sequence was begun 6 h after plating of the myofiber on to a culture dish. The images were taken at 6 min intervals under 32× VAREL objective magnification.
(110 KB AVI).
We thank Phillip Gates for plasmid construction, Tim Landy and Karel Liem for comments and suggestions on the manuscript, and Daniel Ciantar (Confocal Unit, Department of Anatomy, University College London) for his help. This work was funded by a Programme Grant from the Medical Research Council to JPB.
amphibian minimum essential medium
analysis of variance
2,3-butanedionne monoxime
5-deoxy-2′-bromouridine
myosin heavy chain
paraformaldehyde
Texas red-conjugated dextran
variable relief contrast