REPLICATION AND THE ACTIVATION OF MUSCLE DIFFERENTIATION

Biochemical Research Laboratory, The Dow Co., Midland, Michigan*

Abstract

1. Phleomycin,the specific and selective inhibitor of DNA synthesis, drastically curtailed muscle regeneration in mice, as judged biochemically by the suppression of contractile muscle protein synthesis.
2. Inhibition was greater than 90 percent when phleomycin was applied during the first of three wave of DNA synthesis that occur between wounding and the first signs of new muscle.
3. Low grade inhibition (some 20 percent) occurred when phleomycin was applied so as to arrest later waves of replication, reflecting diminution rather than obliteration of the pool of myogenically competent cells, as seen when the critical first wave was inhibited.
4. Variations in phleomycin sensivity were observed among specific cell types, including several transplantable tumors; these variations were correlated with different ontogenic fates of the particular cells.

Regeneration in mouse skeletal muscle (3) was selected as the principal test system because reasonably close approximations could be made of the time of appearance of the first wave of new muscle fibers or myotubes (96-120 h [4]), and the interval during which replication takes place (38-72 h [4]). Also, because regeneration is not spontaneous in healthy skeletal muscle, but ensues as an almost inevitable consequence of non-fatal injury, it was possible to gauge the timing of events from a fairly precise starting point -- much more so than with other specific myogenic systems (for example, the embryo, regenerating amphibian limbs, rhabdomyosarcoma, etc.).

DNA synthesis during the second-third day of regeneration.

There are three waves of thymidine incorporation between 38.5 and 72 hours post-wounding (Fig. 1 I, II and III). Peaks and profiles for I and III were confirmed in two independently executed studies. Wave II also was obtained in two separate series, but its peak and period varied by some 2 h. Wounds for all groups had been inflicted at approximately the same time of day, and, of necessity, the time of day at which tritiated thymidine was introduced varied for each point. This opened the possibility that circadian rhythm would influence wave amplitudes. Therefore, investigations were conducted with wounds (38.5, 42.0 or 62.0 h) receiving tritiated thymidine at different hours of the day. Routinely, all mice are maintained in a windowless, air-conditioned room and are exposed to consecutive, alternating 12-h periods of light and darkness, and this, of course, was maintained in the experiments under immediate consideration. It was learned (Table 1) that wave I (38.5 and 42-h wounds is not influenced by the light rhythm employed in this laboratory. On the other hand, tritiated thymidine uptake at wave III was influenced by the time of day at which pulse labeling was carried out, for 62-h wounds that had been pulsed at the high point in light showed significantly more activity than did wounds of the same age pulsed at the maximal point in darkness.

These results indicated that, in order to approximate the true characteristics of wave III, it would be necessary to correct for the time of day of injection.

Corrections for circadian rhythm

The applicability of the correction factor was determined by computing 40-42h/62-64h values from corrected curves and comparing this ratio with that obtained from cases with contralateral tibialis anteriors presenting I and III wave wounds at pulse labeling. The latter ratio was 0.382 and the corrected one was 0.372. [The latter differences was not statistically significant.]

The corrected wave III curve proved extremely useful in determining the length and period of wave III and also in establishing a base-line for activity of the waves per se; that it was possible to extrapolated through regions of overlap and to obtain a truer estimate of background values. On the basis of information to be described in the next section (cf. Tables 2 and 3) of this article it was learned that wave III terminates a little short of 68 h, and extrapolating to this point from the descending shoulder of the corrected curve for this wave, it was possible to traverse regions that could not be handled empirically. The base-line generated horizontally to the left from 68 h intersected the ascending segment of wave I at about 38.5 h, providing an approximation for the onset of its period. After having dropped the point at 48 h to the base-line (once more on the strength of evidence presented in Tables 2 and 3), it was possible to idealize the curve for wave I. This was done by the least-squares methods with a Burroughs 5000 digital computer for waves I and III which produced the equations represented in the legend of Fig. 1. Areas under corrected curves turned out to be 27,000 units for wave I and 34,000 for III. [The differences are insignificant.] Thus, wave I appears to begin at approximately 38 h and to end just short of 48 h. Wave III spans the interval between 60 and 68 h. The length of wave III, based on corrected expressions, turned out to be almost identical to the time established rather masterfully by others for the s-phase (DNA synthesis interval) in replicating mouse cells (9). The approximately 10-h length of wave I, 3 h in excess of published s-phase values, no doubt represents asynchrony in the population plus inaccuracies of extrapolation. Interestingly, the similar areas under the curves for waves I and III indicated that the total amounts of DNA being synthesized are of roughly the same magnitude during these two periods, despite amplitude differences. Owing to uncertainties about its origin, wave II could not be corrected and thus adequately evaluated.

Effects of phleomycin

Actomyosin synthesis was judged by the incorporation into this protein complex of L-lysine-14C (uniformly labeled 223 mc/mmole) given intraperitoneally in two doses of 10.0 ul each at 94 and 117 h after wounding. At death, seven days after wounding, the entire tibialis anterior muscle was dissected out and actomyosin extracted in buffered 0.6 M potassium chloride (10, 11). Concentration of actomyosin was determined from its ultraviolet absorption maximum (275 mu), after the Folin reaction. Size of starting material was judged by determining non-fibrillar protein content of the sample. Radioactivity was measured by liquid scintillation counting. Uninjured muscle was processed similarly in order to develop a base-line for radioactivity not attributable to newly synthesized actomyosin. Phleomycin (kindly supplied by Bristol Laboratories, Syracuse, New York) was dissolved in physiological saline just prior to use, and its biological activity was confirmed on agar plates seeded with A. aerogenes.

Phleomycin (0.25 mg) applied before wounding (minus-13 h) or at the very end of waves I and III (48 and 68 h, respectively) failed to suppress actomyosin synthesis (Table 2). When the antibiotic was applied around the peak of wave III activity was moderately suppressed. However, phleomycin virtually eliminated actomyosin synthesis when it was administered to animals between 44 and 45 h; that is, on the descending shoulder of wave I.

The effects of phleomycin on DNA synthesis (Table 3) were judged after a single intraperitoneal dose of the antibiotic (0.5 mg) followed 1 h later by 29 uc tritiated thymidine. Animals were killed 40 min later, and tissues were processed in the manner described in the previous section.

As might be expected from the foregoing account, thymidine incorporation was severely depressed when phleomycin was given during wave I. Incorporation was also markedly suppressed when phleomycin was given during wave II, suggesting that, although these cells are influenced by the antibiotic so far as DNA synthesis is concerned, their future capacity to manufacture muscle proteins is unimpaired. When the antibiotic was given at minus-17 h there was no evidence of curtailed thymidine incorporation at wave I. Similarly, phleomycin at 59 h failed to suppress DNA synthesis at wave III. Phleomycin did not suppress activity when given during wave II; but it is emphasized that there is uncertainty concerning the exact period of this wave; consequently the significance of wave II remains obscure, and for reasons of economy and simplicity it has not bee considered further.

Different responses to phleomycin

1. the antibiotic might have restricted access to DNA except in rather circumscribed conditions;
2. the nucleotide sequences where the antibiotic makes its union may have a different operation significance in replication various cell types;

3. the nucleotide sequences where phleomycin attaches to DNA may be dissimilar among unlike cells; that is, different 'genes' may involved in binding in varying circumstances

In a limited number of additional experiments, phleomycin was introduced into animals bearing one of the following transplantable tumours: rhabdomyosarcoma (BW 101139), myeloid leukaemia (C1498), adenocarcinoma (BW102320, melanoma (B16) (ref. 5). Tritiated thymidine was injected into treated and untreated tumour hosts 1 h after introduction of the antibiotic. Animals were killed 40 min later and for each the ratio of specific activity between tumour and splenic DNA was computed and compared. Only the phleomycin-treated rhabdomyosarcoma (muscle tumour) showed inhibition; that is, a significant increase in splenic thymidine incorporation was compared with untreated tumour/spleen rations (Table 4).

In other cases, tumour/spleen ratios, treated versus untreated, showed insignificant differences, suggesting that, like spleens and testes but unlike regenerating muscle and muscle tumor, there was no immediate, frank reaction to phleomycin by leukaemia, carcinoma or melanoma cells. Despite the prediction that this would happen, it must be recognized that testes and spleens are subject to individual physiological variations among animals and the same might very well have been true of tumours; thus, because of this the frame of reference might have been variant. Therefore, more exacting conditions were sought to provide a final test of the proposition in question (that cells of unlike type vary with respect to phleomycin sensitivity).

While cells of both waves I and III are highly susceptible to phleomycin, their different causal relationships to actomyosin synthesis suggested, intuitively: a) they developmentally different cells and, b) would vary quantitatively in response to phleomycin.

Owing to large standard deviations in data from different individual mice (Fig. 1) it was impossible to design a statistically efficient experiment based on anticipated small comparative differences. It was decided, therefore, to use both tibialis anterior muscles in the same animal, to vary the times of wounding and hold all else constant. Wounds were compared on an individual basis for each animal, the information sought being the comparative value between mean ratios of treated versus those for untreated individuals. The premise was that if phleomycin influenced wounds on both sides of the animal in the same way, then, while absolute values might differ animal to animal, the ratios between the two sides would fall within the same statistical range. Conversely, if cells of the two sides responded differently to phleomycin the ratios between contralateral wounds would differ significantly, treated versus untreated.

Two series were conducted, 41/63 and 41/43,***** numerators and denominators representing the ages of the wounds when phleomycin was introduced. Untreated 41/43 ratios showed a mean (four cases) of 0.827 +/- 0.27 S.D.; means (five cases) for 41/43 phleomycin-treated animals were 0.873 +/- 0.27 S.D., indicating no difference in response to the antibiotic at the two points along wave I. In 41/63 h experiments untreated ratios (seven cases) were 0.382 +/- 0.20 S.D., but phleomycin-treated specimens revealed ratios (eight cases) of 0.638 +/- 0.18 S.D.; that is, differences between the two groups of very high significance (F distribution beyond 1 percent).

Thus, phleomycin can virtually eliminate myogenesis but only when it is made available during a critical time-interval. In mouse skeletal muscle regeneration in particular this interval coincides with the period of wave I. As is true with other agents which inhibit regeneration, there is a close correlation between the events going on at the time of application and the mode of inhibitory action (see review of these in Fig. 3). A phase comparable to wave I seems required to set myogenesis into motion, and this appears to entail at least one round of DNA synthesis. Further division may occur but it is not mandatory. It has been reported of tissue-cultured myoblasts that they need not divide in order to accomplish differentiation (12, 13). Judging from the advanced age of embryonic donors in the last-cited experiments, the cultivated cells, doubtless, had passed the point equivalent to wave I prior to explanation.

Obviously the channels to muscle information are not opened merely by genomic duplication (otherwise all dividing cells would produce muscle) and obviously DNA synthesis cannot be regarded as the proximate cause of myogenesis. Indeed, the nature of the wave I contribution remains to be established. During regeneration also, there is a significant time lapse between wave I and m-RNA synthesis (at about 72 h [ref. 14]) and it is fairly well established that when DNA synthesis is going on actomyosin synthesis is not (15). While it is much simpler to regard wave I as the immediate source of myogenic cells the possibility should not be ignored that wave I supplies information that is utilized by other cells.

Differential susceptibility to phleomycin of wave I cells versus III cells (as well as cells of other types) suggests that specificity in the operational features of thymidine incorporation during wave I, at the molecular or chromosomal level or both -- the specific information that may become readable as a result of the uniqueness of factors bestowing differences -- may determine the kind of future the daughter cells shall have, whether they shall divide again or enter into relative dormancy or begin to synthesize, anew, polymeric classes previously not among the cell's complement. It seems quite possible, carrying this line of speculation a little further, that the release of specific information could in principle be a function of detailed firing orders among chromosomes, polynucleotide sequences, or both, with read-out being a consequence of the order in which elements were reproduced, and not the gross fact of duplication. If such a method of release is true of myogenesis, then it seems likely that there are other examples of differentiation in which details of replication play an important part in establishing the cell line's ontogenic fate. This possibility suggests that it would be quite worth-while to gain comprehensive knowledge concerning the physical chemistry of phleomycin-DNA complexes in different cell types. Indeed, the decision to include these few speculations at all was reached only because technology exists for determining the kinds of interactions possible between this very valuable antibiotic and deoxypolynucleotides.

References

• 01 Maeda, K. et al, J. Antibiotics (Tokyo), 9, A, 82 (1956).
• 02 Falaschi, A. and Kornberg, A., Fed. Proc., 23, 940 (1964).
• 03 Lash, J. et al., Anat. Rec. 128, 679 (1957).
• 04 Pietsch, P. Anat. Rec., 139, 167 (1961).
• 05Handbook on Genetically Standardized JAX Mice (Bar Harbor Times, Bar Harbor, Maine, 1962).
• 06 Within a few minutes of intraperitoneal injection of tritiated thymidine into mice, their urines become highly radioactive. Within 1 h activity falls to a fraction of this value, approaching background levels. Splenic biopsies 15 min after injection show a level of activity that is maintained in biopsies taken 30 and 90 min later. Thus, it would appear that label is available for a brief interval after administration and that a single injection closely approximates a pulse. These observations were made by Strossberg and Pietsch, 1962 (unpublished), and also incidental to the present investigation.
• 07 Ogur, M. and Rosen, G. Arch. Biochem., 25, 262 (1950).
• 08 Pizzarello, D., et al., Science, 145, 286 (1964).
• 09 Pilgram, Ch., and Mauer, W., Exp. Cell Res., 37, 183 (1965).
• 10 Mommaerts, W. and Parrish, R., J. Biol. Chem., 188, 545 (1951).
• 11 Tassoni, J. P., et al., Exp. Cell Res., 34, 219 (1964).
• 12 Konigsberg, I. R., Science, 140, 1273 (1963).
• 13 Stockdale, F. and Holtzer, H., Science, 146, 866 (1964).
• 14 Pietsch, P., Nature, 203, 117 (1964).
• 15 Stockdale, F. and Holtzer, H., Exp. Cell Res., 24, 508 (1961).
• 16 Snedecor, G., Statistical Methods (Iowa State Univ. Press, Ames, Iowa, 1950).
• 17 Williams, R., and Pietsch, P., Anat. Rec., 151, 434 (1965).
• 18 Umezawa, H., et al., J. Antibiotics (Tokyo), 15, A, 275 (1962).

FIGURES

y = 8.05069149X-0.00143509987X^*3 -225.365393;

y =31.2881719X -0.00254415843X^*3 -1326.22411.

TABLES

Text Footnotes

Web Contact: pietsch@indiana.edu