The well-documented nerve dependence of limb regeneration in the newt was analysed by study of accumulation of newly synthesized macromolecules following denervation. The specific activity of RNA and DNA in the denervated early regenerate bud was determined following intraperitoneal injection of [3H]-uridine and [3H]-thymidine. Results showed an outburst in the incorporation into RNA and DNA which reached a peak 3 h after denervation for the former and 7 h for the latter. There was then a decline in incorporation to a plateau about 50-60% of the control non-denervated side within 48 h. Combining these results with our previous demonstration of a similar outburst in the accumulation of newly synthesized protein with a peak at 4 h, the sequence of the outbursts was in order RNA, protein and DNA. The results are interpreted to mean that the nerve influences either macromolecular synthesis or macromolecular processing and turnover, and therefore accumulation in the regenerate.

Regeneration of the salamander limb requires the presence of an adequate nerve supply at the amputation wound (review, Singer, 1952). If the stump is denervated at the time of amputation or during early limb growth, regeneration is interrupted and only resumes after nerve fibres have regrown to the amputation site. The nature and the precise effect of the neuronal contribution to regeneration is not known. The agent of the nerve is commonly considered to be chemical, and its effect is to cause accumulation of mesenchymatous cells and their subsequent multiplication to form the blastema of regeneration (review, Thornton, 1970).

There are recent attempts to define, by biochemical means, neurotrophic activity during regeneration. Dresden (1969) reported that the synthesis of protein, DNA and RNA during a late stage of regeneration declined after nerve transection. The greatest rate of change occurred within the first two days and was followed by a levelling off of macromolecular synthesis which reached a plateau at about 60 % of normal. Studies of protein synthesis in the early regenerate, a stage more sensitive to nerve transection than the later stage (Singer & Craven, 1948), showed a more rapid decrease reaching a value of about 50 –60 % of the control within about a day after denervation (Lebowitz & Singer, 1970). However, in these studies the decline was preceded by an initial outburst in protein synthesis which reached a peak at about 3 –5 h after denervation. The studies also showed that crude nerve homogenates when infused into the 48 h denervated regenerate stimulated the recovery of about 50 % or more of the lost protein synthesis. The evidence was interpreted to mean that the neurotrophic agent is indeed chemical in nature and that protein synthesis in the denervated regenerate could serve as an assay of the neuronal effect. The meaning of the initial outburst in protein synthesis after denervation is not known. The present paper reports our continuing analysis of the phenomenon and deals with the influence of nerve transection on overall incorporation of labeled substrates into RNA and DNA in the blastema stage of regeneration with an emphasis on the early hours after denervation.

The forelimb of adult Triturus viridescens, collected in Massachusetts, was amputated bilaterally in the lower third of the upper arm. The left stump was denervated 10 –13 days after amputation; the right served as the non-denervated control. A sham operation was performed on the right side in some instances but not in others; since there was no difference in results, no further mention will be made of the sham comparison. At 10 –13 days postamputation, the regenerate is in the early stage (see stages of Singer, 1952) and consists of a small mound of blastema covered by a thickened epithelium. Denervation at this stage stops further regeneration and the blastema withers and is resorbed. Since significant variation exists among animals in the speed of regrowth, animals in the same stage of limb regeneration were selected for denervation. Little variation is exhibited between the two forelimbs of the same animal. Animals were kept at 25 °C throughout the experimental period. To avoid the possible interference of an anesthetic with neurotrophic activity, the animals were inactivated by wrapping them in moist cotton with only the operation site exposed. The contact and pressure stimuli of the wrapping apparently served to minimize response to painful stimuli. Denervation was performed in the brachial plexus; it involved transection of spinal nerve 3 and the combined nerve trunk of 4 and 5. The sympathetic postganglionic fibers, which in the newt follow the arteries into the limb, were not interrupted; previous studies showed that these fibers by themselves are quantitatively inadequate to sustain regeneration (see review, Singer, 1952).

Except in one experiment, the animals were injected intraperitoneally 3 h before harvesting. In the one exception to the 3 h ‘pulse’ time, harvesting of the regenerates for RNA determination occurred 2 h after denervation and isotope injection. For RNA determinations 90/ μ Ci of [5-3H ]-uridine (25 ·4 Ci/mmole) in 0 ·1 ml aqueous solution was injected into each animal. In one of the 5 h denervation series [2-14C]-uridine (90/μCi/animal; specific activity 59 ·8 mCi/ mmole) was employed; since no difference in results was observed, no further mention will be made of it. For DNA studies [5-3H ]-thymidine (90/ μCi/animal; specific activity 20 Ci/mmole) was used.

At the selected postdenervation time, the regenerates were removed without anesthesia using an iris scissors and avoiding adult stump tissues as much as possible. The left (denervated) and right (innervated) regenerates were separately pooled and homogenized in 0 ·2 ml 5 % cold trichloracetic acid (TCA) with a small ground glass homogenizer at 0 ·2 °C. The homogenate was transferred to another tube, the homogenizer rinsed three times with 0 ·1 ml 5 % TCA, and the washing combined to yield 0 ·5 ml final volume of homogenate. After removing aliquots for determination of total radioactivity, RNA, DNA and protein were separated from each other by the method of Schmidt & Thannhauser (1945) and measured by the colorimetric methods of Mejbaum (1939), Burton (1956) and Lowrey, Rosenbrough, Farr & Randall (1951), respectively. Radioactivity was measured on aliquots of these samples by use of a threechannel liquid scintillation spectrometer equipped with external standardization. The counting fluid was dioxane:anisole:dimethoxyethane (750:125:125) and contained 7 g PPO (2,5-diphenyl-oxazole) and 0 · 5 g POPOP (1,4-bis[2-(4-methyl-5-phenyloxazolyl) ]-benzene) per litre. The efficiency of counting was 30 % for 3H and 87 % for 14C.

In selected cases, the recovery and distribution of labeled substrate from the acid soluble fraction of denervated and innervated blastemas of the several time periods was checked by first separating the bases, nucleosides and nucleoside phosphates on thin-layer chromatography sheets (Randerath & Randerath, 1967) and then measuring the distribution of radioactivity with a liquid scintillation spectrometer.

Since the amount of each nucleic acid in a single early regenerate is too small for reliable determination (approximately 5 μ DNA and 13 μ RNA), the regenerates from a number of animals were pooled for each postdenervation time. The regenerates of five animals were found sufficient for DNA determinations, and seven to nine for RNA (see Table 1 and Table 1 (cont.)). The use of these quantities of blastemas permitted colorimetric measurements by spectral analysis (Schneider, 1957) on at least three different-sized aliquots of each sample and always gave readings of more than 0 · 15 absorbency unit. Hence, a high degree of confidence can be placed on each individual measurement, the variation being less than ±5%. The average wet weight of an early regenerate, based on 10 readings, was 1 · 15 mg, S.D. ±0 · 32.

Table 1.

RNA synthesis in denervated and control regenerates

DNA synthesis in denervated and control regenerates

RNA synthesis in denervated and control regenerates
RNA synthesis in denervated and control regenerates

After correcting for quench, the results were normalized to equivalent volumes and then expressed as specific activity in counts per minutes (cpm) per micro-gram (μ g) of the labeled macromolecule (see Table 1 and Table 1 (cont.)). The specific activity for the left denervated regenerate was then expressed as a fraction of that for the right control side and plotted as a percentage value in Fig. 1 relative to the non-denervated control which was set at 100 %.

Fig. 1.

Specific activity of RNA, DNA and protein of the denervated forelimb regenerate expressed as percentage of the opposite non-denervated control limb. Note the sequence of the outburst in macromolecular syntheses in the early hours after denervation and then the decline to a plateau at about 2 days.

Fig. 1.

Specific activity of RNA, DNA and protein of the denervated forelimb regenerate expressed as percentage of the opposite non-denervated control limb. Note the sequence of the outburst in macromolecular syntheses in the early hours after denervation and then the decline to a plateau at about 2 days.

The effect of denervation on the accumulation of newly synthesized macro‐ molecules in the early limb regenerate is shown in Fig. 1 and the essential data of the experiments are recorded in Table 1 and Table 1 (cont.).

Incorporation of [14C]-leucine into protein of the denervated regenerate

In a previous study (Lebowitz & Singer, 1970) we presented the data for postdenervation protein synthesis using [14C]-leucine as a marker. The curve is reproduced in Fig. 1 for comparison with those for RNA and DNA. It shows a postdenervation rise in the accumulation of newly synthesized protein culminating in a peak at about 4 h then a continuous decline to a plateau at about 50 h reflecting about a 45 % loss in protein synthesis. The plateau persisted with a small decrement to 100 h when the experiment was terminated. We then showed that it is possible to recover partially the loss in ability of the denervated plastema at the 48 h postdenervation time to accumulate labeled protein by infusing crude homogenates of intact nerves directly into the nerveless regenerate.

In the present study we affirmed selected points on the protein synthesis curve, pooling two to four regenerates. We obtained averages of 0-59 ± 0-05 for seven readings at 51 h; 0-68 ± 0 06 for eight readings at 26 h; and 0-85 for two runs at 12 h.

Incorporation of [3H]-uridine into RNA of the denervated regenerate

As already noted, the amount of RNA in an individual blastema of the early regenerate stage was too low for reliable determinations. Dresden (1969) employed the ‘palette’, an advanced stage of development, which weighs about 5 times that of the early regenerate bud; yet, the RNA content of this later stage was also too small and had to be estimated by normalization with the protein content. In our studies it was necessary to pool 7 – 9 regenerating blastemas of each time period to obtain a reliable measurement of RNA. When this was done, the determinations as recorded in Table 1 were relatively close for each postdenervation time. The averages are plotted in Fig. 1. The graph depicts an outburst in the accumulation of newly synthesized RNA within the early hours after denervation. The increase in specific activity of the RNA reached a peak at 3 h, then fell rapidly to a value of the control at 5 h. The rate of change fell off during subsequent hours to a plateau of less than about 60 % of the control value. Initial studies of 72 and 96 h after denervation suggest a further decline to a value of about 30 % of the control. The peak value for specific activity of RNA was almost 40 % above that for the control side. The value at 2 h was based upon a labeling period of 2 rather than 3 h. If the value was normalized to the 3 h labeling time, it would not alter the curve significantly.

Incorporation of [3H]-thymidine into DNA of the denervated regenerate

The specific activity of DNA likewise showed a similar increase after denervation (Fig. 1 and Table 1 (cont.)). However, the peak value which was at least 1 · 5 times that of the control was not reached until about 7 h after denervation, although the onset in the outburst was much earlier. Moreover, the decline was less precipitous than that for RNA and the increase extended over a longer period of time. A plateau was also reached similar to that for protein and RNA. Initial observations at 72 and 96 h suggest that this level of accumulation persisted at least through 4 days after denervation.

Availability of labelled substrates

During the course of these experiments we have measured the distribution of labeled nucleoside and nucleoside triphosphate in the denervated and non‐ denervated blastemas at several time periods following transection of the nerves. From the results summarized in Table 2, it is apparent that both types of blastema contained similar levels of labeled substrates. Also, it appears that the conversion of the nucleoside to the nucleoside triphosphate reached a steady state during the first hour after the labeled nucleoside was injected into the animal.

Table 2.

Recovery of labelled nucleosides and nucleoside triphosphates from denervated and non-denervated blastemas at different time periods after nerve transection*

Recovery of labelled nucleosides and nucleoside triphosphates from denervated and non-denervated blastemas at different time periods after nerve transection*
Recovery of labelled nucleosides and nucleoside triphosphates from denervated and non-denervated blastemas at different time periods after nerve transection*

These results indicate that the differential labeling pattern of macromolecules in the denervated and non-denervated regenerates was probably due to factors other than differential availability of radioactive substrates.

The mitotic rate in denervated regenerates

The relation of the outburst of DNA synthesis to the results of previous studies on mitotic activity in the denervated regenerate (Singer & Craven, 1948) should be remarked upon. In those studies mitotic counts showed an outburst of mitosis within 24 h after denervation followed by a precipitous drop to a low level. Counts were not made before the 24 h period ; it may be that the peak occurred sooner. The outburst was particularly evident in the early regenerate and less so in later stages. The present biochemical results are in accord with these cytological observations. DNA synthesis reached a peak at about 7 h and the mitotic outburst occurred sometimes afterward, not later than 24 h postdenervation. The difference in timing conforms to our present understanding of the relation between DNA synthesis and mitosis, the peak outburst falling within the S phase of the cycle.

Normalization of RNA and DNA synthesis with the protein content

For comparison with the results of Dresden (1969) in which the RNA and DNA counts were normalized with the protein content of the sample, protein determinations were also made on the homogenate pools used in our experiments. Normalization of the nucleic acid counts with the protein content yielded values which are recorded in the last column of Table 1 and Table 1 (cont.). Although somewhat erratic, the results are similar to those for the specific activity of the nucleic acids. Both methods of expressing the results showed an outburst in accumulation of newly synthesized nucleic acids of the same character and timing followed by the same sort of decline to a similar plateau.

A previous work from this laboratory (Lebowitz & Singer, 1970) showed that denervation of the early regenerate bud results in an initial outburst in accumulation of newly synthesized protein. This outburst was followed by a decline to a plateau about 50 % of the control at 48 h. The present work demonstrates a similar response to denervation in the specific activity of RNA and DNA. The results thus support the view that a level of control of RNA and DNA metabolism (and/or accumulation) in some cellular components of the early regenerate is also nerve dependent.

The decline in accumulation of newly synthesized macromolecules follows logically from well-established information on the influence of the nerve on limb regeneration, namely that growth of the young regenerate ceases after denervation. What is less understandable is the initial outburst in synthesis and/or accumulation of newly synthesized macromolecules in the early postdenervation hours. Dresden (1969) reported a decline but not an outburst in protein, RNA and DNA synthesis. However, his results cannot be strictly compared to ours because his earliest postdenervation reading was 7 h, whereas the outbursts seen in our experiment peaked at 3 and 7 h for RNA and DNA respectively ; and by 7 h protein synthesis had returned to a normal level and RNA synthesis was already greatly depressed. Moreover, he used a later regenerate, whereas we employed the early bud which is affected more profoundly by denervation (Singer, 1952); also, his analytical procedures were quite different from ours. An outburst in RNA and protein synthesis in the lateral geniculate nucleus of the monkey following transection of the optic nerve was reported by Kupfer & Downer (1967). It persisted for 2 days and was then followed by a prolonged decline to a subnormal level, in the case of RNA to about 30 %.

In a previous paper (Lebowitz & Singer, 1970) we likened the outburst in accumulation of protein to the supersensitivity of denervated effectors to stimulating agents, a phenomenon embodied in W. B. Cannon’s Law of Denervation (1939) which states: ‘When in a series of efferent neurons a unit is destroyed, an increased irritability to chemical agents develops in the isolated structure or structures, the effects being maximal in the part directly denervated’ (see also Cannon & Rosenblueth, 1949). However, the law defines physiological phenomena and not biochemical changes although Cannon & Rosenblueth speculated upon the chemical basis for the increased sensitivity. Furthermore, the outburst in accumulation of macromolecules develops within hours of denervation whereas the physiological changes are much slower in onset and are pronounced only days later. The difference in time course may mean that it is not the initial outburst but rather the decline in accumulation of macromolecules that is the chemical basis for the subsequent altered sensitivity if, indeed, these biochemical changes are directly related to altered sensitivity. However, it may be that biochemical changes other than those measured here cause altered sensitivity and that they may occur within the first few hours of denervation.

The mechanism whereby nerve interruption alters the synthesis and accumulation of macromolecules is not elucidated in the present experiments. Perhaps cutting the nerve releases a nervous restraint on macromolecular synthesis which in a short time is reasserted by controls within the synthetic mechanisms themselves. Assuming that the neurotrophic agent is chemical in nature, an assumption for which experimental evidence is already presented (Lebowitz & Singer, 1970), it may be that the outburst occurs after exhaustion of the neurotrophic agent from the cut nerve and reflects an ‘overshooting’ of the synthetic mechanisms before a new equilibrium is established. Alternatively, one may imagine that the initial amplification reflects increased release of the chemical agent due to the transection causing a corresponding augmentation in macromolecular synthesis, and that the later decline to a plateau reflects exhaustion of the contribution. If the observed outbursts are due to exhaustion or to hurried emptying of the trophic substance, it is possible to calculate the approximate speed of movement of the neurotrophic agent in the axon based on the observed time of the outburst. The length of the distal segment of the transected nerve from the brachial plexus to the regenerate was about 15 mm and the outburst in RNA synthesis began to rise at about 2 h. Therefore, the velocity in the transected part is in the order of 180 mm per day. Such a velocity places the trophic agent among the faster components of axoplasmic flow (compare Lasek, 1970).

We have avoided the use of the term ‘synthesis’ in reporting the effect of denervation on the macromolecular content of the regenerate because our experiments do not distinguish between synthesis and accumulation. It may be that the primary effect is on control systems which regulate turnover of macromolecules and only indirectly macromolecular synthesis. The neurotrophic influence may be likened in this way to the reported effect of phytohaemagglutinin on the lymphocyte. In the ‘resting’ lymphocyte rRNA wastage appears to be very high; but upon addition of the growth stimulant, the waste is diminished dramatically (Cooper, 1968, 1969). Our data also do not reveal the primary target of the nerve effect. It may be RNA or its polymerase since the RNA response to denervation precedes that of protein and of DNA.

The authors are grateful for the assistance of Mrs Kai-Yu Clara Lin and Charles S. Maier. This work was supported by grants from the National Multiple Sclerosis Society, the American Cancer Society and the National Institutes of Health.

Burton
,
K.
(
1956
).
A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid
.
Biochem. J
.
62
,
315
323
.
Cannon
,
W. B.
(
1939
).
A law of denervation
.
Am. J. med. Sci
.
198
,
737
750
.
Cannon
,
W. B.
&
Rosenblueth
,
A.
1949
.
The Supersensitivity of Denervated Structures. A Law of Denervation
.
New York
:
The Macmillan Co
.
Cooper
,
H. L.
(
1968
).
Ribonucleic acid metabolism in lymphocytes stimulated by phyto-hemagglutinin
.
J. biol. Chem
.
243
,
34
43
.
Cooper
,
H. L.
(
1969
).
Ribosomal ribonucleic acid wastage in resting and growing lymphocytes
.
J. biol. Chem
.
244
,
5590
5596
.
Dresden
,
M. H.
(
1969
).
Denervation effects on newt limb regeneration: DNA, RNA and protein synthesis
.
Devi Biol
.
19
,
311
320
.
Kupfer
,
C.
&
Downer
,
J. L. Dec
. (
1967
).
Ribonucleic acid content and metabolic activity of lateral geniculate nucleus in monkey following afferent denervation
.
J. Neurochem
.
14
,
257
263
.
Lasek
,
R.
(
1970
).
Protein transport in neurons
.
Int. Rev. Neurobiol
.
13
,
289
324
.
Lebowitz
,
P.
&
Singer
,
M.
(
1970
).
Neurotrophic control of protein synthesis in the regenerating limb of the newt, Triturus
.
Nature, Lond
.
225
,
824
827
.
Lowrey
,
O. H.
,
Rosenbrough
,
N. J.
,
Farr
,
A. L.
&
Randall
,
R. J.
(
1951
).
Protein measurement with the Folin phenol reagent
.
J. biol. Chem
.
193
,
265
275
.
Mejbaum
,
W.
(
1939
).
Uber die Bestimmung kleiner Pentosemengen insbensondere in Derivaten der Adenylsaure
.
Hoppe-Seyler’s Z. physiol. Chem
.
258
,
117
120
.
Randerath
,
K.
&
Randerath
,
E.
(
1967
).
Thin-layer separation methods for nucleic acid derivatives
.
In Methods in Enzymology
, vol.
xn
, part A, pp.
323
347
(ed.
L.
Grossman
&
K.
Moldave
).
New York
:
Academic Press
.
Schmidt
,
G.
&
Thannhauser
,
S. J.
(
1945
).
A method for the determination of desoxyribonucleic acid, ribonucleic acid, and phospho-proteins in animal tissues
.
J. biol. Chem
.
161
,
83
89
.
Schneider
,
W. C.
(
1957
).
Determination of nucleic acids in tissues by pentose analysis
.
In Methods in Enzymology
, vol.
III
, pp.
680
684
(ed.
S. P.
Colowick
&
N. O.
Kaplan
).
New York
:
Academic Press
.
Singer
,
M.
, (
1952
).
The influence of the nerve in regeneration of the amphibian extremity
.
Q. Rev. Biol
.
27
,
169
200
.
Singer
,
M.
&
Craven
,
L.
(
1948
).
The growth and morphogenesis of the regenerating forelimb of adult Trituras following denervation at various stages of development
.
J. exp. Zool
.
108
,
279
308
.
Thornton
,
C. S.
(
1970
).
Amphibian limb regeneration and its relation to nerves
.
Am. Zool
.
10
,
113
118
.