/. Embryol. exp. Morph. Vol. 26, 3, pp. 425-441, 1971
425
Printed in Great Britain
The reactions to carcinogens in the
axolotl (Ambystoma mexicanum) in relation to
the 'regeneration field control' hypothesis
By ANDREW J. INGRAM 1
From the Department of Zoology, University of Southampton
SUMMARY
About 200 axolotls, between 5 and 21 months old, were treated in the body wall with
carcinogens or control substances, by subcutaneous or intradermal injection, or by subcutaneous implantation.
In response to an injection of dibenzanthracene in olive oil an initial reaction appeared
within 13 days. This consisted of an epidermal proliferation and a subcutaneous infiltration
of macrophages. The epidermis returned to normal after several weeks, but the subcutaneous
response took 6 months to disappear. The initial reaction appeared to be, at least partially,
a wound healing response; its regression could not have been due to regeneration field control
as it occurred in the posterior body region.
Following the disappearance of the initial reaction a secondary reaction of papilloma-like
outgrowths arose between 7 and 20 months after injection. Of the axolotls surviving for a
sufficient length of time, outgrowths arose in 14 out of 54 animals injected with dibenzanthracene and in 18 out of 57 sites injected with methylcholanthrene. These outgrowths had
some features in common with the papillomata of mice, but none have yet progressed to
carcinomata. In a group of axolotls injected in three different regions the frequency of outgrowths varied according to the sites; however, this was thought to be due to differences in
the difficulty of injecting the three regions.
Two tumours, a sarcoma and a hepatoma, arose in the course of these experiments, 3 and
2 years respectively after carcinogen treatment. From this it is suggested tentatively that
tumour induction by polycyclic hydrocarbons in the axolotl may require a long latent period
and involve a low tumour incidence; however, it is possible that at least one of the tumours
was not induced by the treatment.
INTRODUCTION
The relationship between tumour development and regeneration was discussed
by Waddington (1935) and Needham (1936). They suggested that an 'individuation field', the persistence of which they postulated was responsible for
regeneration, might be capable of preventing tumour development or inducing
tumorous tissue to redifferentiate.
Mizell (1960) claimed to have found support for the regeneration field hypothesis by observing the breakdown of Lucke carcinoma, which he had grafted
1
Author's address: Department of Zoology, University of Southampton, Southampton
SO9 5NH, U.K.
426
A. J. INGRAM
into the regenerating tails of tadpoles. However, when Ruben (1955) grafted
Lucke carcinoma into regenerating and non-regenerating limbs of newts he
found no control of tumour growth; also no control of Xenopus laevis lymphosarcoma was found by Ruben & Balls (1964a) when they transplanted it into
regenerating forelimbs of that species. Ruben & Balls suggested that lymphosarcoma tissue may not have been competent to respond to a limb field.
Ermin & Gordon (1955) found that spontaneous melanomas on the fins of
fish were unaffected by fin regeneration; also Scheremetieva (1965) found that
spontaneous melanomas growing on the tails of axolotls regenerated when the
tails were cut off through the tumours. In both these cases a regeneration field
failed to control tumour tissue.
Thus the balance of evidence is against the possibility of a regeneration field
having a controlling influence on tumour growth.
Against this background Seilern-Aspang & Kratochwil (1962, 1965) injected
solutions of various polycyclic hydrocarbons into the skin glands, in the tail
and sacral regions of the newt Triturus cristatus.
This treatment resulted in an epidermal infiltration originating in the skin
glands, which these workers claimed constituted tumour formation. They
observed that ^differentiation of this infiltration occurred more frequently in
the tail than in the sacral region, where further infiltration and metastasis
occurred. This they claimed was evidence of a regeneration field inducing a
tumour to redifferentiate. Neukomm (1944, 1957, 1959) in his newt test for
carcinogenicity also reported an epidermal infiltration in response to carcinogen
injection; however he did not claim that this was true tumour formation as it
disappeared in 2 months. Arffmann & Christensen (1961) repeated the technique
of Neukomm and confirmed his results.
In order to see whether the reaction, described by Seilern-Aspang & Kratochwil
in newts, was common to other urodele amphibia, it was decided to test the
reaction of the axolotl (Ambystoma mexicanum) to carcinogenic polycyclic
hydrocarbons. It was planned to examine the reaction in the short term (up to
6 months) and in the long term (after 6 months) in order to see, should any
reaction occur, whether it could be attributed to an inflammatory response,
wound healing or true tumour formation.
MATERIALS AND METHODS
The axolotls were reared from eggs obtained from eight pairs of mature
axolotls kept permanently paired.
The main carcinogens used were 1,2-5,6-dibenzanthracene (DBA) and 3methylcholanthrene (MC); however, 3,4-benzpyrene (BP) was used on two
occasions in conjunction with DBA. In most of the experiments the carcinogens
were injected as suspensions of crystals in oil; olive oil was used in the experiments involving DBA and trioctanoin was used in the experiments involving
Reactions to carcinogens in axolotl
427
MC. In injection experiments 1-4 (see Table 1), injections were made in the
posterior body region, laterally, above the hind limb. The tail was found to be
too delicate to inject adequately in young axolotls. In the posterior body region
the diffuse collagen layer of the dermis started to develop, between the compact
collagen layer and the epidermis, at the age of around 6 months. Due to this,
subcutaneous injections were made in injection experiment 1, which involved
axolotls under 6 months old; whereas intradermal injections, between the compact collagen layer and the epidermis, were made in injection experiments 2-5.
The amount injected was usually around 01 ml, but, particularly with intradermal injections, back-pressure often prevented the injection of this amount.
In injection experiment 5 each axolotl which was mature or nearly so received
three injections of MC in trioctanoin. Injections were made into the anterior
dorsal region (above the forelimb), the posterior body region and midway down
the tail. It was hoped that this might show whether the site of injection affected
the incidence of tumours or their subsequent development. It was found that
the tough nature of the skin in the anterior body region made intradermal
injection difficult and that the tail skin was rather delicate, also making injection
difficult. In the posterior body region intradermal injection was easier; however
some of the injected material passed subcutaneously. Olive oil alone was injected
as a control for the first injection experiment.
The initial reaction to polycyclic-hydrocarbon injection in oil was followed
in injection experiment 1 by sacrificing animals at intervals during the first 6
months. The axolotls remaining in this experiment and all the animals in subsequent injection experiments were retained for long-term observations.
To provide additional controls for the initial reaction, crystals of perylene
and picryl chloride were implanted subcutaneously into 6-month-old axolotls.
This was to indicate what reactions were produced by inert crystals and by a
chemical irritant respectively. Dibenzanthracene crystals were also implanted in
the same way to compare with the reactions to perylene and picryl chloride.
In 'sandwich' experiments 1 and 2 crystals of DBA surrounded by a sandwich
of epidermis were implanted subcutaneously in the flank. The epidermis was
obtained by trypsinization of a piece of skin removed from the other side of the
same animal, thus avoiding rejection problems. It was hoped that the closer
apposition between the carcinogen and the epidermal tissue of the 'sandwich'
might result in an epidermal tumour at the site of implantation.
The experiments just described are summarized in Tables 1 and 2.
Histology
The injected regions of the axolotls sacrified in injection experiment 1 were
fixed in formol saline, but in the remaining experiments Helly's fluid (Zenker
formol) was used. Any decalcification required was carried out either with citric
acid-sodium citrate (Andersen & Jorgensen, 1960) or with neutral EDTA
(Drury & Wellington, 1967). Sections cut at 10 jum or 8/«n were stained with
Table 1. Summary of the polycyclic-hydrocarbon injection experiments
Injection details
Experiments
Injection
experiment no.
Further
details
Repeat
2
2
3
3
4
No. of
animals
injected
Carcinogens injected
Age at
injection ,
in months Nature
Form
taneous
or intradermal
DBA in
olive oil
DBA in
olive oil
and DBA
Subcutaneous
45
5-3
10
5-3
and 12
Single
(black)
10
Single
(white)
(black)
(white)
(white)
10
9-4
5
5
7
14-4
12
13-5
(black)
10
11-5
40
BPin °
olive oil
Suspension
of crystals
Suspension
of crystals
DBA in
suspension
BPin
solution
DBA+
01 % BP
in olive
oil
DBA in
suspension
BP in
solution
Position
Posterior
body
region
12
Intradermal
MC in
In solutrioctanoin tion only
18
MCin
Suspension
(mean) trioctanoin of crystals/
Anterior
and posterior body
-h tail
142 = total animals injected
Table 2. Summary of controls and other experiments
Experimental details
Title of
experiment
Other
details
Carcinogen
substance
involved
Age in
animals
treated
at time of
treatment
Injection
experiment 2
Control injection
Olive oil
alone
10
5-3
Crystal
implantation
experiment
Subcutaneous
implantation
DBA crystals
12
6
Perylene
crystals
Picryl chloride
crystals
DBA crystals
5
6
5
6
Sandwich
experiment 1
Subcutaneous
implant of
carcinogen
sandwiched in
epidermis
Control - sandwich without
Sandwich
experiment 2
carcinogen
As sandwich
experiment 1
Control
8
of
treatment
Posterior
body
\
12-21
Posterior
flank
—
DBA crystals
—
8
12-21
10
12-21
4
12-21
Reactions to carcinogens in axolotl
429
Ehrlich's haematoxylin and eosin or with Celestin blue-Mayer's haemalum
counterstained with van Gieson's stain. Alcian blue combined with the PAS
(periodic acid Schiff) reaction was used to demonstrate acid and neutral mucopolysaccharides.
RESULTS
The initial reaction
In injection experiment 1 the oil droplet containing suspended particles of
DBA took up a position between the musculature and the compact collagen
layer of the dermis. As the axolotls were only 5-3 months old, the compact
collagen layer was the only barrier between the oil droplet and the epidermis.
At 3 days there was no reaction; but at 6 days, although the histological
appearance was the same, the epidermal mitotic rate above the oil droplet had
increased from 1 to 2-5 per thousand nuclei. By 13 days (Fig. 1) a strong reaction
was apparent in the skin and subcutaneously. The epidermis above the oil
droplet space was 3 | times as thick as the epidermis on the opposite side of the
animal and the mitotic rate had further increased to 5 per thousand nuclei.
Subcutaneously, the oil droplet space had become nearly filled with a massive
infiltration of macrophages. These contained granules of glycogen, demonstrated
by the PAS reaction with a saliva digested control. Separation of the compact
collagen layer from the epidermis, leading to the formation of the diffuse collagen
layer, was not apparent at the injection site; although it was evident on the
opposite side of the animal. At 17 days the same features were shown as at 13
days but they were all more pronounced, the number of mitoses in the epidermis
having reached 9-2 per thousand nuclei and its thickness six times normal.
By 28 days after injection (Fig. 2) the epidermal thickening was less marked
and it appeared that some of the cells, which had occupied the droplet space,
had migrated into the surrounding tissue. These apparently infiltrating cells
were closely associated with fibrocytes which appeared to be laying down
collagen. By 52 days (Fig. 3) the oil droplet space was devoid of cells but, adjacent
to it, cells were infiltrating the musculature. The presence of fibrocytes producing
collagen between the infiltrating cells was more evident than at 28 days (see
Fig. 4). It can be seen from Fig. 3 that the diffuse collagen of the dermis had
not started to form above the injection site, whereas on the other side of the
animal the diffuse layer was quite clear.
Six months after injection (Fig. 5) the area adjacent to the oil droplet showed
a lower density of cells and a greater preponderance of collagen fibrils; also
eosinophilic cells were apparent among the infiltrated cells. An animal sacrificed
2 years after injection (Fig. 6) showed a massive deposition of collagen at the
site of injection, clearly corresponding in position to the infiltrating cells at
52 days and 6 months.
To summarize the above findings: the epidermal reaction to the carcinogen
reached a peak at 17 days, subsiding over the next 2 months; whereas the sub-
430
A. J. INGRAM
;-.500//m \
FIGURES
1-6
Figs. 1-6. Injection experiment 1 showing the initial reaction, stained with Celestin
blue-Mayer's haemalum counterstained with van Gieson's stain, except Fig. 5
which was stained with haematoxylin and eosin.
Fig. 1. Thirteen days after injection showing (E) thickened epidermis, (O) oil droplet
space, (M) muscular tissue and (C) infiltrated cells lying in oil droplet space.
Fig. 2. Twenty-eight days after injection showing (C) cells lying in oil droplet space
and (/) cells migrating into the surrounding tissues.
Fig. 3. Fifty-two days after injection showing (0) empty oil droplet space and (/)
cells infiltrating musculature.
Fig. 4. Higher power of infiltrating region of Fig. 3 showing (/) infiltrated cells and
(F)fibrocytesproducing collagen.
Fig. 5. Six months after injection showing (O) empty oil droplet space, (E) epidermis
normal in thickness and (/) lower density of infiltrated cells.
Fig. 6. Two years after injection showing (D) diffuse and (CC) compact collagen
layers of the dermis and (C) a massive deposition of collagen in place of the infiltrated
cells.
Reactions to carcinogens in axolotl
431
cutaneous reaction was more prolonged, its regression being associated with the
deposition of collagen by fibrocytes. At no stage did the epidermis succeed in
penetrating the compact collagen layer of the dermis. The carcinogen seemed
to have an inhibitory effect on the development of the diffuse collagen layer of
the dermis throughout the initial reaction.
Controls and the initial reaction
Olive oil injected axolotls showed no reaction at 13 or 52 days; thus the
presence of oil was not responsible for the initial reaction.
Picryl chloride implantation produced pronounced haemorrhage on the day
after implantation, with obvious damage to the epidermis; however the haemorrhage disappeared after 9 days. An animal sacrificed at 15 days (Fig. 7) showed
a marked epidermal proliferation with the epidermis apparently undercutting
the picryl chloride crystal. Subcutaneous damage was evident but there was no
massive cellular infiltration. An animal sacrificed at 31 days (Fig. 8) showed a
virtual return to normal.
Perylene implantation, on the other hand, gave rise to a marked infiltration of
cells subcutaneously, but no epidermal hyperplasia, 15 days after implantation;
however, the reaction had virtually disappeared by 31 days (Figs. 9, 10). With
DBA implantation as with the injection of a suspension of crystals in oil, both
an epidermal hyperplasia and a subcutaneous infiltration of cells occurred.
Unlike the reactions due to perylene and picryl chloride, the reaction to DBA
implantation was maximal at 31 days (Fig. 11).
From these results, the features of the initial reaction which would seem to be
due to the carcinogenic nature of the compound are: the thickening of the
epidermis and the tendency of the infiltrated cells to remain longer at the site
of implantation.
Secondary reaction
Subsequent to the disappearance of the 'initial reaction', a secondary reaction
became apparent in all the injection experiments excepting injection experiment
4, in which the carcinogen was in solution only. The secondary reaction took the
form of opaque papilloma-like outgrowths which were 1-3 mm in diameter
(Figs. 12-14). They occurred in 14 out of 54 animals in the combined DBA
experiments and in 18 out of 57 sites injected with MC in injection experiment 5,
there being usually one outgrowth per site.
Six of these outgrowths have been examined histologically, Fig. 15 showing
the typical general structure. They were composed solely of epidermal cells, of
which both types were represented; however, the proportion of Leydig cells
(large cells with eosinophilic granules) tended to be lower than in normal epidermis. Spaces filled with mucopolysaccharide material, demonstrated by the
alcian blue and the PAS reaction, were present in all the outgrowths examined.
It appeared that these spaces arose from the breakdown of groups of epidermal
432
A. J. INGRAM
500 fim
,
FIGURES
7-11
Figs. 7-11. Crystal implantation experiment showing reactions to picryl chloride,
perylene and dibenzanthracene; stained with haematoxylin and eosin.
Fig. 7. Fifteen days after picryl chloride implantation showing (E) thickened epidermis, (U) undercutting epidermis and (P) remains of picryl chloride.
Fig. 8. Thirty-one days after picryl chloride implantation showing implantation site
returning to normal.
Fig. 9. Fifteen days after perylene implantation showing (is) normal epidermis
and (/) infiltrated cells.
Fig. 10. Thirty-one days after perylene implantation showing the implantation site
returning to normal.
Fig. 11. Thirty-one days after DBA implantation showing (E) thickened epidermis
and (/) infiltrated cells.
Reactions to carcinogens in axolotl
433
cells which had accumulated large amounts of PAS-positive material in their
cytoplasm.
The close correlation between the outgrowths and the injection was demonstrated by the close proximity of empty oil droplet spaces to some of the outgrowths (e.g. Fig. 15). The base of the epidermis appeared undulating in two
of the outgrowths examined and increased vascularization was found in association with this. Also a reduction in thickness of the diffuse collagen layer below
the outgrowths was occasionally observed.
FIGURES
12-16
Figs. 12-14. Examples of papilloma-like outgrowths.
Fig. 15. Section of a papilloma-like outgrowth from injection experiment 1, 19£
months after injection showing (JP) papilloma-like outgrowth, (0) oil droplet space
and (D) diffuse and (CC) compact collagen layer of dermis.
Fig. 16. High power of papilloma-like outgrowth showing (S) space filled with cell
debris and (L) Leydig cell.
Time of appearance ofpapilloma-like outgrowths with MC and DBA
Table 3 shows the time of appearance of the outgrowths in all the injection
experiments involving DBA. Observations for outgrowths were made every 2
months and, as the various experiments were started at different times, 2-monthly
groupings were taken for this table. As animals died or were sacrificed during
the course of these experiments, the number of animals surviving at the first
appearance of outgrowths in each experiment was taken for comparison with
434
A. J. INGRAM
the number of outgrowths developed. If an animal with an outgrowth died or
was sacrificed it was still counted in subsequent observations. It can be seen by
reference to Table 3 that the older the animals were at the start of each experiment
the shorter was the latent period for outgrowth formation (that is: the time
from injection to the first appearance of outgrowths).
Table 4 shows the time of appearance of papilloma-hke outgrowths in injection
experiment 5 (involving MC injection into three sites). As a moderately large
Table 3. Observations for papilloma-like outgrowths in DBA
injection experiments
No. of animals with outgrowths
experiment
in months
Expt 1
9-10-9
11-12-9
13-14-9
15-16-9
17-18-9
19-20-9
21-22-9
23-24-9
Age at time of injection
in months
No. of animals in
group at first
appearance of
outgrowths
Expt 2
repeat
Expt 2
single
Expt 3
Total
—
—
—
2
5
5
5
5-3
(first)
10
. .
1
1
2
2
2
2
2
10-7
(mean)
20
2
2
3
3
3
3
3
3
13-2
(mean)
10
2
3
4
5
10
14
14
14
—
—
—
—
—
3
4
4
4
5-3
14
54
Table 4. Observations for papilloma-like outgrowths in injection
experiment 5
Duration
of experiment in
months
5
7
9
iii
13
Animals Animals
with skin examined
No. of papilloma-like
for outdisease
outgrowths
A
(appear- growths ,
ance of (excluding Anterior Posterior
outgrowths diseased
body
body
masked) animals)
region
region
Tail
8
8
7
6
2
22
22
20
19
15
Maximum no. of outgrowths =
(19 animals)
Total
Total
no. of
no. of
sites
outgrowths examined
—
—
1
3
1
—
3
6
10
7
—
4
4
5
5
—
7
11
18
13
3
10
5
18
66
66
60
57
45
57
Reactions to carcinogens in axolotl
435
number of sites was involved it was thought better, in this table, to relate the
number of outgrowths to the number of injection sites observed for outgrowths
at each time.
A comparison of the development of outgrowths between the combined DBA
experiments and injection experiment 5 (involving MC) is shown in Fig. 17. By
comparing the times for the development of half the outgrowths, it can be seen
that the outgrowths appeared quicker in the MC experiment than in the combined
DBA experiments (see arrows, Fig. 17). Although this could be due to the
differences in carcinogenicity between MC and DBA, it could also be due to
differences in age.
18r
4 -
10
12
14
16
18
22
24
Time after injection (months)
Fig. 17. A comparison of the development of papilloma-like outgrowths in response
to injections of dibenzanthracene and methylcholanthrene (• injection experiment
5 involving methylcholanthrene, O combined dibenzanthracene experiments).
The effect of site of injection on papilloma-like outgrowth formation
It can be seen from Table 4 that in injection experiment 5, in which three sites
were injected, the greatest number of outgrowths occurred at 11£ months in all
three sites. Out of the 19 animals examined for outgrowths at this time, three
had developed outgrowths in the anterior body region, ten had developed outgrowths in the posterior body region and five had developed outgrowths in the
tail.
Although it is possible that the differences in frequency of outgrowth formation
were a result of the posterior body region being a more favourable site, it is
thought more likely that the differences were a result of the difficulty encountered
28
E M B 26
436
A. J. INGRAM
in injecting the anterior body region and the tail. In any case control by a regeneration field could not have been responsible as, if this was the case, the anterior
body region should have had the highest frequency of outgrowths.
1 mm
Fig. 18. A montage of the tumour found in injection experiment 2-repeat, 3 years
after the first injection, showing (7) main body of tumour, (/) infiltration of musculature and between arrows infiltration through the compact collagen layer of the
dermis.
Fig. 19. Montage of hepatoma found in sandwich experiment 1 two years after
implantation, showing (U) undifferentiated cells in the centre, (P) peripheral region
and (//) more or less normal hepatic tissue.
Reactions to carcinogens in axolotl
437
Tumour development in the injection experiments
The only tumour found so far in the injection experiments occurred in an
axolotl from 'injection experiment 2-repeat', 3 years after the first injection.
The tumour was a sarcoma and was found at the site of injection. It measured
5-5 x 1-5 mm in cross-section, the axolotl weighing 82 g. The montage (Fig. 18)
shows the overall structure of the tumour and its infiltration into the dermis and
the surrounding musculature. The main body of the tumour consisted of undifferentiated cells (see Fig. 20); however, where the tumour cells were infiltrating
into the musculature (Fig. 21) they bore a resemblance to fibrocytes. I would
therefore suggest tentatively that the tumour was a fibrosarcoma, though a
mixed cell origin is possible. Figure 22 shows how the tumour cells had penetrated the compact and diffuse collagen layers of the dermis and infiltrated into
the epidermis. Mitoses were very rare and no metastases were found in the
internal organs. As this tumour was found during routine sectioning no attempt
at transplantation was made.
Results of the sandwich experiments
In the sandwich experiments a reaction, probably corresponding to the initial
reaction, was apparent externally as a swelling of the site. In some cases this
persisted for up to a year after implantation, perhaps due to the larger amount
of carcinogen introduced. One of the axolotls (from sandwich experiment 1), in
which the swelling persisted longest, died 2 years after implantation and a
large tumour was found in the liver, the implant having disappeared from the
implantation site. The montage (Fig. 19) shows the central and peripheral
regions of the tumour and a region of more or less normal liver. In the central
region (Fig. 25) cells with nuclei of various sizes were evident together with
occasional giant cells; also groups of blood cells were present. In the peripheral
region (Fig. 24) cords of abnormal hepatic cells were evident interspaced with
fibrous tissue. This suggests that the tumour was an anaplastic hepatoma.
It is possible that macrophages may have engulfed the carcinogen and then
migrated to the liver, where a primary tumour subsequently arose.
DISCUSSION
The initial reaction
The initial reaction in the axolotl appears to correspond in timing to the
reversible reaction to carcinogens reported in the newt Triturus cristatus by
Neukomm (1944, 1957, 1959) and by Arffmann & Christensen (1961). It also
corresponds in timing to the early part of the reaction claimed to be tumorous
by Seilern-Aspang & Kratochwil (1962, 1965) in the same species of newt. The
infiltration of the underlying tissues by the epidermis, observed by these workers,
was never observed in the axolotl; perhaps due to the comparative thickness of
28-2
438
A. J. INGRAM
25
Figs. 20-22. High power of sarcoma (montage Fig. 18).
Fig. 20. Central region of sarcoma, showing undifferentiated cells.
Fig. 21. Infiltration of musculature, showing (/) infiltrating cells and (M) musculature.
Fig. 22. Infiltration into skin showing (D) diffuse and (C) compact collagen of
dermis, (T) tumour cells, (B) breakdown of dermal collagen layers and (/) infiltration
of epidermis by tumour cells.
Figs. 23-25. High power of hepatoma (montage Fig. 19).
Fig. 23. 'Normal' region of liver showing (H) hepatic cells and (BV) blood vessels.
Fig. 24. Peripheral region showing (C) cords of abnormal hepatic cells and (F)
fibrous tissue.
Fig. 25. Central region showing variation in nuclear size, (G) occasional giant cells
and (J?) groups of blood cells.
Reactions to carcinogens in axolotl
439
the dermal collagen in the axolotl. Ruben & Balls (19646), who implanted methylcholanthrene crystals into the forelimbs of Xenopus laevis, reported an undercutting of dead tissue by the epidermis, resulting in its elimination. It may be
that the early infiltration by the epidermis in the newt represents a similar
elimination process.
The infiltration of the oil droplet by macrophages, observed in the axolotl
13 days after injection, was also reported in the newt by ArfTman & Christensen
(1961). In the axolotl this appeared to be due to the mechanical irritation of the
carcinogen crystals, as a similar infiltration was observed following perylene
crystal implantation.
It is therefore possible that the infiltration by the epidermis in the newt and
the infiltration by macrophages, observed in the newt and the axolotl, may be
partially a wound healing response. The deposition of collagen, associated with
the disappearance of the initial reaction in the axolotl, suggests a further parallel
between the initial reaction and wound healing.
If the initial reaction was regarded as tumorous, its regression in the axolotl
could not have been due to a regeneration field as it occurred in the posterior
body region.
The secondary reaction
The papilloma-like outgrowths were clearly induced by the carcinogens, as
they were nearly all at the injection sites or reasonably close. No outgrowths
occurred in the group injected with a saturated solution of MC instead of a
suspension of crystals, perhaps due to the removal of all the carcinogen from
the site during the initial reaction.
Various features of the outgrowths showed similarities with the papillomata
of mammals. Orr (1938) found that, following skin painting in mice, the time
of papilloma formation was twice as long with dibenzanthracene as with methylcholanthrene. This would agree with the observation that the latent period to
papilloma-like outgrowth formation in the axolotl was longer with DBA than
with MC. Increased vascularization and modifications of the dermal collagen
beneath the outgrowths in some cases also suggested a possible homology with
mammalian papillomata.
The decrease in the latent period to outgrowth formation in the axolotl with
increasing age was contrary to the relationship between the rate of epidermal
carcinogenesis and age in mice reported by Cowdry & SuntzefT (1944).
The progression of one of these outgrowths to a carcinoma would be proof
of homology with the papillomata of mammals; however, this has not yet been
observed.
Occurrence of tumours
The two tumours found in the course of these experiments, a sarcoma and a
hepatoma, took 2 and 3 years respectively to become apparent. Although both
these tumours could have been spontaneous, the sarcoma is unlikely to have
been as it was found at the site of injection. The possibility that the sarcoma
440
A. J. INGRAM
was induced is reinforced by the findings of Breedis (1951, 1952) that only two
spindle-cell sarcomas resulted from the injection of MC in olive oil into the
limbs of 500 newts. These took around one year to develop and proved transplantable. Thus the induction of tumours in urodeles with polycyclic hydrocarbons may, in some cases, involve both a long latent period and a low tumour
incidence. This might explain the failure of many workers to induce tumours in
urodeles, for example Finkelstein (1944) in the axolotl and Neukomm (1944)
and ArrTmann & Christensen (1961) in the newt. The finding by ScheremetievaBrunst & Brunst (1948) that melanomas take several years to develop in the
axolotl, suggests that a long latent period to tumour formation in urodeles might
be expected in some cases.
Thanks are extended to the Wellcome Trust for financial support and to Dr F. S. Billett
for his help and advice.
REFERENCES
ANDERSEN, H. & JORGENSEN, J. B. (1960). Decalcification and staining of archaeological
bones with histochemical interpretation of metachromasia. Stain Technol. 35, 91—96.
ARFFMANN, E. & CHRISTENSEN, B. C. (1961). Studies on the newt test for carcinogenicity.
I. Benzo(a)pyrene, dibenz(a,h)anthracene and 3-methylcholanthrene. Act a path, microbiol.
scand. 52, 330-342.
BREEDIS, C. (1951). Transplantable sarcoma of the salamander induced by methylcholanthrene. Cancer Res. 11, 239.
BREEDIS, C. (1952). Induction of accessory limbs and of sarcoma in the newt (Triturus viridescens) with carcinogenic substances. Cancer Res. 12, 861-866.
COWDRY, E. V. & SUNTZEFF, V. (1944). Influence of age on epidermal carcinogenesis induced
by methylcholanthrene in mice. Yale J. Biol. Med. 17, 47-58.
DRURY, R. A. B. & WALLINGTON, E. A. (1967). Carlton's Histological Technique. London:
Oxford University Press.
ERMIN, R. & GORDON, M. (1955). Regeneration of melanomas in fishes. Zoologica, N.Y.
40, 53-84.
FINKELSTEIN, E. A. (1944). Opukholeviy rost n bespozvonotchnikh i nizshikh pozvonotchnikh (Tumour growth in invertebrates and lower vertebrates). Usp. sovrem. Biol. 17, 320347.
MIZELL, M. (1960). Anuran (Lucke) tumour breakdown in regenerating anuran tadpole tails.
Anat. Rec. 137, 382-383.
NEEDHAM, J. (1936). New advances in the chemistry and biology of organized growth. Proc.
R. Soc. Med. 29, 1577-1626.
NEUKOMM, S. (1944). Le probleme de la cancerisation par le goudron et les substances cancerigenes chez les Tritons. Mem. Soc. vaud. Sci. nat. 8, 187-215.
NEUKOMM, S. (1957). Un test sensible et ultra-rapide du pouvoir cancerigene de certaines
substances chimiques. Oncologia, Basel 10, 107-119.
NEUKOMM, S. (1959). The newt test for rapid biological assay of certain carcinogenic substances. Acta Un. int. contra Cancr. 15, 654-658.
ORR, J. W. (1938). The changes antecedent to tumour formation during treatment of mouse
skin with carcinogenic hydrocarbons. /. Path. Bact. 46, 495-515.
RUBEN, L. N. (1955). The effects of implanting anuran cancer into non-regenerating and
regenerating urodele limbs. /. exp. Zool. 128, 29-51.
RUBEN, L. N. & BALLS, M. (1964#). Implantation of lymphosarcoma of Xenopus laevis into
regenerating and non-regenerating forelimbs of that species. J. Morph. 115, 225-237.
RUBEN, L. N. & BALLS, M. (19646). Implantation of methylcholanthrene crystals into regenerating and non-regenerating forelimbs of Xenopus laevis. J. Morph. 115, 239-253.
Reactions to carcinogens in axolotl
441
SCHEREMETIEVA, E. A. (1965). Spontaneous melanoma in regenerating tails of axolotls. / . exp.
Zoo/. 158, 101-111.
SCHEREMETIEVA-BRUNST, E. A. & BRUNST, V. V. (1948). Origin and transplantation of a
melanotic tumour in the axolotl. Spec. Publs N. Y. Acad. Sci. 4, 269-287.
SEILERN-ASPANG, F. & KRATOCHWIL, K. (1962). Induction and differentiation of an epithelial
tumour in the newt (Triturus cristatus). J. Embryo/, exp. Morph. 10, 337-356.
SEILERN-ASPANG, F. & KRATOCHWIL, K. (1965). Relation between regeneration and tumour
growth. In Regeneration in Animals and Related Problems (eds. V. Kiortsis and H. A. L.
Trampusch), pp. 452-473. Amsterdam: North Holland Publishing Company.
WADDINGTON, C. H. (1935). Cancer and the theory of organizers. Nature, Lond. 135, 606-608.
{Manuscript received 15 February 1971, revised 30 May 1971)