/ . Embryol. exp. Morph. Vol. 66,pp. 117-126, 1981
Printed in Great Britain © Company of Biologists Limited 1981
Regeneration from duplicating fragments of the
Drosophila wing disc
By JANE KARLSSON 1 AND R. J. SMITH
From the Zoology Department, University of Edinburgh and the
MRC Laboratory of Molecular Biology, Cambridge University
Postgraduate Medical School
SUMMARY
It is a general rule that of two complementary Drosophila imaginal disc fragments, one
regenerates and the other duplicates. This paper reports an investigation of an exception to
this rule. Duplicating fragments from the periphery of the wing disc which lacked presumptive
notum were found to regenerate notum structures during and after duplication. The propensity for this was greatest in fragments lying close to the presumptive notum, with the
exception of a fragment confined to the posterior compartment, which did not regenerate
notum. Structures were added sequentially, and regeneration stopped once most of the
notum was present.
These results are not easily explained by the polar coordinate model, which states that
regeneration cannot occur from duplicating fragments. Since compartments appear to be
involved in this type of regeneration as in others, it is suggested that a new type of model is
required, one which permits simultaneous regeneration and duplication, and assigns a major
role to compartments.
INTRODUCTION
In general, if the Drosophila imaginal wing disc is cut into two fragments,
one duplicates during culture while the other regenerates (Bryant, 1975; Karlsson,
1981 a). An exception to this rule was found recently; in certain fragments,
duplication of structures expected from the fate map was found to be accompanied by regeneration of others (Karlsson, 1981 a). The structures regenerated
were from the notum, and all fragments which were tested and which lacked
presumptive notum were found capable of regenerating notal bristles while
duplicating. These are probably the 'adventitious bristles' described by Bryant
(1975).
A similar phenomenon has been reported by Schubiger (1971) in the leg disc,
van der Meer & Ouweneel (1974) in the haltere disc, and Bryant & Hsei (1977)
in the genital disc. In all these cases, certain fragments were found to duplicate
some structures and regenerate others. Van der Meer & Ouweneel (1974) have
suggested for this the term 'regenerative duplication'.
1
Author's address: Genetics Laboratory, Department of Biochemistry, South Parks Road,
Oxford 0X1 3QU.
118
J. KARLSSON AND R. J. SMITH
These observations contravene the first rule of the polar coordinate model
(French, Bryant & Bryant, 1976). This rule views the complementarity between
regeneration and duplication as the result of the regenerating fragment having
more than half, and the duplicating fragment less than half, of the values in a
circumferential axis of positional information. Wound healing brings disparate
values into apposition, and intercalary regeneration occurs to fill in the missing
ones by the shorter of the two possible routes round the circumference. Thus if
duplicating fragments regenerate, they must do so by the forbidden longer route.
We have therefore investigated this phenomenon in detail to find whether it
had sufficient regularity to warrant description as a new, non-intercalary, type
of regeneration.
MATERIALS AND METHODS
Host and donor flies were of ebony11 genotype (Lindsley & Grell, 1968) and
were raised at 25 °C on standard cornmeal/agar/syrup medium seeded with live
yeast. Wing discs were removed from late third instar larvae in insect Ringer's
solution. Fragments were cut with tungsten needles and implanted into the body
cavities of well-fed 1 - to 3-day-old fertilized adult females where they remained for
varying periods. They were then removed and reimplanted into the body cavities
of late third instar larvae using the method of Ephrussi & Beadle (Ursprung,
1967). When the hosts emerged as adults, the metamorphosed implants were
removed, mounted in Gurr's Hydramount, and scored for the structures shown
in Fig. 1. Implants cultured for 0 days (Table 1) were implanted directly into
mature third instar larvae.
Incubation
Short-term cultures were incubated at 25 °C and long-term ones (14 days and
over) at 30 °C. It was found that development of Drosophila melanogaster at
30 °C takes 80 % of the time it takes at 25 °C. 'Days of culture' in Table 1 refers
either to actual days at 25 °C or to the number of days a 30 °C incubation would
have taken at 25 °C.
Scoring ofnotal structures
The different structures were identified with the aid of the descriptions of
Bryant (1975). It is possible to identify individual macrochaetes of the notum
in implants where much of the notum is present. However, where only a few
bristles are present, the only one which can be positively identified is anterior
post-alar, which is flanked by two sensilla trichodea. Scutellum is easily recognizable as a vesicle of heavy cuticle with two macrochaetes and no microchaetes.
Scutellum makes a joint with posterior notal wing process and axillary sclerite
4; this is also clearly recognizable in implants and is here termed 'joint'. Prescutum is also usually recognizable as an area of heavy cuticle with about 50
closely spaced microchaetes arranged in rows.
This description agrees with that of Bryant (1975).
Regenerative duplication in Drosophila
119
RESULTS
Table 1 shows the results of scoring the metamorphosed implants of the
fragments depicted in Fig. 1. These results allow the following conclusions to
be drawn.
(1) Duplicating fragments lacking presumptive notum have the
ability to regenerate it
Most implants of fragments a, c and e showed evidence of duplication (Table
1), and none regenerated a complete disc. Fragments b and d, on the other hand,
appeared to belong to the class of fragment which regenerates sometimes and
duplicates sometimes (Karlsson, 1981 a); some implants of these fragments
regenerated a complete disc with no evidence of duplication. For this reason,
only those implants of these two fragments which had at least two duplicated
structures were included in Table 1.
No implant had notal bristles after direct implantation into larvae with the
exception of a few implants of fragments c and d. 4/14 and 4/13 implants
respectively of these two fragments had a few bristles after 0 days of culture;
these two fragments were cut very close indeed to presumptive notum.
In those implants which produced bristles after culture, in many cases these
bristles could not be positively identified as being derived from the notum. In
others, however, they were clearly notal bristles, with a characteristic mixture
of micro- and macrochaetes on well-tanned cuticle. It was therefore concluded
that all bristles were notal bristles. Fragments b, c and d often produced prescutum, and c and d usually produced scutellum and joint. These latter two
structures were often duplicated; notal bristles may have been duplicated as
well, but this would be almost impossible to detect.
(2) The time-course of regenerative duplication is longer than
that of regeneration
Regeneration of wing disc fragments is complete after 3 or 4 days (Karlsson,
1981 b). In fragment a, however, there was a significant (P < 0-05) increase in
bristle number between 7 and 14 days of culture. Duplication in this fragment
was complete after 4 days; no increase in the number of duplicated structures
per implant was observed between 4 and 7 days (1-0 ± 0-2 after 4 days; 1-1 ± 0-2
after 7 days). Fragment c also showed a significant increase in average numbers
of bristles during culture periods longer than 7 days.
(3) Regenerative duplication stops once most of the notum is present
Fragment c appeared to undergo more extensive regenerative duplication than
the others, and was therefore selected for a 30-day culture to find out whether a
whole disc would eventually be produced. These implants were not appreciably
different from those cultured for 14 days, except that more notal bristles were
0
11
_
18
_
18
82
64
100
_
0
0
t
4
12
_
_
_
_
_
100
58
100
42
0-7
0-3
75
0
1
20
_
_
_
_
_
_
_
15
_
_
_
5
85
80
100
40
2-7
1-5
65
0
a
14
13
8
_
_
_
_
_
8
8
8
_
54
23
31
77
54
100
100
12
4-2
62
0
_
31
19
13
63
50
100
100
13
2-5
44
0
22
16
_
_
_
_
_
6
\
<—-^
7
0
16
20
_
30
_
_
_
_
_
_
_
_
_
_
_
_
_
_
85
60
50
63
65
88
90
100
88 100
94 100
70
19
_
60
0 (100)
0
0
b
7
0
14
30
15
14
13
10
_
27
38
50
_
13
21
8
_
60
54
80
_
53
54
30
_
40
7
15
_
60
62
23
100 100 100
20
87
92
86
60
100
93 100
50
_
_
15
10
_
_
_
15
_
_
8
_
_
_
8
_
_
_
100 100 100
90
29
73
69 100
1 0 27
41
70
0-5 9-7 17
13
93 100
7
20
0
33
31
0
c
*
A
d
\
7
0
13 20
_ 20
20
20
_
50
15 40
15
8
25
38
77 100
95
92
100 100
_
5
_
_
_
_
_
_
_
90
85
100 100
85
31
1-9 22
1-9 6-4
0 (100)
0
20
i—
e
-^
A
s
e*
7
0
14
34
12
18
_
_
8
_
_
_
_
_
8
_
_
8
_
_
_
_
_
_
78
71
50
100 92
94
100 75
94
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
94 100 100
8
0
0
9-3 0
0
0
91 0
0 100 79
8
0
-
i—
0
16
_
_
_
_
_
_
_
_
_
_
_
_
_
100
25
0
0
0
0
/
f
7
57
2
_
_
33
23
19
5
7
9
4
100
44
9
0
0
25
0
A
7
13
8
46
23
38
8
15
8
8
100
31
8
0
0
31
0
.
/ , mixed
* These implants of fragment e wen; out slightly smaller than the preceding ones so as to be sure of excluding tissue from the anterior
compartment.
Days of culture
n
Prescutum
APA
Scutellum
Joint
PNWP
AS 4
AS 3
A Cord
A Lobe
ANWP
Tegula
HP
P Costa
YC/PVR
PS/PWP
Wing Blade
AS1/2/UP/PDR
Notal bristles
Number of bristles
S.E.
% D (FM)
% D (non F M )
Fragment
(The figures refer to the percentage of implants where the marker was completely or partially present, singly or in duplicate. All average
bristle numbers above 10 have been rounded off to the nearest whole number, and are shown with standard errors. % D (FM):
percent of implants having at least one structure expected from the fate map duplicated. This figure is bracketed for fragments b and d
because no implant of these fragments which did not duplicate was included in the Table. % D (non F M ) : percent of implants having
at least one structure not expected from the fate map duplicated. P costa, proximal costa. Other abbreviations as in legend to Fig. 1.)
Table 1. Structures differentiated by implants of the fragments shown in Fig. 1, either implanted directly into larval hosts
for metamorphosis (0 days of culture) or cultured in adult females for varying periods before transfer to larval hosts
>—i
<_
on
>
O
r
>
Regenerative duplication in Drosophila
121
D
Fig. 1. Fate map of the wing disc (after Bryant, 1975), and the fragments used.
Compartment borders are shown in broken lines. The position of the anteroposterior
border was found using clones of a new cell marker (Brower, Lawrence & Wilcox,
1981). Presumptive wing blade is stippled. Large dots represent microchaetes of
the notum. A, P, D, V, presumptive anterior, posterior, dorsal, ventral. Abbreviations: APA, anterior post-alar; Jo, joint; Scu, scutellum; PNWP, posterior notal
wing process; AS 1 -4, axillary sclerites 1 -4; A Cord, axillary cord; A Lobe, alar lobe;
ANWP, anterior notal wing process; HP, humeral plate; YC, yellow club; PVR,
proximal ventral radius; PS, pleural sclerite; PWP, pleural wing process; UP,
unnamed plate; PDR, proximal dorsal radius.
present (Table 1). Some structures were lost, as expected in long-term cultures
(Bryant, 1978), and so the frequency of duplication decreased. All implants
had regenerated a fairly complete notum, but virtually nothing else, indicating
that regenerative duplication stops once most of the notum is present.
(4) Regenerative duplication produces first those structures closest to the cut edge,
and adds others in sequence; structures located some distance from the periphery
of the disc in the fate map are not produced
This is most clearly seen in fragment c. If the above is true, no implant of this
fragment should have had prescutum which did not also have scutellum and
joint, and none should have had tegula which did not also have prescutum,
scutellum and joint (see Fig. 1). Also, anterior post-alar, axillary sclerites 1 and
2, and proximal dorsal radius should not have been present, as these markers
do not lie at the periphery of the disc in the fate map (Fig. 1). This was all found
122
J. KARLSSON AND R. J. SMITH
YC
PS
(b)
250 Mm
Fig. 2. Camera-lucida drawings of implants which have duplicated structures
expected from the fate map and regenerated notum. (a) fragment a, cultured for
7 days. Three notal bristles have appeared on the vesicle of mesopleura. (b) fragment
d, cultured for 7 days. This implant has regenerated scutellum and joint, and some
microchaetes which are probably part of the prescutum. Abbreviations as in legend to
Fig. 1.
Regenerative duplication in Drosophila
123
to be true. The two most frequently regenerated structures were scutellum and
joint, which lie close to the cut edge. These usually appeared together; of 19
implants having one or the other or both, 12 had both. Prescutum was the next
most frequent structure, and implants having it almost always (8/9) had
scutellum as well. Two implants in the 14-day set had tegula and anterior notal
wing process (and one of these had humeral plate and proximal costa too), and
both of these had prescutum and either joint or scutellum.
Structures located towards the centre of the disc in the fate map were noticeably lacking. Axillary sclerites 1 and 2, unnamed plate and proximal dorsal
radius were never found in any of the experiments, except of course in fragment
/ , in which these structures are expected from the fate map. The frequency of
anterior post-alar did not increase after culture in either fragment c or d, and
its absence is even more striking since it lies so close to the cut edge.
It seems likely that regenerative duplication in anterior fragments proceeds
sequentially too, in the opposite direction to posterior fragments. Fragment b
never produced scutellum, joint or anterior post-alar, although it often had
prescutum. Anterior notal wing process, which lies near the cut at a short
distance from the edge of the disc in the fate map, was lacking as well, suggesting
that regenerative duplication in this fragment too proceeds round the periphery
of the disc.
Fragment a seems at first glance not to regenerate sequentially. Notum appears
for example before tegula and humeral plate, which lie between this fragment
and notum on the anterior edge of the disc. However, there is a structure which
lies still closer to the edge of the disc at this location. This is mesopleura, which
appears in implants as a semicircular vesicle running from the notum towards
the yellow club. It is this vesicle on which notal bristles first appear (Fig. 2 a).
It is thus possible that mesopleura is regenerated first, followed by notum, in
which case regenerative duplication in this fragment too would be sequential.
(5) Regenerative duplication is most frequent and complete in fragments
lying very close to presumptive notum in the fate map
Implants of fragments c and d, which were cut very close to presumptive
notum, sometimes had a few bristles when implanted directly into larvae. One
of these bristles was usually (6/8 cases) recognizable as anterior post-alar. These
two fragments shared with fragment b, which was also cut very close to presumptive notum but on the anterior side, the highest bristle frequency and average
bristle numbers. The difference in average bristle number between these fragments and fragment a, which lies some distance from presumptive notum in the
fate map, is significant (P < 0-01 in both cases).
124
J. KARLSSON AND R. J. SMITH
(6) Tissue confined to the posterior compartment does not undergo
regenerative duplication
Fragment e was cut so as to exclude tissue from the anterior comparment
(Garcia-Bellido, Ripoll & Morata, 1976), and after 14 days of culture only one
implant out of 12 had produced any notal bristles, although all had duplicated.
Another experiment was done in which this fragment was cut very slightly
smaller so as to be absolutely sure of excluding anterior tissue, and none (0/34)
of these implants produced any notal bristles. This second experiment is marked
with an asterisk in Table 1.
It is probably not possible to cut a fragment confined to the anterior compartment. If compartment borders are closed lines, as at least early ones must
be since clones never cross them, the anteroposterior border probably runs down
the anterior edge of the disc. All anterior fragments would therefore contain
part of it.
(7) Regenerative duplication is a property of the periphery of the disc; the centre
is not involved, nor is the central part of the peripodial membrane
Fragment / is from the centre of the disc and includes presumptive wing
blade and dorsal hinge (proximal dorsal radius, axillary sclerites 1 and 2,
unnamed plate). Only 5 out of 57 implants of this fragment had notal bristles;
none of these had duplicated structures, and all 5 had other regenerated structures as well as notum. These included axillary sclerite 3 (4 cases), axillary cord
(3 cases), proximal costa (2 cases), and humeral plate, tegula, yellow club and
alar lobe (1 case each). None of these structures was found in implants injected
directly into larvae (Table 1). Central fragments do sometimes regenerate
(Bryant, 1975), and it was concluded these 5 implants had regenerated normally,
and that this type of fragment is not capable of regenerative duplication. However the possibility was not ruled out that the peripodial membrane was involved
in regenerative duplication and that it had become detached during culture.
Another experiment was therefore done with the same fragment in which tungsten
needles were used to fold the peripodial membrane inside the epithelial layer
to ensure that they remained together (Haynie & Bryant, 1976). One out of
13 of these implants had notal bristles, and as this implant, like the unmixed
ones, had no duplicated structures, and had other peripheral structures as well
(axillary sclerite 3, humeral plate, proximal costa and yellow club), it was
concluded that this central part is not capable of regenerative duplication.
DISCUSSION
The results show that many duplicating fragments from the periphery of the
wing disc which do not include presumptive notum are able to regenerate it
during and after duplication. This process appears to be an orderly and predictable
Regenerative duplication in Drosophila
125
one, structures being produced in sequence from the cut edge as in the more
usual type of regeneration (Abbott, Karpen & Schubiger, 1981; Karlsson,
1981 b), and appearing in virtually all cases to stop once most of the notum is
present. Perhaps the most striking result is that tissue confined to the posterior
compartment does not produce notum, although slightly larger fragments which
contain a small part of the anteroposterior compartment border do so at high
frequency. The posterior fragment which was tested was necessarily rather small,
but its size clearly did not prevent growth as the frequency of duplication was
very high. This experiment was performed twice, the second time great care
being taken to exclude anterior tissue. One implant in the first experiment did
produce notum; this was virtually complete, again indicating that small size
does not prevent this fragment from regenerating notum. It would be interesting
to know whether tissue confined to the anterior compartment can regenerate
notum, but it is probably not possible to cut such a fragment (see Results).
Since, then, compartments seem to be involved in this phenomenon, it is
perhaps no coincidence that by and large, the structures produced lie along
the anteroposterior compartment border (see Fig. 1). This apparent ability of
tissue at the anteroposterior border to regenerate proximally is reminiscent of
the ability of the other end of this border to regenerate distally. Fragments
lacking the ventral end cannot regenerate distally (Karlsson, 1980), just as
fragments lacking the dorsal end are apparently unable to regenerate notum.
Both phenomena are difficult to explain in terms of the polar coordinate model;
according to the revised second rule of this model (Bryant, French & Bryant,
1981), all proximal tissue should be able to regenerate distally, and neither of
the two rules makes any provision for regeneration which is non-intercalary and
proceeds from distal to proximal. It is quite clear that the regeneration described
in this paper is not intercalary. The fragments studied duplicate, and their
complementary fragments regenerate (Bryant, 1975; Karlsson, 1981a). They
cannot thus have a sufficient number of values for complete regeneration, and
in any case complete regeneration never occurs. If, for example, notum regeneration were initiated by a transdetermination-like event which provided the fragment with over half the values, then intercalary regeneration would be expected
to produce a complete disc; also, the newly regenerated structures should not
be duplicated, which they often are.
There is at present no model which makes provision for epimorphic regeneration which is not intercalary. The present results therefore suggest that a
completely new type of model is required, one in which new positional values
can be generated by non-intercalary means, and which assigns a major role to
compartment borders.
This work was supported by the MRC and by a Beit Memorial Fellowship to JK.
EMB 66
126
J. KARLSSON AND R. J. SMITH
REFERENCES
L. C , KARPEN, G. H. & SCHUBIGER, G. (1981). Compartmental restrictions and
blastema formation during pattern regulation in Drosophila imaginal leg discs. Devi Biol.
(In press.)
BRYANT, P. J. (1975). Pattern formation in the imaginal wing disc of Drosophila melanogaster:
fate map, regeneration and duplication. J. exp. Zool. 193, 49-77.
BRYANT, P. J. (1978). Pattern formation in imaginal discs. In The Genetics and Biology of
Drosophila (ed. M. Ashburner & T. R. F. Wright), vol. 2c. London: Academic Press.
BRYANT, P. J. & HSEI, B. W. (1977). Pattern formation in asymmetrical and symmetrical
imaginal discs of Drosophila melanogaster. Amer. Zool. 17, 595-611.
BRYANT, S. V., FRENCH, V. & BRYANT, P. J. (1981). Distal regeneration and symmetry in
epimorphic fields. Science (in press).
BROWER, D. L., LAWRENCE, P. A. & WILCOX, M. (1981). Clonal analysis of the undifferentiated wing disc of Drosophila. Devi Biol. (in press).
FRENCH, V., BRYANT, P. J. & BRYANT, S. V. (1976). Pattern regulation in epimorphic fields.
Science 193, 969-981.
GARCIA-BELLIDO, A., RIPOLL, P. & MORATA, G. (1976). Developmental compartmentalisation
in the dorsal mesothoracic disc of Drosophila. Devi Biol. 48, 132-147.
HAYNIE, J. L. & BRYANT, P. J. (1976). Intercalary regeneration in the imaginal wing disc of
Drosophila melanogaster. Nature 259, 659-662.
KARLSSON, J. (1980). Distal regeneration in proximal fragments of the wing disc of Drosophila.
J. Embryol. exp. Morph. 59, 315-323.
KARLSSON, J. (1981 a). The distribution of regenerative potential in the wing disc of Drosophila.
J. Embryol. exp. Morph. 61, 303-316.
KARLSSON, J. (1981 b). Sequence of regeneration in the Drosophila wing disc. J. Embryol. exp.
Morph. 65, Supplement 37-47.
LINDSLEY, D. L. & GRELL, E. H. (1968). Genetic variations of Drosophila melanogaster.
Carnegie Inst. Washington Publ. No. 627.
SCHUBIGER, G. (1971). Regeneration, duplication and transdetermination in fragments of the
leg disc of Drosophila melanogaster. Devi Biol. 26, 277-295.
URSPRUNG, H. (1967). In vivo culture of Drosophila imaginal discs. In Methods in Developmental Biology (ed. F. M. Wilt & N. K. Wessels). New York: Crowell.
VAN DER MEER, J. M. & OUWENEEL, W. J. (1974). Differentiation capacities of the dorsal
metathoracic (haltere) disc of Drosophila melanogaster. II. Regeneration and duplication.
Wilhelm Roux Arch. EntwMech. Org. 174, 361-373.
ABBOTT,
(Received 4 November 1980, revised 18 June 1981)