/. Embryo/, exp. Morph. Vol. 33, 2, pp. 459-469, 1975
Printed in Great Britain
459
Expression of homeotic mutations
in duplicated and regenerated antennae of
Drosophila melanogaster
By WALTER J. GEHRING 1 AND GEROLD SCHUBIGER 2
From Biozentmm, University of Basel and the Department of Zoology,
University of Washington
SUMMARY
Three homeotic mutants, aristapedia (ssa and ssa'UCI) and Nasobemia (Ns) which involve
antenna-leg transformations were analyzed with respect to their time of expression. In
particular we studied the question of whether these mutations are expressed when the mutant
cells pass through additional cell divisions in culture. Mutant antennal discs were cultured
in vivo and allowed to duplicate the antennal anlage. Furthermore, regeneration of the
mutant antennal anlage was obtained by culturing eye discs and a particular fragment of the
eye disc. Both duplicated and regenerated antennae showed at least a partial transformation
into leg structures which indicates that the mutant gene is expressed during proliferation in
culture.
INTRODUCTION
Extensive studies on transplantation and exstirpation of imaginal discs of
Drosophila melanogaster have shown that each disc during metamorphosis gives
rise to a specific area of the adult epidermis with its cuticular structures. By
transplanting defined fragments of discs from mature larvae into host larvae of
the same age and allowing them to metamorphose, the various adult structures
can be 'mapped' within each disc in great detail (for a review see Gehring &
Nothiger, 1973). Each fragment forms its particular area of the adult cuticle.
However, if the fragment is allowed to proliferate, it can exhibit a larger
developmental potential: it may give rise to duplicated adult structures, it may
regenerate structures which normally are derived from other areas of the respective disc, or after extensive growth, it may undergo transdetermination,
whereby it can form additional structures which are normally derived from
another kind of disc (allotypic structures). Proliferation of disc fragments can
best be achieved by culturing them in adult hosts and subsequent transplantation into host larvae in order to obtain metamorphosis (Hadorn, 1963). The
developmental response depends essentially on the kind of fragment which is
1
Author's address: Biozentrum, University of Basel, 4056, Switzerland.
Author's address: Department of Zoology, University of Washington, Seattle, Washington 98105, U.S.A.
2
460
W. J. GEHRING AND G. SCHUBIGER
••
100//in
100//m
A II
Tu
. ( « ) •
100 ftm
\W/.im
Fc
Cl
{d)
Fig. 1. Normal and homeotic antennae in situ, (a) Aristapedia (ssa); (6) aristapediaUCI (^- T C / ); (c) wild-type (ss+); (d) Nasobemia (Ns); A I, A II, A III = first,
second and third antennal segment. Ar, Arista; Cl, claws; Co, coxa; Fe, femur;
St, sternopleura; Ta, tarsus; 77, tibia; Tr, trochanter.
Mutations in antennae o/Drosophila
461
Fig. 2. Fragments of the eye-antennal disc used for the experiments. A, Antennal
disc; E, eye disc; F, fragment of the eye disc.
used. Certain kinds of fragments tend to give rise to duplicated structures, others
are capable of regeneration (Gehring, \966a; Schubiger, 1971; Bryant, 1971;
Gehring & Nothiger, 1973). For transdetermination usually more extensive
proliferation is needed.
The kind of allotypic structures which are obtained after transdetermination
again depends on the specific fragment which is cultured (Hadorn, 1966;
Gehring, ]966a). Each fragment can give rise to one or a few types of allotypic
structures. Transdetermination of one cell type into another has also been
demonstrated in clones of genetically marked cells (Gehring, 1967), for example
palpus cells can give rise to wing cells, or cells of the third antennal segment can
become transdetermined and form tarsal cells.
Similar 'transformations' are also observed in situ in homeotic mutants.
However, the relationship between transdetermination and homeotic mutation
is not understood. In the present study we have analyzed discs from three
homeotic mutants with respect to their capacity to duplicate and regenerate
homeotic structures. Two of the mutants are alleles of spineless-aristapedia, ssa
(Balkaschina, 1929) and ssa'UCI (Vyse & James, 1972), which transform the
arista and adjacent parts of the third antennal segment into a four-jointed tarsus
(Fig. 1). These two mutants have complete penetrance and constant expressivity
which makes them highly suitable for such studies. Experiments on a temperature-sensitive allele of aristapedia will be reported separately (Schubiger, 1975).
The third mutant Nasobemia (Gehring, 19666) in the extreme case transforms
the antenna and adjacent parts of the head into a complete mesothoracic leg.
This mutant also has complete penetrance, but it is much more variable in its
expression and normally gives only partial transformation (Fig. 1).
The head structures of the fly are derived from three pairs of imaginal discs:
the eye-antennal discs, the imaginal cells of the clypeo-labrum, and the labial
discs (Gehring & Seippel, 1967). The eye-antennal disc (Fig. 2) is composed of
two morphologically distinct parts, the eye disc (E in Fig. 2) which gives rise to
462
W. J. GEHRING AND G. SCHUBIGER
one half of the head capsule including one compound eye, and the antennal disc
(A in Fig. 2) which forms the antenna and the maxillary palpus. The two parts
can easily be separated mechanically with a tungsten needle. The wild-type
antennal disc has been studied extensively with respect to its capacity for
duplication and regeneration (Gehring, 1966 a). It was found that in cultured
antennal discs, the antennal anlage frequently duplicates and forms two approximately symmetrical antennae. A disc fragment which contains only the antennal
anlage is capable of regenerating a palpus, but in no case was it found that the
antennal disc can regenerate eye structures. On the other hand, cultured eye discs
frequently regenerate an antenna (Gehring & Nothiger, 1973). Therefore, by
using the homeotic mutants mentioned previously, we can study the duplication
of a homeotic antenna from a preexisting antennal anlage and its regeneration
from cells of the eye disc. Previous work with in vivo cultures of antennal discs
homozygous for aristapedia has shown that after prolonged culturing the mutant
tissue is capable of producing both tarsal structures as well as aristae (Gehring,
1966a). This observation indicates that ssa cells contain the genetic information
for forming an arista which normally is not expressed in situ. Since tarsal structures can also arise by transdetermination, the earlier work did not answer the
question of whether the ssa mutation is also expressed in disc cells which undergo
additional divisions in culture. This latter question will be examined in the
present paper.
MATERIALS AND METHODS
The donor larvae were taken from homozygous aristapedia (ssa and ssa'UCI),
Nasobemia (Ns), and Oregon-R wild-type stocks (Lindsley & Grell, 1967; Vyse
& James, 1972). As hosts ssa and two wild-type stocks {Oregon-R and Sevelen)
were used. The Sevelen stock was marked with multiple wing hairs (mwh) and
ebony (e). The stocks were reared on standard Drosophila medium at 25 °C.
In the control experiments the disc material was taken from donor larvae at
about 100 h after egg deposition and transplanted into larvae of the same age,
using the technique developed by Ephrussi & Beadle (1936). The eye-antennal
disc was dissected from the donor larva in tricine-buffered Ringer's solution or
balanced saline (Chan & Gehring, 1971) and cut with fine tungsten needles
(Fig. 2).
In the duplication and regeneration experiments the discs from 'mature'
larvae (100 h) were transplanted into newly eclosed adult females, cultured for
five days or about 2 weeks, and subsequently transfered into larval hosts (80 h
after egg deposition). The metamorphosed cuticular structures were dissected in
Ringer's solution and mounted directly in Gurr's water mounting medium or
Faure's solution.
Mutations in antennae 0/Drosophila
463
RESULTS
Control experiments
The effect of transplantation on developing antennal discs from aristapedia
and Nasobemia mutants was examined by transplanting them into wild-type
larvae (see Materials and Methods). It has been shown previously by several
authors that aristapedia is expressed autonomously in wild-type hosts (Braun,
1940; Vogt, 1946) and also in genetic mosaics (Roberts, 1964). This was confirmed by our control experiments summarized in Table 1: both ssa and ssa'UCI
transplants form antennal tarsi. However, the number of tarsal bristles is
reduced particularly in ssa'UCI and to a lesser degree in ssa. This effect might be
due to a partial non-autonomy of the mutant discs in the wild-type hosts.
However, a reduction in the number of tarsal bristles is observed, when ssa discs
are transplanted into ssa hosts. Both in wild-type and in ssa hosts, the number of
tarsal bristles is not only reduced, but it also is much more variable as indicated
by the large standard errors. However, the first and second antennal segments
do not show any reduction in the number of bristles and only slightly larger
standard errors. Furthermore, the ssa transplants, especially in the area of the
third antennal segment and the tarsus, show signs of incomplete metamorphosis
and necrosis much more often than is observed in the corresponding area of
wild-type discs. Therefore, we may conclude that the aristapedia discs are more
sensitive to handling in vitro and transplantation than wild-type discs. In addition, the mutant expression may be somewhat reduced under the experimental
condition, i.e. in transplants there probably are more cells which differentiate
into arista structures than in situ. However, the mutant expression is constant
enough for the study of duplication and regeneration of the homeotic antenna.
In this context it is interesting to note that in situ the most distal part of the
arista is not always transformed into tarsal structures.
In starved animals from crowded cultures of ssa'UCI a strong reduction of
the aristapedia phenotype is observed. Under these conditions fewer leg bristles
are formed (Table 1) and the distal part of the appendage differentiates as an
arista. In transplantation experiments nerves and tracheas are disconnected,
which might lead to a similar effect as starvation.
The mutant expression in Nasobemia is much more variable, but it also
behaves autonomously in transplants (Gehring, 1970). Ns shows a similar
sensitivity to experimental manipulation as mentioned previously for the
aristapedia alleles.
Duplication experiments
When wild-type antennal discs (A in Fig. 2) are cultured first in an adult host
for 13-17 days and subsequently transplanted into a larva of the early third
instar, the antennal anlage frequently duplicates and forms two symmetrical
antennae (Gehring, 1966tf). Under these conditions transdetermination to tarsal
ssa-UCl
ssa-UCI
ssa-UCl
ssa-UCl
ssa-UCl
ss*
ss"
ssa
ss"
Donor
genotype
In situ
Transplanted
Transplanted
Duplicated
Regenerated
In situ
In situ starved
Transplanted
Transplanted
Regenerated
Experimental
condition
ss+
ssa-UOl
ss+
ss"
ss+
ss+
—
—
ss+
Host
genotype
—
—
17
16
—
—
—
—
5
Culture period
(days)
46
22
27
16
6
18
6
9
12
11
Number
of cases
4-3 + 0-4
3-4+1-1
3-2 + 0-8
3-8+1-8
1-7 ±0-6
5-2 ±0-7
5-4 + 0-7
4-3 ±0-9
4-3 ±1-4
3-6±0-5
1st antennal
segment
21-9 + 0-9
200 ±4-9
24-4 ±3-0
20-7 ± 5-4
17-4 ±2-8
26-7 ±2-8
24-3 ±2-1
25-9 ±4-2
26-8 ±2-2
24-8 ±4-5
2nd antennal
segment
K
Number of bristles
Table 1. Number of bristles on homeotic antennae formed under given experimental conditions
^—1
57-8 ± 3-9 o
44-2 ±22-4 >
490 ±10-3
21-9±13-6
15-2± 6-5 p
98-0 ±7-5
O
41-1+8-7
55-0±17-7
66-4 ±10-2
38-0 ±23-7
Homeotic
tarsus
O
w
IGER
Mutations in antennae o/Drosophila
465
Table 2. Phenotype of duplicated antennae in cultures offragment A
(in Fig. 2)
Genotype
of cultured
fragment
N u m b e r of
cultures
Culture period
(days)
N u m b e r of
duplications
Wild-type
SS*
Ns
24
30
29
16
17
14
9
12
14
Phenotype of duplicated
antennae
K
,
^
Normal
Homeotic*
9
0
10
0
12
4
* A n a n t e n n a is classified as h o m e o t i c if it carries at least 2 tarsal bristles.
structures is rare and in our experiments it was never observed. In 9 out of 24
cultures the antenna duplicated and formed two normal antennae (Table 2). The
same experiments performed with ssa yielded 12 duplicated antennae, all of
which showed the homeotic phenotype. However, in some cases only a few
tarsal bristles were found at the base of a fairly normal arista. Nevertheless, this
result clearly indicates that the ssa mutant phenotype is also expressed in duplicated antennae. In Ns four duplications were obtained in which both partners
were homeotic, and in 10 pairs only one of the partners was homeotic whereas
the other appeared to be normal. This observation can be explained by the
variable expression of Ns, and the 4 positive cases are in agreement with the
result obtained with ssa.
Regeneration experiments
The eye disc (fragment E in Fig. 2) when transplanted into a host larva of the
same stage as the donor, gives rise to structures of the head capsule and the
compound eye only. However, if we allow it to proliferate extensively in an
adult host, it will regenerate a complete antennal disc with an antenna and a
palpus anlage. If we compare wild-type and homeotic mutants with respect to
their capacity for this type of regeneration (Table 3), we find that wild-type discs
always regenerate normal antennae, whereas aristapedia (ssa and ssa'UCI) discs
give rise to homeotic antennae only (Fig. 3). In ssa'UCI, antennal regeneration is
obtained after five days in culture, a culturing period which does not allow
transdetermination. Ns is again intermediate presumably because of its variable
expression.
In order to make sure that the cultured fragment does not contain any
antennal cells which accidentally have not been removed, we cultured F fragments (Fig. 2) in the same way as E fragments. Regeneration in F fragments was
much rarer, but again the regenerated antennae differentiated according to their
genotype (Table 3).
466
W. J. GEHRING AND G. SCHUBIGER
100//m
<*•*
t\
Fig. 3. Regenerated antennae, (a) Aristapedia (ss"~UCl); (b) wild-type (ss+). All,
A III, Second and third antennal segment; Ar, arista; Ta, tarsal structures.
DISCUSSION
Homeotic mutants alter the determination of particular groups of cells such
that they differentiate according to a different developmental pathway. In the
present study we are primarily interested in the time of gene action of these
mutants or their wild-type alleles.
Several lines of evidence indicate that the bithorax (bx) gene complex which is
Mutations in antennae 0/Drosophila
467
Table 3. Phenotype of regenerated antennae in cultures of fragments
E and F (Fig. 2)
Genotype
cultures
Culture period
(days)
Wild-type
40
31
36
28
16
28
f~Niltiirf*H
fragment
E
ssa
ss*-uci
Ns
F
Wild-type
Phenotype of regenerated
antenna
Normal
Homeotic
13
16
5
14
17
0
0
10
0
il
7
6
16
16
2
0
0
2
involved in the determination of the thoracic and abdominal segments is already
expressed in embryonic stages of development. First, it was observed that homozygous Ultrabithorax larvae and several mutant combinations of the bx complex have, in addition to the normally present mesothoracic pair of spiracles,
both a metathoracic and a first abdominal spiracle pair (Lewis, 1964). Since these
spiracles are formed during embryogenesis, this observation suggests an early
action of these genes. Furthermore, the sensitive period for phenocopying bx
was found to be the blastoderm stage around 3 h of development (Henke &
Maas, 1946; Gloor, 1947), which strongly suggests that the bx gene complex is
involved in the initial determination of the thoracic segments during embryogenesis. Using the technique of induced somatic recombination, Lewis (1964)
has been able to show that the wild-type allele of bx is still functioning during
larval stages, as late as the late third instar, which indicates that determination
is under continued genetic control until metamorphosis sets in.
In aristapedia there is no evidence available on gene action in early development. Somatic recombination studies indicate that the wild-type allele must
function at least until 20 h before puparium formation (Roberts, 1964; Postlethwait & Schneiderman, 1973; Postlethwait & Girton, 1974). The temperaturesensitive period (TSP) for various aristapedia alleles also falls into the third
larval instar (Villee, 1943; Vogt, 1946; Grigliatti & Suzuki, 1971). The TSP
presumably corresponds to the time interval during which the gene product
controlled by the temperature-sensitive allele is biologically active, and does not
indicate the time of transcription or translation (Suzuki, 1970), but we can
assume that transcription and translation occur prior and/or during the TSP.
Whether the gene ceases to function after the TSP is an open question.
Our experiments aimed at the question of whether the homeotic genes are
active when metamorphosis is delayed and the imaginal disc cells go through
additional mitoses in culture. Both in duplicated and in regenerated antennae,
the antennae are at least partially transformed into leg structures. For the
30
EMB 33
468
W. J. GEHRING AND G. SCHUBIGER
duplicated antennae one might argue that the homeotic leg cells pass their
determination on to their daughter cells in the course of duplication (cell heredity). However, a recent study on the duplicating antennal disc of a temperaturesensitive homeotic mutant (Schubiger, 1975) clearly rules out cell heredity of the
state of determination as the mechanism for duplication in these experiments.
Furthermore, the interpretation which assumes cell heredity cannot be applied to
the regeneration experiments in which the cultured fragment does not contain
any homeotic leg cells. In this case we have to assume that determination is
under continued genetic control by the homeotic genes or their wild-type alleles
during proliferation in the adult host. The hypothesis of stable determinative
factors synthesized early in development and passed on in the course of proliferation is very unlikely since the regenerated antenna contains several thousand cells and such factors would be greatly diluted in the course of cell
multiplication. It seems much more probable that the gene product is synthesized
during regeneration.
The biochemical nature of the gene product remains unknown. The observation that after prolonged culture in vivo, ssa antennal discs besides producing
tarsal structures, can also produce normal antennae with aristae (Gehring,
1966fl) indicates that the mutant cells do not lack any of the structural genes for
forming an antenna, but rather that ssa is a controlling gene involved in the
alternative differentiation of antenna or leg structures. The present experiments
indicate that the ssa gene is expressed during proliferation in culture, but its
expression becomes more variable, and antennal as well as leg structures are
produced.
This work has been supported by the following grants: GB-17267 from the National
Science Foundation and 258 from the Jane Coffin Childs Memorial Fund for Medical
Research to W. Gehring and GB-38375 from NSF to G. Schubiger.
REFERENCES
BALKASCHINA, E. (1929). Ein Fall der Erbhomoosis (Die Genovariation 'aristopedia') bei
Drosophila melanogaster. Wilhelm Roux Arch. EntwMech. Org. 115, 448-463.
BRAUN, W. (1940). Experimental evidence on the production of the mutant 'aristapedia' by
a change of developmental velocities. Genetics Princeton 25, 143-149.
BRYANT, P. (1971). Regeneration and duplication following operations in situ on the imaginal
discs of Drosophila melanogaster. Devi Biol. 26, 606-615.
CHAN, L. N. & GEHRING, W. (1971). Determination of blastoderm cells in Drosophila
melanogaster. Proc. natn. Acad. Sci. U.S.A. 68, 2217-2221.
EPHRUSSI, B. & BEADLE, G. W. (1936). A technique for transplantation for Drosophila.
Am. Nat. 70, 218-225.
GEHRING, W. (1966a). Uebertragung und Aenderung der Determinationsqualitaten in
Antennenscheiben-Kulturen von Drosophila melanogaster. J. Embryol. exp. Morph. 15,
77-111.
GEHRING, W. (19666). Bildung eines vollstandigen Mittelbeines mit Sternopleura in der
Antennenregion bei der Mutante Nasobemia (Ns) von Drosophila melanogaster. Arch.
Julius Klaus-Stift. Vererb-Forsch. 41, 44-54.
Mutations in antennae 0/Drosophila
469
GEHRING, W. (1967). Clonal analysis of determination dynamics in cultures of imaginal discs
in Drosophila melanogaster. Devi Biol. 16, 438-456.
GEHRING, W. J. (1970). Problems of cell determination and differentiation in Drosophila. In
Problems of Biology: RNA in Development (ed. E. Hanly), pp. 231-244. Salt Lake City:
University of Utah Press.
GEHRING, W. J. & NOTHIGER, R. (1973). The imaginal discs of Drosophila. In Developmental
Systems: Insects. Vol. 2 (eds. S. Counce & C. Waddington), pp. 211-290. London and
New York: Academic Press.
GEHRING, W. & SEIPPEL, S. (1967). Die Imaginalzellen des Clypeo-Labrums und die Bildung
des Riissels von Drosophila melanogaster. Revue suisse Zool. 74, 589-596.
GLOOR, H. (1947). Phanokopie-Versuche mit Aether an Drosophila. Revue suisse Zool. 54,
637-712.
GRIGLIATTI, T. & SUZUKI, D. T. (1971). Temperature-sensitive mutations in Drosophila
melanogaster. VIII. The homoeotic mutant ss."4Oa Proc. natn. Acad. Sci. U.S.A. 68,
1307-1311.
HADORN, E. (1963). Differenzierungsleistungen wiederholt fragmentierter Teilstiicke mannlicher Genitalscheiben von Drosophila melanogaster nach Kultur in vivo. Devi Biol. 7,
617-629.
HADORN, E. (1966). Dynamics of determination. In Major Problems in Developmental Biology
(ed. M. Locke), pp. 85-104, New York and London: Academic Press.
HENKE, K, & MAAS, H. (1946). Ueber sensible Perioden der allgemeinen Korpergliederung
von Drosophila. Nachr. Akad. Wiss. Gottingen. Math. phys. Kl. 1, 3-4.
LEWIS, E. B. (1964). Genetic control and regulation of developmental pathways. In The Role
of Chromosomes in Development (ed. M. Locke), pp. 231-252. New York: Academic
Press.
LINDSLEY, D. L. & GRELL, E. H. (1967). Genetic variations of Drosophila melanogaster.
Carnegie Instn Wash. Publ. 627.
POSTLETHWAIT, J. H. & SCHNEIDERMAN, H. A. (1973). Developmental genetics of imaginal
discs. A. Rev. Genetics 7, 381-433.
POSTLETHWAIT, J. H. & GIRTON, J. R. (1974). Development of genetic mosaics of aristapedia,
a homoeotic mutant of Drosophila melanogaster. Genetics Princeton 76, 161-11 A.
ROBERTS, P. (1964). Mosaics involving aristapedia, a homoeotic mutant of Drosophila
melanogaster. Genetics Princeton 49, 593-598.
SCHUBIGER, G. (197.1). Regeneration, duplication and transdetermination in fragments of the
leg disc of Drosophila melanogaster. Devi Biol. 26, 277-295.
SCHUBIGER, G. (1975). Regeneration and duplication in a temperature-sensitive homeotic
mutant of Drosophila melanogaster. (In preparation.)
SUZUKI, D. T. (1970). Temperature-sensitive mutations in Drosophila melanogaster. Science
N. Y. 170, 695-706.
VILLEE, C. (1943). Phenogenetic studies of the homoeotic mutants of Drosophila melanogaster. I. The effects of temperature on the expression of aristapedia. J. exp. Zool. 93,
75-98.
VOGT, M. (1946). Zur labilen Determination der Imaginalscheiben von Drosophila. II.
Unwandlung praesumptiven Fiihlergewebes in Beingewebe. Biol. Zbl. 65, 238-254.
VYSE, E. R. & JAMES, A. (1972). ss"'ocl: spineless aristapedia. Drosoph. Inf. Serv. 49, 39.
(Received 14 July 1974)