Download the genetic and cytogenetic localization of the three structural genes

THE GENETIC AND CYTOGENETIC LOCALIZATION OF THE THREE
STRUCTURAL GENES CODING FOR THE MAJOR PROTEIN
O F DROSOPHILA LARVAL SERUM
DAVID B. ROBERTS
AND
SUSAN EVANS-ROBERTS
Genetics Labonatory, Biochemistry Department,
South Parks Road, Oxford OX1 3QU England
Manuscript received August 1, 1979
Revised copy received October 11, 1979
ABSTRACT
The a, p and y polypeptides that make up Drosophila Larval Serum
Protein-I seem to be coded for by genes that have evolved by duplication of
a common ancestral gene. We have found variants of all three polypeptides,
and these are variants of the coding sequences. The a-chain variant mapped to
39.5 on the X chromosome and to the polytene interval 1IA7-11B9. The
p-chain variant mapped to 1.9 on chromosome 2L and to 21D2-22A1. The
y-chain variant was mapped as 0.13 map units from the tip of chromosome
3L or to -1.41 with respect to ru, which has been defined as 0.0, and to
61AI -6 1A6.
HE larval serum proteins (LSP-1 and LSP-2) are the major protein comTponents of the haemolymph of Drosophila larvae shortly before pupariation
(ROBERTS.
WOLFEand AKAM1977). These proteins are being studied in order
to understand the mechanism of their control. As part of this program, we have
mapped the coding sequence for LSP-2 to 37.0 and 6832-4 on chromosome 3
(AKAM
et al. 1978) ;in this paper, we describe the genetics of the coding sequences
for the LSP-1 polypeptide chains.
LSP-1 is a family of proteins made up of an apparently random association into
hexamers of three polypeptides that have similar amino acid compositions and
share common antigenic sites (WOLFE,AKAMand ROBERTS
1977). The common
antigenic sites must be due to common amino acid sequences as this protein has
no associated carbohydrate ( CRUMPTON,
personal communication) or lipid (ROBERTS, unpublished results). The three polypeptides differ slightly in molecular
weight, and the heaviest is designated as the a chain, the middle molecular weight
form the chain, and the lightest the y chain. The similar amino acid composition and the common antigenic sites suggest that these three chains have evolved
by duplication of an ancestral gene, followed by mutation. The similarity of both
one- and two-dimensional peptide maps of the subunits following partial and complete digestion of the 01, @ and y polypeptides support this view. BROCKand ROBERTS, in preparation).
Our first step was to identify variants of the three polypeptide chains. These
were shown to be variants of the coding sequence by quantitative studies (ROBGenetics 93 : GG3-679 November. 1979
664
DAVID B. ROBERTS A N D S U S A N EVANS-ROBERTS
and EVANS-ROBERTS
1979; AKAMet al., in preparation). The variant alleles
have been mapped using standard genetic techniques and have been mapped
cytogenetically, using existing deficiencies or by generating segmental aneuploids
(LINDSLEY
et al. 1972).
ERTS
MATERIALS A N D METHODS
Stocks: The Drosophila strain Oregon-R was used as the reference wild-type strain in this
work and, unless indicated otherwise, the LSP-1 of all stocks was the same as Oregon-R. Stocks
were maintained on standard yeast-cornmeal-agar medium in half-pint milk bottles at 25". The
other wild-type stocks were either raught locally or obtained from stock centers around the
world as listed in Drosophila Information Service. The wild-type stocks that eventually yielded
variants were obtained as follows: Bacup, Genetics Institute, Groningen University, The Netherlands; Formosa, Zoological Institute, Munich University, Germany; Kochi-R, Department of
Genetics, U m e l University, Sweden; Ponza, Biological Institute, Vienna University, Austria;
Stromsvreten-11 caught in Akersberga, Sweden.
The mutant stocks used in this work were obtained as follows: C y / P m ; Sb/D, v 2 f w , w y ,
Df(l)mz59-4 ras f / v f B S - Y y f / y w f : = , Df(l)JA26/FM7, Df(1)NlZ ras v/FM6. Df(l)C246/
FM6, Df(l)HA92/FM7, Df(l)HF368/FM7, Df(ZL)al/In(gL)Cy, In(ZR)Cy, E(S)Cy cnz,
Df(Zt)Sz/In(2L)Cy In(ZR)Cy, E(S)Cy cnz, Df(ZL)SS, S / S M j , T(1;Z ) v 6 5 b lz50e/y w f : = ,
T(Y;3)P3, T(Y;3)H141 (California Institute of Technology), sc ec cv cPv gz f , ru h th st cu sr e8
ca (rucuca, Bowling Green, Ohio), y ; Dp(1)scJ~iuira red/TMl mrvh ve (MRC Cambridge),
and the segmental aneuploid stocks: R132, R108, B71 and S50 (LINDSLEY
et al. 1972) (UCSD,
La Jolla, California). The other stocks used have been in our possession for over five years.
For details of all mutant stocks, see LINDSLEY
and GRELL(19G8).
Haemolymph collection: In order to collect haemolymph for analysis from india idual larvae,
flies were allowed to lay eggs on yeast-glucose-agar medium (10% yeast, 10% glucose, 3%
agar
0.1% Nipagin) seeded with fresh yeast suspension. Haemolymph was collected from
larvae reared on this medium, as described previously (ROBERTS,
WOLFEand AKAM1977). In
some experiments it was necessary to analyze the haemolymph of female larvae only. in which
case the larvae were sexed according to the size d the gonads (BODENSTEIN
1950). For the experiments described here, wandering larvae (within 12 hr of pupariation) were used.
Cellogel electrophoresis: Cellogcl clectrophoresis was carried out according to the manufacturers' instructions (Reeve Angel Scientific, Limited). The buffers used were: 0.05 M p H 7.0
phosphate; and 0.05 M pH 9.3, barbitone/HCl. A potential difference of 6v/cm was applied to
the gel for 3 hr at 4". The gels were stained with amido schwarz 10B and washed in 5% acetic
acid in 47.5% methanol.
Polyacrylamide gel electrophoresis: Dodecyl sulphate polyacrylamide gel electrophoresis was
carried out according to the technique described by L ~ E M M L(1970)
I
with a 1 cm 3% acrylamide
stacking gel and 8% acrylamide resolving gel in the BioRad model 220 vertical slab gel apparatus. The apparatus was cooled during electrophoresis by running tap water. and the gels
were run at a constant current of lOmA per gel until the bromophenol blue tracking dye reached
the end. They were stained mith 0.1% coomassie blue in 50% trichloracetic acid and destained
i n a 25% ethanol, 7% acetic acid mixture. Tris-borate polyacrylamide gel electrophoresis was
carried out i n the Biorad model 220 vertical slab gel apparatus. The 30% acrylamide (29%
acrylamide, 1% N,N'-methylene-bisacrylamide) in 0.06 M tris adjusted to p H 8.6 with 2M
citric acid was polymerized with 0.05% persulphate and 5 mM N,N,N,N-tetramethylethylenediamine. The electrode buffer was 30 mM sodium tetraborate. The gels were run at a constant
current of 25 mA for 2 h r at 4".
For two-dimensional electrophoresis, haemolymph samples were first electrophoresed in
cellogel. A strip of the cellogel 2 mm wide was cui out and the remaining frame stained to show
the distribution of bands in the first dimension The strip was layered in molter. agar (1.2%)
on top of the stacking gel in the slab gel apparatus and electrophoresis was carried out as
described above.
+
GENETICS O F DROSOPHILA
LSP-1
665
RESULTS
Search for variants: When wild-type larval haemolymph is electrophoresed
on Cellogel at pH 7.0, LSP-1 migrates as a single band (ROBERTS,
WOLFEand
AKAM1977), but at pH 9.3 it migrates as a triplet of bands that can be identified
by cross electrophoresis into a dodecyl sulphate polyacrylamide gel (Figure 1).
The most rapidly migrating band on cellogel is the a chain, and the most slowly
migrating band the y chain, and we have shown that, at high pH, LSP-1 dissociates into monomers (see below and BROCKand ROBERTS,
unpublished results).
We analyzed haemolymph from late third-instar larvae of 48 different wildtype stocks on Cellogel in 0.05 M pH 7.0 phosphate buffer (LSP-1 a single band)
and in 0.05 M pH 9.3 barbitone buffer (LSP-1 a triplet). Because of the likelihood
that the stocks were heterogeneous, four larvae from each stock were tested at each
pH. In the first analysis, a number of stocks gave patterns unlike Oregon-R, but
on retesting only five stocks consistently gave some larvae that did not have the
Oregon-R pattern in one or other of the buffer systems. I n no case did all the
larvae from a variant stock give the same pattern, so that it was necessary to inbreed the variant stocks. This was done by setting up a number of pair matings
and keeping the sibs of ten larvae, all of which gave the same variant pattern.
These stocks were retested many times before they were actually established and
used for further study. In this way we established six variant stocks, Ponza G8,
Ponza Gla, Stomsvreten 11A, Formosa M, Bacup 3D and Kochi-RB1.
I n a search for variation in the amount of LSP-1 produced by different stocks
(ROBERTS,
BLACKWELL
and AKAM,unpublished results), a low LSP-1 producer
was found to lack the y chain. This stock carried the recessive chromosome 3
marker, eyg. During the course of this study, we found that the balancer chromosome TM6 in the stock S50/TM6 also carried the y-chain null allele. When
tested, both these stocks behaved, with respect to the y chain, like the Ponza G8
stock.
The electrophoretic pattern of the variant stocks and of the variant Oregon-R
heterozygotes are shown in Figures 2, 3 and 4 compared with the reference pattern of Oregon-R. With the exception of Bacup 3D, all of the variant stocks heterozygous with Oregon-R gave a pattern on Cellogel unlike the variant parent. The
pattern of the Bacup/Oregon heterozygote is, under optimum conditions, unlike
either parental pattern, but as the two parental patterns are similar under the
standard gel conditions, it is frequently difficult to distinguish the heterozygote
in anything other than the best gels. Thus, the alternative tris-borate acrylamide
gel was developed to analyze the Bacup variant.
The Kochi-RB1 and Stromsvreten 11A stocks carry slow electrophoretic variants of the 01 chain, while Formosa NI carries a fast electrophoretic variant. The
Ponza G8 stock clearly lacks the y chain; it is a y-chain null allele. Ponza G8/
Oregon-R heterozygotes synthesize only 50% of the y chain synthesized by
Oregon-R (unpublished results). Ponza G8 also carries a fast allele of the 01 chain,
but this is not analyzed here. The Ponza Gla stock is complicated, and we believe
that it might be heterozygous for the y-chain null allele, the heterozygotes being
666
-
DAVID B. ROBERTS A N D SUSAN EVANS-ROBERTS
p~ 9.3 Cellogel
r
n
N
FIGURE1.-Correlation of LSP-I bands on Cellogel with LSP-1 bands on SDS acrylamide
gels. Haemolymph samples from larvae were run on Cellogel at pH 9.3 and cross electrophoresed into SDS gels. The a, p and y polypeptides of LSP-1 and the LSP-2 polypeptide are
indicated. (a) Oregon-R; (b) Kochi-RBI. (Note the reduced amount of the slowmigrating a
chain of Kochi-RB1.)
maintained by a balanced lethal system. We do not consider this stock further
here.
The /3 chain in Bacup 3D is a fast-moving allele of the wild-type j3 chain, and
in the tris-borate gel system the Bacup/Oregon heterozygote shows both /3 chains
(Figure 3).
Genetic localization of the a-chain structural gene: Using standard genetic
667
GENETICS OF DROSOPHILA LSP-I
hl
t
i
tn
i
i
OR-R
PONZA G8
OR-R
PONZA G l a
OR- R
-
KOC H I R 61
OR-R
Sll-A
OR-R
BACUP 3 D
OR-R
p
#
10
OR-R
OR-R
/
S11-A
Sll-A
OR-R
1 -/
OR R P O N ' Z A G ~
PONZAG8
0 R-R
OR-R
/
FORMOSA-M
FORMOSA- M
L S G &2
FIGURE2.-The electrophoretic pattern of variant and heterozygous stocks. (a) and (b)
Larval haemolymph run on Cellogel at pH 9.3. (c) Larval haemolymph run on Cellogel at
pH 7.0. OR-R is Oregon-R and Sll-A is Stromsvreten-l1A. Note in Figure 2a that the slow a
band of Kochi-RB1 ovcrlaps the p band, giving a diffuse band.
crosses, the a-chain variants of Kochi-RBl, Strijmsvreten l l A and Formosa M
were localized on the X chromosome (Figure 4). For the more precise localization
of the Kochi-RB1 variant, the protocol described in Figure 5 was used. The first
experiment made use of a series of recombinants in which the recombination
event had taken place in different intervals. The results are shown in Figure 5.
This established that the variant locus was between U and g'. I n the second experi-
668
DAVID B. ROBERTS A N D SUSAN EVANS-ROBERTS
s P1
-
4L S P 2
FIGURE3.-Electrophoretic pattern of the Bacup 3D variant. Haemolymph from Oregon-R,
Bacup 3D and Oregon/Bacup heterozygous larvae was electrophoresed in the tris-borate gel
system. Under these conditions, LSP-1 is dissociated into its three polypeptides. The figure
shows the fast allele of the /3 chain in Bacup 3D and both forms of the /3 chain in the heterozygote.
ment, twenty males with crossovers in the v-g2 interval were crossed to KochiRBI females, and 10 larvae from each cross were examined in Cellogels. The results of these twenty crosses and the eight from the previous experiment are given
in Figure 5.
I n a third experiment involving the Formosa M variant, nineteen recombinant
males were tested as before, and the results are also given in Figure 5. These results are consistent with the suggestion that both Kochi-RBI and Formosa M
are variants of the same structural gene so that the data from all three experiments have been grouped to give the map position for the variant locus as 39.5 on
669
GENETICS O F DROSOPHILA LSP-1
$
1)
Fld
KOCH I - R B 1
FIQURE
$.-Localization of the U chain variant on the X chromosome. Kochi-RBl females
were crossed to Oregon-R males. Haemolymph from male and female larvae was electrophoresed
on Cellogel at pH 9.3. Male larvae had the Kochi-RB1 pattern, establishing the X-linkage of
the a chain.
the X chromosome, with 95% confidence limits from the binomial distribution of
37.8 and 41.2 map units, taking the map positions of v and g' to be 33.0 and 44.4
map units, respectively (LINDSLEY
and GRELL1968).
Cytogenetic localization of the a-chain structural gene: A series of overlapping
deficiencies of the region between v and g2 was supplied by the stock center a t
Pasadena from the collection of X-chromosome deficiencies deposited by G. LEFEVRE. These were tested to see whether the Df/Kochi-RBl heterozygote uncovered the variant pattern or gave the heterozygous pattern. The protocol was as
shown in Figure 6.
flies of the genotype Df(l)mPSs-b
ras f/v+B"Yy+
I n the case of Df(I)mpss-b,
were crossed to Kochi-RBl females. The female progeny will have the Kochi-RBl
pattern if the variant locus is uncovered by the deficiency; otherwise they will
have the heterozygous pattern. Of the six deficiencies tested, only JA26 and
HF368 uncovered the variant locus.
We analyzed Kochi-RBl against the deficiencies generated in the two insertional translocation stocks, T(I;2)vGSb
and T ( 2 ; 3 ) ~ + ~T(I;2)vGSb
4~.
did not uncover the variant locus, but T ( Z ; ~ ) V +did.
~ ~The
" results giving the cytological
position of the a-chain structural gene are shown in Figure 7. We placed the achain structural gene between i iA7 and 1lB9, an interval of 14 bands.
I n a second series of experiments, the variants Striimsvreten 11A and Formosa
M were tested against Of(2)HF368, and both were uncovered by this deficiency.
Genetic localization of the 8-chain structural gene: Using standard genetic
crosses, the &chain variant in the Bacup 3D stock was localized on chromosome 2.
Using techniques similar to those described for mapping the a-chain variant, we
mapped the /3-chain variant between a1 and dp. For a more precise localization,
670
DAVID
B. ROBERTS A N D SUSAN EVANS-ROBERTS
E eCCg6yq2! db
9 9 Kochi-R 91 x
K o c h c x sc ec C V Ct6y a’!
99
sc e_c c_vct6ysz;
9 9 Kochi-RB1
+
bb
I
66
x recombinant
e.g. +++++g2!
1 0 3rd. instar larvae tested
L S P l p a t t e r n of
male phenotype
99 Kochi-RE1 x sccc_vct6vgzf
+
+
+
+
+
+
+
c_v Cte; g2i
+ +
9 progeny
heterozygous
heterozygous
heterozygo u s
heterozygous
heterozygous
h e t e r o z y g o u s & Kochi-RB1
Kochi-RB1
Kochi-R 9 1
g’i
& g25
+ + + Ig2i
+ + + +g2i
+ + + ++!
+ + + + *++
L S P l pattern of 1 0 p r o g e n y
-
male
recombinants a l l
C.O.
variant
C.O.
phenotype
tested
variant interval heterozygous interval
Kochi-RBlx + + + + +&I
6
+
-sceccvcty+
__-
14
14
4
FormosaMx+ + + + + &
sceccv&+;
14
5
6
1
2
sc ec cv ct6 v
+ +
1
7
+ +
1
10
I
I
7
II
1
8
I
3
I
II
a+@$
+var
27
+ +
20
crossovers
v = 3 3 . 0 map u n i t s
g2= 4 4 . 4 map u n i t s
var.= 3 9 . 5 map units
95% limits 37.8- 41.2 m a p u n i t s
FIGURE
5.-Protocol for mapping the a chain variant. Haemalymph from the larvae was
tested on Cellogcl at pH 9.3.
the protocol described in Figure 8 was used. Twenty-eight crossovers in the al-dp
region were crossed to Bacup 3D females, and 10 larvae from each cross were examined on Cellogels. The results of these crosses are given in Figure 8. The ,&
chain variant was mapped to 1.86 on the left arm of chromosome 2, with 95%
confidence limits from the binomial distribution of 0.52 and 4.24 map units,
taking the map positions of a1 and d p as 0.1 and 13.0 map units, respectively
(LINDSLEY
and GRELL1968).
Cytogenetic localization of the @-chainstructural gene: Deficiencies close to the
p-chain structural gene (based on the mapping data) were made heterozygous
with the variant chromosome, and ten larvae were analyzed on tris-borate gels to
see whether a variant or heterozygous pattern was formed. Deficiencies Df (2L)Sz
671
GENETICS O F DROSOPHILA LSP- 1
FM7
Df(1)
Koc h i - R 6 1
Y
Y
d'
let
ha I
9
Expected p a t t e r n g i v e n by 1 0 9 l a r v a e
Kochi-R 6 1
FM 7
5 variant p a t t e r n
5 heterozygous p a t t e r n
if variant is uncovered
b y deficiency
10 heterozygous p a t t e r n
i f variant is not
uncovered by deficiency
FIGURE
6.-Protocol for the cytogenetic localization of the a-chain gene. Haemolymph from
the larvae was tested on Cellogel at pH 9.3.
and Df(2L)S3both uncovered the variant gene. This placed the &chain structural
gene between 21D2 and 22A1, an eleven band interval (Figure 9.)
Genetic localization of the y-chain structural gene: The null allele of the y
chain in the variant Ponza G8 stock was localized on chromosome 3, using standard genetic crosses. For a more precise localization, the protocol described in
Figure 10 was used.
The first experiment used a series of recombinant males in which the recombination event had taken place in different intervals. The results placed the variant
locus to the left of h. In a second experiment, forty-three recombinants in the
CZ46
1IA;~I4~2.1;
I I/\ A
l;l
10
1%
AI! I A I A ; ~ A I i i t A A A
1 0 11
i
l!;AIk;t! ;
--+
I
1 $I
,
~
.
~
,
~
~
~
1112
LSPl(Y
FTGURE
7.-Cytological positions of the a-chain gene and of the breakpoints of the deficiencies (LEFEVRE,
personal communication). (After BRIDGES
1938.)
,
~
~
~
~
6 72
B. ROBERTS A N D SUSAN EVANS-ROBERTS
DAVID
9 9 Bacup 3 0 x al d
-p 66
9 9 Bacup 3D x -a l dp 66
a l dp
+
a
-
--
I
+
99Bacup3D x a l + o r + dp bb
+I
10 3rd. instar larvae tested
2 8 recombinants
al
+
dP
+
va r
+
4
24
c r os sove r s
a l = O . l map units
dp= 13.0 map units
var =1.86map units
%%limits 0 . 5 2 - 4.24 m.u.
FIGURE
S.-Protocol
for mapping the p-chain variant. Haemolymph from the larvae was
tested on Cellogel at pH 9.3.
interval hu-h were examined. Ten of these were of the class 4-h th st sr es ca
and all carried the variant allele. The remaining thirty-three were all IU
4and none carried the variant allele. TU maps at 0.0 and h at 26.5
map units on the left arm of chromosome 3 (LINDSLEY
and GRELL1968), which
places the variant locus either to the left of TU or, with 95 % confidence limits
from the binomal distribution, with 2.1 map units to the right of ru.
+ ++++
673
GENETICS O F DROSOPHILA LSP-1
s3
I
I
4
LSP1
p
FIGURE
9.-Cytological positions of the p-chain gene and of the breakpoints of the deficiencies. (After BRIDGES
1938.)
Although ru is defined as 0.0 on the standard chromosome 3 map, it is known
from cytological and genetic data to be some distance from the tip of 3L, a minimum of 23 bands and four map units, respectively (LINDSLEY
and GRELL1968).
In order to map the variant locus to the left of +U, we constructed the stock y ;
Dp(Z)sd4ue/TM1 mwh, which marks the tip of 3L with yf ue. In constructing
this stock, the frequency of crossovers between the tip of 3L and ue was 1.7%
PP
PonzaG8 x rucuca
99
rucuca
dd
ss
x recombinant
e.g. ru-h-t h + + + + +
1
lolarvae tested for L S P 1 p a t t e r n
FIGURElO.-Protocol for mappicg the y-chain variant. Haemolymph from the larvae was
tested on Cellogel at pH 9.3.
6 74
DAVID
B. ROBERTS A N D SUSAN EVANS-ROBERTS
- 0.5
-0.3
0.0
0.0
0.0
1.04
+
1.24
+
1-54
0.2
0.2
1.7
fap
mwh
ru
ve
- 4.0
0
+
+
+
Robertson&Riviera(l972)
Lindsley&Grel l(1968)
present study
FIGURE
11.-Revised genetic map of the tip of chromosome 3L.
(51/3004). The distance between the tip and mwh from the same study was 1.24
map units, and between m w h and w ,0.46 map units. (cf., 0.5 map units, ROBERTSON and RIVIERA1972). Combining the data of ROBERTSON
and RIVIERAwith
the present data allows us to redraw the tip of 3L as in Figure 11. This result,
while in good agreement with ROBERTSON
and RIVIERA,
does not agree with the
distance of 4.0 map units between the tip of 3L and ru ched by LINDSLEY
and
GRELL(1968).
The protocol for mapping the y-chain variant to the left of r u is shown in Figure
12. These results place the y-chain structural gene at 0.13 map units from the tip
of 3L, with 95% confidence limits of 0.01 and 0.41 map units. The cytological
position f o r ue is given as 61E2-62A6, but as it maps to the right of T U , whose
cytological position is 61F5-62A3, the true cytological position of ue must be
between 61F5 and 62A6. The recombination frequency in this region lies between
0.043 and 0.053 map units per band. This compares with a value of 0.1 map units
and
per band calculated for the ru to h interval from the data given in LINDSLEY
GRELL(1968).
Cytogenetic localization of the y-chain structural gene: To localize the 7-chain
structural gene cytologically, we made use of the segmental aneuploid stocks gen-
.)
y+p+or-ve+ or
J"+oyve
dorQ
v
v+ ve
(phenotype y've')
ve
x PONZA G 8 ( Y - )
I
(phenotype yve)
1
10 larvae examined
Results of 27 recombinants
Parental
phenotype
Heterozygous
pattern
Ponza
pattern
y+ve+
13
2
Y ve
12
-
FIGURE
12.-Protocol for mapping the y chain variant to the left of ue. Haemolymph from
the larvae was tested on Cellogel at pH 9.3.
GENETICS O F DROSOPHILA
LSP-1
675
erated by LINDSLEY
et al. 1972. This approach was made particularly easy by the
fact that heterozygous deficiencies for a considerable length of 3L are viable. The
technique used for this analysis is illustrated in Figure 13. All of the segmental
aneuploid stocks used, including the stock generating the shortest deletion of the
tip of 3L, S50, uncovered the y-chain null allele. This places the structural gene
between the tip (61Al) and 61A6, which is in the 6-band interval, and is in
good agreement with the genetic localization. These results are summarized in
Figure 14.
In a second series of experiments, we used two duplications, T(Y;3)H242 and
T(Y;3)P3, which have 61B-62D and tip of 3L to 61E-F inserted into the Y
chromosome. The experimental design is illustrated in Figure 15. I n this experiment, we demonstrated that the y-chain coding sequence must lie between the tip
T(Y;S)P3-but not in the region 61B-62D7 i.e., not in
and 61EF,-on
T(Y;3)H24Z7which agrees with the results using deficiencies. These results are
summarized in Figure 14.
DISCUSSION
The genetic data presented here show that the genes for the LSP-1 ,a,fi and y
chains map to different chromosomes, although we believe that these genes have
descended from a common ancestral gene. Moreover, quantitative studies on the
synthesis of these proteins suggest that they are coordinately synthesized and that
in the female they are produced in equimolar amounts (POWELL,
SATOand ROBERTS, in preparation), whereas in the male only 50% of the amount of the X Bs
9 9 PonzaG8
Y+
-L
x XY
-TM6
1
assuming adjacent I disjunct ion
x"v
L
XP
-G8
- X
-
- pattern
if null
allele is uncovered
examine
lo 99
' larvae
PG8
4 -TM6
X
9
y
____o___
PGB
d
FIGURE
13.-Protocol f o r the cytogenetic localization of the y-chain gene using segmental
aneuploids. Haemolymph from the larvae was tested on Cellogel at p H 9.3. Note that the TM6
chromosome used in this experiment carries the y+ allele.
676
DAVID
B. ROBERTS A N D SUSAN EVANS-ROBERTS
Dp( T ( Y :3)H141)
D-
61162
FIGURE
14.-Cytological position of the y-chain gene and of the breakpoints of the deficien1941.)
cies and duplications. (After BRIDGES
linked a: chain found in females is produced (ROBERTS
and EVANS-ROBERTS
1979).
Since gene duplication was first observed in Drosophila (BRIDGES
1936; MULLER 1936), its significance in evolution has been discussed on many occasions
(HOROWITZ
1945; OHNO1970). Duplicated genes fall into two classes, those that
have been duplicated by some mechanism such as unequal crossing over and have
remained closely linked, forming a gene cluster, and those that have been duplicated but are scattered through the genome. The scattering of these duplicates
may have been brought about by translocation after duplication, or they may
have been scattered as a consequence of the duplication, for example by the incorporation of a duplicated part or all of a chromosome into the genome. Another possible mechanism for gene duplication in Drosophila would be if a gene were
and RAMEL1976) to a
moved in association with a transposing element (ISING
second chromosome; from this fly, a stock with four copies of the gene could be
selected. As there is no obvious similarity in the banding pattern of the regions
carrying the LSP-1 genes, it is unlikely that gross chromosomal translocations
or duplications account for this scattering. Thus, this last suggestion involving
transposing elements may be particularly appropriate in this case.
Gene clusters have been studied extensively (see review by BODMER
1979);
perhaps the best example at the molecular level is the cluster of genes coding
for the 8, y, 6 and E chains of haemoglobin. Haemoglobin is in fact the best example for comparison with the cy, @ and y chains of LSP-1 because the gene coding
GENETICS O F DROSOPHILA LSP-1
677
-
YDp. Ill&f
X ’ 111
X
-
9 9 X. PG8
x YD? PG86Cf
X ’ PG8
X PG8
-3-
1
test 5 9
5 d larvae
FIGURE
15.-Protocol for the cytogenetic localization of the y-chain gene using duplications.
Haemolymph from larvae was tested on Cellogel at pH 9.3.
for the a chain of haemoglobin, which associates with one of the other chains to
form an active molecule, is in a different linkage group. In discussing this, BODMER
(1979) pointed out that “in all those cases where the genetics of dimeric proteins
that are made up of different subunits is sufficiently well advanced, the genes for
the two different chains are on different chromosomes, o r at least very loosely
linked.” The 01, /3 and y chains of LSP-1 fall into this category.
We have suggested that the dispersal of the LSP-1 genes may be “recent,” as the
a-chain gene on the X chromosome has not yet acquired dosage compensation
(ROBERTSand EVANS-ROBERTS
1979), and we are studying the evolution of these
proteins by comparing the proteins of different Drosophila species and comparing
the distribution of their coding sequences by in situ hybridization with cDNA
prepared from the mRNAs.
It is interesting t o speculate why the coding sequences for polypeptides that
associate to form an active protein are dispersed. Especially interesting is the case
678
DAVID B. ROBERTS A N D SUSAN
EVANS-ROBERTS
of the LSP-1 a, ,8 and y chain genes, as we believe that their polypeptides are coordinately synthesized. The coordination of synthesis of these polypeptides is not
achieved as in bacteria by the close association of their coding sequences. BODMER
(1979) has observed that in many cases only one gene in a cluster is expressed at
any one time, and that for two related genes to be expressed simultaneously, the
strategy for control in higher eukaryotes may demand that they be unlinked (or
no more than loosely linked). If this is so, then a study of the control of the LSP-1
polypeptides in Drosophila, an organism that is amenable both to classical genetic
analysis and to techniques involving recombinant DNA, will go a long way to test
this hypothesis.
We thank the stock centers mentioned for supplying LIS with stocks, and in particular we
for anticipating our needs. We should also like to thank
would like to thank LORINGCRAYMER
JEANMATTHEWS
for preparing the Drosophila medium, MICHAELAKAMfor reading through
the first draft of this paper, our colleagues for allowing us to mention their unpublished work
and ANN HARRIS
for the initial screening of wild-type stocks for the variants. This work was
supported in part by a grant from the Science Research Council.
LITERATURE CITED
AKAM,M. E., D. B. ROBERTS,
G. P. RICHARDSand M. ASHBURNER,
1978 Drosophila: The
genetics of two major larval proteins. Cell 13: 215-225.
BODENSTEIN,
D., 1950 The postembryonic development of Drosophila. Chapter 4.In: Biology of
Drosophilu. Edited by M. DEMEREC.
John Wiley and Sons, New York.
BODMER,W. F., 1979' Gene clusters and the HLA system. CIBA Foundation Symp. 66.
C. B., 1936 The Bar gene duplication. Science 83: 210-211.
BRIDGES,
BRIDGES,
C. B., 1938 A revised map of the salivary gland X chromosome of Drosophila Enelanoguster. J. Heredity 29: 11-13.
P. N., 1941 A revised map of the left limb of the third chromosome of Drosophila
BRIDGES,
melanogaster. J. Heredity 32: 64-65.
HOROWITZ,
N. H., 1945 On the evolution of biochemical synthesis. Proc. Nat. Acad. Sci., U.S.
31: 153-157.
ISING,
G. and C. RAMEL,1976 The behaviour 01 a transposing element in Drosophila melanogaster. pp. 947-954. In: The Genetics and Biology of Drosophila, Vol. Ib. Edited by M.
ASHBURNER
and E. NOVITSKI.
Academic Press, London.
LAEMMLI,
U. K., 1970 Cleavage of structural proteins during the assembly of the head of
bacteriophage TS. Nature 227: 680-685.
LINDSLEY;D. L. and E. H GRELL: 1968 Genetic Variutions of Drosophila melanogaster.
Carnegie Inst. Wash. Publ. No. 627.
LINDSLEY,
D. L., L. SANDLER,
B. S. BIIKER, A. T. C. CIRPENTER,
R. E. DENELL,J. C. HALL,P.A.
J.4COBS, G . L. G. MIKLOS,
B. K. DAVIS,R. C. GETHMANN, R. W. HARDY,
A. HESSLER,
s. M.
MILLER,H. NOZAWA,
D. M. PARRY
and M. GOULD-SOMERO,
1972 Segmental aneuploidy and
the genetic gross structure of the Drosophila genome. Genetics 71: 157-184.
MULLER,
H. J., 1936 Bar duplication. Science 83: 528-530.
OHNO,S., 1970 Euoiution by Gene Duplication. Springer-Verlag, Berlin.
1979 The X-linked 01 chain gene of Drosophila
ROBERTS,
D. B. and S. M. EVANS-ROBERTS,
LSP-1 does not show dosage compensation. Nature 280: 691-692.
GENETICS OF DROSOPHILA LSP-1
679
ROBERTS,
D. B., J. WOLFEand,lM. E. AKAM,1977 The developmental profiles of two major
haemolymph proteins from Drosophila melanogaster. J. Insect Physiol. 23 : 871-878.
ROBERTSON,
A. and M. RIVIERA,1972 Drosophila melanogaster linkage data. Dros. Inf. Ser.
48: 21.
WOLFE,J., M. E. AKAMand D. B. ROBERTS,
1977 Biochemical and immunological studies on
larval serum protein 1, the major haemalymph protein of Drosophila melanogmter third
instar larvae. Eur. J. Biochem. 79: 47-53.
Corresponding editor: A. CHOVNICK