1
Homology of dipteran bristles and lepidopteran scales:
requirement for the Bombyx mori achaete-scute homologue
ASH2
Qingxiang Zhou1, 2, Linlin Yu1, Xingjia Shen2, Yinü Li1, Weihua Xu3, Yongzhu Yi2,
Zhifang Zhang1, §
1The Biotechnology Research Institute, National Engineering of crop germplasm and
genetic improvement, Chinese Academy of Agricultural Sciences, Beijing, 100081,
China
2The Sericultural Research Institute, Chinese Academy of Agricultural Sciences,
Zhenjiang City, Jiangsu Province, 212018, China
3State Key Laboratory for Biocontrol and Institute of Entomology, School of Life
Sciences, SUN YAT-SEN University, Guangzhou 510275, China
Genetics: Published Articles Ahead of Print, published on August 10, 2009 as 10.1534/genetics.109.102848
2
Running title: wing scales and bristles are homologous
Key words: silkworm, scaleless, Bm-ASH2, scales, bristles
§Corresponding author:
Zhifang Zhang
Biotechnology Research Institute, Chinese Academy of Agricultural Sciences, 12
Zhongguancun South Street, Beijing, China, 100081
Tel: +86-10-82109854, Fax: +86-10-82105136, E-mail: zhifangzhang@yahoo.com
3
ABSTRACT
Lepidopteran wing scales and Drosophila bristles are considered homologous
structures based on similarities in their cell lineages. However, the molecular
mechanisms underlying scale development are essentially unknown as analysis of gene
function in Lepidoptera is sorely limited. In the present study, we used the Bombyx mori
mutant scaleless (sl), displaying a nearly complete loss of wing scales, to explore the
mechanism of lepidopteran wing scale formation. We found that Bm-ASH2, one of four
Bombyx achaete-scute homologues, is highly expressed in early pupal wings of wild
type (WT) silkworms, but its expression is severely reduced in sl pupal wings. Through
molecular characterization of the mutant locus using luciferase and gel shift assays,
genetic analysis of recombining populations and in vivo rescue experiments, we provide
evidence that a 26 bp deletion within the Bm-ASH2 promoter is closely linked to the sl
locus, and leads to loss of Bm-ASH2 expression and the scaleless wings phenotype.
Thus, the Bm-ASH2 appears to play a critical role in scale formation in B. mori. This
finding supports the proposed homology of lepidopteran scales and dipteran bristles,
and provides evidence for conservation of the genetic pathway in scale/bristle
development at the level of gene function.
INTRODUCTION
The wing surface of lepidopteran adults is covered by wing scales, which function
in heat preservation, mimesis, touch, etc (Nijhout, 1991). Lepidopteran scales and
Drosophila bristles are considered homologous structures based on similarities in their
cell lineages and the expression of a few molecular markers (Overton, 1966; Overton,
1967; Galant et al., 1998). Development of the Drosophila bristle is regulated by the
bHLH transcription factors of the Achaete-Scute Complex (AS-C). Within the clusters of
cells expressing achaete (ac) and scute (sc), some are selected to become sensory organ
precursors which then form and innervate the mature bristles (Skeath and Carroll, 1991;
4
Jan and Jan, 1994). In ac/sc double mutant flies, the majority of the bristles are lost; on
the contrary, ectopic expression of ac and/or sc can induce extra bristles (Garcia-Bellido,
1979; Campuzano et al., 1986; Balcells et al., 1988; Rodriguez et al., 1990). In the
butterfly P. coenia, the AS-C homolog B-ASH1 is indeed expressed in scale-forming
cells during pupation (Galant et al., 1998) and as we reported, the four AS-C homologs
of the silkmoth B. mori are also expressed during wing development, particularly at the
early pupal stage when scale formation begins (Zhou et al., 2008). ). By studying the
genomic organization and evolution of ac/sc genes of multiple distant insect species,
Negre and Simpson (2009) have recently argued that independent evolution of the
homologues might have contributed to morphological diversity in Diptera and
Lepidoptera.
We show here that a small deletion in the Bm-ASH2 gene underlies the scaleless (sl)
mutant phenotype in B. mori. As previously described, this mutant displays a severe
reduction in wing scales (Zhou et al., 2004; Zhou et al., 2006). We find that Bm-ASH2
expression is strongly reduced in the sl pupal wing. At the molecular level, we identify a
26 bp deletion within the promoter region of Bm-ASH2 which is linked to the mutant
phenotype. Using an electrophoresis mobility shift assay, we show that the 26 bp region
contains a cis-regulatory element. Lastly, we show that targeted expression of Bm-ASH2
in the sl pupal wing can partially rescue the lack of wing scales, confirming the central
role of Bm-ASH2 downregulation in the etiology of this mutant. Thus, at least one ASH
factor is fundamentally involved in scale formation in B. mori. This finding supports the
proposed homology of lepidopteran scales and Drosophila bristles and provides
evidence for conservation of the genetic pathway in the formation of these structures.
MATERIALS AND METHODS
Animals
Silkworms used in the experiments were the wild type strains 7532 and Furong,
and the transparent wings mutant strain scaleless (Zhou et al., 2004). The larvae were
5
cultured on mulberry leaves under 25 °C with 70-80% relative humidity.
In situ hybridization
DIG-labeled RNA probes were generated by in vitro transcription, using the ORF
region of a gene as the template. One day (1 d) old pupal wings were dissected out in
cold 0.75% NaCl and fixed in fresh 4% formaldehyde in 100 mM Hepes (pH 7.9), 2
mM MgCl2 and 1 mM EDTA for 2-3 h at room temperature or overnight at 4 °C. The
tissues were washed for 3x5 min in PBST0.2 (0.2% Tween20), then digested with
Proteinase K (20 μg/ml) for 5 min, followed by treatment with 0.2% glycine in PBS.
The wings were refixed in 4% polyformaldehyde for 30 min, and washed 2x5 min in
PBST0.2. Prehybridization was processed in Hybridization Solution (50% formamide,
5xSSC, 2% blocking powder, 10 mg/ml yeast tRNA, 5 mg/ml salmon sperm DNA and
20 mg/ml heparin) for 2 h. 100-500 ng anti-sense RNA probe was used for each
experiment and hybridization was done overnight. Then the tissues were washed 4 times
with PBST0.2, for 10 min, 20 min, 30 min and 30 min respectively. Subsequently, they
were dipped in GB-PBST0.2 (5% goat serum and 2% BSA in PBST0.2) for 2 h and anti-
DIG-AP Fab fragment for 2 h. The samples were washed 3x10 min in PBST0.2 and
stained with NBT/BCIP in the dark until signal appeared. The issues were dehydrated
successively with 30%, 50%, 70% and 100% methanol for 15 min each, followed by
successive 1 h treatment with 50% and 80% glycerol. Sense-strand RNA was used as
control for each experiment.
Cell culture and Dual-luciferase Reporter Assay
Cell culture and Dual-luciferase Reporter Assay were processed as described (Zhou
et al., 2008). One pupal wing was dissected out and lysed with 50 μl of the passive lysis
buffer, and 10 μl of the lysate was used for the Dual-luciferase reporter assay according
to the manufacturer’s protocol (Dual-luciferase® Reporter (DLRTM) Assay System,
Promega). Luciferase activity was determined with a 20/20n Luminometer (Turner
6
BioSystems, Inc.).
Electrophoretic mobility shift assay (EMSA)
Probes for EMSA were prepared by annealing two partially overlapping oligos
(Supplemental Table 1), synthesized to have two sticky ends to introduce dATP by
filling the gaps. The probes were labeled with [α-32P]dATP. Nuclear extracts were
prepared from 1 d pupal wings as described (Feng et al., 1998; Blough et al., 1999), and
the protein concentration was determined by the Bradford Assay (Bradford, 1976). 5 μg
of nuclear proteins were incubated with probe at a concentration of 104 cpm in 0.04
pmol DNA, in a final volume of 20 μl containing 50 mM Hepes-KOH (pH 7.9), 250
mM KCl, 20 mM MgCl2, 5 mM DTT, 5 mM EDTA, 0.5 mg/ml BAS, 0.25 μg/μl Poly
(dI-dC) (Sigma) and 50% Glycerol. After incubation for 30 min at 25 °C, the binding
reactions were analyzed on 7% polyacrylamide gels. The gels were dried and exposed to
a phosphoscreen (Bio-Rad) for 3 h. For competition analysis, 100-fold excess of
unlabeled double-stranded oligos was used.
Genomic DNA extraction from silkmoth
The head and thorax of a moth was cut and homogenized in 2 ml Solution A (0.25
M Sucrose, 10 mM EDTA, 30 mM Tris-HCl (pH 7.5)) on ice. After centrifugation for 3
min at 1,500xg, the supernate was discarded and the pellet was resuspended in 600 μl
Solution B (10 mM Tris-HCl (pH 7.5), 10 mM EDTA, 0.15 M NaCl, 1% Sarcosyl) and
incubated on ice for 15 min. The sample was treated successively with 1 vol
phenol/chloroform to denature the proteins. Genomic DNA was precipitated with 2 vol
ethanol, and the pellet was washed with 70% ethanol, dried, and dissolved in TER (20
µg/ml RnaseA in TE) and stored at -20 °C for further use.
Administrator
Rectangle
7
Ectopic expression of a foreign gene in the silkworm pupal wing by transient
expression system
The ORF region of a candidate gene was cloned into a modified pBacPAK8 vector
with an ie-1 promoter and a hr3 enhancer (Chen et al., 2004). 10 μl of the recombinant
plasmid (5 μg) with 5 μl of lipofectin were micro-injected into the pupal wing. Injection
of 10 μl of double distilled water with 5 μl of lipofectin was used as control.
RESULTS AND DISCUSSION
Expression of Bm-ASH2 is severely reduced in scaleless pupal wings
The scaleless mutant nearly losses all wing scales. Given that AS-C genes play
critical roles in the development of Drosophila bristles and that AS-C homologues (ASH)
are expressed in the developing wings of butterflies and silkmoths (Jan and Jan, 1994;
Galant et al., 1998; Zhou et al., 2008), we decided to investigate the potential
involvement of these bHLH-type of transcription factors in the etiology of this mutant.
We used semi-quantitative RT-PCR to assess the expression level of Bm-ASH genes in
WT and sl wings. The levels of Bm-ASH1, Bm-ASH3 and Bm-ase in sl pupal wings were
similar to those from two different WT lines, Furong and 7532. On the contrary, Bm-
ASH2 level was severely reduced in sl compared to WT (Zhou et al., 2006; data not
shown).
We further used in situ hybridization to confirm these findings and to investigate
the pattern of Bm-ASH2 expression. In the WT, Bm-ASH2 was robustly expressed across
1 d pupal wings in regularly spaced clusters of scale mother cells (Fig.1 A & C).
Consistent with the reduction in mRNA levels detected by RT-PCR, only a few cells
expressing Bm-ASH2 were detected on sl wing surface (Fig.1 B, D). By contrast, and in
agreement with the RT-PCR data, clusters of Bm-ASH1 expressing cells were present
across the mutant wings in a pattern essentially indistinguishable from WT (Fig. 2 A-D).
The scaleless wings phenotype is caused by a single locus and correlates with the
8
presence of a 26 bp deletion in the Bm-ASH2 genomic region
To understand the defect leading to loss of Bm-ASH2 expression in the sl, we
decided to characterize the locus at the molecular level. Although in situ analysis (Fig. 1)
suggested a defect in transcriptional regulation, we nonetheless sequenced the entire
Bm-ASH2 coding region. Unsurprisingly, sequences from the mutant and several WT
strains had no significant differences (data not shown). Since transcription is often
controlled through the chromosomal region flanking a gene on the 5’ side, we decided
to analyze a 1,175 bp of genomic DNA just upstream of the Bm-ASH2 translation start
site. Based on in silico prediction, this fragment spans ~1,010 bp upstream of the start of
transcription (Fig. 3). A number of isolated single bp changes were identified
throughout the region. However, the most striking sequence alteration occurred ~850 bp
upstream of the putative transcription start site, where a 28 bp sequence present in the
WT was replaced by only 2 bp in the sl (Fig. 3).
We used this deletion as a molecular marker to investigate whether the sl
phenotype segregated with this Bm-ASH2 mutant allele. Two primers flanking the
region deleted in sl (Supplemental Table 1) were used to amplify a 242 bp fragment
from the WT genomic DNA (Fig. 4, Lane 1; Supplemental Figure 1, Box 1) and a 216
bp fragment from sl (Fig. 4, Lane 2; Supplemental Figure 1, Box 1). To ensure our
findings are reasonable, we studied individuals with multiple different genetic
backgrounds. In the series of crosses shown in Supplemental Figure 1, individuals were
scored for their wing phenotype and then selected for PCR testing. In all cases,
phenotypically sl individuals carried the deletion allele but not the WT allele (Fig. 4,
Lanes 3-4; Supplemental Figure 1, Box 2-3). On the contrary, phenotypically WT
individuals either carried two copies of the WT DNA or one copy of the WT and one
copy of the deletion alleles (Fig. 4, lane 5; Supplemental Figure 1, Box 3; and data not
shown).
Because the original strain in which the mutant was discovered was unknown, a
more extensive analysis of linkage was carried out from a cross of sl to another WT
9
strain Furong (Fig. 4, lane 6; Supplemental Figure 1, Box 4). Heterozygous F1 progeny
were intercrossed to generate a mix of phenotypically sl and WT F2 progeny. Among
581 F2 adults, the proportion of phenotypically sl to WT individuals was ~ 1:3 (Table 1,
χ
2
= 0.014, χ20.05, 1=3.84). We randomly picked some of the moths to do molecular
analysis. All 40 F2 silkmoths with sl wings carried exclusively the Bm-ASH2 deletion
allele. Among 112 F2 individuals with a WT phenotype analyzed, 34 carried the
homozygous WT genotype, and the other 78 individuals were heterozygous (Table 1).
The ratio of homozygous to heterozygous individuals approximated the predicted 1:2
ratio expected for recessive inheritance at a single locus in a hybrid cross (χ2=0.322,
χ
2
0.05, 1=3.84). The results are consistent with a single gene hypothesis, i.e., that the sl
phenotype is caused by a recessive allele at a single locus.
These data are consistent with simple inheritance of sl as a recessive trait and
support a close linkage between the 26 bp deletion and the mutant locus. However, this
evidence is not sufficient to establish that Bm-ASH2 is the locus responsible for the sl
phenotype. Therefore, we proceeded to investigate: 1) whether the 26 bp sequence
could be involved in regulating Bm-ASH2 transcription, and 2) whether restoring
expression of Bm-ASH2 in the developing wing could rescue the mutant defect.
The 26 bp region contains sequences that contribute to normal transcriptional
regulation of Bm-ASH2
To test whether the 26 bp region was involved in transcriptional regulation, we
relied on transient expression assays in cultured cells.
We had previously shown that the 1,175 bp DNA fragment upstream of the Bm-
ASH2 translation start site (Fig. 5) is sufficient to drive robust luciferase expression in
the presence of the putative upstream regulators Bm-ASH1 and Daughterless and that
two E-boxes at positions 194 and 797 likely mediate this regulation (Zhou et al., 2008).
As expected, reporters bearing 5’ truncations that eliminate either the 797 E-box or both
797 and 194 E-boxes displayed much lower transcriptional activity, below 20% of the
10
full length ASH2P. However, expression did not rely solely on these regulatory elements
because a 979 bp reporter that retained both E-boxes and only lacked 169 bp at the 5’
end was also severely downregulated and expressed at only ~20% of the ASH2P.
Interestingly, this truncated 979 bp lacked the 26 bp region deleted in the sl.
To investigate whether the lower activity of the 979 bp was due to loss of the 26 bp
region (Fig. 3, shaded), we assayed expression of a ‘full length ASH2P’ reporter
generated from the sl genomic DNA (sl-ASH2P) and thus lacking the 26 bp sequence.
Noticeably, the activity of sl-ASH2P was significantly lower than that of WT ASH2P,
and similar to 979 bp.
Based on these findings, we hypothesized that the 26 bp region contains one or
more cis-regulatory elements required for normal expression of Bm-ASH2. Consistent
with this proposal, an activator present in nuclear protein extracts from 1 d pupal wings
bound a 49 bp double strand oligos that contained the 26 bp sequence in EMSA.
Binding was competed by either cold 49 bp DNA or a smaller one spanning the 26 bp
sequence, whereas it was unaffected by a cold 23 bp DNA that excluded the 26 bp (Fig.
6).
These findings, together with the loss of Bm-ASH2 expression in sl tissue, strongly
support a role for the 26 bp region in transcriptional regulation of Bm-ASH2.
Ectopic expression of Bm-ASH2 in the pupal wing can rescue the scaleless wing
phenotype
The findings above strongly support a model whereby decreased expression of Bm-
ASH2 results in loss of scales in the sl. To test this hypothesis we attempted to rescue
the mutant phenotype by providing exogenous Bm-ASH2 in the wing.
To achieve this goal, we developed a novel method for transient gene expression in
vivo. Previous studies have shown that B. mori baculovirus promoters are functional in
expression vectors and display expected transcriptional activity when injected into
silkworm larvae (Zhou et al., 2002; Chen et al., 2004; Tang et al., 2005). To assess the
11
efficiency of this method in the pupal wing, we generated a constitutive reporter
encoding the luciferase gene under the control of the previously described ie-1 promoter
and hr3 enhancer (pBacPAK8-ie-1-luc-hr3; Chen et al., 2004). The reporter plasmid
was micro-injected into the left forewings of 12 hour old pupae and the wings were
dissected 36 h later, lysed and tested for luciferase activity. Among a total of 14
individuals tested, only 2 (14%) lacked luciferase activity. In the positive samples
(n=12), luciferase activity ranged from 1,500 RLU (10 or 71.4%) to more than 10,000
RLU (3 or 21.4%), the highest reaching 248,064 RLU (Table 2).
Having validated this method for the transient induction of exogenous gene
expression in the wing, an expression plasmid encoding the Bm-ASH2 was micro-
injected into the left forewing of 12 h sl pupae. Out of 50 pupae treated in this fashion,
37 developed into moths, but 16 were excluded because their left forewings were folded
and difficult to score. The remaining 21 moths were evaluated for presence/absence of
scales. Nine (42.9%) displayed the sl phenotype, whereas 12 (57.1%) displayed
significant rescue. In these 12 individuals, the number of scales (as assessed by the
presence of scale sockets) on the left forewing was significantly higher than the number
of scales on the untreated right one (t-test, p<0.01, Table 3, Fig. 7). Controls subjected
to mock injections (33 sl moths) had wings with scale socket totals indistinguishable
between treated and untreated sides (p>0.05, Table 3).
Interestingly, expression of exogenous Bm-ASH1 did not rescue the mutant
phenotype indicating that loss of scales in the mutant was not due to a reduction of AS-
C type factors in general, but to the specific loss of Bm-ASH2 protein. We cannot
exclude, however, the possibility that Bm-ASH2 co-operates with Bm-ASH1 in the
selection of scale mother cells. Moreover, we did not observe formation of
supernumerary scales in WT wings injected with either Bm-ASH2 or Bm-ASH1
expression vectors (Tables 3, 4). This is in contrast to the induction of ectopic bristles in
Dros