Bmmp influences wing shape by regulating anterior-posterior and proximal-distal axis development

Insect wings are subject to strong selective pressure, resulting in the evolution of remarkably diverse wing shapes that largely determine flight capacity. However, the genetic basis and regulatory mechanisms underlying wing shape development are not well understood. The silkworm Bombyx mori micropterous (mp) mutant exhibits shortened wing length and enlarged vein spacings, albeit without changes in total wing area. Thus, the mp mutant comprises a valuable genetic resource for studying wing shape development. In this study, we used molecular mapping to identify the gene responsible for the mp phenotype and designated it Bmmp. Phenotype-causing mutations were identified as indels and single nucleotide polymorphisms in non-coding regions. These mutations resulted in decreased Bmmp mRNA levels and changes in transcript isoform composition. Bmmp null mutants were generated by CRISPR/Cas9 and exhibited significantly smaller wings. By examining the expression of genes critical to wing development in wildtype and Bmmp null mutants, we found that Bmmp exerts its function by coordinately modulating anterior-posterior and proximal-distal axis development. We also studied a Drosophila mp mutant and found that Bmmp is functionally conserved in Drosophila. The Drosophila mp mutant strain exhibits curly wings of reduced size and a complete loss of flight capacity. Our results increase our understanding of the mechanisms underpinning insect wing development and reveal potential targets for pest control.


65
Wings endow insects with tremendous adaptive advantages because they enhance 66 survival and fitness by making it possible to change environments rapidly. Insect wings 67 are constantly subject to adaptive evolution and exhibit remarkable diversity in shape. 68 Changes in wing shape result in differences in flight capacity, leading to variations in 69 insect lifestyle [1,2]. For example, dimorphism in wing shape occurs in a wide range 70 in insects, such as rice planthoppers [3,4] and aphids [5]. Long-winged morphs can fly, 71 which allows them to escape adverse habitats and track changing resources, whereas 72 short-winged morphs are flightless, but usually possess higher fecundity [1,2]. In the 73 order Lepidoptera, wing shapes are distinctly different between migratory species and 74 non-migratory species. Typically, migratory moths and butterflies have relatively 75 narrower forewings with straighter costal margins compared to those of non-migratory 76 species [6]. the most distal regions of the wing [11,12]. Reduced levels of Dpp affect both the width 88 and length of the resulting wing and significantly decrease total wing area [13]. 89 Moderate and uniform amounts of exogenous Wnt1 stimulate proliferative wing growth, 90 leading to enlargement of the prospective wing [14]. 91 The identification of new factors that influence wing shape will expand our 92 understanding of the genetic basis of wing diversity. We hypothesized that as-yet 93 uncharacterized key regulators coordinately regulate both A-P and P-D axis signals 94 during wing development. To examine this developmental process more closely, we 95 used the silkworm Bombyx mori (Lepidoptera, Bombycidae) micropterous (mp) mutant, 96 which exhibits shortened wing length and enlarged vein spacings. We identified the 97 gene responsible for the mp phenotype and designated it Bmmp. We found mutations in 98 the noncoding regions of Bmmp that result in decreased Bmmp mRNA levels and 99 changes in transcript isoform composition. In addition, we generated a Bmmp null 100 mutant and determined that Bmmp exerts its effect on wing shape by regulating wing 101 A-P and P-D axis development.  (Fig 1A). Further examination demonstrated that the wing length of mp moths was 112 significantly shorter than that of WT moths within each sex, although there was not a 113 significant difference in total wing area (Fig 1B, 1C, 1D). In addition, there was 114 significantly greater spacing between adjacent longitudinal veins in the wings of mp 115 moths compared to those of WT moths within each sex (Fig 1E). These results 116 demonstrate that the mp phenotype is not associated with a specific gender. To reflect 117 the overall changes in wing morphology, we divided the wing length by the sum of 118 longitudinal veins spacings. The resulting value is significantly smaller for mp moths 119 than for WT moths within each sex (Fig 1F). To identify candidate gene(s) responsible for the mp phenotype, we performed a 124 genetic linkage analysis using B. mori simple sequence repeat (SSR) markers and newly 125 designed markers polymorphic between WT and mp silkworms. Initially, we roughly 126 mapped the mp phenotype using 456 BC1M individuals and SSR markers on the 127 eleventh linkage group. The results indicate that the gene responsible for the mp 128 phenotype is located within a 12.1-cM region linked to SSR marker S1146 (Fig 2A and   129 2B). Subsequent fine mapping with 320 BC1M and newly designed primer sets 130 narrowed the mp locus to an approximately 260-kb region between markers 2810A and 131 2810C on the nscaf2810 scaffold. The 2810M marker was tightly linked with the mp 132 locus (Fig 2C). Two candidate genes (KWMTBOMO06923 and KWMTBOMO06924) 133 were identified within the 260 kb region, based on annotated sequences obtained from 134 the SilkBase database [15] (Fig 2D). 135 Because the functions of KWMTBOMO06923 and KWMTBOMO06924 are   identified in the mp silkworms (Fig 3). An intact BTB domain, encoded by exons [1][2][3]159 was present in all transcript isoforms identified in both silkworm strains. Quantitative

160
RT-PCR analysis of wing discs from silkworms at the initiation of the wandering stage 161 revealed that total KWMTBOMO06924 mRNA levels were significantly lower in mp 162 vs. WT silkworms during this critical period of wing development (Fig 4). 163 Together, these results suggest that KWMTBOMO06924 is responsible for the mp  we utilized the CRISPR/Cas9 system to disrupt Bmmp. We selected four genomic 173 targets spanning 130 bp in exon 1 to generate large fragment deletions (Fig 5A). Since this region is shared across isoforms, any frame-shift mutations would be predicted to 175 abolish all functional transcripts. sgRNAs were synthesized in vitro for the genomic 176 targets, mixed with Cas9 protein, and injected into the preblastoderm of Dazao embryos.

177
In total, 110 injected embryos hatched, and 81 individuals survived to an adult stage.

178
Out of these 81 silkworms, 67 exhibited markedly smaller wings in pupal and adult 179 stages, compared to uninjected WT controls (Fig 5B and 5C). To confirm that the 180 Bmmp deletions caused the decrease in wing size, genomic DNA was extracted from 181 three moths with small wings. Regions spanning the four sgRNA targets were amplified 182 by PCR, subcloned, and sequenced. As expected, the three selected moths contained 183 Bmmp deletions and no wildtype sequences (Fig S1). Notably, five distinct mutations 184 were identified in moth #11 (Fig S1), demonstrating the presence of mosaicism in 185 silkworms of the injected generation (generation 0, G0).

186
To further confirm the function of Bmmp in a uniform genetic background, 187 homozygous or compound mutant silkworms were obtained by crossing mosaic 188 knockouts. We randomly surveyed 3 egg batches of generation 1 (G1). All individuals 189 surveyed were homozygous or compound mutants (Fig 5D), demonstrating that 190 germline transmission of the mutations was highly efficient. Compared with the WT 191 control, homozygous or compound mutant silkworms all exhibited significantly smaller 192 wings (Fig 5E and 5F), consistent with the phenotype observed in G0 mosaics.

193
It is noteworthy that we obtained homozygous knockout silkworms with an in-194 frame 108 bp deletion in the coding region of exon 1 by crossing G1-mp-24♀ with G1-195 mp-15♂ (Fig 5D). Presumably this mutation disrupts the functional BTB domain without affecting the downstream BACK and TLDc domains (Fig 5D). However, these 197 knockout silkworms were identical in wing phenotype to silkworms harboring engrailed, hedgehog, dpp, and gbb, which are responsible for wing A-P axis 217 development (Fig 6). Likewise, mRNA levels in knockouts were reduced for apterous 218 A, apterous B, vestigial, Wingless (wnt1), and distal-less, which participate in wing P-219 D axis development (Fig 6). These results suggest that Bmmp directs wing morphology 220 by regulating genes responsible for wing A-P and P-D axis development. Bmmp knockout and WT silkworms (Fig 7A and 7B). We speculate the milder 235 phenotype may be due to genetic differences as the Dropsophila mp mutant contains an      Table S3. Eukaryotic translation initiation factor 4A (silkworm microarray 357 probe ID sw22934) was used as the internal control.  Figure 4A.

362
The DNA template for the T7 promoter used to drive in vitro transcription was 363 constructed by PCR as described [28]. Briefly, an oligonucleotide containing the T7 364 promoter and the sgRNA target sequence (N20) was designed as a forward primer with 365 the sequence 5'-TAATACGACTCACTATAGG(N20)GTTTTAGAGCTAGAAATAGC.