spalt is functionally conserved in Locusta and Drosophila to promote wing growth

Locusta has strong fly wings to ensure its long distance migration, but the molecular mechanism that regulates the Locusta wing development is poorly understood. To address the developmental mechanism of the Locusta flying wing, we cloned the Dpp target gene spalt (sal) and analyzed its function in wing growth in the Locusta. The Locusta wing size is apparently reduced with vein defects when sal is interfered by injection of dsRNA, indicating that sal is required for locust wing growth and vein formation. This function is conserved during the Drosophila wing development. To better understand sal’s function in wing growth, we then used Drosophila wing disc as a model for further study. We found that sal promotes cell proliferation in the whole wing disc via positive regulation of a microRNA bantam. Our results firstly unravel sal’s function in the Locusta wing growth and confirm a highly conserved function of sal in Locusta and Drosophila.

and wing hinge 27 and the melanin pigmentation of wing eyespots in butterflies [28][29][30] . Therefore, sal may play a conserved role in insect wing development to generate the fast oscillating wings and whether sal functions through ban remains to be illustrated.  To better understand the developmental mechanism of Locusta flying wings, we cloned Locusta sal cDNAs and investigated the role of sal during wing blade and vein formation. By RNA interference (RNAi), we found that sal is required for wing growth and vein formation, which is conserved between the Locusta and Drosophila. Then, using the Drosophila wing disc model we dissected sal function in wing growth and found that sal promotes cell proliferation rate during larval wing development. Furthermore, mircoRNA bantam mediates the role of sal in cell proliferation regulation.

Materials and Methods
Drosophila strains and Locusta. All fruit flies used were Drosophila melanogaster. Flies were raised at 25 °C. Three transgenes UAS-salm 27 , UAS-salmRNAi (Tsinghua Fly Center THU3581) and UAS-salrRNAi (THU3147) were used to manipulate the expression level of sal. The efficiency of these RNAi lines has been recently verified by Tang et al. 31 . The Gal4 lines used were ci-Gal4 32 , sal-Gal4 33 , and ap-Gal4 34 to drive the expression of the UAS transgene. The bantam enhancer reporter line was br-C12-lacZ 35 .
Locusta migratoria was raised at 28 °C with 60% relative humidity.
Identification and sequence analysis of spalt from L. migratoria transcriptome. The available salm and salr gene sequences from Drosophila were used as references to screen L. migratoria transcriptome database we previously established 6 . The potential candidates of L. migratoria sal genes were confirmed by searching the BLASTX algorithm against the non-redundant NCBI nucleotide database using a cut-off E-value of 10 −5 . Based on the Drosophila Spalt protein sequences, the deduced Locusta Spalt protein domains were determined by using DNAMAN software (Lynnon Biosoft). Other insect Spalt and Spalt-like proteins are obtained from In each genotype, five nymphs were injected which give similar results as shown. *, ** and *** represent significant difference with the GFP-dsRNA control (pairwise comparison of t-tests, p < 0.05, 0.01, and 0.001, respectively). ns means no significant difference. Scale bar represents 4 mm.
Scientific RepoRts | 7:44393 | DOI: 10.1038/srep44393 NCBI (https://www.ncbi.nlm.nih.gov/protein). Global protein alignment and neighbor-joining phylogenic tree were performed by Geneious R9. RNA interference. The primers to make the double strand RNA probes were as follows: Lmsal411-dsRNA, Lmsal468-dsRNA, and GFP-dsRNA were injected into the 4 th instar nymphs using 20 μ g dsRNAs per insect. Nymphs were raised at 28 °C for 10 days until they grew up to the 5 th instar. Imaging the whole body and wing discs using SONY DSC-HX1camera before the wing discs were lysed.
Quantitative Real-Time RT-PCR Analysis. Total RNAs were extracted with Trizol reagent from the 5 th nymph wing discs, digested with DNase I, and reverse transcribed using the FastQuant cDNA Synthesis kit (Tiangen Biotech), then quantitative real-time RT-PCR was performed using Go Taq qPCR Master Mix (Promega). The primers for quantitative RT-PCR were as follows: Relative transcript levels were assessed using the Comparative CT method. β -actin was used as an internal control.
BrdU staining. BrdU staining was performed as previously described 20,36 . Dissected third instar wing discs were co-cultured with BrdU (1:100) in Schneider's medium for 40-50 min at 25 °C. Then, samples were fixed in 4% formaldehyde and washed in PBT before immunostaining.   Wing size measurement. Image-J program was used to measure the wing areas and vein distances. High resolution images were opened in Image-J program, a straight line was drawn at the wanted sites between L2 and L4 veins to calculate the distance; and a wanted region as outlined along the L2 and L4 veins (for Drosophila wing) or the wing margin (for Locusta wing) to calculate the areas.

Results and Discussion
Identification of Locusta sal genes. Based on the transcriptome of the Locusta wing disc in our previous study 6 , we blasted the cDNA sequences with salm and salr in Drosophila and identified two orthologous unigenes a12971 and a6922, which are 2087 bp and 1073 bp, respectively, in the Locusta (Sup. Fig. 1). We named these two putative genes as Lmsal411 and Lmsal468, respectively. Although these two unigenes were incomplete spalt gene sequences, both putative translations were predicted (Sup. Fig. 1). Protein sequence alignment showed that these two putative translations were 23.8% identity. Lmsal411, which was aligned with the anterior part of Drosophila Sal, was 33.1% identical to both Salm and Salr, while Lmsal468, aligned with the posterior part, showed a little higher identity (42.1% and 35.5%, respectively) (Fig. 1). The ZnF-C2H2 motifs showed a highly conservation in Lmsal411 (87.0% identity) and Lmsal468 (72.4% identity), indicating a conservative DNA binding motif of transcription factors. In addition, the phylogenetic analysis from different insect species revealed that Lmsal411 clustered with Drosophila sal, while Lmsal468 with bees (Fig. 2). Therefore, due to the less identity between these two translations and higher similarity to Drosophila salm, we speculate that these two unigenes may be different loci of the same Locusta spalt.
Above analysis implies that Locusta sal may share partial functions with that of Drosophila. In the Drosophila wing, the transcription factors Salm and Salr participate in wing growth 21 and vein formation 18,25 . However, the regional growth effects of Sal on wing growth and cell proliferation rate have not yet intensively investigated.

Locusta sal regulates wing growth.
To test the functional conservation of sal, we performed RNA interference experiments in Locusta using Lmsal genes we identified. After 10 days of injection, the mRNA levels of each gene were measured using quantitative real-time RT-PCR. The mRNA expression levels of Lmsal411 and Lmsal468 were significantly reduced to 62.7 ± 9.1% (p < 0.05) and 36.1 ± 8.5% (p < 0.01), respectively, when the corresponding gene was knocked-down ( Fig. 3A and B). We noted that Lmsal411 mRNA was reduced to 30.3 ± 21.8% when interfering with Lmsal468, and vise versa, Lmsal468 mRNA was reduced to 33.5 ± 11.3% when interfering with another ( Fig. 3A and B). Thus, the expression of these two unigenes are dependent on each other, implying that they belong to either the same gene or one gene complex. As when knocking-down either gene, the efficiency of interference of both genes were similar and had no statistic difference (Fig. 3A and B), and taken together with the above analysis of sequence identity, it is more likely that they are belonging to the same gene. When Lmsal411 was interfered in 4 th instar, the wings of early 5 th instar nymphs were remarkably reduced to 45% of the control (15.4 vs 27.5 mm 2 ) ( Fig. 3D and F). Consistently, interfering with Lmsal468 resulted in 69% loss of wing area (8.6 vs 27.5 mm 2 ) ( Fig. 3E and F). The phenotypes of Lmsal468-RNAi were much more obvious than that of Lmsal411-RNAi (Fig. 3D and E) probably due to the higher RNAi efficiency of the former. Moreover, the hindwing concomitantly reduced with the forewings, while the body size was not apparently affected ( Fig. 3D and E). The veins in the posterior part was slightly affected in Lmsal468-RNAi nymphs (arrow in Fig. 3Eʼ). This result indicates that Locusta sal regulates wing growth and vein pattern formation.
Drosophila sal promotes adult wing growth. To further confirm sal is functionally conserved in other insects to promote wing growth, we regionally up or down-regulated salm and salr in Drosophila wings and examined the adult wing size. Drosophila wing has five longitudinal veins, which are L1-L5 from the anterior to posterior part. Between L1 and L4 is the anterior part and other areas are the posterior wing. The distance from L2 to L4, which is indicated by a dashed line in Fig. 4A, was measured to monitor the size of the anterior part of the wing. The flies containing ci-Gal4, which is expressed in the anterior part of the wing, was used as a control and showed normal size and morphology. When salm was overexpressed, the relative distance was significantly increased (1.07 ± 0.01), while when knocked-down salm or salr, the distance was strikingly reduced (0.85 ± 0.01 and 0.94 ± 0.01, respectively) ( Fig. 4B-D and I). To confirm this result, we used another driver sal-Gal4, which is expressed in the medial pouch to repeat the manipulation. Again, the relative area enclosed by L2 and L4 (sal-Gal4 expression region) was increased in salm overexpression wings and reduced in salm and salr knockdown wings (Fig. 4F-H and J). In addition, sal regulated the vein formation in all tested lines except ci-Gal4> salm which is consistent with previous studies 21,25,26 . These data confirm that Drosophila sal regulates wing growth and vein pattern formation. Thus, sal function in promoting wing growth is conserved in the fruitfly and locust.
sal promotes cell proliferation in the wing disc. To unravel the mechanism of wing growth regulated by sal, we use the Drosophila wing disc as the research model because it is easy of molecular labeling. We examined the cell proliferation rate in Drosophila wing discs by Bromodeoxyuridine (BrdU) incorporation which is widely used in cell proliferation studies. BrdU is a thymidine analog that can be incorporated into DNA during DNA synthesis in culture, and then be detected by specific anti-BrdU antibodies following with quantitative analysis of the anti-BrdU fluorescent intensities. In wild-type control, BrdU level is evenly distributed in the wing discs 20 . Elevating Salm level by expressing UAS-salm in the ci-Gal4 or ap-Gal4 domain, the level of BrdU level was increased (Fig. 5A and B). Within the pouch, the fluorescence intensity of BrdU was increased in the salm expressing regions, indicating an enhanced proliferation rate ( Fig. 5A and B). Vice versa, compromising salm and salr by expressing UAS-salm-RNAi or UAS-salr-RNAi in the sal-Gal4 or ci-Gal4 domain, the level of BrdU staining was reduced which means that the cell proliferation rate was decreased. These data demonstrate that Drosophila sal promotes cell proliferation in the early wing developmental stages. sal promotes microRNA bantam expression. We previously found that one of the Dpp target genes omb represses ban expression in medial regions of the Drosophila wing discs while plays an opposite role in lateral regions 20 . Subsequently we took the advantages of transgenic lines available for Drosophila to do further investigation, because we are lacking of genetic tools to do it on the Locusta wing to test whether and how sal regulates the expression of ban. ban transcription was monitored by br-C12-lacZ 35 which was mainly expressed in the hinge/ blade folds surrounding the wing pouch (Fig. 6A). Expression of UAS-salm in the medial region of the wing disc driven by dpp-Gal4 apparently upregulated ban transcription level (arrows in Fig. 6B and Bʼ ). To confirm this result, clones were generated in the wing disc. Consistently, ban was upregulated in both medial regions and lateral region (Fig. 6C-C"). Therefore, sal promotes ban expression in the wing disc in a non-regional specific manner, unlike the manner of omb.
Previous studies have illustrated several models to explain how Dpp gradient controls a uniform cell proliferation rate in the wing disc 37 . Dpp restricts the transcription factor Brinker (Brk) domain to the lateral wing disc where suppresses the growth promoter ban 15 . While Brk activity is not essential for the medial wing disc growth [38][39][40] . Omb is a main mediator of Dpp signaling in the control of cell proliferation rate. Omb represses ban in medial regions, however, promotes it in lateral regions 20 . Distinct to omb, sal promotes ban expression in the whole wing disc (Fig. 6). Therefore, sal may mediate partial functions of Dpp in growth control. Other possibility is that sal may mediate the roles of other upstream factors such as Lines 41 , Wingless 42 , and Ubx in Drosophila 43 as well as Tribolium 44 .