The insect development is guaranteed by incorporated nutrients, which become the source of energy for adult development. Numerous studies have clarified the molecular mechanism of insect growth. After the final larval ecdysis insect takes food, it is digested and secreted as nutrients into the hemolymph (Riddiford, 1985), which stimulates the production of insulin-like peptides (ILPs) from the brain and other tissues (Masumura et al., 2000; Ikeya et al., 2002), resulting in the activation of the insulin cascade in the peripheral organs (Nijhout, 2003). Then, the insect transfers stored nutrients and energy for adult development and reproduction during metamorphosis. Involvement of juvenile hormone (JH), insulin cascade and ecdysone in the body size or wing disc size have been discussed (Shingleton et al., 2008; Tober and Nyjhout, 2010; Mirth and Shingleton, 2014). Nevertheless, this transfer has not yet been clearly explained, since an adequate experimental system has not been found and used.

Cell division is suggested to be a suitable marker for judging growth and differentiation. The enhancement of cell division was observed in Bombyx mori wing discs in the early feeding stage and the wandering stage of the fifth (last) larval instar (Kurushima and Ohtaki, 1975). The enhancement of cell division in the early feeding stage is observed when the hemolymph ecdysteroid titer (Sakurai et al., 1998) is less than the concentrations that induce cell division in wing disc cells in vitro (Kawasaki, 1995, Moriyama et al., 2016). Cell division in the early feeding stage is brought out by the disappearance of JH (Kurushima and Ohtaki, 1975). The enhanced cell division and growth of wing discs in this early stage are parallel with the growth of the whole body (Nijhout and Grunert, 2002). At this stage, nutrition-dependent factors, such as ILPs (Nijhout and Grunert, 2002; Nijhout et al., 2007), imaginal disc growth factors (Li et al., 2009; Wang et al., 2009b), and nutrients (Mirth and Riddiford, 2007), are likely involved in the control of wing disc growth instead of ecdysone. Therefore, the PI3K/Tor cascade is suggested to function in the wing disc at the feeding stage of the last larval instar. Through the PI3K/Tor cascade, the activation of Myc has also been reported (Michaelson et al., 2010; Grewal et al., 2009). Myc has been reported as a key gene for cell division (Johnston et al., 1999; Duman-Scheel et al., 2004; Adhinkary and Eilers, 2005; Siddall et al., 2009; Hu et al., 2012), and it also activates a number of genes (Adhinkary and Eilers, 2005; Tu et al., 2015).

The latter enhanced cell division at the wandering stage was induced by ecdysone through the activation of Myc (Kawasaki, 1995, Moriyama et al., 2016) when the hemolymph ecdysteroid concentrations are those that induced cell division of wing disc cells in vitro (Kawasaki, 1995, Moriyama et al., 2016). These low concentrations of ecdysone also accelerated the cell cycle in the accessory gland of Tenebrio molitor (Yaginuma et al., 1988). Thus, wing disc cells of B. mori are suggested to show cell division in response to different hormones. Therefore, we believe the change from insect growth to metamorphosis will be clarified through the analysis of the differences in the two faces of cell divisions. Nevertheless, a clear explanation has not yet been given as to what kind of signaling cascades work during insect development and metamorphosis. As described above, Myc expression is correlated to cell cycle; therefore, we selected Myc expression as a marker of cell-cycle progression.

We (Moriyama et al., 2016) found that Bmwnt1 was induced by 20-hydroxyecdysone (20E), and Wnt has been reported to be involved in the induction of differentiation (Loh et al., 2014; Stevens et al., 2017, Davidson et al., 2012; König et al., 2015; Bejsovec et al., 2018; Upadhyay et al., 2018), proliferation (Nakamura et al., 2007; Theisen et al., 2007; Parker et al., 2020), or chromatin remodeling (Parker et al., 2008; Cadigan et al., 2012). From this, we suggest that Wnt is a possible candidate for inducing the differentiation and making a difference before and after ecdysteroid secretion in the hemolymph. Therefore, we examined Wnt expression together with Myc expression in the present study.

The morphogen opposite of Wnt (Theisen et al., 2007, Bakker et al., 2020; Zecca and Struhl, 2021), Dpp (Szȕts et al., 1998; Pyrowolakis, 2004; O’Keefe et al., 2014), also regulates a number of genes and interacts with several signal cascades. In addition, Ras is known as a key factor of the Ras/MAPK signal cascade (Migliaccio et al., 1998, Halfar et al., 2001; Sudaram, 2005) and is related to proliferation and differentiation (Jiang and Edgar, 2009; Halfar et al., 2001, Nakamura et al., 2021) and to rRNA synthesis and Myc activation (Sears, 2000; Sriskanthadevan-Pirahas et al., 2018; Grewal, 2011). Therefore, we examined Dpp and Ras expression together with Myc and Wnt expression.

Wing discs of B. mori elongate and evaginate after hemolymph ecdysteroid titer increases before pupation, and matrix metalloproteinase (MMP) genes (BmMMPs: BmMMP1, BmMMP2, BmADAMTS1, and BmADAM-like) that are related to morphological change showed upregulation (Kawasaki et al., 2018). Therefore, we examined BmMMPs as differentiation markers.

Wing disc showed through three stages of differentiation, one is for morphogenesis for pupal wing; described above and tubulin/actin related protein were involved as reported before (Kawasaki et al., 2004). Second one is for pupal character and Broad Complex (BR-C) are concerned (Murata et al., 1985, Riddiford et al., 2003, Wang et al., 2009; Shahin et al., 2016). BR-C expression was reported to be induced by the disappearance of JH and the addition of insulin in vitro (Koyama et al., 2008), and also ecdysone inducible (Chen et al., 2004; Reza et al., 2004). BR-C also related with pupal commitment (Matsumura et al., 2008). Thus, BR-C are concerned with differentiation, therefore we also examined BR-C expression. The third one is for adult differentiation such as scale formation (Kawasaki et al., 1988), which occurs after pupation and also was induced by the culture in the continuous low concentration 20E (Kawasaki 1995).

Thus, Myc, Wnt, Dpp, and Ras have relation with several genes and signaling cascades interact, therefore the interpretation of experimental results in vivo would be complicated. Therefore, in the present paper, we applied the wing disc culture system used in previous reports (Kawasaki, 1995; Noji et al., 2009; Wang et al., 2009; Moriyama et al., 2016), and Kawasaki, 1995, Moriyama et al., 2016 observed cell division by the addition of 20E using this system. Thus, the system used in the present study is suitable for analyzing the direct effects of hormones or inhibitors. We tried to analyze differences in the cell-division cycle between the feeding and wandering stages and make clear what initiates the difference between two cell divisions by analyzing Myc and Wnt genes. Through our analysis, we found that the PI3K/TOR cascade enhanced division at the feeding stage and maintained cell division at the feeding and wandering stage together, while ecdysone signaling brought about cell division through the Myc and Wnt signaling cascade, which precipitated differentiation and metamorphosis.