Development of polyploidy of scale-building cells in the wings of Manduca sexta

https://doi.org/10.1016/j.asd.2012.09.003Get rights and content

Abstract

The developing wings of butterflies and moths are composed of two epithelial monolayers. Each epithelial sheet is made up of two kinds of cells, diploid cells that make up the epidermal surface and body of the wing, and large polyploid cells that become the scale-building cells whose cytoplasmic projections develop into the scales that will cover the adult wing and bear the pigment pattern. We studied the development of polyploidization of the scale-building cells during the pupal stage of the tobacco hornworm moth, Manduca sexta. The endomitotic divisions of the presumptive scale-building cells and the mitotic divisions of the diploid epithelial cells begin on day 3 of the pupal stage and continue until day 7. We show that scales of different colors and positions on the wing differ in size, and that the size of the scale is proportional to the ploidy of the scale-building cell. Scale-building cells are arranged in irregular rows and within each row there is an alternation of ploidy levels, with the lower ploidy cells giving rise to the underscales and the higher ploidy cells giving rise to the cover scales that carry the color pattern. Along the wing there is a proximo-distal decreasing gradient of average ploidy and scale size. Scale-building cells of high ploidy are surrounded by fewer epidermal cells than those of low ploidy. This inverse relationship is known as Henke's compensation principle, which posits that the number of endomitoses of a pre-polyploid cell and the number of mitotic divisions of its diploid daughter cell add up to a constant. We show that the inverse relationship fits the predictions of the compensation principle and does not fit constraints imposed by packing density, and we discuss mechanisms that could give rise to the inverse relationship.

Highlights

► A sub-population of cells becomes polyploid during wing development. ► Polyploid cells will form the scales on the wing. ► Ploidy level alternates within a row of scale-forming cells. ► Ploidy level is associated with the size of the scale that develops. ► High-ploidy cells have fewer diploid neighbors than low-ploidy cells.

Introduction

The wings of insects develop from two monolayers of epithelial cells that make up the dorsal and ventral surfaces, respectively. Because of its two-dimensional nature, the insect wing has long been an attractive study system for morphogenesis and pattern formation (Henke, 1933, 1946; Stossberg, 1938; Esser, 1961; Sondhi, 1963; Nardi and Kafatos, 1976; Nijhout, 1980, 1991, 2010; Nardi and Magee-Adams, 1986).

The wings of holometabolous insects develop internally during the larval stage as imaginal disks, which originate as invaginations of the body wall (Kuntze, 1935; Nijhout, 1991). In the Lepidoptera, the wing imaginal disks are extruded through a pore in the body wall during pupation and come to lie on the surface of the pupa. It is during the pupal stage that the undifferentiated epithelial cells of the developing wing become differentiated into epidermal cells, which will secrete the general cuticle of the adult wing, and scale-building cells, which will manufacture the large scales that cover the adult wing (Stossberg, 1938; Henke and Pohley, 1952). The epidermal cells undergo several rounds of proliferative mitosis and remain diploid, whereas the scale-building cells undergo several rounds of endomitosis and become highly polyploid.

The cell-level development of the lepidopteran wing has been most extensively described in the meal moth, Anagasta (formerly Ephestia) kühniella (Stossberg, 1938; Henke, 1946; Henke and Pohley, 1952; Esser, 1961). This early work forms the foundation of our understanding of wing development in the Lepidoptera. These studies used light microscopy of whole-mounts and sectioned material to describe temporal and spatial patterns of cell division, to estimate the ploidy level of the scale-building cells, and to describe the growth and differentiation of the scales. In Anagasta, the scale-building cells achieve different degrees of ploidy (4n to 32n) depending on their location on the wing. Each scale arises as a flattened outgrowth of a single scale-building cell (Stossberg, 1938). The scales begin to grow at about the time that the scale-building cells reach their final ploidy level. The final sizes of scales also vary, depending on their location on the wing, and are larger in region of the wing where the ploidy level of the scale-building cells is high (Henke, 1946; Henke and Pohley, 1952). This correlation provoked Henke (1946) and Henke and Pohley (1952) to suggest that the size of a scale is determined by the ploidy of the scale-building cell from which it develops. Because it has not been possible to firmly link a given scale with a specific scale-building cell, this hypothesis has remained untested.

In Manduca, as in many other Lepidoptera, the scale-building cells are arranged in more-or-less regular, and evenly-spaced parallel rows that are perpendicular to the long axis of the wing (Köhler, 1932; Braun, 1936; Yoshida and Aoki, 1989; Nijhout, 1991). Nardi and Magee-Adams (1986) have shown that the scale-building cells in Manduca initially appear in a random pattern but gradually arrange into parallel rows during days 3–5 of the pupal stage. They hypothesized that this orderly arrangement is mediated by epidermal feet (basal filopodia) that extend several cell diameters from each cell in the early pupal wing epithelium. The fine structure of the wing epidermis and the ultrastructure of the epidermal cells during the later stages of development of the pupal wing have been studied by Nardi and Ujhelyi (2001), particularly in relation to changes in the basal lamina that separates the dorsal and ventral cell monolayers and changes in cellular shape that enable later expansion of the wing at adult emergence from the pupal cuticle. The cellular changes that occur in the developing wing in the pupal stage have been described at the general anatomical and at the ultrastructural level (Nardi and Magee-Adams, 1986; Nardi and Ujhelyi, 2001), but little is known about the exact timing and rate of cell division, nor about the timing and pattern of endomitoses and the resulting degree of polyploidization of the presumptive scale-building cells.

In the present paper we focus on the development of polyploidy in the scale-building cells in Manduca sexta, which occurs during the first week of the pupal stage. We show that nuclear shape is exceedingly variable. By reconstructing the volume of nuclei we were able to get better estimates of ploidy level than was possible before. We show that there are gradients of ploidy level across the wing. We were also able to associate scale-building cells with specific scales and confirm the hypothesis that the size of the scale is related to the ploidy level of the scale-building cell. In addition, we show that the number of diploid cells in a given area of the wing is inversely proportional to the ploidy level of the scale-building cells in that region, in a pattern that supports Henke's compensation principle (Henke, 1946), and we discuss the mechanisms that might give rise to this relationship.

Section snippets

Animals

Larvae of M. sexta were reared on an artificial diet (Davidowitz et al., 2003) at a constant temperature of 25 °C and a long-day photoperiod of 16 h light and 8 h darkness. Under these conditions, larvae pupate between 6 h and 8 h after lights off. The day of pupation was designated as day zero, and pupae were synchronized relative to this time point. The duration of the pupal stage is 18–20 days.

Dissections and wing preparation

Pupae were anesthetized with CO2, their wings were dissected under lepidopteran saline (4 mM NaCl,

Results

Immediately after pupation the cells of the pupal wing are undifferentiated and are tightly attached to the pupal cuticle. Apolysis (separation of cells from the pupal cuticle) occurs late on day 2 of the pupal stage and signals the initiation of adult development of the wing. Our observations on cell division and the development of polyploidy thus begin on day 3.

Nuclear size and ploidy

Separation of the wing epithelium from the pupal cuticle occurred early on day 3 of the pupal stage. This event (called apolysis) signals the onset of adult development. Polyploid nuclei first appeared in the wing epithelium on day 4 of the pupal stage. The development of polyploidy was preceded by nuclear enlargement of a subset of the epithelial cell population on day 3. The DNA content of the enlarged nuclei on day 3 was not different from that of the much smaller nuclei of the other

Conclusions

The cells in the developing wings of Manduca differentiate into two populations: one undergoes endoreplication and will form the scales of the adult wing, and the other continues to multiply by mitosis and will deposit the general cuticle of the adult wing. These cellular events take place over a 4-day period at the very beginning of adult development in the pupal stage. The results described in this paper raise several interesting questions for future research. First, what is the mechanism by

Acknowledgments

We thank Laura Grunert for technical assistance with confocal microscopy, and Viviane Callier and Inder Jalli for helpful discussions. This work was supported by Grant IOS-0744952 from the National Science Foundation and by a grant from the Howard Hughes Medical Institute to Trinity College of Duke University.

References (29)

  • K. Henke

    Über die verschiedenen Zellteilungsvorgänge in der Entwicklung des beschuppten Flügelepithels der Mehlmotte Ephestia kühniella Z

    Biologisches Zentralblatt

    (1946)
  • K. Henke

    Einfache Grundvorgänge in der tierischen Entwicklung I. Über Zellteilung, Wachstum und Formbildung in der Organentwicklung der Insekten

    Naturwissenschaften

    (1947)
  • K. Henke et al.

    Differentielle Zellteilungen und Polyploidie bei der Schuppenbildung der Mehlmotte Ephestia kühniella Z.

    Zeitschrift für Naturforschung – B

    (1952)
  • J.M. Janssen et al.

    Correlations between scale structure and pigmentation in butterfly wings

    Evolution & Development

    (2001)
  • Cited by (25)

    • Insights into eyespot color-pattern formation mechanisms from color gradients, boundary scales, and rudimentary eyespots in butterfly wings

      2019, Journal of Insect Physiology
      Citation Excerpt :

      Accumulation of knowledge on the adult and pupal wings is expected in due course. Histological analysis (Nardi and Magee-Adams, 1986; Otaki et al., 2005; Cho and Nijhout, 2012) of the pupal wing tissue and morphological examinations of adult wings at the scale level (Kusaba and Otaki, 2009; Iwata and Otaki, 2016a,b; Taira and Otaki, 2016) are important. Furthermore, recent advances in real-time imaging techniques together with the forewing-lift method made it possible to observe living cells in vivo (Ohno and Otaki, 2015a,b; Iwata et al., 2014; Iwasaki et al., 2017) as well as the whole-wing development (Kusaba and Otaki, 2009; Iwata et al., 2014, 2018).

    • Wing morphogenesis in Lepidoptera

      2018, Progress in Biophysics and Molecular Biology
      Citation Excerpt :

      Irrespective of their final morphology, all wings begin their development as nearly identical crescent-shaped imaginal disks. The number of cells in the wing can be accurately counted at all stages of development, and with confocal z-sections it is possible to measure cell volumes and nuclear volumes accurately (e.g. (Cho and Nijhout, 2013)). It is therefore possible to ascribe local changes in shape and size explicitly to changes in cell number, cell volume and ploidy level of the nucleus.

    • Why are morpho blue?

      2018, Biodiversity and Evolution
    • Spatial patterns of correlated scale size and scale color in relation to color pattern elements in butterfly wings

      2016, Journal of Insect Physiology
      Citation Excerpt :

      We speculate that a group of cells from a single stem cell may be physically connected with each other with epidermal feet and with cytonemes or nanotubes, forming a cluster of cells (Ohno and Otaki, 2015a) to regulate the total number of genomic amplifications. It has been demonstrated that polyploidization causes a cellular volume increase in scale cells in Manduca sexta (Cho and Nijhout, 2013). Except E. kühniella and M. sexta, polyploidy has not been demonstrated in lepidopteran wings.

    • When bigger is better: The role of polyploidy in organogenesis

      2015, Trends in Genetics
      Citation Excerpt :

      In the moth Manduca sexta pigmentation patterns on the wing are controlled by the size of the cells that build the scales bearing the pigment colors. A recent study confirms the hypothesis that ploidy of the scale-building cells correlates with their size, and thus with the size of pigmented areas on the wings [35]. Ploidy of scale-building cells varies by position across the wing, being highest (32–64C) at the distal margin.

    View all citing articles on Scopus
    View full text