Cell Structure and Function
Online ISSN : 1347-3700
Print ISSN : 0386-7196
ISSN-L : 0386-7196
Temporal and Spatial Pattern of Dref Expression during Drosophila Bristle Development
Akihito KawamoriKouhei ShimajiMasamitsu Yamaguchi
著者情報
ジャーナル フリー HTML

2013 年 38 巻 2 号 p. 169-181

詳細
Abstract

The DNA replication-related element-binding factor (DREF) is a BED finger-type transcription factor that has important roles in cell cycle progression. In an earlier study, we showed that DREF is required for endoreplication during posterior scutellar macrochaete development. However, dynamic change in the dref expression in the cell lineage is unclear. In this study, we focused on the spatio-temporal pattern of expression of the dref gene during bristle development. Gene expression analysis using GAL4 enhancer trap lines of dref and the upstream activation sequence-green fluorescent protein with nuclear localization signals (UAS-GFPnls) in combination with immunostaining revealed the half-life of GFPnls in vivo (<6 hours) is short enough to monitor the dref gene expression. The analysis revealed that the dref expression occurs in clusters that include cells consisting of a bristle as well as surrounding epidermal cells. The intensity of GFP signals was almost the same in those cells, suggesting expression of the dref gene in bristle cell lineages occurs simultaneously in clusters. Further analysis showed that GFP signals increased twice during sensory organ precursor development as well as in bristle development at 9 hours and 15 hours after pupal formation, respectively. However, its expression was barely detectable in the cell lineages in and around asymmetric cell division or at other stages of development. For the first time, we clarified a spatio-temporal pattern of expression of the dref gene in vivo and revealed that expression of the dref gene occurs in clusters and is temporally regulated at specific times during bristle development.

Introduction

The transcription regulatory factor DNA replication-related element-binding factor (DREF) was first discovered in Drosophila and was later found to have orthologs in other species, including human (Matsukage et al., 2008). DREF forms homodimers that bind to the DNA replication-related element (DRE; 5′-TATCGATA) sequence to activate various target genes. (Emberly et al., 2008; Matsukage et al., 2008; Suyari et al., 2009; Park et al., 2010; Trong-Tue et al., 2010; Fujiwara et al., 2012; Killip and Grewal, 2012; Valadez-Graham et al., 2012; Yoshioka et al., 2012). It was shown that DREF forms a complex with many chromatin regulatory proteins, including the TATA box-binding protein (TBP)-related factor 2 and activates the promoter of the PCNA gene (Hochheimer et al., 2002). Earlier studies provided lines of evidence showing DREF is required in vivo for endoreplication in salivary gland cells (Hirose et al., 1999) and in shaft cells of mechano-sensory bristle cell lineage (Kawamori and Yamaguchi, 2011), efficient proliferation in mitotic wing disc cells (Hyun et al., 2005) and cell metabolism (Killip and Grewal, 2012). Moreover, the human ortholog of Drosophila DREF (hDREF) is required for proliferation of HeLa cells by regulating transcription of the histone H1 gene (Ohshima et al., 2003) and ribosomal protein genes (Yamashita et al., 2007). These studies therefore suggest that DREF has common important roles in cell cycle progression in a variety of metazoans.

The distributions of DREF and mRNA differ in different types of cells and developmental contexts. For example, in early embryos during the first 7 nuclear division cycles, DREF was present in ooplasm but not in nuclei, whereas after cycle 8, DREF is localized mainly in nuclei (Yamaguchi et al., 1995; Hirose et al., 1996). In somatic cycling cells, for example the mitotic cells of the wing disc, the distribution of DREF was observed to be similar to those of DNA replication enzymes such as PCNA, DNA polymerase α and DNA polymerase ɛ in the nuclei of cells (Yamaguchi et al., 1995). Elevated levels of DREF and mRNA in eye discs were observed in the cells just posterior to the morphogenetic furrow (MF), which correspond to the S-phase zone in the second mitotic wave and low levels of DREF were present even in the post-mitotic cells undergoing differentiation (Hirose et al., 2001; Hyun et al., 2005). In addition, relatively high levels of expression of DREF were observed in nuclei at the apical tips of testes (Ida et al., 2007). In endocycling cells of salivary glands, strong nuclear staining of anti-DREF, anti-PCNA, anti-DNA polymerase α and anti-DNA polymerase ɛ was observed in most of the cells in salivary glands of the second instar larva, whereas no significant staining was observed in the salivary gland cells of the late third instar larva, except those in and around the imaginal rings (Yamaguchi et al., 1995). These observations indicate that, in general, DREF is expressed at a high level in actively dividing or endocycling cells. Although there are many studies of DREF expression in the literature, detailed analysis of the dref gene expression pattern in a single cell lineage is not available.

The mechano-sensory bristle system, with which cell lineage analysis can be performed, is an excellent model system for the study of various cell biological events (Jan and Jan, 1998; Audibert et al., 2005; Tilney and DeRosier, 2005). There are two types of mechano-sensory bristle in the Drosophila adult thorax. There are ∼200 small bristles of microchaete and ∼22 large bristles of macrochaete on the adult thorax (Simpson, 1990). Each type of bristle consists of four types of differentiated cells, the shaft, socket, sheath and neuron, which are differentiated via a series of asymmetric cell division (Hartenstein and Posakony, 1989; Gho et al., 1999; Reddy and Rodrigues, 1999; Audibert et al., 2005). Sensory organ precursors (SOPs) of macrochaetes emerge in groups of cells called proneural clusters (PNCs) in the wing disc notum at the third instar larval stage (Huang et al., 1991). Timings of SOP differentiations and subsequent asymmetric cell division depend largely on the type of macrochaete. For example, an SOP of a posterior scutellar (pSC) macrochaete is formed at 30 hours before puparium formation (BPF) and begins its asymmetric cell division at around time zero after puparium formation (APF). While a SOP of an anterior scutellar (aSC) macrochaete is formed at 6 hours BPF and begins its division at ∼3 hours APF (Huang et al., 1991). After differentiation ends, several successive rounds of endoreplication take place for a shaft cell and a socket cell of micro- and macrochaete cell lineages (Huang et al., 1991; Audibert et al., 2005; Kawamori et al., 2012; Salle et al., 2012). At 25°C, bristles begin to emerge from pupae 32 hours APF (Tilney et al., 1996).

Earlier, we described the detailed roles of DREF during pSC macrochaete development; however, little is known about how the dref gene is regulated during development. Clarification of the expression patterns will allow deeper understanding of the roles of DREF in vivo. Here, we utilized the GAL4-enhancer trap line for the dref gene and the upstream activation sequence-green fluorescent protein (UAS-GFP) reporter system in combination with immunostaining, and quantitatively examined the spatio-temporal expression patterns of the dref gene during pSC macrochaete development. For the first time, our analysis revealed that expression of the dref gene occurs in clusters and is temporally regulated at specific time points during bristle development.

Materials and Methods

Fly stocks

The following fly stocks were used in this study. Canton S, neurA101 (Huang et al., 1991) (a marker for imaginal disc SOPs and its daughter cells), w*; P{GawB}DrefNP4719/CyO (DGRC, 113503), y1 w*; P{GawB}sca109-68/CyO, y1w1118, UAS-GFPnls (DGRC, 107870), UAS-DREF-RNAi15 (Yoshida et al., 2004).

Creation of the DREF reporter line

The DREF reporter line P{GawB}DrefNP4719, UAS-GFPnls/Cyo, green balancer was created by recombination of P{GawB} DrefNP4719 and UAS-GFPnls on the second chromosome. Flies with the genotype P{GawB}DrefNP4719/UAS-GFPnls were crossed with flies with the genotype Gla/CyO, green balancer. A wandering larva with strong green fluorescence in the salivary gland, wing disc (Fig. 1A and B) and gut (the green fluorescent protein (GFP) is from green balancer) was selected under a fluorescence microscope (Olympus SZX12).

Fig. 1

The GAL4 enhancer tarp line of dref expresses GAL4 in the same pattern as expression of endogenous DREF. The genotypes of flies were as follows: w*/+ and P{GawB}DrefNP4719,UAS-GFPnls/+ (A–B); w*/+ and P{GawB}DrefNP4719/UAS-GFPnls + (C–E). Expression patterns of the dref genes monitored by GFPnls (the dref reporter) at ∼1 hour APF in a pupa (A and B), larval wondering stage wing disc (C–E). Note that C (whole wing disc), D (wing pouch) and E (notum) are differently magnified images in the same sample. Endogenous DREF signals in a wing disc at the third instar stage (C–E). Larvae were cultured at 25°C and their tissues were stained with rabbit anti-GFP antibodies (green in D and E), mouse anti-GFP antibodies (green in C′, C″, D′, D″, E, E″) and rabbit anti-DREF antibodies (red in C, C″, D, D″, E, E″).

Immunohistochemistry

Collection and dissection of pupae and larvae were done as described (Kawamori and Yamaguchi, 2011). The dissected larvae or pupae were fixed in 4% (v/v) paraformaldehyde (PFA) for 30 min. After washing with PBS-T (PBS containing 0.3% (v/v) Triton X-100), samples were incubated with the following primary antibodies for 16 hours at 4°C or for 2 hours at 25°C: mouse anti-lacZ (Developmental Studies Hybridoma Bank, DSHB, 1:500); mouse anti-achaete (DSHB, 1:100) ; rat anti-Su(H) (a kind gift from Dr F. Schweisguth, 1:250) (Gho et al., 1996); rabbit anti-GFP (MBL, 1:500); mouse anti-GFP (Invitrogen, 1:500); and rabbit anti-DREF (Hirose et al., 1996, 1:100). After washing with PBS-T, samples were incubated for 2 hours at 25°C with the following secondary antibodies: anti-rabbit IgG conjugated with Alexa 488 or anti-mouse IgG conjugated with Alexa 594 (Molecular Probe, 1:600). After washing, the samples were mounted with Vactashield mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI; Vector Laboratories). Images were obtained with a fluorescence microscope (Olympus BX-50) equipped with a cooled CCD camera (ORCA-ER; Hamamatsu Photonics, Japan) and analyzed with AQUACOSMOS Ver.2.5 software (Hamamatsu photonics, Japan).

Temperature shifting and quantitative analysis of bristle length

The dref knockdown flies (UAS-DREF-RNAi15/UAS-DREF-RNAi15: sca-GAL4/+) were cultured until 15, 24 and 48 hours APF at 18°C. Pupae were collected, transferred to 28°C and cultured until adult emergence. Bristle length was measured as described (Kawamori and Yamaguchi, 2011).

Imaging of GFP and quantification of gene expressions in bristle cell lineages

Imaging of GFP in pupae was done with a microscope (SZX10®, Olympus) equipped with filters for GFP fluorescence (Fig. 1A and B). For quantification of gene expression in bristle cell lineages, the number of SOPs or bristle cell lineages that express GFP at a particular developmental timing was counted. GFP expression ratios at those times were determined by dividing the number by the total number of SOPs or cell lineages observed.

Statistical analysis

All statistical analysis used Microsoft Excel Toukei (version 6.0; Esumi) software. Scheffe's post hoc test was used for comparison of more than two groups. Statistically significant difference was set at P<0.01. All data are reported as mean±SEM.

Results

GAL4 enhancer tarp line of dref expresses GAL4 in a pattern similar to that of the endogenous DREF as a whole in wing discs

The GAL4 enhancer trap lines combined with the GAL4/UAS system (Brand and Perrimon, 1993) has been used to identify gene expression patterns in vivo (Gerlitz et al., 2002; Hayashi et al., 2002; Seroude, 2002; Ward et al., 2002). To better understand the important roles the DRE/DREF system has during bristle development, we examined the gene expression pattern of dref using a GAL4 enhancer trap line of the dref gene combined with the DREF reporter line UAS-GFPnls (P{GawB}DrefNP4719, UAS-GFPnls/Cyo, green balancer) (see Materials and Methods). The GAL4 enhance tarp line of dref is one of the 6966 enhancer trap lines that have been well characterized (Hayashi et al., 2002). The line has the P{GawB} element inserted into the 5′ noncoding region of the dref gene. The DREF reporter lines expressed strong green fluorescence in salivary gland cells during larval and pupal development (Fig. 1A, B) as reported (Hayashi et al., 2002). The endogenous DREF was detected ubiquitously in the wing disc with significantly higher expression in the wing pouches than in the notum (Fig. 1C) (Yoshida et al., 2004). Similarly, relatively stronger expression of GFP in wing pouches was detected with spotted and randomly distributed weaker expression of GFP in the wing notum (Fig. 1C–C″ and E–E″). The merged image with higher magnification revealed that slightly stronger expression of GFP in the anterior region of the wing pouch (Fig. 1D–D″). The slight difference between GFP expression and DREF expression may be due to difference in protein stability between GFP and DREF, since we found the half-life of GFPnls in vivo is relatively short (less than 6 hours) at this stage as described below. Thus, the results suggest that the GFP expressions of reporter lines likely represent the endogenous pattern of dref expression.

Characterization of the DREF reporter for gene expression analysis

Earlier studies with mammalian cultured cells reported that GFP is stable with a half-life of 26 hours (Corish and Tyler-Smith, 1999). GFP can be destabilized to a half-life of 2, 5.8 or 9.8 hours by addition of the signal sequence (Li et al., 1998; Corish and Tyler-Smith, 1999). GFP with nuclear localization signals (GFPnls), a variant of GFP (Shiga et al., 1996), is reported to be unstable but, to our knowledge, the precise half-life of this GFP variant in vivo is not determined. To examine the turnover of GFPnls during bristle development, we first determined the precise timing of asymmetric cell division in aSC bristle cell lineages by counting the number of cells in both pSC and aSC cell lineages at 0, 1, 2, 3, 4, 5 and 6 hours APF. The cell lineages were identified with a lacZ enhancer trap line of neurA101 (Huang et al., 1991) as a maker for imaginal disc SOPs and its daughter cells. Immunostainng with anti-lacZ showed 80% of the aSC cell lineage is in the SOP state at 2 hours APF; however, this value decreased to 43% at 3 hours APF and to 11% at 4 hours APF (Table I), suggesting that a major peak of asymmetric cell division in the aSC lineage is present at 3–4 hours APF. Next, we monitored changes in the GFPnls signals, focusing on the aSC macrochaete lineage (see Materials and Methods) (Fig. 5K). GFP signals were detected at 24% at time zero APF. However, at 1 hour later, apparent GFP signals were detected at 65% (Fig. 5B, B′ and K). After this stage, the GFP expression ratio became 13, 16, 0, 32 and 18% for 2, 3, 4, 5 and 6 hours APF, respectively (Fig. 5D–G and K). The results given in Table I indicate that the half-life of GFPnls is <1 hour; moreover, two peaks of GFP expression were observed in pSC SOPs at 90 and 108 hours after egg laying (AEL) (Fig. 3G), indicating that stability of the GFPnls is also short, half-life <6 hours in this case. Together, these data indicate that the DREF reporter can be used to monitor expression of the dref gene during bristle development.

Table I. Number of cells in pSC or aSC macrochaete cell lineage from 0–6 h APF
No. of cells in PSC lineages 1 2 3 3 4 4 4 5 5 5 5 4 4 No. of samples
No. of cells in ASC lineages 1 1 1 2 1 2 3 1 2 3 4 4 5
Hours APF % of observations
0 0 100 0 0 0 0 0 0 0 0 0 0 0 9
1 0 79 21 0 0 0 0 0 0 0 0 0 0 28
2 0 17 33 3 30 7 0 0 7 3 0 0 0 30
3 0 7 21 4 11 21 0 4 32 0 0 0 0 28
4 0 0 0 0 0 33 11 11 44 0 0 0 0 9
5 0 0 0 0 0 31 0 0 19 50 0 0 0 16
6 0 0 0 0 0 0 0 0 0 38 23 15 23 13

Timings of asymmetric cell division of aSC or pSC cell lineage from 0–6 hours APF were examined at intervals of 1 hour. The numbers of cells in both cell lineages were counted and classified into different groups according to the number of cells in both cell lineages. For example, at 1 hour APF, 79% of flies had 2 cells in pSC and 1 cell in aSC cell lineage, and 21% of flies had 3 cells in pSC and 1 cell in aSC cell lineages.

Expression of the dref gene in bristle cell lineages occurs simultaneously in clusters including the cell lineages and surrounding cells

We determined the pattern of GFPnls expression during bristle development. The immunofluorescence data showed that the DREF reporter expresses GFPnls in clusters of cells, which run from anterior to posterior during thorax development (Fig. 2A–A‴). The expression profiles appear to be similar until at least 30 hours APF (Fig. 2, Fig. 3, Fig. 4, Fig. 5, Fig. 6). Microchaete (white arrowheads in Fig. 2A′) and macrochaete (yellow arrowheads in Fig. 2A′) cell lineages can easily be distinguished by nucleus size, cell number and relative positions of the lineages in the thorax, since positions of those bristle cell lineages in the thorax are fixed and the cell differentiation times are different between microchaete and macrochaetes. Surprisingly, a SOP of a microchaete is included in the clusters (Fig. 2A–A‴ and B–B‴, see arrowheads). The intensity of GFPnls expression in the SOP is almost the same as those in surrounding epidermal cells (Fig. 2A–A‴ and B–B‴). The data suggest that expression of the dref gene in a SOP of microchaete and the surrounding cells occurs simultaneously during development.

Fig. 2

Expression of the dref gene in bristle cell lineages occurs simultaneously in clusters, including the cell lineages and surrounding cells. The expression pattern of the dref reporter in a whole thorax at 15 hours APF (A–A‴) and at 21 hours APF (B–B‴). Microchaete (white arrowheads in A′) and macrochaete (yellow arrowheads in A′) cell lineages are distinguished by nucleus size, cell number and relative positions of the lineages in the thorax. The pattern of expression of the dref reporter in a shaft cell nucleus (arrowheads in C–C‴) and nuclei of a socket cell (arrows in D–D″), sheath cell and neuron (arrowheads in D′–D″, the larger appears to be a sheath cell) of pSC macrochaete cell lineages at 21 hours APF. Pupae with genotype w*/+; P{GawB}DrefNP4719,UAS-GFPnls/+; neurA101/+ (A–D‴) (neurA101 is a marker for imaginal disc SOPs and its daughter cells which express lacZ in the cells) were cultured at 25°C and their thoraxes were stained with DAPI (cyan), anti-lacZ antibodies (magenta) or anti-GFP antibodies (green).

Fig. 3

Expression of the dref gene in an SOP of a pSC macrochaete. The expression pattern of the dref reporter in a pSC SOP from 90–120 hours AEL. Arrowheads indicate pSC SOPs (A–F). Larvae with genotype w*/+; P{GawB}DrefNP4719,UAS-GFPnls/+; neurA101/+ were cultured at 25°C and larval wing discs were stained with DAPI (cyan), anti-lacZ antibodies (magenta) or anti-GFP antibodies (green). Data for expression of the dref reporter (G). The Y-axis indicates the ratio of GFP-positive cells observed in an SOP at a particular developmental time point (X-axis).

Fig. 4

Pupariation occurs at 121∼126 hours AEL. Flies with appropriate genotype were mated for a day and left to lay eggs for 6 hours. Eggs and larvae with the genotype w*/+; P{GawB}DrefNP4719,UAS-GFPnls/+; neurA101/+ were cultured at 25°C until pupariation and the number of female pupae was counted. The Y-axis indicates the number of female pupae; the X-axis indicates time (h) AEL when the pupae were collected.

Fig. 5

Expression of the dref gene in a pSC macrochaete cell lineage during asymmetric cell division. The expression pattern of the dref reporter in a pSC macrochaete cell lineage from 0–6 hours APF (A–J′). Larvae with the genotype w*/+; P{GawB}DrefNP4719,UAS-GFPnls/+; neurA101/+ were cultured at 25°C and pupal thoraxes were stained with anti-lacZ (magenta), anti-GFP (green) or DAPI (data not shown). Arrowheads indicate nuclei of pSC cell lineages. Arrows indicate nuclei of aSC cell lineages. Note that D, D′ and E, E′ are images with different foci of the same sample. Similarly, G, G′ and H, H′ are images with the different foci of the same sample. Data for the expression of the dref gene (K). The Y-axis indicates the ratio of GFP-positive cells observed in pSC bristles (dark gray line) or aSC bristles (gray line) at a particular developmental time point; the X-axis indicates time (h) APF.

Fig. 6

Expression of the dref gene in a pSC macrochaete cell lineage after asymmetric cell division. Expression pattern of the dref reporter in a pSC macrochaete cell lineage from 7–30 hours APF (A–R). Larvae with the genotype w*/+; P{GawB}DrefNP4719,UAS-GFPnls/+; neurA101/+ were cultured at 25°C and stained with anti-lacZ (magenta), anti-GFP (green) or DAPI (data not shown) (A–Q). Data for expression of the dref reporter (R). The X-axis indicates time (hour) APF; the Y-axis indicates the GFPnls expression ratio in pSC bristles at a particular developmental time point.

The data described above prompted us to determine whether a similar expression profile is observed in macrochaete cell lineages. We monitored the GFPnls expression in the pSC bristle cell lineages. As expected, the GFPnls signals were observed in all types of differentiated cells (Fig. 2C–C‴ and D–D‴). The intensity of GFPnls expression in a shaft cell, a socket cell and two other types of cell appeared to be similar, suggesting expression of the dref gene occurs to the same extent at the same time. In addition, similar expression profiles were observed at the different developmental time points (Fig. 3A, A′, D, D′, 5C, C′, I, I′, 6B, B′, C, C′, F, F′, G, G′, I, I′, J, J′, K and K′). These data indicate that expression of the dref gene in bristle cell lineages occurs simultaneously in clusters, including the cell lineages and surrounding cells.

Expression of the dref gene occurs at least twice during SOP development

We monitored expression of the dref gene, focusing on a pSC bristle lineage from SOP formation to the onset of asymmetric cell division. The formation of pSC SOPs occurs about 30 hours BPF (Huang et al., 1991). Formation of the puparia begins at ∼120 hours AEL (Fig. 4). So, 30 hours BPF corresponds to ∼90 hours AEL. We collected wing discs of flies with DREF reporters at 90, 96, 102, 108, 114 and 120 hours AEL and determined GFPnls expression ratios in pSC SOPs (see Materials and Methods). The results showed peaks of GFPnls expression ratios of 44% at 90 hours AEL (Fig. 3A, A′, G) and 55% at 108 hours AEL (Fig. 3D, D′, G); however, the ratios were less than 13% at the other phases (Fig. 3B, B′, C, C′, E, E′, F, F′ and G). The data suggest there are at least two time points of dref activation during SOP development.

Expression of the dref gene in a pSC macrochaete cell lineage in and around asymmetric cell division

We monitored expression of the dref gene from 0 to 6 hours APF in the pSC bristle cell lineages at intervals of 1 hour (see Materials and Methods). The results showed that low expression ratios of 14, 19 and 10% were observed for GFPnls at 0, 1 and 5 hours APF, respectively (Fig. 5) and no signal was observed at any other developmental stage. The results suggest the expression of the dref gene occurs at around the beginning of asymmetric cell division, although we do not know why a clear peak of expression was not observed.

Most of the dref gene expression ends within a few hours after the end of asymmetric cell division

We monitored expression of the dref gene from 7 hours APF until 30 hours APF at intervals of 1 hours (Fig. 6; and see Materials and Methods). The results showed that GFP signals were not detectable at 7 hours APF (Fig. 6R). The expression ratio increased gradually to 17% at 8 hours APF (Fig. 6A) and 38% at 9 hours APF (Fig. 6B and C) and decreased to zero at 10 hours APF. The results suggest that the dref gene might be activated at ∼9 hours APF. A relatively high GFP expression ratio was observed only at ∼15 hours APF (Fig. 6F, G and K). However, low GFP expression ratios were observed at other developmental stages until 24 hours APF (Fig. 6D–M and R). Surprisingly, the GFP signals in the cell lineages disappeared after 24 hours APF (Fig. 6N–R). These results suggest that the dref gene is actively expressed up until 15 hours APF but not later than that.

Expression of the dref gene until 12 hours APF at the latest is required for endoreplication of shaft cells

To obtain further evidence that most of the dref gene expression ends a few hours after the end of asymmetric cell division (Fig. 6), we examined the effects of dref knockdown at a specific time during bristle development. We compared the pSC bristle length of dref knockdown flies (UAS-DREF-RNAi15/UAS-DREF-RNAi15: sca-GAL4/+) that were cultured at 18°C (control) and those transferred to 28°C after 12, 15, 24 or 48 hours APF at 18°C. We confirmed that the average length of the bristle in the controls was almost the same as that of wild type flies (Fig. 7D), suggesting that GAL4 and/or its expression is not significantly active in flies grown at 18°C. The average bristle length of pupae that were cultured for 12 or 15 hours APF at 18°C and then shifted to 28°C was significantly shorter (Scheffe's test, P<0.01) compared to the control (Fig. 7D). However, there was no significant difference in bristle length between the control and pupae that were cultured for 24 or 48 hours APF at 18°C (Fig. 7D). The data suggest that most of the dref gene expression at 18°C ended by 24 hours APF. However, it is still not known which time points APF at 25°C correspond to 15 and 24 hours APF at 18°C.

Fig. 7

Effects of dref knockdown on the endoreplication process at a specific time point of bristle development. Shaft and socket cells of flies grown at 18°C are specified at least by 15 hours APF (A–C). White arrowheads indicate pSC shaft cell nuclei. White arrows indicate pSC socket cell nuclei. Larvae with the genotype UAS-DREF-RNAi15/UAS-DREF-RNAi15; P{GawB}sca109-68/+; neurA101/+ were cultured at 18°C until 15 hours APF and stained with anti-lacZ (green) or anti-Su(H) (magenta). The effects of dref knockdown on pSC bristle length after transferring flies from 18°C to 28°C at specific developmental time points (D). Control flies were cultured at 18°C throughout development. The genotype of the flies was as follows: UAS-DREF-RNAi15/UAS-DREF-RNAi15; P{GawB}sca109-68/+; +/+. The X-axis indicates time (hour) APF when flies were transferred to 28°C. The Y-axis indicates the mean pSC bristle length. Statistical tests were done as described in Materials and Methods. Significant differences are indicated by letters above bars. n=13, 22, 13, 20 and 13 corresponds to sample numbers of the control at 12, 15, 24 and 48 hours APF.

To clarify this, we carried out immunostaining experiments with anti-Su(H) antibodies, a marker for socket cells (Gho et al., 1996). The Su(H) signal was detected in one among five cells at 15 hours APF at 18°C (Fig. 7A–C), suggesting that cells in the pSC bristle lineages have already differentiated into the five cells state. These results, together with the data given in Table I, suggest that 15 hours APF at 18°C corresponds to ∼6 hours APF at 25°C. Moreover, we noticed that the oral armature of the larva that is expelled ∼12 hours APF at 25°C (Bainbridge and Bownes, 1981) remains until ∼32 hours APF at 18°C (data not shown). The data suggest that 24 hours APF at 18°C might correspond to <12 hours APF, or earlier, at 25°C. Taken together, the data support the suggestion that most expression of the dref gene ends a few hours after the end of asymmetric cell division and expression of the dref gene is required for endoreplication of shaft cells.

Discussion

Expression of the dref gene is regulated at specific developmental time points during the development of pSC bristles

Our study revealed a spatio–temporal pattern of the expression of the dref gene during development of pSC macrochaetes. We used DREF reporter lines (Fig. 1 and Fig. 2) and found that the dref gene is activated at least twice during SOP development (Fig. 3). However, no clear peak of dref gene expression in and around asymmetric cell division was obtained (Fig. 5). The dref gene is activated after the end of asymmetric cell division, possibly twice at ∼9 hours and ∼15 hours APF (Fig. 6). Surprisingly, the dref gene is not active during the later stage after asymmetric cell division and the expression of the dref gene is finished 24 hours APF (Fig. 6). The reason for the low percentage of GFP signals in this stage is unclear (Fig. 5 and Fig. 6). The asynchronous age of pupae, influenced by the growing conditions, could be the technical reason. However, it is likely that almost inactive but temporal gene activation occurs during development. If some other nuclear protein(s) can substitute function of DREF for pSC bristle development, the stochastic expression of dref could be allowed. The apparently stochastic expression of dref may thus imply a non-essential role of DREF for pSC bristle development. In addition, we also cannot exclude the possibility that the periods of dref expression may significantly vary among individuals. We also found that dref knockdown by <12 hours APF at 25°C results in short bristle phenotypes (Fig. 7). Therefore, expression of the dref gene, required for endoreplication in shaft cells, is probably activated by the early stage of development of the pSC macrochaete. In the present study, we used A101 as a SOP marker. Using markers for early SOP development or using live imaging system may allow us to determine more precise timing of DREF expression.

We found evidence that expression of the dref gene is temporally regulated in a non-cell autonomous manner (Fig. 2 and Fig. 8). The idea is further supported by the observation that its expression occurs simultaneously in clusters (Fig. 2, Fig. 3, Fig. 5 and Fig. 6). It is reported that cell cycle progression in the wing disc occurs in evenly spaced clusters (Hartenstein and Posakony, 1989; Milan et al., 1996). Thus, the manner of expression of the dref gene might be correlated to this pattern of cell cycle progression. Possible synchronization of cell cycle progression in bristle cell lineage and those in surrounding epidermal cells may suggest link between cell cycle regulation and some positional information in tissues. Further studies are necessary to explore this possibility.

Fig. 8

A model of temporal and spatial expression pattern of the dref gene during pSC macrochaete development. All data were combined into this graph. The X-axis indicates time (hour) APF; the Y-axis indicates GFP expression ratios in the pSC bristle lineage. Socket cell, shaft cell, sheath cell and neuron are indicated by so, sf, sh and n, respectively (magenta). Cells marked with green indicate DREF-expressing cells. Cells marked with blue indicate surrounding cells.

Which upstream factors regulate expression of the dref gene is the next question to be answered. Transcription factor dMyc is required for normal expression of the dref gene in Drosophila (Thao et al., 2008). It is reported that a high level of dMyc protein is preset in the developing wing pouch (Prober and Edgar, 2002), where expression of the dref gene is relatively active (Fig. 1). Furthermore, over-expression of dMyc can up-regulate levels of the DREF protein in wing discs (Thao et al., 2008). Therefore, dMyc is likely to be involved in regulation of the expression of the dref gene in the development of pSC macrochaetes. Future analysis will be required to clarify these points.

Is expression of the dref gene inactive during endocycling stages?

The results of the present study suggest that most dref gene expression ends in a few hours after the end of asymmetric cell division when endocycling is still active (Fig. 6 and Fig. 7). This observation is consistent with the report that the level of expression of cell cycle-related genes is relatively low in endocycling cells (Maqbool et al., 2010). However, Our additional experiment indicated expression of the dref gene in the endocycling salivary gland is continuous during development (Data not shown). This discrepancy could be due to the presence of the salivary gland enhancer present in the hsp70 sequence upstream of the GAL4 coding region, as described (Duffy, 2002; Gerlitz et al., 2002). However, we cannot exclude the possibility that the regulation of dref expression in bristle cell lineages is different from those in salivary gland cells, because multiple gene loci are amplified extensively in the salivary gland. Interestingly, our preliminary data indicated that the DREF protein exists in shaft cells during macrochaete development even after expression of the dref gene has stopped (data not shown). This observation raises the possibility that expression of the DREF protein during endocycling stage is controlled in a post-transcriptional manner and DREF might have a role in the regulation of target gene expression at this endocycling stage. Determination of the dynamics of DREF mRNA and protein expression pattern during development would be necessary to determine when DREF activates expression of its target gene during bristle development.

Acknowledgments

We thank the Developmental Studies Hybridoma Bank and Dr F Schweisguth for antibodies; the Bloomington Stock Center and Drosophila Genetic Resource Center for fly lines; Drs Hajime Mori and Eiji Kotani for access to a SZX16 microscope and useful advice.

References
 
© 2013 by Japan Society for Cell Biology
feedback
Top