The objective of this study is to examine the functions of Dll in developing butterfly wings. To do this, a Dll-gfp fusion gene was transferred to living wing tissues through a baculovirus vector. Our hypothesis was that if Dll is sufficient for eyespot development, ectopically expressed Dll should be able to construct ectopic eyespots or similar elemental color patterns.
We used the blue pansy butterfly J. orithya (Linnaeus, 1758). Female adult individuals were field-caught in Okinawa-jima or Ishigaki-jima in the Ryukyu Archipelago, Japan. This is a common butterfly in this region, and no permission is required to catch them in the field. Eggs were collected from these females. Alternatively, larvae were caught on these islands. Larvae were fed their natural host plants at ambient temperature (25–27 °C). Pupae were also placed at the same ambient temperature (25–27 °C) before and after experimental treatments.
Baculovirus design and production
We designed a recombinant Dll-gfp baculovirus vector that contained an expression unit for Dll (J. coenia Dll sequence: GenBank Accession No. AF404110.1) and green fluorescent protein (Aequorea victoria gfp sequence: GenBank Accession No. L29345.1). The construct was Dll-spacer-gfp-His6-stop (1131 + 24 + 714 + 18 + 3 = 1890 bp; GenBank Accession No. KP748528).
The entire baculovirus production process based on this sequence information was performed by Wako Pure Chemical Industries, Ltd. (Osaka, Japan). First, the entire construct was chemically synthesized with the flanking Xba I sequence at the 5’ end and Bgl II sequence at the 3’ end as a part of a plasmid pBMH. The construct was excised and subcloned into the cloning site of Xba I and Kpn I of a transfer vector pPSC8 (Protein Sciences, Meriden, CT, USA). Purified transfer vector (2 μg), linear baculovirus (AcNPV) DNA (85 ng), and Insect GeneJuice Transfection Reagent (5 μL) (Merck, Darmstadt, Germany) were mixed with Sf900II SFM (200 μL) (Gibco, Life Technologies, Carlsbad, CA, USA). The mixture was added to a 25 cm2 flask with 1.0 × 106 Sf9 cells. The cells were incubated at 28 °C for 6 days. The supernatant was collected as co-transfection medium. This co-transfection medium (1/200 of the culture volume) was added to infect express SF+ cells (1.5 × 106 cells/mL in Sf900II SFM) in a 100 mL culture in a 250 mL flask. This was incubated for 72 h at 28 °C with shaking (130 rpm). The culture medium was collected and centrifuged (3000 × g, 4 °C for 30 min). The supernatant, approximately 1 × 107 pfu/mL as estimated by a conventional plaque assay, was stocked for pupal injections.
The supernatant and pellets were subjected to SDS-PAGE and Western blot analysis using an anti-His antibody conjugated with horseradish peroxidase, Penta · His HRP (QIAGEN, Hilden, Germany). The blot signals were detected using Immobilon Western Chemiluminescent HRP Substrate (Millipore, Billerica, MA, USA). As predicted, the expressed Dll-GFP protein was clearly detected from the pellet and not from the supernatant (not shown). Dll-GFP was not secreted to liquid media from the infected cells.
In addition to the Dll-gfp baculovirus, we used a control gfp baculovirus that was obtained from AB Vector (San Diego, CA, USA) at the original baculovirus titer of 1 × 108 pfu/mL. In these gfp and Dll-gfp baculovirus vectors, gene expression was driven by the strong polyhedrin promoter [30, 32], and thus we assumed that GFP or Dll-GFP was expressed immediately after infection as early as 12 h post-infection .
For each baculovirus vector, the baculovirus dilution factor for injection and the post-infection time for antibody injection were optimized for the present study. This is partly because the Dll-gfp baculovirus vector appeared to be more toxic than the gfp baculovirus vector. Injection site was always located at the abdominal segments 5 or 6, which are considerably remote from pupal wings. This ensured that no physical damage on pupal wings was elicited during an injection process. We note that only heavy physical damage can induce ectopic patterns; accidental physical damage, if any, on wings during the injection process does not induce ectopic patterns.
For the Dll-gfp baculovirus vector, pupae were injected through the abdominal cuticle as mentioned above with 2.0 μL of a solution containing the recombinant baculovirus vector within 18–24 h after pupation using an Ito microsyringe (Fuji, Shizuoka, Japan). At the same site, 6 h post-infection, we injected 2.0 μL of mouse monoclonal anti-gp64 antibody IgG2a (200 μg/mL in PBS) against the baculovirus gp64 (AcV1) of extracellular nonoccluded AcNPV (Autographa californica nucleopolyhedrovirus) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) using an Ito microsyringe (Fuji, Shizuoka, Japan). The 18–24 h post-infection injection of antibody caused 100 % pupal mortality with 2, 5, 10, 100, and 1000 fold dilutions of baculovirus (although not with 10,000-fold dilutions). However, using 100- and 500-fold dilutions of the Dll-gfp baculovirus (1 × 105 and 2 × 104 pfu/mL; 2.0 μL) and anti-gp64 antibody injection (2.0 μL) 6 h post-infection, we were able to obtain GFP fluorescence from adult wings.
For the gfp baculovirus, we followed the original protocol  with some modifications; the diluted gfp baculovirus (1 × 106 pfu/mL or less; 2.0 μL) was injected 18-24 h post-pupation, followed by an injection of anti-gp64 antibody (2.0 μL) 18–24 h post-infection.
These gene transfer experiments were permitted by the Safety Committee for Genetic Recombination Experiments of the University of the Ryukyus.
Visualization of GFP fluorescent signals
When necessary, pupal wings from 4-day-old pupae were dissected following a published protocol with some modifications . The pupa was lightly anesthetized on ice. The cuticle around the wing margin was cut using a scalpel and lifted up to cut through the trachea connecting the wings to the thorax. Dissected wing tissues were placed on glass slides and then directly subjected to the fluorescent microscope to examine GFP fluorescence.
Whole pupae, whole adults, isolated pupal wings, or isolated adult wings were placed on the ATTO illuminator VISIRAYS-B (Tokyo, Japan), a blue-LED light unit with emission wavelengths λ = 440–500 nm and λmax = 470 nm. Under this illuminator, low magnification GFP fluorescence images were observed and recorded using a Canon digital single-lens reflex camera EOS 50D (Tokyo, Japan) with an ATTO filter SCF515.
For high magnification images of GFP fluorescence, we used a Nikon inverted epifluorescence microscope Eclipse Ti-U (Tokyo, Japan) equipped with a Nikon Intensilight C-HGFI (a mercury pre-centered fiber illumination system with a 130-W Hg lamp), a Nikon Epi-Fl Filter Cube GFP-B (EX480/40, DM505, and BA535/50), and a Hamamatsu Photonics ImagEM EM-CCD camera (Hamamatsu, Japan). This microscope hardware system was controlled with a Hamamatsu Photonics AQUACOSMOS/RATIO analysis system. For these observations, we either isolated pupal wings or lifted the forewing to expose the surface of the hindwing as described elsewhere [22, 25].
For bright-field low-magnification images, we used a Canon digital single-lens reflex camera EOS 50D (Tokyo, Japan) and a Saitou Kougaku microscope SKM-S30-PC (Yokohama, Japan). For bright-field high-magnification images, we used a Keyence high-resolution digital microscope VHX-1000 (Osaka, Japan) and the Nikon microscope system described above.
Detection of transcripts
Pupae were injected with the Dll-gfp baculovirus followed by anti-gp64 antibody. Three days after anti-gp64 antibody injection, pupal wings were dissected according to a standard protocol . To compare Dll gene expression levels between infected and non-infected individuals, we used 4-day-old infected and non-infected pupae. We also used non-infected first-day pupae for comparison. Isolated wings were readily frozen at –80 °C. The RNeasy Kit (QIAGEN) was used for RNA isolation. Total RNA was isolated from both the right and left dissected fore- and hindwings of three treated pupae (with 100-fold or 500-fold diluted baculovirus vectors) or three non-treated pupae. The isolated total RNA (340 ng per reaction) was subjected to RT-PCR using the AccessQuick RT-PCR System (Promega, Madison, USA) with AMV reverse transcriptase (Promega) and Tfl DNA polymerase (Promega).
The thermal cycling conditions for detecting the Dll-gfp transcript were 45 °C for 45 min, 95 °C for 2 min, 45 cycles of 95 °C for 1 min, 50 °C for 30 s, and 72 °C for 2 min, and lastly 72 °C for 5 min, using DLL2UP primer 5’-AAGTCTGCGTTCATAGAGTTACAGC-3’ and GFPDOWN primer 5’-GTATAGTTCATCCATGCCATGTGTAATC-3’. The expected size of the amplified DNA was 1708 bp.
To detect the Dll mRNA transcripts without gfp transcribed from the endogenous genomic DNA and from the baculovirus-mediated transgenes, RT-PCR was performed under thermal cycling conditions as follows: 45 °C for 45 min, 95 °C for 2 min, and 20 cycles of 95 °C for 1 min, 50 °C for 30 s, and 72 °C for 1 min 20 s, and the last incubation at 72 °C for 5 min. For Dll, DLL1UP primer 5’-ATGACCACCCAGGAGCTAGATCACC-3’ and DLL1DOWN primer 5’-AGGGTTGGCATCAGCCTGGTACCAG-3’ were used for the first round of PCR. Nested PCR was then performed with Tfl DNA polymerase using the DLL2UP primer described above and DLL2DOWN primer 5’-TACTGCGGCACGTAGGGCGGGTGCG-3’. Thermal cycling conditions for the nested PCR were set as follows: 95 °C for 2 min, 25 cycles of 95 °C for 1 min, 50 °C for 1 min, and 72 °C for 1 min 20 s, and the last incubation at 72 °C for 5 min. The expected size of the amplified DNA was 855 bp. The PCR products were subjected to electrophoresis using 1 % Agarose S (NIPPON GENE, Tokyo, Japan) with ethidium bromide (Promega).
After subjecting the PCR products to electrophoresis, the band intensities were measured and compared semi-quantitatively. Agarose gel images were taken with Image Quant LAS 400 (GE Healthcare Life Sciences, Piscataway, USA) and were used for image analysis using Image Quant TL 7.0 400 (GE Healthcare Life Sciences). A given gel was imaged three times, the band intensities were measured for each image, and their mean values were used as a final value for that gel. To compare Dll gene expression levels between infected and non-infected individuals, we used 4-day-old infected and non-infected pupae. We also used non-infected first-day pupae for comparison.
To examine the difference in expression levels, we performed two-sided Student’s t-test using IBM SPSS Statistics 19 (2010). To examine the difference in occurrence of ectopic color patterns between the two baculovirus constructs, Fisher’s exact test was performed using JSTAT 13.0 (2012).