RNA In Situ Hybridization for Detecting Gene Expression Patterns in the Abdomens and Wings of Drosophila Species

RNA in situ hybridization (ISH) is used to visualize spatio-temporal gene expression patterns with broad applications in biology and biomedicine. Here we provide a protocol for mRNA ISH in developing pupal wings and abdomens for model and non-model Drosophila species. We describe best practices in pupal staging, tissue preparation, probe design and synthesis, imaging of gene expression patterns, and image-editing techniques. This protocol has been successfully used to investigate the roles of genes underlying the evolution of novel color patterns in non-model Drosophila species.


Introduction
RNA in situ hybridization (ISH) is a method used to detect and localize specific mRNA transcripts in cells, tissues, and whole organisms. Early ISH procedures used radioactively labeled RNA probes that hybridized with denatured DNA in tissues. The RNA-DNA hybrids were then detected by autoradiography [1,2]. A significant technical advancement to this method was the development of non-radioactive labeling systems that facilitated colorimetric visualization, which allowed for gene expression patterns to be observed in intact Drosophila embryos [3]. mRNA distribution patterns can now be detected by treating fixed tissues with digoxigenin-labeled RNA probes, to which F ab fragments of anti-digoxigenin attached to alkaline phosphatase (anti-DIG-AP) are bound. The addition of 5-bromo-4-chloro-3-indoyl-phosphate (BCIP) and 4-nitroblue tetrazolium chloride (NBT) chromogenic solutions then results in the formation of a purple crystalline deposit, wherever the probe has bound to its target mRNA. The results, indicative of the gene expression patterns of interest, can be observed under a regular stereo microscope [3,4]. The non-radioactive ISH method is sensitive, easier, and safer.
ISH is a very powerful molecular tool used in research and diagnostics. The nonradioactive RNA ISH technique has been applied to studies in developmental biology, evolutionary biology, and cancer biology [5][6][7][8]. It has played a significant role in the detection of mRNA expression in humans, mice, insects, and in viral RNA detection [9][10][11][12][13][14].
Detailed protocols to perform RNA ISH in Drosophila embryos can be found in the published literature [4,15,16]. However, these protocols cannot be adapted for pupal wings and abdomens because additional critical steps are required for processing pupal tissues. Especially during the early pupal stage, RNA ISH can be very difficult to perform due to the fragility of the newly forming adult tissues. Although several studies have used the RNA ISH technique to detect and characterize gene expression patterns on pupal wings and abdomens of various Drosophila species [6,7,[17][18][19], no protocol describing this technique in Drosophila pupae has been published in any scholarly journal.
Here we provide a protocol suitable for model and non-model Drosophila species alike, detailing the steps of probe design and synthesis, pupal tissue dissection, the ISH process, and imaging techniques. We have applied this protocol to investigate the gene-regulatory networks governing the development and evolution of pigmentation patterns in pupal abdomens and wings of a variety of Drosophila species, such as D. melanogaster, D. guttifera, D. deflecta, D. palustris, and D. subpalustris [5,6,20]. The users of this protocol should pay close attention to the pupal tissue processing steps, as the outcome of an RNA ISH critically depends on the tissue quality.

Overview
The major steps involved in ISH for Drosophila abdomens and wings are outlined in the flowchart ( Figure 1). First, species-specific PCR primers are designed to amplify a partial coding region of the gene of interest, using genomic DNA as a template. The PCR product is then cloned into the vector pGEM ® -T Easy. A region containing the PCR product with flanking T7/Sp6 sites is PCR-amplified from this vector and used as a template for the synthesis of the DIG-labeled antisense RNA probe by in vitro transcription. Abdominal epidermis preparations involve pupal dissection, the removal of unwanted tissues by pipetting, fixation of the epidermal cell layer, and tissue storage. For wing preparations, the order of wing dissection and fixation depends on the pupal stage. In younger pupae, the wings are fixed before they are separated from the body, while in older pupae, the fixation step follows the separation of the wings. The important steps in the ISH procedure provided here are xylene treatment, tissue rehydration, fixation, proteinase K treatment, second fixation, prehybridization, probe addition, anti-DIG-AP F ab treatment, and NBT/BCIP staining. Overview of the ISH procedure in the wings and abdomens of Drosophila species. The major steps are shown sequentially in the boxes linked by arrows. The duration of each step is indicated in parentheses inside the boxes. Boxes on the right side describe the probe preparation steps, and boxes on the left side illustrate the sample preparation and ISH steps.

Application of This Protocol
ISH has been integral to the field of evolutionary developmental biology in that it is a valuable technique to study the emergence of novel traits, such as the color patterns of butterflies and fruit flies [17,18,21,22]. This protocol allows for the detection of gene expression patterns in abdomens and wings of different Drosophila species. We have used this procedure in model and non-model Drosophila species to study toolkit and terminal pigmentation genes involved in color pattern formation, such as D. melanogaster, D. guttifera, D. deflecta, D. palustris, and D. subpalustris [5,6,20]. No specialized skills are required to use this protocol; we have had undergraduate students generate high-quality results in our lab. We believe that this protocol will facilitate the study of novel gene expression patterns in rare and unstudied fruit flies, which can be collected and brought into the lab using the two new field guides to drosophilid species [23,24].

Advantages and Limitations
Our protocol allows for the detection of developmental toolkit genes' involvement in color patterning in Drosophila species. The most important advantage of this ISH protocol is that it enables the detection of gene expression patterns in very early pupal abdomens (as soon as the epidermal layer has formed, i.e., from pupal stage P7 onwards), as well as early wings that cannot be dissected without prior fixation (from stage P5ii onwards) ( Figure 2). This procedure requires basic laboratory equipment to generate high-quality ISH results [5,20]. However, this method is not without limitations. One of the limitations of the ISH technique is that it is semi-quantitative, and although it can detect and visualize spatial gene expression patterns, it is less accurate in determining quantitative gene expression differences than RNA-seq-based methods. Also, the final staining outcome of an ISH technique depends on the duration of the staining reaction, probe concentration, and fixation conditions, therefore making it sometimes difficult to generate the desired result in one attempt. Furthermore, the development of background staining is a limiting factor for how long a staining reaction can continue, which causes problems when a gene is weakly expressed.  [25] for D. melanogaster and adopted for D. guttifera by Werner et al. [6]. (MaITu stands for Malpighian tubule).

Probe Design and Synthesis
The process starts with designing PCR primers to amplify a partial protein-coding region from a single exon, using the GenePalette software [26] as described in Appendix A. Primers are 18-25 bases in length and are designed to yield products of 200-500 bp in size. We prepared and used external primers to amplify the PCR product and internal primers to amplify a shorter "internal PCR product". When comparing expression patterns of a gene among different Drosophila species, we used the multiple alignment tool in GenePalette to identify the most highly conserved regions of the open reading frame. For the probe design, we chose a region that did not contain indels among the species; thus, the PCR products for the same gene for different species would have the same length. It should be noted that the use of the same PCR primers that are designed from a highly conserved exon to amplify the amplicon in different Drosophila species may lead to mismatches in the sequence, which can reduce the hybridization efficiency of the probe. The PCR products (inserts) were then cloned into the vector pGEM ® -T Easy, which contains the Sp6 and T7 promoters required for in vitro synthesis of antisense RNA probes. The E. coli DH5-α competent cells were transformed with the cloned vector in order to generate several clones, after which we performed the colony PCR to isolate the positive colonies. The positive colonies were cultured overnight, and the cloned plasmids were extracted (mini-prep DNA). In order to determine the orientation of the insert in the vector, we carried out insertion direction PCR. Depending on the direction of insertion, we used the Sp6/T7 RNA polymerase to synthesize antisense RNA probes.

Pupal Staging
We collected wandering third-instar (L3) larvae in a 10 cm Petri dish with moist tissue paper covering the bottom. After the larvae began to pupate, we followed the progression of the pupal stages under a dissection scope. We adopted the description by Bainbridge and Bownes [25] and Fukutomi et al. [27] for Drosophila pupal stage determination ( Figure 2).

Abdominal Epidermis Preparation
Once the pupae reached the desired stage, they were placed in groups of ten onto a double-sided tape fixed to a microscope slide to cut them and clean the epidermal tissue. We performed two types of cuts: (1) the pupae were placed with their ventral side facing the tape and then cut longitudinally between both eyes (dorsal cut) and (2) one of their lateral sides faced the tape, and the cut ran longitudinally through the pupae, separating the dorsal from the ventral half (lateral cut) ( Figure S1). Dorsal cuts are best suited to examine lateral patterns of gene expression, while lateral cuts allow for the visualization of the dorsal and ventral regions of the abdomen. The cut pupae were then washed with phosphate-buffered saline (PBS), fixed in paraformaldehyde, and stored in 100% ethanol at −20 • C until further processing by ISH.

Pupal Wing Preparation
Wing preparations are possible from stage P5ii onward. It is important to note that wings from stages P5ii to P8 are too fragile to be dissected without prior fixation. These early pupae were pulled out of their puparia in PBS with a pair of fine forceps, followed by cutting off the head and the tip of the abdomen with a small pair of surgical scissors. The resulting carcass tubes were cleaned by pipetting PBS through them to remove the inner organs. The carcasses were then fixed in paraformaldehyde (see Section 6 for reagent preparation), overnight at 4 • C, after which the hardened wings were dissected from the pupae with two pairs of fine forceps. Pupae of stage P9 and older have sturdier wings, which allows the fixation process to follow dissection, instead of preceding it. These older pupae were placed with the ventral side down on a glass slide with double-sided tape. The pupae were pulled out of their puparia by the head and submerged in a glass-viewing dish in distilled water. The wings were then carefully extracted from the pupal membrane and allowed to inflate, followed by fixing them with paraformaldehyde at room temperature for 30 min. Regardless of the pupal stage, the extracted wings were washed twice with methanol and twice with 100% ethanol after the fixation step and then stored in ethanol at −20 • C until the ISH procedure was performed.

ISH of Drosophila Abdomens and Wings
On the first day of the ISH, the processed tissues (pupal abdomens or wings) were carefully transferred into the wells of a glass-viewing dish, using a cut 1 mL pipette tip. They were then treated with xylenes to remove any fatty tissue, re-hydrated, fixed, washed, treated with proteinase K, washed, post-fixed, and prehybridized. After the prehybridization, DIG-labeled RNA probes were added, and the samples were incubated at 65 • C for 18 h to 3 days (d). On the second day, any unhybridized probe molecules were washed away to reduce background staining. The washed tissues were incubated in anti-DIG-AP F ab fragments at 4 • C overnight. On the third and final day, the tissues were treated with NBT/BCIP staining solution to detect the dark-purple hybridization pattern. This reaction took place in the dark. The staining progress was observed under a dissecting scope every 20 min to avoid overstaining. After the staining reached its desired intensity, the staining solution was washed off with staining buffer, then with PBT, and the gene expression patterns were observed under a stereo microscope. Several images of the tissues were taken at slightly different focal planes and Z-stacked with Helicon Focus software. A minimum of three days is required to perform the ISH experiment in Drosophila using this protocol. (1) Prepare a reaction mix, according to the Table 1 below.

Materials
(2) Amplify the PCR product according to the appropriate cycling conditions (Table 2).
(3) Run the PCR product through a 1% (w/v) agarose gel in 1× TAE buffer by electrophoresis and visualize it under UV light. (4) If the size of the PCR product matches the expected size, perform a gel extraction.

Gel Extraction and Purification of PCR Products
Timing 30 min (5) On a table-top UV light, cut out the gel slice containing the DNA fragment, using a clean razor blade and place it into a 2 mL Eppendorf tube. (6) Add 1:1 volume of binding buffer to the gel slice (w/v). (7) Incubate at 60 • C until the gel slice is completely dissolved.

PAUSE STEP
Purified PCR products can be stored at −20 • C for one month. CRITICAL STEP MeanGreen master mix adds A-tails to the PCR products. In case you use a PCR master mix that does not add A-tails, add the A-tails after per forming the gel extraction, according to the Table 3 below. Also, note that A-tails de-grade after one month of storage at −20 • C.

Ligation
Timing 18 h (overnight) (15) To ligate the A-tailed PCR product with the pGEM ® -T Easy vector, use the reaction mix in the Table 4 below. Table 4. Reaction mix for ligation of A-tailed PCR product with the pGEM ® -T Easy vector.

CRITICAL STEP
The pGEM ® -T Easy vector features T-overhangs essential for T/A cloning of an A-tailed DNA fragment. Also, this vector contains the SP6/T7 promoters, which will later flank the insert after an additional PCR reaction.  (Table 5). Use the cycling conditions described in (Table 6).

Plasmid Extraction from a Positive Clone (Mini-Prep) Using the Plasmid Mini-Prep Kit by Thermo Scientific
Timing 40 min

Insertion Direction PCR
Timing 3 h (44) Carry out the insertion direction PCR for each mini-prep DNA to determine the orientation of the insert in the pGEM ® -TEasy vector to choose the correct RNA polymerase that will synthesize an antisense probe, according to Table 7. (45) Set up two PCR reactions simultaneously for each DNA clone, using the following primer pairs: (i) the M13F primer plus the gene-specific internal forward primer; and (ii) the M13F primer plus the gene-specific internal reverse primer. (46) Perform the PCR reactions according to the cycling conditions described in (Table 8).
(47) Perform gel electrophoresis of the PCR products, using a 1% (w/v) agarose gel in 1× TAE and visualize the bands under UV light.  CRITICAL STEP If the primer pair M13F/internal reverse shows a PCR band, use Sp6 polymerase to make an anti-sense probe. However, if the primer pair M13F/internal forward shows a PCR band, use T7 polymerase to make an anti-sense probe. Only one of the primer pairs should produce a clear PCR band.

RNA Probe Synthesis
Timing 5 h (48) PCR-amplify the cloned insert, using the mini-prep DNA as a template (Step 43). Use the M13F and M13R primer pair, as tabulated below (Table 9). (49) Amplify according to the cycling conditions described in Table 8. (53) Prepare the anti-sense RNA probe reaction mix as tabulated below (Table 10). CRITICAL STEP Remember to use the correct RNA polymerase based on the insertion direction PCR result (Steps 44-47). (Recall: If the primer pair M13F/internal reverse shows a PCR band, use Sp6 polymerase to make an anti-sense probe. However, if the primer pair M13F/internal forward shows a PCR band, use T7 polymerase to make an anti-sense probe).  (59) Precipitate the remaining 9 µL of the synthesized probe as shown below (Table 11). Table 11. Reaction mix for probe precipitation.

Reagent Volume Per Reaction (µL)
Probe (from step 54) 9 Linear acrylamide (5 µg/µL)  (71) Use blunt forceps (type #2) to gently remove the pupae from the Petri dish (one at a time) and immediately transfer them onto the dissection platform. (72) Lay the pupae with their ventral side facing the tape and cut longitudinally between both eyes (dorsal cut) or lay them on their lateral side and cut longitudinally through the pupae, separating the dorsal from the ventral half (lateral cut) ( Figure S1). Perform only one type of cut in a session. (73) With a razor blade, immediately cut each pupa lengthwise (starting with the one first placed on the tape). This is best accomplished with a single rapid cut from the anterior to the posterior end of the pupae. Ensure that the puparium is intact after cutting. (74) Using a medium-sized paintbrush, transfer a small amount of 1× PBS from the glass-viewing dish to each cut pupa to dissolve them from the tape. (75) Transfer the pupal halves with the brush into the well of a glass-viewing dish filled with 1× PBS.
CRITICAL STEP Dissect 10 pupae within 2 min and then transfer them into 1× PBS immediately. Label each end of the razor blade and use each end to dissect 50-60 pupae, after which the blade is too blunt and should be discarded.
(76) With a pair of surgical (sharp-pointed) forceps (type #5), grasp an individual pupa half anteriorly (by the head) and gently wash away the internal organs with 1× PBS. (77) Use a pipettor to gently flush 1× PBS over the internal organs without touching the epithelial layer of the pupa with the pipette tip ( Figure S2). Prevent the epidermal tissue from becoming detached from the puparium at this time, as the puparium provides mechanical protection throughout the process.
CRITICAL STEP You must ensure to keep the epithelial cell layer intact throughout the washing steps. Therefore, apply low pressure from the pipettor by setting a 20-µL pipettor to 8.5 µL for pupal stages P7, P8 or to 15 µL for pupal stage P9. For stages P10 and older, use a 200-µL pipettor set to 25 µL. Too much pressure or excessive washing will lead to the loss of epithelial cells. CAUTION Allow methanol waste to evaporate in the hood or pour it down the drain and flush for 1 min. Do not inhale, swallow, or allow methanol to touch your skin.

ISH of Drosophila Abdomens and Wings
Timing 3 d

Day 1
(106) Take the processed wings and the pupa halves from the −20 • C freezer. (107) Transfer the pupa halves with a pipette and a cut 1-mL tip into a glass-viewing dish. (108) Transfer the wings with a pipette and a cut 200-µL tip into a glass-viewing dish.
Ensure that the glass-viewing dish is gel slick coated (Appendix C). (118) Incubate the tissues at room temperature for 10 min.
CRITICAL STEP Incubate the wings and abdomens in Proteinase K for 10 min for pupal stages P6 to P10 and 20 min for P11 to P15 to increase the tissue permeability and reduce background staining.

Day 3
(145) Transfer the samples back into a clean glass-viewing dish by pipetting, using a cut 1-mL tip for abdomens and a cut 200-µL tip for wings.

Expected Results
The images of the gene expression patterns on Drosophila wings and abdomens generated using this protocol are shown in Figures 4-7. This protocol has been used to generate quality ISH images for toolkit genes during early pupal stages on wings and abdomens of non-model Drosophila species [6,20] (Figures 4 and 5).    The pupal dissection steps determine the final orientation of the pupa, whereby improper dissection can lead to the loss of important features, which might affect the image quality and interpretation. Through pupa abdominal ISH, we have shown that the yellow gene is expressed in all six rows of spots foreshadowing the adult D. guttifera and D. quinaria patterns ( Figure 6).
It may be technically challenging to use one ISH image to show gene expression patterns in all the six rows of spots on the abdomen of some Drosophila species in the quinaria species group. Therefore, the mastery of the anatomy of the pupa and proficiency in making different types of cuts are very important to ensure that all spot rows are revealed in at least two ISH images. For example, the mid-cut pupa used for abdominal ISH shows the lateral, median, and/or dorsal rows of spots, which prefigure the spots in the lateral region of the adult fly's abdomen (Figures 6a and 7c), while the lateral cut shows the dorsal and median rows of spots foreshadowing the spots in the dorsal region of the adult fly's abdomen (Figures 6b and 7a). However, the lateral cut may occasionally reveal all six rows of spots as shown in (Figure 7b). Interestingly, we have shown this protocol to work in a wide range of rarely studied fruit flies in the quinaria species group, such as D. deflecta, D. guttifera, D. recens, D. quinaria, D. subpalustris, and D. palustris. This protocol contains the necessary information to facilitate the study of novel gene expression patterns in rare and unstudied fruit flies in the future.

Troubleshooting
This protocol is accompanied by an important troubleshooting table that summarizes a variety of problems that we have encountered in the past and offers appropriate solutions (Table 12).  Figure S1: D. guttifera pupae lined up for lateral and dorsal cuts, Figure S2: A sketch of a Drosophila pupal abdomen showing the internal epithelial cell layer and the cuticular lining holding the cells, Figure S3: A graphical representation of the conserved region on the ebony gene of D. melanogaster, Table S1: List of primers used to prepare probes for ISH.

4.
Take several images by slowly tuning the fine focus adjustment knob between the photo shots, starting from the top-most piece of tissue that comes into focus and progressing downwards, until the lowest-laying part of the tissue goes out of focus. 5.
Z-stack the raw images with Helicon Focus software. 6.
Use the "curves" function in Adobe Photoshop to reduce the background noise and maintain the natural colors of the image. Do not bend the "line" to avoid altering the ratios of the original result.