Drosophila stocks

CPC marker: w; P{y[+t7.7] w[+mC] = GMR61G12.GAL4}attP2 (BDSC #41286). dSmad2 mutant MARCM clones: yw P{w[+mC] = UAS.mCD8.GFP.L}Ptp4E^LL4 dSmad2^MB388 P{ry[+t7.2] = neoFRT}19A / FM7C referred to as dSmad2 mutant [11] (BDSC #44384), w P{ry[+t7.2] = hs.FLP}1 P{w[+mC] = tubulinP.GAL80}LL1 P{ry[+t7.2] = neoFRT}19A referred to as Tub.GAL80 [12] (BDSC #5132) and w; P{w[+mC] = dCORL.GAL4(-).PT} P{w[+mC] = UAS.GFP.S65T} / SM6A referred to as dCORL.GAL4 [13]. Rescue of dSmad2 mutant MARCM clones with UAS.dSmad2: dSmad2 mutant, Tub.GAL80, dCORL.GAL4, and yw; P{w[+mC] = UAS.dSmad2} referred to as UAS.dSmad2 [11]. Expression of dSmad2 RNAi: dCORL.GAL4 and yw; P{w[+mC] = UAS.dSmad2 RNAi} P{w[+mC] = UAS.dicer2} / TM6 referred to as UAS.dSmad2 RNAi [14]. Rescue of dSmad2 RNAi with UAS.dSmad2: dCORL.GAL4, UAS.dSmad2 RNAi and UAS.dSmad2.

dCORL RNAi, upd2Δ and Df(4)dCORL: Parental stocks for genotypes containing various combinations of OK107.GAL4 (BDSC #854), UAS.dCORL RNAi, UAS.lacZ and dCORL.AH.lacZ are described in [3]. The deletions w upd2Δ (BDSC #55727) and yw; Df(4)dCORL were originally reported in [4, 15], respectively.

Stock backgrounds contained either a white mutation or yellow and white mutations as noted above. Stocks were maintained in four narrow glass vials with roughly 5ml of dextrose media each, in an incubator maintaining 25°C with 68% humidity. Media was prepared in 40 bottle batches as follows: agar 20g, yeast 69g, cornmeal 130g, anhydrous dextrose 275g in 2125ml water plus 29ml Tegosept (100 g/l in 95% ethanol). In preparation for experiments, stocks were grown up in six ounce polyethylene bottles for virgin female collection. Collected females were maintained at 18°C with 68% humidity until needed for crosses. Males were taken from stock bottles maintained at 25°C for crosses.

Drosophila genetics

dSmad2 mutant MARCM clones and clone rescue with UAS.dSmad2: The experimental genotype for dSmad2 mutant MARCM clones was dSmad2^MB388 / Tub.GAL80; dCORL.GAL4/+. The complete genotype was yw P{w[+mC] = UAS.mCD8.GFP.L}Ptp4E^LL4 dSmad2^MB388 P{ry[+t7.2] = neoFRT}19A / w P{ry[+t7.2] = hs.FLP}1 P{w[+mC] = tubulinP.GAL80}LL1 P{ry[+t7.2] = neoFRT}19A; P{w[+mC] = dCORL.GAL4(-).PT} P{w[+mC] = UAS.GFP.S65T}/+. The experimental genotype for rescued dSmad2 mutant MARCM clones was dSmad2^MB388 / Tub.GAL80; dCORL.GAL4 / UAS.dSmad2. The complete genotype was yw P{w[+mC] = UAS.mCD8.GFP.L}Ptp4E^LL4 dSmad2^MB388 P{ry[+t7.2] = neoFRT}19A / w P{ry[+t7.2] = hs.FLP}1 P{w[+mC] = tubulinP.GAL80}LL1 P{ry[+t7.2] = neoFRT}19A; P{w[+mC] = dCORL.GAL4(-).PT} P{w[+mC] = UAS.GFP.S65T} / P{w[+mC] = UAS.dSmad2}. The control genotype in the same brain did not express UAS.dSmad2. Crosses were maintained at 25°C. Embryos from these crosses aged 4 to 14 hours were heat-shocked for 1 hour at 37°C (to express FLP recombinase initiating mitotic recombination), allowed to recover at 18°C for 1 hour, then returned to 25°C to eclosion.

dSmad2 RNAi and RNAi rescue with UAS.dSmad2: The experimental genotype for dSmad2 RNAi in IPCs was dCORL.GAL4/+; UAS.dSmad2 RNAi/+. The complete genotype was yw/w; P{w[+mC] = dCORL.GAL4(-).PT} P{w[+mC] = UAS.GFP.S65T}/+; P{w[+mC] = UAS.dSmad2 RNAi} P{w[+mC] = UAS.dicer2}/+. The control genotype in the same brain did not express UAS.dSmad2 RNAi. The experimental genotype for rescued dSmad2 RNAi was dCORL.GAL4 / UAS.dSmad2; UAS.dSmad2 RNAi/+. The complete genotype was yw/w; P{w[+mC] = dCORL.GAL4(-).PT} P{w[+mC] = UAS.GFP.S65T} / P{w[+mC] = UAS.dSmad2}; P{w[+mC] = UAS.dSmad2 RNAi} P{w[+mC] = UAS.dicer2}/+. The control genotype in the same brain did not express UAS.dSmad2.

dSmad2 overexpression: The experimental genotype for UAS.dSmad2 overexpression was UAS.dSmad2 / dCORL.GAL4. The complete genotype was yw/w; P{w[+mC] = dCORL.GAL4(-).PT} P{w[+mC] = UAS.GFP.S65T} / P{w[+mC] = UAS.dSmad2}. The control genotype in the same brain did not express UAS.dSmad2.

To maintain a common culture density across experiments, each cross was conducted with 20 females and 10 males in a glass vial. The vial was passed after four days, then the second vial passed after four days and the adults cleared from the third vial after four days. All vials contained dextrose media in an incubator maintaining 25°C with 68% humidity. Just prior to eclosion of the first flies in each cross, the vials were moved to 18°C with 68% humidity overnight. The next morning virgin females were collected from 18°C between 10:00AM and 12:00PM. The cross was placed back at 18°C, and females of the appropriate genotype identified from those collected. These females were placed, as an age matched group, into glass vials in the 25°C incubator. After 24 hours their brains were dissected and brains in good condition were fixed then stored at -20°C in methanol. The process was repeated until roughly 20 fixed brains were obtained. During each experiment the lab environment provided a roughly 9 hour light and 15 hour dark cycle. This cycle was not a formal variable as light was ambient and dark was not total. In every experiment mutant and wild type cells were from the same brain. The compiled pairs of mutant and wild type cells analyzed for each experiment shared genetic background, cell lineage and environment (including light:dark cycle) within the two hour collection window.

dCORL RNAi and Df(4)dCORL: Original confocal images of 1 day old unmated adult female IPCs previously assayed by counting dILP2 expressing cells [3] were reanalyzed for single cell dILP2 pixel intensity employing the parameters for dSmad2 mutant clones. Single cells viewed in a single slice from two pairs of genotypes were compared in FIJI: Df(4)dCORL versus dCORL.AH.lacZ and OK107.GAL4 driven UAS.dCORL RNAi versus OK107.GAL4 driven UAS.lacZ, with the lacZ genotype in both pairs serving as wild type.

upd2Δ: 1 day old unmated adult females and males were analyzed side by side with w¹¹¹⁸ serving as wild type. Flies were dissected, fixed, stained with dILP2 and dILP5, then imaged as described [10]. Single cell pixel intensity comparisons in FIJI [16] followed the parameters employed in the analysis of dSmad2 mutant clones.

Immunofluorescence

One day old unmated female adult brains were dissected, fixed, stained for GFP, FasII, and either dILP2 or dILP5 then mounted and imaged as described [13]. Primary antibodies: DSHB mouse α-FasII 1D4 (RRID: AB_528235), Abcam rabbit α-GFP ab6556 (RRID: AB_305564), Abcam chicken α-GFP ab13970 (RRID: AB_300798), mouse α-dILP2 and rabbit α-dILP5 (Pierre Leopold, Institute Curie, Paris). Secondary antibodies: α-GFP was detected with AlexaFluor488 (green; excitation 495nm, emission 519nm), α-FasII shown as blue was detected with AlexaFluor546 (yellow; excitation 556nm, emission 573nm), α-dILP2 and α-dILP5 were detected with AlexaFluor633 (far-red; excitation 632nm, emission 647nm). Common settings were utilized for all dSmad2 experiments and each channel was collected sequentially to minimize bleed-through.

All confocal images from the same dSmad2 experiment were analyzed in FIJI with the same parameters. Mutant and wild type cells were imaged in the same brain then identified as wild type or mutant by the presence or absence of GFP. For analysis we prioritized mutant and wild type pairs where both cells were in a single slice. A single slice was preferred because a 2μm slice is thinner than the diameter of the cell body of a fly brain neuron. When imaging a single slice we were confident of measuring the pixel intensity from a single cell. Single slices chosen for analysis were those that displayed the nucleus of the mutant and wild type cells at their widest point. This guideline ensured that the chosen slice captured the largest area of fluorescence in each cell and that both cells were in the same plane in the same orientation.

For upd2Δ and dCORL studies of unmated adult females, a single mutant cell from a single slice (following the parameters for dSmad2 mutant clones) was compared to a single wild type cell in a different brain. Here it was not possible to compare cells in the same brain as the mutations were genomic and not clonal. In both studies mutants and wild type brains were collected, fixed, stained, imaged, and analyzed side by side to maximize confidence in the comparisons. upd2Δ unmated males were included since their dILP2 phenotype in IPCs was known [8].

Statistics

Within an appropriate image (single slice or small stack) each cell to be analyzed was traced at the plasma membrane and their dILP2 or dILP5 pixel intensities (red channel) measured in FIJI. The Mean and Standard Error of the Mean (Mean ± SEM) for the pixel intensities of a set of cells were calculated in Excel (minimum set was six cells). Statistical analyses of pixel intensities from pairs of mutant and wild type cells in each experiment were conducted in Excel.

Two-tailed paired Student’s t-test (also called a type 1 t-test) evaluates whether the mean of the differences for each pair in two groups significantly deviates from zero, the mean of the null hypothesis. The standard probability for significance was applied (p ≥ 0.05). The rationale for employing a type 1 test is that the cells being compared are in the same brain with just a single recombination event separating them genotypically. Their developmental lineage, their cellular environment and their organismal life history are the same. The alternative type 3 test evaluates whether the difference of the means of two groups significantly deviates from zero, the mean of the null hypothesis. To increase confidence, all statistical analyses reported with a type 1 test were repeated with a type 3 test. While the type 3 p-values were not identical, outcomes were unchanged (i.e., no significant differences became not significant and vice versa).

The formula for the illustrative statistic Δ% is clone pixel intensity minus wild type pixel intensity divided by wild type pixel intensity. The sign of the Δ% was positive if clone pixel intensity was more than wild type and negative if the clone was less. Note that Δ% shown in figures provides a measure of the size and sign of the difference in pixel intensity between the individual mutant and wild type cells shown.

dILP5 is coexpressed with Pigment dispersing factor (Pdf) in circadian pacemaker cells

dCORL.GAL4 was shown to be expressed in a subset of IPCs, the mushroom body and four pairs of neurons in a ventral lateral region of the central brain that also express dILP5 (Fig 1A; [13]). We previously thought the lateral cells were located in the lobula plate and thus formed part of the visual system. To confirm the identity of the lateral cells, we performed co-expression experiments with dILP5 and GFP from two well-characterized GAL4 drivers. We expected to see co-expression of dILP5 with the visual system line GawB-3A [17]. We chose Pdf.GAL4 expressed in circadian pacemaker neurons (CPCs; also known by their anatomical location—small ventral Lateral Neurons sLNv; [18]) as a negative control. CPCs were considered a negative control because no connection between dILP5 and CPCs was known. Instead IPCs were shown to be electrically connected to a distinct set of circadian cells in a dorsal lateral region [19].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. dILP5 is coexpressed with Pdf.GAL4 in circadian pacemaker cells.

A) Illustration of a Drosophila adult brain in anterior view indicating clusters of insulin producing cells (IPCs; red express dILP2 and dILP5), circadian pacemaker cells (CPCs; yellow express dILP5), and the mushroom body (blue; for orientation). B-E) Brain in anterior view from a one day old unmated adult female with Pdf.GAL4 driven UAS.GFP. The secreted neuropeptide Pdf is a marker for small ventral Lateral Neurons sLNv also known as circadian pacemaker cells (CPCs; [18]). Pdf.GAL4 neurons express GFP (green) and dILP5 (red). B) Stack of CPCs (8 slices). Arrow indicates one of the two symmetrical clusters of CPCs that coexpress Pdf.GAL4 and dILP5. C-E) Single confocal slices of the indicated cluster. Two color images on the left and the red channel (dILP5) on the right. Slices together show that all four CPCs in the cluster coexpress Pdf.GAL4 and dILP5.

https://doi.org/10.1371/journal.pone.0280529.g001

We observed that dILP5 (Fig 1B) was co-expressed with Pdf.GAL4 in CPCs. All four cells of the CPC cluster co-express Pdf.GAL4 and dILP5 (Fig 1C–1E). In checking the literature again, we noted studies showing that mammals display a circadian rhythm for glucose metabolism (e.g., [20]), but we could not find any reports of an intracellular connection between the clock and insulin production. The presence of dILP5 in CPCs in unmated adult female brains provides a new avenue for investigation of this connection.

dSmad2 mutant clones display increased dILP2 in IPCs and decreased dILP5 in CPCs

To examine the role of dSmad2 in adult IPCs we analyzed dILP2 expression in dSmad2 mutant MARCM clones. We employed dCORL.GAL4 driven UAS.GFP (together abbreviated as dCORL.GAL4) to identify mutant clones in otherwise phenotypically wild type (genotypically heterozygous) unmated adult females. The absence of haploinsufficiency for dSmad2 is attested to by its location on the X chromosome where it is hemizygous in all males. The fact that dCORL.GAL4 is expressed in a subset of IPCs [13] and dILP2 is expressed in all IPCs [5] ensures that mutant and wild type cells are present in the same brain, providing a highly rigorous experiment.

We expanded the analysis to include dILP5 in IPCs and CPCs as dCORL.GAL4 is able to identify dSmad2 mutant MARCM clones in CPCs. We added dSmad2 RNAi to the study as a second loss of function approach. To confirm the results of the MARCM and RNAi experiments we conducted UAS.dSmad2 rescue of both phenotypes. This set of four loss of function studies was complemented by a gain of function (overexpression) experiment. Together the five experiments revealed distinct outcomes and modes of dSmad2 regulation of dILP2 and dILP5.

Homozygous mutant MARCM clones were generated in the IPCs of unmated adult females heterozygous for dSmad2^MB388 (a null allele in the protein-protein interaction domain due to a charge altering mutation in a universally conserved residue E300K; [11]). dSmad2 mutant IPC clones displayed a significant increase in dILP2 expression compared to adjacent wild type cells (paired t-test of wild type and mutant cells in the same brain and as often as possible the same confocal slice; p = 0.043 see Table 1 for mean ± SEM; Fig 2A and 2D). In contrast, dSmad2 mutant IPC clones displayed no significant difference in dILP5 expression (p = 0.333; Table 1, Fig 2B and 2D). In contrast to dILP5 IPC clones, dSmad2 mutant CPC clones displayed a significant decrease in dILP5 expression (p = 0.001; Table 1, Fig 2C and 2D). Note p-values in figure tables do not refer to Δ% (an illustrative value for each panel reflecting the comparison of the two cells shown). Instead p-values derive from a type 1 Student’s t-test of significance for the Mean of the differences between pairs of wild type and mutant cells (Mean ± SEM shown in Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. dSmad2 mutant clones display increased dILP2 in IPCs and decreased dILP5 in CPCs.

A-C) Brains in anterior view from one day old unmated adult females with the dSmad2 mutant MARCM clone genotype. Homozygous dSmad2 clones express GFP (green) in IPCs and CPCs. IPCs are visualized with dILP2 or dILP5 (red as indicated) and CPCs with dILP5 (red). Left side two color and right side the red channel (dILPs). Paired cells for pixel intensity comparisons are indicated by arrows with white indicating a mutant and red a wild type IPC or CPC. A,A’) Single slice of IPCs. dSmad2 mutant IPC clone displays significantly increased dILP2 compared to wild type. B,B’) Single slice of IPCs. dSmad2 mutant IPC clone displays wild type dILP5 expression. C,C’) Stack of CPCs (4 slices). dSmad2 mutant CPC clones display significantly reduced dILP5 expression compared to wild type. Two comparisons are illustrated with an asterisk marking one pair. D) Table summarizing dSmad2 mutant clone pixel intensity comparisons in IPCs and CPCs. Note p-values do not refer to Δ% (an illustrative value for the cells compared in each panel), but are derived from a type 1 t-test of the mean of the differences between pairs of wild type and mutant cells shown in Table 1.

https://doi.org/10.1371/journal.pone.0280529.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of mutant phenotypes for dILP2 & dILP5 pixel intensity in unmated adult female IPCs and CPCs.

https://doi.org/10.1371/journal.pone.0280529.t001

The results from rescue experiments were consistent with the MARCM data providing additional confidence. UAS.dSmad2 rescued both mutant phenotypes. For dILP2 in IPC clones, UAS.dSmad2 rescued overexpression to wild type (p = 0.671; Table 1, S1A, S1D Fig in S1 File). For dILP5 in IPC clones, UAS.dSmad2 had no effect with all cells remaining wild type (p = 0.375; Table 1, S1B, S1D Fig in S1 File). For dILP5 in CPC clones, UAS.dSmad2 rescued reduced expression to wild type (p = 0.106; Table 1, S1C, S1D Fig in S1 File).

In summary, dSmad2 mutant MARCM clones reveal two types of dSmad2 differential regulation of dILP2 and dILP5. The first difference is in IPCs between dILP2 (increased expression revealing a loss of negative regulation) and dILP5 (no difference). The second distinction is between dILP5 in IPCs (no difference) and in CPCs (decreased expression revealing a loss of positive regulation). The rescue of both mutant phenotypes with UAS.dSmad2 reinforces their connection to dSmad2 rather than a second mutation distal to the FRT.

dSmad2 RNAi matches mutant clones for dILP2 but suggests dosage effects for dILP5

To further support our findings we expressed UAS.dSmad2 RNAi via dCORL.GAL4. In IPCs, dSmad2 RNAi phenocopied dSmad2 mutant clones with overexpression of dILP2 (p = 0.013; Table 1, Fig 3A and 3D). In IPCs there continued to be no change in the expression of dILP5 (p = 0.116; Table 1, Fig 3B and 3D). Interestingly, for dILP5 in CPCs there was an increase in expression (p = 0.004; Table 1, Fig 3C and 3D). Distinct phenotypes for a dSmad2 mutation in a CPC clone and dSmad2 RNAi in the same cells could have two causes. The first is the influence of dSmad2 dosage. Complete loss of function in clones could be interpreted by CPCs distinctly from an RNAi induced partial loss of function. The second is the influence of the circadian pacemaker, a variable we did not formally control in these experiments.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. dSmad2 RNAi matches mutant clones for dILP2 but suggests a dosage effect for dILP5 in CPCs.

A-C) Brains as shown in Fig 2 with dCORL.GAL4 driving UAS.dSmad2 RNAi and UAS.GFP. A,A’) Single slice of IPCs. dSmad2 RNAi IPC displays significantly increased dILP2 expression compared to wild type. B,B’) Single slice of IPCs. dSmad2 RNAi IPC displays wild type dILP5 expression C,C’) Single slice of CPCs. dSmad2 RNAi CPC displays significantly increased dILP5 expression compared to wild type. D) Table as described in Fig 2 summarizing dSmad2 RNAi pixel intensity comparisons in IPCs and CPCs.

https://doi.org/10.1371/journal.pone.0280529.g003

We then added UAS.dSmad2 to the dSmad2 RNAi genotype for rescue studies. For dILP2 in IPCs, UAS.dSmad2 rescued overexpression to wild type (p = 0.661; Table 1, S2A, S2D Fig in S1 File). For dILP5 in IPCs, adding UAS.dSmad2 to the genotype led to an unexpected increase in expression (p = 0.005; Table 1, S2B, S2D Fig in S1 File). However, the increase is consistent with the influence of dosage. Partial loss due to dSmad2 RNAi is overcompensated with UAS.dSmad2, whereas in dSmad2 mutant clones complete loss was simply compensated. For dILP5 in CPCs, UAS.dSmad2 rescued overexpression to wild type (p = 0.154; Table 1, S2C, S2D Fig in S1 File).

In summary, the dSmad2 RNAi phenotype for dILP2 in IPCs was consistent with phenotypes for MARCM clones and both rescue studies. The RNAi phenotype for dILP5 in IPCs was consistent with MARCM clones and the MARCM rescue study. The RNAi rescue study suggested a dosage effect for dILP5 in IPCs with expression increasing from wild type to above wild type. For dILP5 in CPCs, the RNAi suggestion of a dosage effect is strengthened by the presence of increased dILP5 versus a decrease in MARCM clones. Rescue of CPC phenotypes with UAS.dSmad2 reinforces their connection to dSmad2 rather than a second mutation.

dSmad2 overexpression phenotype complements the dSmad2 loss phenotype

For completeness, we conducted a gain of function study expressing UAS.dSmad2 in wild type IPCs. This led to dILP2 phenotypes that were consistent with loss of function studies. For dILP2 in IPCs, complementary to the significant increase seen with dSmad2 loss, expressing UAS.dSmad2 significantly reduced expression below wild type (p = 0.003; Table 1, Fig 4C and 4D). For dILP5 in IPCs, consistent with dSmad2 loss, expressing UAS.dSmad2 had no effect with all cells remaining wild type (p = 0.607; Table 1, S4B, S4D Fig in S1 File). For dILP5 in CPCs, complementary to the significant decrease seen with dSmad2 loss, expressing UAS.dSmad2 significantly increased expression above wild type (p = 0.008; Table 1, Fig 4C and 4D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. dSmad2 overexpression phenotype complements the dSmad2 loss phenotype.

A-C) Brains as shown in Fig 2 with dCORL.GAL4 driving UAS.dSmad2 and UAS.GFP. A,A’) Single slice of IPCs. dSmad2 overexpressing IPC displays significantly decreased dILP2 expression compared to wild type. B,B’) Single slice of IPCs. dSmad2 overexpressing IPC displays wild type dILP5 expression. C,C’) Stack of CPCs (2 slices). dSmad2 overexpressing CPC displays significantly increased dILP5 expression compared to wild type. D) Table as described in Fig 2 summarizing dSmad2 overexpression pixel intensity comparisons in IPCs and CPCs.

https://doi.org/10.1371/journal.pone.0280529.g004

In summary, all five experiments for dILP2 in IPCs are consistent with dSmad2 acting as an antagonist (MARCM clones, RNAi, both rescues and overexpression). For dILP5 in IPCs, four of five experiments indicate no role for dSmad2 in dILP5 IPC regulation. The anomalous result, an increase with dSmad2 RNAi rescue, we consider evidence of a dosage effect of dSmad2 on dILP5 IPC expression. For dILP5 in CPCs, four of five experiments indicate dSmad2 acts as an agonist in dILP5 CPC regulation. The anomalous result, an increase with dSmad2 RNAi, we consider evidence of a dosage effect of dSmad2 on dILP5 CPC regulation, though we are unable to rule out an effect of the circadian clock. Regardless of why, the simple regulatory interaction between dSmad2 and dILP2 stands in contrast to the more complicated interaction between dSmad2 and dILP5.

Neither dCORL nor upd2 mutants phenocopy loss of dSmad2 in adult female IPCs

We then addressed whether dSmad2 cooperates with dCORL and whether dSmad2 cooperates with upd2 in IPCs of unmated adult females. To begin we reanalyzed Df(4)dCORL and dCORL RNAi images from [3]. The logic was that the impact of a dCORL mutation on IPC number, as reported there, is likely independent of dSmad2 but the remaining IPCs could have increased dILP2 expression resulting from the loss of a second role for dCORL in cooperation with dSmad2. The reanalysis of both genotypes showed there is no difference in dILP2 expression levels in IPCs between Df(4)dCORL and wild type (p = 0.344; Table 1, S3 Fig in S1 File) or between dCORL RNAi and wild type (p = 0.785; Table 1, S4 Fig in S1 File). We conclude from these studies that dCORL and dSmad2 function independently upstream of dILP2 in IPCs.

The dCORL data also counter an alternative explanation for our dSmad2 data in IPCs. The alternative posits that reliance on dCORL.GAL4 to identify clones led to an artificial result because dCORL.GAL4 expressing IPCs and dCORL.GAL4 non-expressing IPCs are somehow different in a way that impacts dILP2. Data from S3 and S4 Figs in S1 File showing no effect of dCORL mutants (either genomic or RNAi) on dILP2 expression in IPCs reveals that dCORL.GAL4 expressing and non-expressing cells are equivalent with regards to dILP2.

For upd2, we examined unmated adult male brains where it has been shown that dILP2 and dILP5 expression is increased in upd2 mutants [8]. We reproduced that data (dILP2 p = 0.001 and dILP5 p = 0.002; Table 1, Fig 5A, 5B and 5E). We then analyzed unmated adult female brains for dILP2 and dILP5. For both there was a modest increase that was not significant in upd2 mutants versus wild type (dILP2 p = 0.179 and dILP5 p = 0.129; Table 1, Fig 5C–5E). We conclude that upd2 does not cooperate with dSmad2 to regulate dILP2 expression in unmated adult female IPCs. Interestingly, the data suggest that upd2 regulates dILP2 in IPCs in a gender-specific fashion. For dILP5 in unmated adult female IPCs, loss of neither dSmad2 nor upd2 had any significant impact suggesting they are either uninvolved in dILP5 regulation or that additional experiments are needed.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. upd2 mutant unmated adult female IPCs do not phenocopy dSmad2 mutants.

Brains from unmated adult males and females as shown in Fig 2. A-C) One day old unmated adult male brains with IPCs labeled for dILP2 (green) and dILP5 (red). Each panel group shows two colors on the left and each single channel on the right. White arrows indicate IPCs analyzed for pixel intensity. A,A”) upd2 mutant male IPCs. B,B”) Wild type male IPCs. upd2 mutant male IPCs display significantly increased dILP2 and dILP5 pixel intensity compared to wild type males. C) Table as described in Fig 2 summarizing upd2 unmated mutant male pixel intensity comparisons in IPCs. D-F) One day old unmated adult female brains with IPCs labeled as above. D-D”) upd2 mutant female IPCs. E-E”) Wild type female IPCs. upd2 mutant female IPCs display wild type dILP2 and dILP5 expression. F) Table as described in Fig 2 summarizing upd2 mutant female pixel intensity comparisons in IPCs.

https://doi.org/10.1371/journal.pone.0280529.g005