Drosophila strains and culture

Synchronized larvae at the early third instar foraging stage were obtained as described [6] and grown at 25°C unless otherwise stated, in light/dark cycles in food supplemented with vitamin A (1.25 g/L). Mutations were kept over a CyO(y⁺) balancer in a yw background for the identification of homozygous mutants by the mouth hook phenotype.

The RanBPM ^k05201 allele carrying a P{lacW} insertion in the second exon of the RanBPM was identified as responding poorly in the ON/OFF assay [28]. RanBPM ^s135 is a deletion that eliminates a maximum of 4.1 kb from the N-terminus to the middle of the gene (64747 to 68882 in AE003829), generated by excision of the element in RanBPM ^k05201 [29], [30]. RanBPM ^ts7 carries a deletion from within 10 bp of the predicted start site extending upstream toward but not including CG12896 [8]. RanBPM^revertant is a precise excision of the element in the RanBPM ^k05201 as determined by sequencing (data not shown). Strains used include elav-GAL4 (Bloomington Stock Center (BSC) stock #8765), Dmef2-GAL4 (BSC stock #27390), tub-GAL4 (BSC stock # 5138), 247-GAL4 (Schulz et al Oncogene 12:1827 96),386Y-GAL4 [31], 201Y-GAL4 (BSC stock # 4440), UAS-GFP:LacZ.nls (BSC stock # 6452) and UAD-CD8-GFP (BSC stock # 5130). The stock carrying the UAS-RanBPM ^short construct was provided by Paul Lasko (McGill University, Canada). The UAS-RanBPM ^long line was generated from the cDNA clone RH61511 (FBcl0270865, http://flybase.org/). All constructs were sequenced. fmr1^EP3517 (BSC, stock #6928), is a hypomorphic allele, [32] and fmr1^Δ50M (BSC #6930) is a deletion [32].

Behavioral Assays

Statistical Analyses

We used SPSS Version 17 and The SAS System Version 9.2. The statistical tests included one-way analysis of variance (ANOVAs) and Tukey's-pairwise comparisons. In SAS we used the GENMOD Procedure with a Binomial distribution and a Link Function called ‘Logit’ to conduct a form of chi square analysis for nonparametric data. Normality tests on the residuals of the ANOVAs were conducted using the Shapiro-Wilk test. Verification of samples was performed by the F-test or Bartlett's test. The level of significance α in all tests was 0.05. All measurements are shown as mean values and SEM, and * indicates samples significantly different from control genotypes.

Immunohistochemistry and imaging

Mutations in the RanBPM gene cause disruption of the larval response to light and locomotion

The third instar larva light avoidance response is characterized by interruption of forward peristalsis, vigorous head-swinging followed by direction change, leading to a reduction in locomotion ([6], [34] and Figure 1C). Light avoidance can be measured in the ON/OFF assay, as changes to different aspects of locomotion of a single larva exposed to intermittent pulses of light [33], [34].

Download:

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

Figure 1. Mutations in RanBPM gene impair larval response to light and locomotion.

Response to light is seen in the ON/OFF assay as light-induced changes to quantitative aspects of locomotion. This response was quantitated in the semi-automated system as a response index (RI =  [Distance traveled (pixels) in dark (OFF) pulses - distance traveled (pixels) in light (ON)]/Total distance traveled) (A). DIAS was used to calculate direction change (B) or response of an individual larva during the course of the assay (C and D). The latter is shown as empty larval outlines (perimeter stacks) depicting behavior during the light (ON) pulse, and shaded outlines behavior during the dark (OFF) pulse. Quantitative analysis of perimeter stacks are shown in the graphs below as centroid translocation (mm) and direction change (deg). Points below the 20° threshold (dashed line) indicate linear movement. Reduced RanBPM function disrupts the larval response to light as seen by the significant lower RI exhibited by homozygous RanBPM^K05201, RanBPM^TS7 and heterozygous RanBPM^K05201/RanBPM^{TS7 or S135} (p<0.05). The response measured for homozygous RanBPM^S135 is markedly reduced but not statistically significant (p = 0.102) nevertheless this allele does not complement the RanBPM^K05201 (ANOVA F(7,131) = 7,242, p<0.0001, A). The control genotypes (OR, yw and RanBPM^revertant) show the increased change of direction triggered by light, that characterizes the photophobic behavior of the foraging larva (B and C). In contrast, in the RanBPM^K05201 mutants not only is locomotion markedly reduced but it is also uncoordinated as seen by the significant higher change of direction that occurs during the dark and light pulses (B p<0.001 and D), (ANOVAs for direction change during light F_(3,30) = 5.934, p<0.03 and during dark F_(3,30) = 25.351, p<0.0001, B). In all genotypes N≥12.

https://doi.org/10.1371/journal.pone.0010652.g001

Larvae homozygous for a P-element insertion in the predicted second exon of the RanBPM gene [8], henceforth named RanBPM^k05201 or for alleles generated by imprecise excision of this element, displayed a response index in the ON/OFF assay (RI =  [(path length in dark – path length in light)/total path length during the assay]) which was significantly lower than that of controls (Figure 1A). DIAS-assisted analysis of larval locomotion during the assay revealed that in these mutants locomotion is reduced and uncoordinated as seen by the equally high degree of direction change during the lights on and off pulses (Figure 1B, C and D). RanBPM^K05201 mutant larvae also displayed reduced locomotion, in constant dark (Figure S1).

The RanBPM gene was previously characterized as a vital gene (Table S1) playing a role in ovary development [8]. A deletion of the upstream region, (RanBPM^ts7) generated by these authors was used in behavioral assays as described above. Our results suggest that RanBPM^ts7 allele represents a less severe disruption of the RanBPM gene function. While larvae homozygous for the RanBPM^TS7 allele show a reduction in the response to light similar to that exhibited by larvae carrying the RanBPM^k05201 allele (Figure 1A), these mutants move more vigourously than the others (143.88 pixels +/−12.41 versus 26+/−3.96 for RanBPM^k05201 N>15). We conclude that RanBPM gene function is required for larval response to light and for coordinated locomotion.

RanBPM function is required for larval feeding behavior

Larvae homozygous for either the RanBPM^k05201 or RanBPM^s135 alleles, but not for the RanBPM^ts7 allele, were smaller than yw control specimens at chronologically the same stage (Figure S2 and data not shown) This size difference was apparent from the early third instar stage but not before.

Reduced larval growth can occur as a consequence of decreased food intake, arrest of cellular growth or a decrease in endoreplication [3], [41], [42], [43], [44]. In order to address whether insufficient food intake is associated with the reduced size of RanBPM mutants, we conducted feeding assays essentially as described previously [38]. Following a 2-hour starvation period, early foraging third instar RanBPM mutants were placed on yeast paste containing food dye for 30 min. While circa 88% of the control larvae ingested food within 30 minutes of being placed on the yeast paste, only 14–30% of RanBPM mutants (RanBPM^k05201 and RanBPM^s135 homozygous or heteroallelic combinations) displayed the characteristic blue gut (Table 1 and Table S2). As expected, given its normal size, mutant larvae homozygous for the RanBPM^ts7 allele ate like the control genotypes (Table S2).

Download:

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

Table 1. Feeding Rescue.

https://doi.org/10.1371/journal.pone.0010652.t001

To further investigate the feeding phenotype of RanBPM mutants, we allowed larvae to feed for a longer period (1.5 hours) The fraction remaining in the food and the fraction displaying a blue gut were determined as above. Control larvae, vigorously ingested food and remained for the most part immersed in the yeast for the duration of the assay. In contrast, a large fraction of RanBPM mutant larvae left the yeast drop and were found at the end of the assay outside the food source (Figure 2A).

Download:

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

Figure 2. RanBPM^K05201 mutants feed less and are not attracted to food.

Batches of 25 3^rd instar foraging larvae were starved for 2 hours and then placed inside (A, Food Dispersal) or outside the food (B, Food Attraction) laced with blue dye for 1.5 hours. The proportion of larvae that were in the food was determined at the end of the assay. All larvae were examined for the presence of blue matter in the gut indicative of food intake. RanBPM^k05201mutants when placed inside the food in the beginning of the assay ingest food significantly less than control larvae (A shaded bars, x² = 94.68, DF = 3, p = 0.0001) and are found immersed in the food in a significantly smaller number than control (A empty bars, x² = 83.31 DF = 3, p = 0001). Similarly, when placed outside the food plug RanBPM^k05201 mutants feed significantly less (B shaded bars, x² = 120.89, DF = 3,) and are attracted to the food less (B empty bars x² = 121.09 DF = 3) when compared to control larvae. In all experiments * p<0.0001, N>100.

https://doi.org/10.1371/journal.pone.0010652.g002

In order to determine whether the RanBPM mutants were attracted to yeast we placed the larvae outside the centrally located yeast plug. After 1.5 hours the fraction that moved into and remained in the food and the fraction displaying a blue gut was determined as above (Figure 2B). Larval behaviour during the assay was captured and a movie of a representative assay is available as supplementary material (Movies S1 and S2).

As expected, at the end of the assay, a large fraction of the control larvae were found immersed in the food and displayed the characteristic blue gut of fed larvae. In contrast, a large fraction of the RanBPM mutants did not move into the food source or did not eat (Figure 2B). The behavior of RanBPM mutants captured during the assay showed that these larvae moved around the plate and came close to the yeast drop but rarely entered and/or remained immersed in it (Movie S2). Interestingly, lack of RanBPM function does not impair the larva's repulsion to high concentrations of NaCl (Figure S3).

The apparent lack of interest in food displayed by RanBPM mutant larvae is reminiscent of precocious wandering behavior typical of the older third instar larvae, suggesting that lack of RanBPM function disrupts processes that underlie food intake. The multiple phenotypes displayed by RanBPM mutants suggest that this is a pleiotropic gene required in a variety of tissues. Alternatively, these phenotypes may be due to lack of gene function in a defined group of cells.

RanBPM mutations cause reduction in cell proliferation but no apparent disruption in differentiation of identified larval neurons

We asked whether reduced larval growth was associated with reduced cell proliferation as seen by phosphorylated Histone3 (phosphoH3) immunolabelling of whole-mount RanBPM mutant larval brains. Indeed, smaller than wild-type brains of larvae homozygous for RanBPM^k05201 showed a dramatic reduction in phosphoH3 immunolabelling (Figure 3). Surprisingly, we did not detect major morphological defects in these mutant larvae. Labeling of dissected larval brains with a number of different neuronal markers such as 5-HT [45], FasII [46], sNPF [47], [48], 386-GAL4 [31], and 247-GAL4 [49], [50], NPF [51] indicated that RanBPM is not required for the differentiation and/or maintenance of larval neurons (Figure S4 and data not shown). The smaller size of the CNS is reflected in the apparent smaller volume of the MB neuropil (Figure S4 A, A'). Moreover, we detected a small but statistically significant reduction in the number of serotonergic cell bodies with no apparent disruption in the pattern of projection of these neurons (Figure S4 H–J).

Download:

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

Figure 3. RanBPM function is required for cell proliferation.

Confocal micrographs of larval brains immunolabeled with anti-phosphoH3, detected with Alexa 488 secondary antibody. Images shown are projected Z-stacks of 10–13 (wild type) or 5–7 sections (RanBPM^K05201) sections at 2 µm intervals. The RanBPM^K05201 mutant CNS (A) is smaller than the wild type control (B) and show reduced phosphoH3 labeling, indicating that lack of RanBPM gene function disrupts proliferation.

https://doi.org/10.1371/journal.pone.0010652.g003

The long isoform of RanBPM is highly expressed in the Kenyon neurons of the mushroom body

The RanBPM gene encodes two protein isoforms with predicted masses of 67 and 140 kD (named RanBPM^short RanBPM^long). Both proteins contain the SPRY domain, LisH motif, CTLH motif and the CRA domain. The long isoform differs by the presence on the N terminus of a non-conserved glutamine rich segment [8].

To investigate the expression of the RanBPM gene we used an antibody directed against the N-terminus unique to the 105 kDa isoform kindly provided by Paul Lasko (McGill University,[8]). In the brain lobes and ventral nerve cord (VNC), RanBPM was expressed widely but not in all neurons labeled by the pan-neural marker elav [52], [53]. No staining was observed in the proliferating centers of the optic lobe or in the photoreceptors (Figure 4A). Consistent with this observation, no co-localization was detected with phosphoH3 immuno-labeling (Figure S5A–F). Similarly, RanBPM immuno-labeling did not co-localize with that of the glial marker, repo (Figure S5G–L), [54]. In the CNS, RanBPM labeling appeared to be cytoplasmic but not restricted to the neuronal cell bodies (e.g. Figure 4G–I). We did not detect RanBPM labeling in the neuromuscular junction (NMJ) [28]. Some, but not all motoneurons labeled by the motoneuron reporter D42-Gal4 [55], also expressed RanBPM (data not shown).

Download:

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

Figure 4. RanBPM^long is highly expressed in the Kenyon cells of the mushroom body.

Confocal micrographs of wild type larval CNSs immunolabeled with anti-RanBPM^long detected by Cy3-conjugated secondary antibody (red) and carrying a MB reporter construct 247-GAL4/UAS-CD8-GFP (green, 247:GFP, A–F) or double labeled with a pan neuronal antibody (anti-elav, green, G–I). RanBPM labeling is found in the ventral cord (arrowhead in A), the ring glands (inset in A) and in two bilateral clusters of neurons located in the brain hemispheres (box in A–C). Boxed area in A–C is magnified in D–F. These neurons are the Kenyon cells of the MB as seen by extensive co-localization with the MB reporter 247-GFP. In the abdominal portion of the ventral cord, RanBPM expression is found widely but not at the same level in all cell bodies (arrow in G and I). CNS dissected from homozygous mutant larvae (J, RanBPM^K05201; K, RanBPM^S135), and the RanBPM^K05201 mutant expressing the long isoform under the regulation of the MB specific driver 247-GAL4 (L, RanBPM^K05201; 247:RanBPM^long) were labeled with anti-RanBPM antibody (red) that recognizes the long isoform. The CNSs of larvae homozygous for the two most severe alleles show reduction in the volume of the CNS and absence of the characteristic RanBPM expression pattern in the brain (J and K). In the rescued sample (L, RanBPM^K05201; 247:RanBPM^long), RanBPM immunolabelling is restricted to the MB cell bodies and neuropil, the latter represents ectopic expression due to the high level of expression of this GAL4 driver (compare A with L).

https://doi.org/10.1371/journal.pone.0010652.g004

In the brain hemispheres RanBPM was highly, but not exclusively expressed in a bilateral dorsal cluster of neurons. This labeling co-localized with the expression of a mushroom body marker (247-GAL4;UAS-CD8-GFP, [50]; Figure 4A–F). At this level of resolution it appeared that all RanBPM-expressing cells present in this dorsal cluster co-expressed the mushroom body-specific element 247-GAL4. While we did not detect RanBPM expression in the imaginal discs, we did find expression in the muscle fibers attached to the mouth hooks and in the cytoplasm of the ring gland (data not shown and insert in Figure 4A). This pattern of expression was not detected in RanBPM mutant larvae, demonstrating that it is not due to cross reactivity of this antibody (Figure 4J and K). Moreover, targeted expression of the long isoform in the mutant background restored RanBPM expression, as detected by immunolabelling with this antibody, in a pattern consistent with that of the GAL4 driver employed (Figure 4D–E). No signal was detected in wild type larval brain specimens when we used an antibody generated against a portion of the RanBPM open reading frame common to both proteins (kindly provided by Paul Lasko, McGill University [8]). Therefore, we do not know whether in the larval CNS expression of the short isoform differs from that of the long isoform. Nevertheless, as described below, targeted rescue experiments indicate that these two isoforms are functionally equivalent.

Nervous system expression of either RanBPM isoforms rescues all RanBPM larval behavioral phenotypes

In order to identify the cells or group(s) of cells in which RanBPM expression plays a role in larval behavior, we generated RanBPM constructs whose expression can be targeted to particular cells and tissues under the control of tissue-specific GAL4 transgenic constructs [56]. Using standard genetic crosses we created homozygous RanBPM^k05201mutant flies carrying the RanBPM^long or RanBPM^short cDNA under the control of the GAL4-responsive DNA binding site (UAS) as well as different GAL4 drivers.

Expression of either RanBPM isoform under the control of the tub-GAL4 driver reduced the recessive lethality associated with RanBPM mutations from 100% to 61–69% (N>200). demonstrating that the both constructs contained all the information required to restore RanBPM gene function and that the two isoforms are functionally equivalent.

Next, we asked whether expression in the nervous system was sufficient to rescue RanBPM phenotypes. Targeted expression of either the short or long RanBPM isoform using the pan-neural driver embryonic lethal abnormal visual system (elav)-GAL4 effectively rescued the response to light phenotype of RanBPM mutants (Figure 5A) as well as the feeding phenotype (Table 1). Using the elav-GAL4 driver overall locomotion of RanBPM mutant larvae markedly improved but was only significantly above background when the long isoform construct was used. (Figure 5B). Interestingly lethality was partially rescued as seen by the survival of a small fraction (∼14%) of RanBPM^k05201 mutants flies to adulthood (Supplementary Table S1 and S3). These flies displayed a spread wing phenotype similar to that of Dichaete (D) mutants also seen in all escapers of heteroallelic combinations that supported survival to adulthood.

Download:

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

Figure 5. Targeted expression of RanBPM in subsets of neurons rescues the control of locomotion by light and the locomotion phenotypes of RanBPM^k05201 mutants.

A- The presence of any of the GAL4 drivers or either one of the UAS constructs did not significantly change the larval response to light. Pan-neural (elav-GAL4) expression of either RanBPM^long or RanBPM^short constructs significantly increased the response to light of RanBPM^k05201 mutants relative to that of RanBPM^k05201 mutants alone or carrying one copy of either one the GAL4 drivers or UAS constructs. Expression of the short RanBPM isoform under the control of 247-GAL4 (29°C) and Dmef2-GAL4, significantly increased the response to light phenotype of RanBPM^k05201 mutants. RanBPM mutant larvae expressing the UAS-RanBPM^long construct in the MB under the control of the 247-GAL4 driver displayed significantly higher response to light only when grown at 29°C. * p<0.05 N≥10 (ANOVA: F_(10.082) = 20,256, p<0.001). B-Locomotion of RanBPM^k05201 mutant for RanBPM^K05201 mutants expressing the long or short isoform under the control of various GAL4 drivers is shown as the number of pixels (mean ± SEM) in 30 sec of the assay. Locomotion was significantly reduced in RanBPM^k05201 mutant larvae relative to revertant control and did not improve significantly when these mutants carried one copy of any of the GAL4 drivers or one copy of the long isoform construct. Expression of the long RanBPM isoform under the control of elav-Gal4 and 247-GAL4 (29°C) significantly increased locomotion above that of all mutant controls to levels still below that of the RanBPM^K05201 revertant control. Presence of the short RanBPM isoform construct significantly increased the locomotion of RanBPM^K05201 mutant larvae and that was not increased when a GAL4 driver was introduced. *p<0.05, N≥11 (ANOVA: F_(19.491) = 20,257, p<0.001).

https://doi.org/10.1371/journal.pone.0010652.g005

The elav gene and the elav-GAL4 line used in this study are highly expressed in postmitotic neurons of the central and peripheral nervous system [52] and transiently expressed in glia and neuroblasts [53]. Therefore, it is possible that expression in cells other than differentiated neurons contributed to the observed rescue of the mutant phenotypes. Targeted expression of either RanBPM isoform using the glia-specific driver repo-GAL4 did not rescue any of the phenotypes (data not shown) and co-localization with the anti-Repo antibody was not observed (Figure S5 G-L). However, the inability to rescue a given mutant phenotype when using a heterologous promoter may be due to target gene expression that is below the threshold required, rather that inappropriate tissue-specific expression. Taken together, our results demonstrate a role in larval behavior, for RanBPM gene function in the nervous system, perhaps in postmitotic neurons, and suggest that the lack of RanBPM gene function in this tissue contributes to the observed lethality.

Expression of RanBPM in the Kenyon neurons of the MB is sufficient to rescue the response to light and feeding phenotypes

Given that the long isoform of RanBPM is highly expressed in the MB neurons, we asked whether expression in these neurons was sufficient to rescue RanBPM mutant phenotypes. To that end, we used GAL4 drivers whose expression overlapped in the MB. 247-GAL4 is derived from an enhancer found in the Dmef2 promoter and is highly and nearly exclusively expressed in the MB neurons throughout larval development until adulthood ([50] and data not shown). Dmef2-GAL4 is expressed in several neurons, in addition to those of the MB, as well as in somatic muscles [49]. The pattern of expression of 386-GAL4 extends to numerous peptidergic neurons [31] and includes MB neurons (Figure S4 C-D').

The performance of RanBPM^k05201 mutants expressing either isoform under the regulation of MB GAL4 drivers in the ON/OFF assay, is shown in Figure 5 panel A. Targeted expression of at least one of the two RanBPM isoforms under the regulation of two (247 and Dmef2) of the three MB GAL4 drivers was sufficient to restore the response index of RanBPM mutants to levels that are not statistically different from that of control strains. Targeted expression of the long isoform using the 386-GAL4 driver markedly improved the response of mutants but this was not statistically significant. In the case of the 247-GAL4 driver, marked and significant rescue occurred using either isoform when larvae were grown at 29°C, the optimal temperature for the function of the yeast transcription factor GAL4

In rescued larvae (RanBPM^k05201; 247:RanBPM^long) expression of the long isoform of RanBPM is restored in the MB neurons (Figure 4L). The subcellular localization within the MB structure however, differs from that of wild type samples. In the rescued specimens RanBPM immunolabelling is found in the cell bodies, the calyx and in the pedunculi, while in wild type specimens it is restricted to the cell bodies (Figure 4 compare A to L). This is likely due to the heterologous promoter, in this case the 247-GAL4 driver, driving the expression of the UAS-RanBPM target construct to levels above that of wild type, thereby causing the ectopic expression of this protein in the pedunculi.

Significant improvement of locomotion occurred in mutants expressing the long isoform in the MB neurons when grown at 29°C (RanBPM^k05201; 247:RanBPM^long, Figure 5B). While marked increased in locomotion was also observed when the other MB GAL4 drivers were used, they were not statistically significant. Lethality was marginally rescued by expression of RanBPM^long under the control of Dmef2-GAL4 and not at all when 247-GAL4 and 386-GAL4 were employed (Table S1 and S3). In contrast feeding was completely restored when all three MB GAL4 drivers were employed (Table 1). We concluded that RanBPM expression in the Kenyon neurons is sufficient to rescue the behavioral phenotypes of RanBPM mutants.

RanBPM gene function contributes to FMRP-dependent processes

RanBPM was previously identified in a yeast two-hybrid screen as a putative partner of FMRP [18]. FMRP is an RNA binding protein implicated in the transport, translational control and metabolism of mRNAs encoding proteins involved in synaptic plasticity (reviewed by [23], [57], [58]. The Drosophila orthologue, dfmr1, has been used extensively as a model for FMRP function in the regulation of synaptic structure and function (reviewed by [59]). dfmr1 mutants display over -elaboration of neuronal structure exemplified by midline crossing of intrinsic MB neurons as well as increased branching and bouton number at the larval neuromuscular junction (NMJ)[20], [21], [32], [60]. Moreover dfmr1 function is required for activity dependent pruning of MB axons [24].

We asked whether RanBPM function is required for FMRP-dependent processes. To that end, we conducted genetic epistasis analysis to determine whether dfmr1 mutant phenotypes are modified (enhanced or suppressed) by changes in the level of RanBPM function. RanBPM^k05201 mutations cause lethality during late larval development while the dfmr1 mutant alleles used here, dfmr1^EP3517, dfmr1^Δ50M cause lethality at the pharate adult stage and are hypomorphic and amorphic mutations respectively. Interestingly RanBPM;dfmr1 double homozygotes die during embryogenesis. We asked whether reduction in RanBPM function modifies the NMJ overgrowth phenotype of dfmr1 mutants. dfmr1 mutant NMJ is characterized by pronounced synaptic overgrowth as seen by nearly 50% increase in branching and in the total number of boutons ([32], [60]; and Figure 6). While the NMJ of double mutants does not appear entirely normal, introduction of one copy of the lack of function allele RanBPM^k05201 is sufficient to suppress the overgrowth NMJ phenotype. Increased branching and bouton number is reduced in these double mutants to levels comparable to that of wild type controls. These observations indicate that RanBPM function is required for the expression of dfmr1phenotype and suggest that RanBPM and dfmr1function converge in the mechanisms underlying synaptic growth.

Download:

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

Figure 6. RanBPM function is required for FMRP-dependent development of larval NMJs.

Confocal micrographs of larval NMJs (muscles 6 and 7) labeled with FITC-conjugated anti-HRP (A–E), show the overgrowth phenotype characteristic of dmfr1⁵⁰ (B) and dmfr1^EP3517 (D). Introduction of one copy of the lack-of-function allele RanBPM^K05201 partially suppresses this phenotype (compare B to C and D to E). These observations are supported by quantitative analysis, in which total number of branches and boutons are counted (F, G). Reduction of RanBPM function significantly reduces the number of branches (F, ANOVA: F_(4,42) = 9.713 p<0.001) and boutons (G, ANOVA: F_(F4,41) = 28.382 p<0.001) of both dfmr mutants (N≥8).

https://doi.org/10.1371/journal.pone.0010652.g006