Fig 1.
Conserved properties of Drosophila FoxP.
(A) Schematic representation of human FOXP and Drosophila FoxP proteins. FOXP protein domains: glutamate-rich region (Q-rich, in blue), zinc finger (ZF, in yellow), leucine zipper (LZ, in red) and forkhead domain (Forkhead, in green) are indicated. % indicates aa similarity between human FOXP1, 2, 4 and Drosophila FoxP protein domains. (B) The Drosophila FoxP genomic region (exons in black, untranslated regions (UTRs) and introns in grey, and intron 6 who’s retention gives rise to FoxP-I3 in a striped pattern, START and alternative STOP codons in red) and the three encoded transcripts (FoxP-I1 to–I3). Protein domain-encoding regions are highlighted using the color code used in (A). Primers for RT-PCR analysis are indicated with numbers in the FoxP genomic region. (C) Agarose gel analysis of RT-PCR products amplified with primers 1–2 in lane 1 (PCR products corresponding to FoxP-I1 (I1, 1329bp) and -I3 (I3, 1701bp)) and primers 1–3 in lane 2 (lower band corresponds to FoxP-I2 (I2, 2999bp), upper band (*) corresponds to an amplicon derived either from an unspliced FoxP pre-mRNA or amplification of genomic DNA present in the sample (2824bp)). Lane 3: negative control (primers, no template). Lane 4: molecular weight marker. (D) FoxP-FoxP dimerization in the yeast two-hybrid assay. The utilized construct (light grey) and isolated FoxP fragment (prey, dark grey) are depicted. The yeast two-hybrid bait alone shows no autoactivation and growth. When yeast are co-transformed with both, bait and prey induce colony growth and β-galactosidase activity, demonstrating FoxP dimerization.
Fig 2.
FoxP is expressed in approximately 1000 neurons of the Drosophila adult brain.
(A) Graph represents relative FoxP expression levels over several developmental stages in wildtype flies. (B) Bars represent average relative FoxP expression in neural tissues (striped bars) and non-neural tissues (black bars) over different developmental stages. Data are represented as average and SEM of at least 3 biological replicates per developmental stage. (A, B) For the underlying numerical data see S2 Table. (C) Percentage of co-localization between anti-FoxP and anti-Elav signal or anti-Repo signal in wildtype male brains at 0–2 hours post-eclosion. Bars represent averages with SEM of a minimum of 4 brains. For the underlying numerical data see S3 Table. (D) Frontal brain schematic illustration of FoxP expressing neurons (green). The positions of clusters 1, 2 and 3 investigated for co-localization are highlighted by red squares. Clusters 1, 2 and 3 neurons of w; Cha-GAL4, UAS-GFP, w; Gad1-GAL4/UAS-GFPnls and Vglut-GAL4, w; UAS-GFPnls/+ flies co-immunostained with anti-FoxP (magenta) and anti-GFP (green). While maximum projections are shown, co-localisation was assessed on single optical sections. Arrows indicate co-localization. Scale bar corresponds to 10 μm (E-G) Maximum projection of Drosophila frontal brain image stacks. Scale bar corresponds to 50 μm. (E) w;;GFP-FoxP flies co-immunostained with anti-FoxP (magenta) and anti-GFP (green). (F) Wildtype flies co-immunostained with anti-FoxP (magenta) and anti-Elav (green) labeling neurons and (G) anti-Repo (green) labeling glial cells. (E’-G’) Magnification of E, F and G highlighted with a yellow square in the original images. Scale bar corresponds to 10 μm. (H) Maximum projection of image stack over a range of different brain depths showing the distribution of FoxP expressing neurons (green) together with the anatomical marker anti-nc82 (magenta) to visualize the different neuropils of w;;GFP-FoxP flies. Scale bar corresponds to 100 μm. (I) Schematic illustration of FoxP-expressing neurons (green) over the indicated brain sections. (Image stack is provided as S1 Video). Arrowheads indicate co-localization. Images were obtained from male brains at 0–2 hours post-eclosion.
Fig 3.
Characterization of FoxP mutants and RNAi lines.
(A) Schematic representation of the FoxP genomic region and extension of the FoxP deletion. The transcriptional start site is indicated with an arrow, the two alternative stop sites with red lines. The original location of the P-element insertion GS22100 is depicted with a black triangle. The FoxP sequence targeted by RNAi1 and RNAi2 is depicted as a blue line. Genes flanking FoxP on each side are also indicated. All three genes are oriented in the same direction. (B) Adult brain hemisphere of wildtype (Wt), FoxP null (FoxP-/-), w, UAS-Dcr2/Y; Actin-GAL4/+; UAS-FoxP-RNAi1 (FoxP-RNAi1)/+ and w, UAS-Dcr2/Y; Actin-GAL4/+; UAS-FoxP-RNAi2/+ (FoxP-RNAi2) stained with anti-FoxP antibody. Scale bar corresponds to 50 μm. Images were obtained from male brains at 0–2 hours post-eclosion.
Fig 4.
FoxP depletion leads to reduced fitness.
(A, C) Fraction of dead pupa (in %). A minimum of 6 experimental replicates were analyzed per genotype. (B, D) Survival of males (in %) over days post-eclosion. A minimum of 4 experimental replicates were analyzed per genotype, with 15 male flies per experiment. (E, F) Locomotion trajectories of representative flies of the indicated genotypes. Male flies were recorded for 7 minutes at 10 frames per second in a circular arena (37 mm diameter). (G, I) Total distance (in cm) of walk in the 7 minutes of locomotion tracking. Data are represented as average and SEM of a minimum of 2 independent biological replicates per genotype. (H, J) Drosophila escape responses, assessed in the island assay. Graphs show % of flies that remain on the platform over time (10 seconds). Data are represented as average and SEM of 4 independent experimental replicates. The genotypes depicted in the graphs are: FoxP homozygous mutant (FoxP-/-, red), FoxP heterozygous mutant (FoxP+/-, pink), wildtype (Wt, black), w, UAS-Dcr2/Y; Actin-GAL4/+ (Controls, grey), w, UAS-Dcr2/Y; Actin-GAL4/+; UAS-FoxP-RNAi1/+ (FoxP-RNAi1, dark blue) and w, UAS-Dcr2/Y; Actin-GAL4/+; UAS-FoxP-RNAi2/+ (FoxP-RNAi2, light blue). One-way ANOVAs with Tukey’s multiple comparison test were used to compare each condition and determine significant differences (*p<0.05, **p<0.01 and ***p<0.001). For the underlying numerical data see S4–S7 Tables.
Fig 5.
MB α-lobe morphology is affected in FoxP-/- flies.
(A-B) Maximum projection of MB image stacks of fly brains stained with anti-Fasll. Scale bar corresponds to 20 μm. (A) Wildtype and (B) FoxP mutants (FoxP-/-), arrowheads indicate MB α-lobes. (C) MB α-lobes area. Maximum projection of MB Kenyon cells, (D) Wildtype flies co-immunostained with anti-FoxP (magenta) and anti-Dac (green), (E) w;;GFP-FoxP co-immunostained with anti-GFP (green) and anti-Dac (magenta). Scale bar corresponds to 20 µm. Data are represented as average and SEM of a minimum of 31 α-lobes. T-test between conditions was performed to determine significance (***, p<0.001). Images were obtained from male brains at 0–2 hours post-eclosion. For the underlying numerical data see S8 Table.
Fig 6.
FoxP regulates NMJ postsynaptic morphology.
Muscle four type 1b NMJs of FoxP-/- mutant and wildtype wandering L3 male larvae. (A-B) Co-immunostaining of Dlg1 and Hrp. Scale bar: 10μm. Dlg1 staining showing a honeycomb-like pattern, disorganization and covering a wider region at FoxP-/- mutant synaptic terminals compared to wildtype (Wt) terminals. (C) Dlg1 synaptic area is significantly increased in FoxP-/- mutants (wt n = 56, FoxP-/- n = 60). (D) Hrp-labelled synaptic area does not differ between FoxP-/- and wildtype (wt n = 27, FoxP-/- n = 23). (E-F) Co-immunostainings of Syt and Dlg1. Scale bar: 10μm. (G) The number of synaptic boutons (wt n = 55, FoxP-/- n = 35) and (H) the number of active zones (wt n = 18, FoxP-/- n = 19) do not differ between FoxP-/- and wildtype. (I, J) Co-immunostainings of Futsch and Dlg1. Scale bar: 2μm and (K, L) Co-immunostainings of Brp, GluRllC and Dlg1. Scale bar: 2μm. Syt, Futsch, Brp and GluRllC NMJ immunolabeling appear normal in FoxP-/-. Electron micrographs of third instar larvae NMJ synaptic boutons (muscle 6/7), (M) wildtype and (N) FoxP mutants. SSR surrounding synaptic boutons was shaded in pale red using Adobe illustrator. Asterisks indicate non-tubular structures located in the SSR, arrowheads indicate defective mitochondria. Scale bar: 1 μm. (O) SSR area is significantly increased in FoxP-/- (wt n = 39, FoxP-/- n = 30). (P) The area occupied by non-tubular structures is significantly increased in FoxP-/- (wt n = 40, FoxP-/- n = 30). (Q-R) Mitochondria surrounding the SSR present ultrastructural defects in FoxP-/- mutants. Arrowheads indicate defective cristae; arrows indicate multilobar mitochondria; asterisks indicate membranes folds around the mitochondria resembling autophagosomal structures, circles indicate collapsed mitochondria. Scale bar: 250 nm. (S-T) The conformation of neuronal mitochondria is unaffected in the FoxP-/-. Arrowheads indicate mitochondria. Scale bar: 500 nm. Bars represent the mean, error bars indicate SEM, t-test between conditions was performed for each parameter to determine significance (***, p<0.001). For the underlying numerical data see S9 and S10 Tables.
Fig 7.
FoxP is required for dendritic growth and negatively regulates branching of type IV da neurons.
(A-D) Confocal projections of class IV da neurons within segment A3 of third instar larvae, visualized with the class IV da-specific GFP expression (477-GAL4>UAS-mCD8::GFP). Reconstructions are represented in blue and are overlapping the original neuron. Scale bar: 100μm. (A’-D’) Magnification of dendrites as highlighted in the original image. Scale bar: 20μm. The following genotypes are depicted in the panels (A) w w/Y; 477-GAL4<UAS-mCD8::GFP/+; +/+ (wildtype), (B) w/Y; 477-GAL4>UAS-mCD8::GFP; FoxP- (FoxP-/-) (C) w/Y; 477-GAL4>UAS-mCD8::GFP/+ (controls) and (D) w/Y; 477-GAL4>UAS-mCD8::GFP/+; UAS-FoxP-I2/+ (UAS-FoxP-I2). (E-N) Quantitative analysis of dendritic trees, FoxP-/- presents a decrease in (E) dendritic field area and (F) average branch length. (G) Cumulative branch length and (H) number of endings are unaffected. Wt (n = 5), FoxP-/- (n = 5). (I) Dendritic endings density (number of endings in 100μm2) is increased in FoxP-/-. Wt (n = 9), FoxP-/- (n = 9). Wt are depicted in black versus FoxP-/- depicted in red. UAS-FoxP-I2 presents unaffected (J) dendritic field area and (K) average branch length. (L) Cumulative branch length and (M) number of endings are decreased. Controls (n = 5), UAS-FoxP-I2 (n = 5). (N) Dendritic endings density (number of endings in 100μm2) is decreased in UAS-FoxP-I2. Controls (n = 10, in grey), UAS-FoxP-I2 (n = 10, in green). (O, Q) Sholl analysis of cumulative dendritic length; graph indicates the sum of dendritic length in concentric circles from the soma situated every 10μm. (O) Wt versus FoxP-/- and (Q) control versus UAS-FoxP-I2. (P, R) Sholl analysis of cumulative number of branching points; graph indicates the sum of branching points located in concentric circles from the soma situated every 10μm. (P) Wt versus FoxP-/- and (R) control (w/Y; 477-GAL4>UAS-mCD8::GFP/+) versus UAS-FoxP-I2. Wt (n = 5), FoxP-/- (n = 5) controls (n = 5) and UAS-FoxP-I2 (n = 5). Data are presented as averages with SEM. T-tests between conditions were performed for each parameter to determine significance (** p<0.01 and *** p<0.001). For the underlying numerical data see S13 and S14 Tables.
Fig 8.
FoxP regulates light-off jump habituation and social distance.
(A-C) Average jump responses (y axis) of three to six days old male flies from 3 independent experiments, with a minimum of 16 flies tested per experiment, plotted over 100 light off trails (x axis). (A) FoxP-/5-SZ-3955 mutant flies (w; 2xGMR-wIR; FoxP-/5-SZ-3955, in red) and wildtype flies (w; 2xGMR-wIR; +/+, in black) (B) FoxP panneuronal knockdown with FoxP-RNAi1 (light blue) and FoxP-RNAi2 (dark blue) (w/Y; 2xGMR-wIR; elav-GAL4, UAS-Dcr2/FoxP-RNAi) and control flies (w/Y; 2xGMR-wIR; elav-GAL4, UAS-Dcr2/+; in grey), (C) panneuronal overexpression of FoxP-I1 (dark green) and FoxP-I2 (light green) (w/Y; 2xGMR-wIR/+; elav-GAL4, UAS-Dcr2/ UAS-FoxP) and the respective controls (w/Y; 2xGMR-wIR/+; elav-GAL4, UAS-Dcr2/+; in black). (A’- C’) Mean number of trials to criterion (TTC) ± SEM of a minimum of 3 experimental replicates. T-tests or one-way ANOVAs with Dunn’s multiple comparisons were performed to assess differences between TTC of the different conditions (* p<0.05, ** p<0.01 and *** p<0.001). For the underlying numerical data see S17 Table. (D-F) Data of the social space assay are represented as cumulative relative frequency of the distance to the closest neighbor (Freq. of interfly distance). (D) FoxP-/- mutants position themselves closer to each other than their Wt controls (Mann-Whitney, n = 136 Wt and n = 86 FoxP-/- flies). (E) Panneuronal FoxP downregulation decreases social space (Mann-Whitney, n = 122 w;UAS-Dcr2/+; elav-GAL4/+ and n = 108 w;UAS-Dcr2/+; elav-GAL4/UAS-FoxP-RNAi1 flies). (F) Panneuronal overexpression of FoxP-I1, but not Fox-I2, increases social space at a distance >0.5 cm away from each other (grey rectangle). (Mann-Whitney, w;;elav-GAL4/UAS-FoxP-I1 and w;;elav-GAL4/UAS-FoxP-I2 n = 58 flies each, and w;;elav-GAL4/+ n = 45 flies).