2mit, an intronic gene of Drosophila melanogaster timeless2, is involved in behavioral plasticity

Abstract

Background: Intronic genes represent ~6% of the total gene complement in Drosophila melanogaster and ~85% of them encode for proteins. We recently characterized the D. melanogaster timeless2 (tim2) gene, showing its active involvement in chromosomal stability and light synchronization of the adult circadian clock. The protein coding gene named 2mit maps on the 11(th) tim2 intron in the opposite transcriptional orientation.

Methodology/principal findings: Here we report the molecular and functional characterization of 2mit. The 2mit gene is expressed throughout Drosophila development, localizing mainly in the nervous system during embryogenesis and mostly in the mushroom bodies and ellipsoid body of the central complex in the adult brain. In silico analyses revealed that 2mit encodes a putative leucine-Rich Repeat transmembrane receptor with intrinsically disordered regions, harboring several fully conserved functional interaction motifs in the cytosolic side. Using insertional mutations, tissue-specific over-expression, and down-regulation approaches, it was found that 2mit is implicated in adult short-term memory, assessed by a courtship conditioning assay. In D. melanogaster, tim2 and 2mit do not seem to be functionally related. Bioinformatic analyses identified 2MIT orthologs in 21 Drosophilidae, 4 Lepidoptera and in Apis mellifera. In addition, the tim2-2mit host-nested gene organization was shown to be present in A. mellifera and maintained among Drosophila species. Within the Drosophilidae 2mit-hosting tim2 intron, in silico approaches detected a neuronal specific transcriptional binding site which might have contributed to preserve the specific host-nested gene association across Drosophila species.

Conclusions/significance: Taken together, these results indicate that 2mit, a gene mainly expressed in the nervous system, has a role in the behavioral plasticity of the adult Drosophila. The presence of a putative 2mit regulatory enhancer within the 2mit-hosting tim2 intron could be considered an evolutionary constraint potentially involved in maintaining the tim2-2mit host-nested chromosomal architecture during the evolution of Drosophila species.

Figure 1. tim2-2mit host-nested gene organization in D. melanogaster.

(A) Schematic representation of the tim2 locus showing positions and intron-exon structures for the four nested sequences (CG34308, BK002510, 2mit and AY118619) with only 2mit representing a protein coding gene. Arrowheads show positions of PB (indicated with c or f) or MB transposons in different insertional strains. (B) Diagram showing the organization of protein domains contained in 2MIT. The signal peptide (yellow), the LRR domain with 16 repeat units (blue) and their N- and C- flanking regions (orange), the transmembrane region (TM, purple), and the presence of Ala-, Thr-, Ser-rich domains (green) are shown both for the extracellular and cytoplasmic portions of the protein (drawing not to scale). The 3D-structure at the top represents the 2MIT LRR repeats of the Toll-like receptor 3 crystal structure (PDB code: 2A0Z, chain A). The predictions of sequence features (low complexity and disorder) are listed at the bottom as grey and black rectangles. Linear motifs found with ELM with their approximate position are indicated by black vertical bars.

Figure 2. Structural analysis of the D. melanogaster 2MIT LRR domain.

(A) Identification and structural alignment of the LRR units. The sequence of the LRR N-flanking region forming three hypothetical α-helices is outlined at the top. The main LRR repeats are aligned based on the structural correspondence between residues. A schematic cartoon used to depict the β-strand and the α-helix is shown above, and the consensus sequence is highlighted below the repeats. The bottom shows the LRR C-flanking region sequence with connecting lines between cysteines forming disulfide bonds. (B) Two dimensional representation of LRR repeats with β-strand and α-helix from N- to C-terminus. The consensus sequence is shown inside circles representing single residues. Conserved hydrophobic positions (pink circles) face the internal side of the repeat, while conserved polar residues (blue circles) map on the surface.

Figure 3. 2mit mRNA expression in w ¹¹¹⁸ flies.

(A-D) Embryos at different developmental stages hybridized with the antisense 2mit probe. (A) Stage 5. (B) Stage 9. (C-D) lateral and frontal views of stage 15. (E) Negative control showing a stage 15 embryo hybridized with the 2mit sense probe. Bar in (E) represents 25 µm for (A)-(E). (F) Northern blot from wild-type adult heads sampled every 3 h in 12:12 LD conditions. 2mit indicates the single 2mit transcript revealed in adult heads. rp49 represents the rp49 housekeeping mRNA. M: RNA Ladder molecular marker. (G) 2mit mRNA levels [mean ± standard error of the mean (SEM)] sampled every 4 h in 12:12 LD (white bars) and DD (black bars) conditions. For each condition, 3 replicates were performed. In 12:12 LD, significant variations in 2mit mRNA levels were detected (F_5,10=7.89, p <0.01). In DD, no significant modifications in 2mit mRNA levels were detected (F_5,12= 0.56 p= 0.76). (H-N) 2mit mRNA localization in whole-mount adult brains sampled at ZT 0. (H-L) Wild-type adult brains hybridized with the 2mit antisense probe. (H) 2mit mRNA signals are visible at the level of the Kenyon cells (arrow) and in the different lobes of the MBs (image shows a ~6 µm Z-projection along the antero-posterior axis). (I) and (L) ~5 µm Z-projections obtained from the same brain showing 2mit mRNA signals in the Kenyon cells (arrows; I) and in the EB (L). (M) 2mit ^c03963 adult brain hybridized with the 2mit antisense probe. Weak signals are detected in the MB lobes and EB. (N) Wild-type adult brain hybridized with the 2mit sense probe (negative control). (M) and (N) are ~10 µm Z-projections. The following abbreviations are used: ol: optic lobe; al: antennal lobe; mb: mushroom bodies; mbl: mushroom bodies lobes; eb: ellipsoid body; sg: subesophageal ganglion; cg: central ganglion. α/ α’: vertical mushroom bodies lobes; β, β’, γ: medial mushroom bodies lobes. Bar in (N) represents 50 µm for (H), (M) and (N), and 25 µm for (I), (L), and the 2X magnification inset in (M).

Figure 4. Memory and learning in 2mit ^c03963 homozygous mutant flies.

(A) Comparison between Courtship Indices in sham (white bars) and conditioned (black bars) males for different genotypes. w ¹¹¹⁸ and OR-R represent control strains; 2mit ^- : 2mit ^c03963 homozygous mutant flies; dnc ¹: dunce ¹ memory mutant flies; tim2 ^c06976 /+ and tim2 ^f00297 /+ heterozygous flies for two different tim2 ^- insertional mutant alleles. Data are expressed as mean ± SEM, with the number of tested flies indicated above each bar. Mann-Whitney U test revealed significant differences between the CIs of the conditioned and sham males in w ¹¹¹⁸, OR-R, tim2 ^c06976 /+, tim2 ^f00297 /+: p< 0.05 for all genotypes. No significant differences were identified comparing the CIs of the conditioned and sham males in 2mit ^c03963 and dunce ¹ flies (2mit ^c03963: p=0.38; dunce ¹: p=0.32). (B) Learning evaluated as a training index (ratio between CIs for the last (CI_f) and the first (CI_i) 10 min of the training period) in conditioned 2mit ^c03963 and w ¹¹¹⁸ males (Mann-Whitney U test: p=0.60). 2mit ^- : 2mit ^c03963 homozygous mutant flies. *: significant difference; ns: not significant difference.

Figure 5. 2mit over-expression in adult flies.

(A) 2MIT-HA Western blot on adult bodies of ActGal4>2mitO ^F8, ActGal4>2mitO ^M4, ActGal4>2mitO ^M14 and negative control ActGal4>+, in a wild-type 2mit ⁺ background. 2MIT-HA indicates the ~150 KDa 2MIT-HA form revealed in all ActGal4>2mitO over-expressing lines. HSP70 represents ~70 KDa Heat Shock Protein 70, used as loading control. Four replicates were performed. (B) 2mit mRNA relative levels (mean ± SEM) in dissected brains of three independent elavGal4>2mitO; 2mit ^c03963 lines (F8, M4 and M14) and in w ¹¹¹⁸ controls. Plot of 9 replicates. F_3,31= 552 p< 0.0001; Newman-Keuls post-hoc test: each elavGal4>2mitO; 2mit ^c03963 line vs w ¹¹¹⁸ controls p< 0.001. (C) Courtship Indices in sham (white bars) and conditioned (black bars) males for the elavGal4>2mitO, 2mit ^c03963 lines (F8, M4 and M14) and relative negative controls elavGal4>+; 2mit ^c03963; and + > 2mitO ; 2mit ^c03963 (F8, M4 and M14). 2mit ^- indicates the 2mit ^c03963 allele. Data are expressed as mean ± SEM, with the number of tested flies indicated above each bar. The CIs of conditioned flies were not significantly different from those of sham males in the negative control lines (Mann-Whitney U test: elavGal4>+; 2mit ^c03963: p= 0.24 ; + >2mitO ^F8 , 2mit ^c03963: p= 0.81; + >2mitO ^M4 ; 2mit ^c03963: p= 0.33; + >2mitO ^M14 ; 2mit ^c03963: p= 0.43), but were significantly reduced in elavGal4>2mitO, 2mit ^c03963 lines. The number of asterisks indicates the significance level: **: p < 0.005; ***: p < 0.0001; ns: not significant.

Figure 6. 2mit knockdown in adult flies.

(A) 2mit mRNA relative levels (mean ± SEM) in the dissected brains of three independent elavGal4>2mitKD (6.1; 16.2; 61.1) and w ¹¹¹⁸ controls. Plot of 6 replicates. F_3,20= 1571, p<0.0001; Newman-Keuls post-hoc test: each elavGal4>KD line vs w ¹¹¹⁸ controls: p< 0.001. (B) Courtship Indices in sham (white bars) and conditioned (black bars) males for elavGal4>2mitKD (6.1; 16.2; 61.1) and appropriate controls [elavGal4>+; +> 2mit KD (6.1; 16.2; 61.1)]. Mann-Whitney U test showed no significant differences between the CIs of sham and conditioned males in all the elavGal4>2mit KD (elavGal4>2mit KD^6.1.: p= 0.17; elavGal4>2mit KD^16.2: p= 0.13; elavGal4>2mit KD^61.1: p= 0.88). Significant differences between the CIs of conditioned and sham males were found in the control lines elavGal4> +; +> 2mit KD⁶¹. ; +> 2mit KD^16.2 ; +> 2mit KD^61.1 ; +> 2mit KD^61.1. (C) Courtship Indices in sham (white bars) and conditioned (black bars) males for OK107Gal4>2mitKD (6.1; 16.2; 61.1), c772Gal4>2mitKD (6.1; 16.2; 61.1), MB247Gal4>2mitKD (6.1; 16.2; 61.1) lines, and the control lines OK107Gal4>+, c772Gal4>+, MB247Gal4>+. Mann-Whitney U test showed no significant differences between the CIs of sham and conditioned males in two out of three OK107Gal4>2mit KD lines (OK107Gal4>2mit KD^16.2: p= 0.18; OK107Gal4>2mit KD^61.1: p= 0.36), in all the c772Gal4>2mit KD flies (c772Gal4>2mit KD⁶¹. : p= 0.67; c772Gal4>2mit KD^16.2: p= 0.72; c772Gal4>2mit KD^61.1: p= 0.42) and in all the three MB247Gal4>2mit KD lines (M247Gal4>2mit KD⁶¹. : p= 0.88; MB247Gal4>2mit KD^16.2: p= 0.12; MB247Gal4>2mit KD^61.1: p= 0.27). Significant differences between the CIs of conditioned and sham males were found in the control lines OK107Gal4>+, c772Gal4>+, MB247Gal4>+ and in the OK107Gal4>2mit KD⁶¹. transgenic line. In (B) and (C) data are expressed as mean ± SEM, with the number of tested flies indicated above each bar. The number of asterisks indicates the significance level: *: p < 0.05; **: p < 0.005; ***: p < 0.0001; ns: not significant.

Figure 7. Phase response curve of 2mit ^c03963 and w ¹¹¹⁸ flies.

Analysis of variance genotype X time interactions (ZT): F_5,434=1.76 p=0.12, not significant. Advance and delay phase shift responses are represented respectively as positive and negative values. 2mit ^-: 2mit ^c03963 homozygous mutant flies. Data are expressed as mean ± SEM; (Ns) and [Ns] indicate the number of 2mit ^c03963 and w ¹¹¹⁸ flies analyzed, respectively.

Figure 8. 2MIT Phylogenetic analysis.

(A) Unrooted phylogenetic tree of 2MIT protein in 21 Drosophilidae, 4 Lepidoptera and Apis mellifera. Statistical support for nodes on the trees was evaluated by the bootstrapping values (×500) shown in each branch point obtained by the Maximum-Likelihood method included in the MEGA 5.0 Software. (B) Colored bars at the right side of the phylogenetic tree represent binding motifs recognized by different factors in 2MIT ortholog cytoplasmic regions. CLV_NDR_NDR1: N-arginine dibasic convertase site; CLV_PCSK_KEX2_1: subtilisin-like proprotein convertases cleavage site; LIG_14-3-3_3: 14-3-3 ligand site; MOD_PKA_2: PKA Phosphorylation site; LIG_FHA_1 and LIG_FHA_2: FHA phosphopeptide ligands; MOD_N-GLC_1: N-glycosylation site; TRG_ER_diArg_1: di Arginine retention/retrieving signal; TRG_ENDOCYTIC_2: sorting signal motif; LIG_EVH1_1: EVH1 ligands.