Bruchpilot in ribbon-like axonal agglomerates, behavioral defects, and early death in SRPK79D kinase mutants of Drosophila

Abstract

Defining the molecular structure and function of synapses is a central theme in brain research. In Drosophila the Bruchpilot (BRP) protein is associated with T-shaped ribbons ("T-bars") at presynaptic active zones (AZs). BRP is required for intact AZ structure and normal evoked neurotransmitter release. By screening for mutations that affect the tissue distribution of Bruchpilot, we have identified a P-transposon insertion in gene CG11489 (location 79D) which shows high homology to mammalian genes for SR protein kinases (SRPKs). SRPKs phosphorylate serine-arginine rich splicing factors (SR proteins). Since proteins expressed from CG11489 cDNAs phosphorylate a peptide from a human SR protein in vitro, we name CG11489 the Drosophila Srpk79D gene. We have characterized Srpk79D transcripts and generated a null mutant. Mutation of the Srpk79D gene causes conspicuous accumulations of BRP in larval and adult nerves. At the ultrastructural level, these correspond to extensive axonal agglomerates of electron-dense ribbons surrounded by clear vesicles. Basic synaptic structure and function at larval neuromuscular junctions appears normal, whereas life expectancy and locomotor behavior of adult mutants are significantly impaired. All phenotypes of the mutant can be largely or completely rescued by panneural expression of SRPK79D isoforms. Isoform-specific antibodies recognize panneurally overexpressed GFP-tagged SRPK79D-PC isoform co-localized with BRP at presynaptic active zones while the tagged -PB isoform is found in spots within neuronal perikarya. SRPK79D concentrations in wild type apparently are too low to be revealed by these antisera. We propose that the Drosophila Srpk79D gene characterized here may be expressed at low levels throughout the nervous system to prevent the assembly of BRP containing agglomerates in axons and maintain intact brain function. The discovery of an SR protein kinase required for normal BRP distribution calls for the identification of its substrate and the detailed analysis of SRPK function for the maintenance of nervous system integrity.

Figure 1. Structure of the Srpk79D gene, transcripts, and homology of kinase domain.

(A) Exon-intron structure of the Srpk79D gene (CG11489) according to flybase modified due to the detection of the new exon 7 added (pink rectangle). The positions of the forward (f) and reverse (r) primers used for RT–PCR are indicated. The position of the P(lacW) insertion of the line Srpk79D^P1 is shown. It destroys the open reading frame of the RC/RF transcripts but leaves the RB/RE transcripts intact. This line is therefore considered a hypomorph. The P-jump-out mutant line Srpk79D^VN suffered a deletion of 3861 bp indicated by the horizontal bar and contains a 23 bp remnant of the P-element. It is considered a null mutant. Sequences coding for the bipartite kinase domain are depicted in green. (B) The four transcripts inferred from RT–PCR with adult mRNA using the primer pairs indicated in (A). The two groups of transcripts RC/RF and RB/RE use different transcription and translation start sites (vertical bars marked by arrows) and are alternatively spliced. Stop codons are marked by arrowheads. The transcript RD of flybase is shown in a different color as it could not be verified using primer pair 4. The corresponding cDNA probably has been generated from an incompletely spliced RNA. (C) The two parts of the kinase domain of Dm SRPK79D (coding regions green in A) are separated by non-conserved spacers. The kinase domain is identical for all four protein isoforms (PC, PF, PB, PE) and shows high homology to human SR protein kinases SRPK3 and SRPK2 (and SRPK1, not shown). Identical amino acids are in red. The active center of the kinases is underlined.

Figure 2. A peptide from a human SR protein (SRPK1tide) is a substrate for Drosophila SRPK79D in vitro.

The two isoforms PC and PB of the SRPK79D differ in their N-terminal domain but are both able to phosphorylate the synthetic peptide SRPK1tide significantly. Measurements in all experimental groups were repeated five times independently. Statistical significance (Student's t-test with Bonferroni correction, **: p<0.01; ***: p<0.001) was tested against the corresponding groups without substrate, as indicated in the figure. Western blot analyses with anti-Myc antibody revealed that SRPK79D isoforms were loaded in all experimental groups, except in the negative control.

Figure 3. Mutation of the Srpk79D gene leads to the selective accumulation of the active zone protein BRP in axons of larval and adult nerves, but does not qualitatively change BRP distribution in synaptic boutons or neuropil.

(A–I). Bruchpilot accumulations are observed in larval nerves of the hypomorph Srpk79D^P1 (A), the null mutant Srpk79D^VN (B), an elav-driven RNAi line elav-Gal4;;UAS-Srpk79D-RNAi (C), the parental lines for the P1-rescue with the PF isoform, w,elav-Gal4;;Srpk79D^P1 (D) and w;UAS-cDNA-RF;Srpk79D^P1 (E), the parental lines for the null-rescue with the PF isoform, w,elav-Gal4;;Srpk79D^VN (G) and w;UAS-cDNA-RF;Srpk79D^VN (H), and the parental lines for the null-rescue with the PB isoform (G) and w;UAS-cDNA-RB-eGFP;Srpk79D^VN (K). The revertant Srpk79D^REV (J) and wild type (w¹¹¹⁸, not shown) lack these structures. In the P1-rescue line w,elav-Gal4;UAS-cDNA-RF;Srpk79D^P1 (F) and the null-rescue lines w,elav-Gal4;UAS-cDNA-RF;Srpk79D^VN (I) and w,elav-Gal4;UAS-cDNA-RB-eGFP;Srpk79D^VN (L) the density of BRP accumulations is strongly reduced. (M–P). Typical chains of synaptic boutons type Ib stained with anti-BRP antibody nc82 on larval body wall muscles 6/7 of the hypomorphic mutant Srpk79D^P1 (M), wild type w¹¹¹⁸ (N), null mutant Srpk79D^VN (O), and revertant Srpk79D^REV (P). No difference in the distribution and qualitative expression levels of the presynaptic active zone protein BRP is seen between Srpk79D mutants and wild type. (Q–R). CSP does not accumulate in the larval nerves of the hypomorphic mutant Srpk79D^P1 (Q) or the null mutant (not shown). (R) w¹¹¹⁸ control. (S) Overexpression of BRP also leads to BRP accumulations in larval nerves. (T–U) BRP immunohistochemistry on frozen head sections of the null mutant Srpk79D^VN (T) and wild type (U) reveals that massive spot-like BRP accumulations are observed in adult Srpk79D^VN null-mutant antennal nerves (arrows in (T)). Larval preparations in (A–P) and (S) and the head sections in (T) and (U) were stained with monoclonal antibody (MAB) nc82 (anti-Bruchpilot). Larval nerves in (Q) and (R) were stained with MAB ab49 (anti-CSP). Scale bar in (A) for (A L), (Q–S): 10 µm. Scale bar in M for (M–P): 2 µm. Scale bar in (T) for (T–U): 50 µm.

Figure 4. Basic synaptic transmission at the larval neuromuscular junction is normal in Srpk79D^VN null mutants.

Amplitude and time course of evoked EJP responses of w¹¹¹⁸ and Srpk79D^VN null-mutant larvae are not significantly different ((A), upper diagrams and (C) left diagram) (w¹¹¹⁸ = 34.2±4.0 mV n = 9; Srpk79D^VN = 33.8±3.7 mV; n = 7; p>0.05). Spontaneous single vesicle fusion events (mEJPs) have similar amplitude distributions (B) and similar mean amplitudes (w¹¹¹⁸ = 0.92±.1 mV and Srpk79D^VN = 0.92±0.07 mV; n = 8; p>0.05) and frequencies (w¹¹¹⁸ = 1.5±0.2 Hz and Srpk79D^VN = 1.4±0.2 Hz; n = 8; p>0.05) (D). Thus, quantal content is also not significantly different ((C), right diagram) (w¹¹¹⁸ = 116±19 vesicles and Srpk79D^VN = 119±27 vesicles; n = 8; p>0.05). All data are means±SEM.

Figure 5. Mutation of the Srpk79D gene causes accumulation of BRP at extensive electron-dense ribbon structures in larval nerves.

(A) Ultra-thin sections of a larval nerve of the Srpk79D^VN null mutant. (B) Higher magnification of the conspicuous electron-dense structure in (A) (arrow). (C) For comparison a typical synaptic ribbon (T-bar, arrowhead) from a synaptic bouton of an Srpk79D⁺ larval motor neuron terminal is shown. (D–K). Examples of large electron-dense structures observed in the mutant. (L–O) Silver-enhanced immuno-gold labelling using MAB nc82 (anti-BRP) as first antibody demonstrates that the electron-dense ribbon structures contain BRP. The silver precipitates (white circles) can be clearly recognized at enhanced brightness as shown for (L) in Figure S4. Scale bars in (A): 2 µm, in (B–O): 300 nm.

Figure 6. Overexpressed SRPK79D-PC and Bruchpilot co-localize at presynaptic active zones while overexpressed SRPK79D-PB accumulate in discrete perikaryal regions.

The synaptic distribution of overexpressed SRPK79D-PC-GFP protein, stained with anti-GFP (A) or anti-PC (B) antisera (green, left) largely overlap with endogenous Bruchpilot stained with anti-BRP (nc82) (red, middle) at presynaptic active zones. Control stainings in the Srpk79D^VN null mutant with anti-PC (C) demonstrate the specificity of the staining. No reliable synaptic staining is obtained with anti-PB antiserum (D). Both overexpressed isoforms are detected in the perikaryon, PC homogeneously ((E), left column), PB in discrete spots (arrows in (F), left column). BRP is not detectably localized in perikarya ((E), (F), middle column). Right column: overlay. Scale bar in (D) for (A–D): 5 µm. Scale bar in (F) for (E) and (F): 20 µm.

Figure 7. Mutation of Srpk79D gene results in behavioral deficits and reduced life span.

(A) Compared to wild-type controls (WT) a significantly larger percentage of adult Srpk79D^P1 (P1) and Srpk79D^VN (VN) mutants and parental flies for the rescue w,elav-Gal4;;P1 (G-P1) and w;UAS-RF;P1 (U-P1) drop to the bottom of a 500 ml glass cylinder when released at the top of the cylinder. Revertants (REV) and rescue flies w,elav-Gal4;UAS-RF;P1 (RES) are not significantly different from wild-type. (B) Both mutants and the parental flies for the rescue also show lower spontaneous walking activity on a horizontal surface (wings clipped, number of lines of a 2×2 cm grid crossed during a 30 s period) whereas revertants and rescue flies are not impaired. (C) Longevity of both mutants and the parental flies for the rescue is significantly reduced. Rescue flies live significantly longer than their parents but not as long as wild type or revertants. Thus for longevity rescue is only partial. (Significance levels: ***: p<0.001; **: p<0.01; *: p<0.05, paired t-tests with Bonferroni correction of each column against WT).