Myofibril diameter is set by a finely tuned mechanism of protein oligomerization in Drosophila

Figure 1.. Aggregate formation upon Zasp overexpression. (A–H) Confocal images of IFM in control or Zasp overexpression conditions using Act88F-Gal4; actin filaments are marked in magenta and Z-discs in green. (A) In control flies, sarcomeres are regular with Z-discs located in the center of the actin signal. (B) Upon GFP-Zasp52-PR overexpression sarcomere structure is severely affected and big Zasp aggregates appear. (C) The overexpression of Flag-Zasp52-PP, an isoform lacking all LIM domains, has a modest sarcomere phenotype and aggregation is not observed. (D) Overexpression of a mutated form of Zasp52-PR with a deletion in the ZM domain does not form aggregates. (E) Overexpression of a mutated form of Zasp52-PR carrying a deletion in the PDZ domain causes small aggregates and sarcomere phenotypes. (F and G) The overexpression of the other Zasp paralogs, Zasp66 or Zasp67, does not cause aggregates. (H) Overexpression of GFP-Zasp52-PK, a smaller isoform of Zasp52 lacking LIM2, 3, and 4 domains affects sarcomere structure, but aggregation is infrequent. (I) Estimation of Z-disc aggregates in all overexpression conditions, n = 10 muscle fibers. (J) Confocal images of IFM in control or in…

Figure 1—figure supplement 1.. Differences in Zasp isoforms. Cartoon of selected Zasp protein isoforms with annotated protein domains. Isoforms are named according to FlyBase. Numbers at each end represent the length in amino acids. ZM, Zasp-like motif.

Figure 1—figure supplement 2.. Zasp-mediated aggregation depends only partially on α-actinin. (A, B) Confocal images of IFM in control or α-actinin depleted conditions; actin filaments are marked in magenta and Z-discs marked by Zasp66-GFP are shown in green. (A) Control sarcomeres are regular with Z-discs located in the center of the actin signal. (B) In α-actinin-depleted muscles sarcomere structure is severely perturbed, and Z-discs are rarely visible. (C, D) Confocal images of IFM with aggregates caused by the overexpression of Zasp52-PR. (C) Overexpression of Zasp52-PR produces aggregates. (D) Overexpression of Zasp52-PR in α-actinin-depleted muscles produces smaller aggregates. (E) Estimation of Z-disc aggregates in all overexpression conditions, n = 10 muscle fibers. Act88F-Gal4 was used in all panels. Scale bars, 10 µm. p-Values in panel E were calculated using Welch’s two-sample t-test.

Figure 2.. Y2H assays reveal domains involved in homo/heterodimerization. (A) Yeast two-hybrid assays between the three Zasp paralogs. Images of double-transformed yeast grown in selective -Ade/-His/-Leu/-Trp plates. Serial dilutions are shown from left to right (OD: 0.1, OD: 0.01, and OD: 0.001). Zasp52-PK dimerizes with itself and with the longer isoform Zasp52-PE. Zasp66 interacts with Zasp52-PK and Zasp52-PE. Zasp67 interacts with Zasp52-PK and Zasp52-PE. Negative controls using the DNA-binding domain (bait, pGAD-T7) or the Activating domain of Gal4 (prey, pGBK-T7) are shown. (B) Y2H assays testing the interaction between all protein domains encoded by the Zasp52 gene and Zasp52-PK, Zasp66-PK, or Zasp67-PD proteins. Zasp52-PK interacts with isolated ZM and LIM domains, Zasp66-PK and Zasp67-PD interact only with some LIM domains. LIM1A, LIM1B, LIM2A, and LIM2B are different splice isoforms of LIM1 and LIM2 domains. (C) Proposed model of Zasp homo/heterodimerization. The LIM domains bind the ZM domain. Zasp52-PK dimerization occurs through two ZM/LIM pairs. The heterodimerization between Zasp52-PK and Zasp66-PK or Zasp67-PD occurs through only one ZM/LIM-binding site. Zasp66-PH is a small…

Figure 2—figure supplement 1.. Zasp self-interaction by co-IP and chemical crosslinking. (A) Endogenous Zasp52 tagged with GFP was purified from thorax extracts via GFP beads. GFP-Zasp52 co-immunoprecipitates Flag-Zasp52-PK. GFP alone was used as negative control. (B) Purified His-Zasp52-PK-Flag from bacteria was treated with the chemical crosslinker ethylene glycol bis-sulfosuccinimidyl succinate (EGS) and then analyzed by SDS-Page. In addition to monomers (M, 50 kD), dimers (D) are observed upon EGS treatment.

Figure 3.. Zasp interaction in vivo at the Z-disc. (A–C) Confocal images of Zasp66-GFP IFM in control and Zasp52 mutant backgrounds. Zasp66-GFP levels are lower in Zasp52MI00979 mutant than in the control or in Zasp52MI02988 mutants. (D) Boxplot of Zasp66-GFP intensities in control and Zasp52 mutant backgrounds. (E, F) Examples of a negative BiFC control (F) and a positive BiFC signal (E) suggesting Zasp52-PK dimerizes at the Z-disc. (G, H) Plots of the BiFC fluorescence intensity values relative to background noise. Positive BiFC fluorescence is detected between Zasp52-PK and Zasp52-PK (G) and between Zasp52-PK and Zasp66-PH (H). Act88F-Gal4 was used to drive expression of NYFP- or CYFP-tagged proteins. Scale bar, 5 µm. p-Values in panels D, G, and H were calculated using Welch’s two-sample t-test. Figure 3—source data 1.Zasp66 levels and BiFC values.

Figure 3—figure supplement 1.. Zasp52 gene map and BiFC. (A) Cartoon representing the Zasp52 genomic locus with selected transcripts, two alternative transcription start sites, protein domains, and three MiMIC lines introducing an artificial exon which consists of a splice acceptor followed by stop codons for all three reading frames. (B) Confocal images of Actn-GFP IFM in control and Zasp52 mutant backgrounds. Actn-GFP levels are comparable between the control and Zasp52 mutants. (C) Boxplot of Actn-GFP intensities in control and Zasp52 mutant backgrounds. p-Values were calculated using one-sided Welch’s two-sample t-test. (D) Representation of the BiFC principle. (E) The ZM-only Zasp66-PH isoform tagged with a Venus fluorescent protein localizes to the Z-disc, marked by Zasp52-mCherry.

Figure 4.. The evolution of the ZM domain. (A) Protein sequence alignment between the three Zasp paralogs in flies and the ZM consensus sequence SMART:SM00735. Similar amino acids to the consensus are highlighted in gray. (B) Radial phylogenetic tree of selected eukaryotic lineages with annotated presence of the ZM domain from the PFAM domain database (PFAM: PF15936). The ZM domain is restricted to bilateral animals.

Figure 5.. Balance of ZM to LIM domains sets the diameter of the Z-disc. (A) Proposed model of Z-disc growth and aggregate formation. In nascent Z-bodies, Zasp binds α-actinin and forms a homodimer between its ZM and LIM domains. During the early growth phase, Zasp growing isoforms use their free LIM domains to recruit more Zasp molecules through the ZM domain, thus growing the Z-disc. At later stages, growth is downregulated by the incorporation of blocking isoforms of Zasp, which are recruited to the Z-disc but cannot recruit further proteins because they lack LIM domains. (B) RNAseq data plots showing the expression of selected Zasp isoforms at different developmental timepoints. The y-axis corresponds to the log of the number of transcripts per million (TPM) and the x-axis to hours after pupa formation. Isoforms were classified as blocking isoforms (0 or 1 LIM domain) or growing isoforms (two or more LIM domains). (C) Confocal microscopy image of IFM cross sections expressing Zasp52-GFP (green) and stained for actin (magenta). (D) Differential distribution of Zasp growing and blocking isoforms throughout the Z-disc. Images of false-colored denoised Z-discs. Zasp52-growing isoforms are…

Figure 5—figure supplement 1.. Rotation smoothing of cross section Z-disc images. (A–C) Steps for obtaining relative Zasp concentrations throughout single Z-discs. First, a single properly oriented Z-disc is selected from a confocal image (raw image). Then, we changed the display from gray look-up table mode to physics look-up table mode. Higher concentration of Zasp52-GFP at the center of the disc is observed at this stage (A; false colored). Higher concentration of Zasp67-GFP at the periphery of the disc is observed at this stage (B; false colored). Then, we rotated the Z-disc image by 2 degrees 180 times, and we calculated the average from all the resulting images (average image). Finally, we calculated the relative fluorescence intensities. (C) An artificially generated disc image from random noise treated in the same way as panels A and B does not show any obvious pattern.

Figure 6.. Balance of isoforms dictates Z-disc size. (A and B) Confocal microscopy images of control and Zasp52MI00979 mutant. Actin filaments are marked in magenta and Z-discs in green. The Zasp52MI00979 mutant has smaller and frayed Z-discs. (C) Frequency plot of Z-disc sizes in control, Zasp52MI02988, and Zasp52MI00979 mutants. Small Z-discs are only observed in the Zasp52MI00979 mutant. (D) Overexpression of the Zasp52-PP blocking isoform also results in smaller Z-discs. The small Z-disc phenotype is not observed in Zasp52-Stop143, lacking the ZM domain. (E) Small Z-disc phenotypes are observed upon overexpression of Zasp66 and Zasp67. (F and G) Boxplots of the Zasp52 fluorescence intensities upon overexpression of different blocking isoforms. (F) Zasp52-mCherry levels decrease upon overexpression of Zasp52-PP but are restored if the ZM domain of Zasp52-PP is deleted (Zasp52-*143). (G) Zasp52-mCherry levels also decrease upon overexpression of Zasp66 and Zasp67. Act88F-Gal4 was used for overexpression experiments in panels D-G. Scale bars, 5 µm. p-Values in panels C-E were calculated using Fisher's exact test for count data. p-Values in panels F and G were calculated using Welch’s…

Figure 6—figure supplement 1.. Actinin levels in different Zasp overexpression backgrounds. (A) Confocal images of Zasp52-mCherry in control and different Zasp overexpression backgrounds. Quantification shown in Figure 6F,G. (B) Confocal images of Actn-GFP in control and different Zasp overexpression backgrounds. Relative fluorescence levels are similar in all conditions. (C) Boxplot of the Actn-GFP fluorescence intensities upon overexpression of different blocking isoforms. Act88F-Gal4 was used for overexpression experiments. Scale bars, 5 µm. p-Values were calculated using Welch’s two-sample t-test with Bonferroni correction for multiple comparisons.