MRTF specifies a muscle-like contractile module in Porifera

Abstract

Muscle-based movement is a hallmark of animal biology, but the evolutionary origins of myocytes are unknown. Although believed to lack muscles, sponges (Porifera) are capable of coordinated whole-body contractions that purge debris from internal water canals. This behavior has been observed for decades, but their contractile tissues remain uncharacterized with respect to their ultrastructure, regulation, and development. We examine the sponge Ephydatia muelleri and find tissue-wide organization of a contractile module composed of actin, striated-muscle myosin II, and transgelin, and that contractions are regulated by the release of internal Ca²⁺ stores upstream of the myosin-light-chain-kinase (MLCK) pathway. The development of this contractile module appears to involve myocardin-related transcription factor (MRTF) as part of an environmentally inducible transcriptional complex that also functions in muscle development, plasticity, and regeneration. As an actin-regulated force-sensor, MRTF-activity offers a mechanism for how the contractile tissues that line water canals can dynamically remodel in response to flow and can re-form normally from stem-cells in the absence of the intrinsic spatial cues typical of animal embryogenesis. We conclude that the contractile module of sponge tissues shares elements of homology with contractile tissues in other animals, including muscles, indicating descent from a common, multifunctional tissue in the animal stem-lineage.

Fig. 1. Sponge body plan and contractions.

A The tent-like outer layer is supported by spicules (light blue) and is composed of two epithelia, which house a thin extracellular matrix (pink) containing migratory cells. The incurrent system (Inc) comprises the atrium and incurrent canals. Water is drawn into the atrium through incurrent pores (ostia; Ost) and into choanocyte chambers (orange), then into excurrent system (Exc) and through the osculum where it exits the sponge. B Still images from a mechanically induced contraction show that a ~40 min contraction cycle involves at least two phases. During phase I (t = 15 min), incurrent canals (red) narrow as excurrent canals (blue) widen. During phase II (t = 36 min), incurrent canals widen as excurrent canals narrow.

Fig. 2. Organization of actomyosin bundles.

A Organization of actin-bundles (yellow) in incurrent tissues, with cell-junctions marked by vinculin staining (magenta). B Duplication and divergence of the Myosin II Heavy Chain. C The type II MyHC proteins found in E. muelleri (labeled with arrows) fall into either the stMyHC clade or nmMyHC clade. Phylogeny determined by the maximum likelihood method. Tree rooted with MyHC from Dictyostelium. Support values represent 1000 bootstrap iterations. Bottom panels show the basic structure of two type II MyHCs found in E. muelleri and the alignment of epitope region used to generate stMyHC antibody. D stMyHC immunostaining in the basopinacoderm (left), a choanocyte chamber (middle), and incurrent tissues (right). Samples were stained for DNA (magenta), stMyHC (cyan; top), and actin (yellow; bottom). Yellow arrow shows cell boundary staining in basopinacoderm. Immunostaings were performed on multiple sponges over three independent experiments with consistent results. E Developmental series of actin-bundles in incurrent pinacocytes stained for DNA (magenta), stMyHC (cyan; top), and actin (yellow; bottom). Immunostaining and phalloidin stainings were performed on multiple sponges over two independent experiments with consistent results. Scale bars 25 μm in A and 10 μm in D, E.

Fig. 3. Contractions are regulated by calcium.

A–C Time-lapse images depicting sponge contractions A stimulated mechanically, B by 300 nM Ionomycin, or C by treatment with 1 nM thapsigargin. Below each image, an excurrent canal is highlighted in white to show how its diameter changed over the treatment time-course. Graphs (right) plot phase I and II contraction dynamics (n = 5 or 6 individual measurements for incurrent and excurrent canals per sample). D Simplified workflow, as sensory cells are known to require external calcium2 does raising intracellular calcium concentration downstream of them still elicit a contraction? E Top: time points of a sponge treated with 20 µM L-NAME and challenged to contract with 1:1000 Sumi Ink. Bottom: the same sponge treated with 1 nM thapsigargin. Graphs show the average canal diameter relative to the first frame for both incurrent (red) and excurrent (blue) canals (n = 6 individual measurements for incurrent and excurrent canals per sample). Analysis was also performed on sponges from independent experiments and showed consistent canal dynamics. Data are presented as mean values +/− SEM. Scale bars 500 μm.

Fig. 4. Contractions require MLCK activity.

A Fast-contracting myocytes are regulated by the Troponin C/Tropomyosin complex, whereas slow-contracting myocytes are regulated by the MLCK pathway. B Predicted domain structure of E. muelleri RLC; a 23 amino acid N-terminal extension (NTE) exists based on the location of phosphorylatable residues and EF-hand domain. C 1 nM thapsigargin had limited effect on sponges pretreated with 1 µM ML-7 to inhibit MLCK activity (n = 6 individual measurements) Data are presented as mean values +/− SEM. D Sponges were treated with Sumi Ink to block water flow, then treated with DiI to test for restored flow as an indicator of contractile activity. The ratio of sponges that stained positive for DiI was significantly lower following treatment with 50 μg/mL L-NAME and 1 µM ML-7 compared with DMSO treated controls (χ2(df2, N = 70) = 13.08, P-value = 0.001). Contraction assays were performed three independent times with consistent results. E pRLC (grayscale and magenta) immunostaining of the apical pinacoderm, counterstained with phalloidin (yellow) and Hoechst (cyan). F Scatter-plot generated by overlaying pRLC and F-actin images and measuring pixel intensity in the pRLC channel along the length of the actin-bundle between two adhesion plaques. Simple regression analysis was performed (n = 118, r² = 0.218) dark blue represents 90% confidence band and light blue represents 90% prediction band. Below are example images with raw pRLC channel and a merged image showing a brightly pRLC-stained, short actin-bundle. Immunostainings for pRLC were performed on multiple sponges over three independent experiments with consistent results. Scale bars 500 μm in C, D, 10 μm in E, and 5 μm in F.

Fig. 5. Development of contractile structures depends on MRTF-activity.

A Myocytes are developmentally specified by a transcriptional complex that includes MRTF interactions with SRF or Mef2. B Predicted domain organization of E. muelleri MRTF compared with vertebrate homologs (NLS; nuclear localization signal, B1, basic rich domain; SAP, SAP domain; LZ, leucine zipper; TAD, transcription activation domain). C Confocal images of pinacocytes (top) and archeocytes (bottom) immunostained for MRTF (grayscale and magenta), phalloidin (yellow), and Hoechst (cyan). Immunostainings for MRTF were performed on multiple sponges over five independent experiments with consistent results. D Top: confocal images of an actin-bundle immunostained for TAGLN2 (grayscale and magenta), phalloidin (yellow) and Hoechst (cyan). Bottom: qPCR of TAGLN2 in response to treatment with CCG-203971 (20 μM), ISX (50 μM), and cytochalasin D (10 μM) (n = 3 technical replicates and 3 biological replicates). Data are presented as mean values (bars) with individual experimental values overlaid (crosses). Immunostainings for TAGLN2 were performed on multiple sponges over four independent experiments with consistent results. E Sponges pretreated with CCG-203971 were unable to clear ink-blockages by contraction (χ2(df2, N = 77) = 12.64, P-value = 0.002). A redrawn image from The evolutionary origin of bilaterian smooth and striated myocytes by Brunet T, et al., modified to generalize protein families and highlight central interaction, expressing relationships based on our phylogenetic analysis of MRTF family proteins, Creative Commons—Attribution 4.0 International—CC BY 4.0. Scale bars 500 μm (E), 5 μm in C, and 2 μm in D. Contraction assays were performed three independent times with consistent results.

Fig. 6. MRTF drives contractile tissue differentiation in primmorphs.

A Primmorphs treated with 50 µM ISX or DMSO and stained with phalloidin (yellow) and Hoechst (cyan). B Western Blot showing elevated stMyHC levels in ISX-treated samples. Raw and uncropped scans in source data file. C Contraction dynamics of ISX-treated primmorphs in response to thapsigargin. Data are presented as mean values +/− SEM. (n = 3 biological replicates). D Heatmap of select transcripts that were differentially expressed in ISX-treated samples. Primmorphs treated with ISX and vehicle controls were maintained in an attachment-free environment in nine independent experiments. Phalloidin stainings were performed twice, total protein extraction was performed on one sample, and total RNA was extracted from three independent experiments with consistent results. Scale bars 10 μm.