We have previously shown that Golgi stacks and recycling endosomes (REs) exist as Golgi/RE units in sea urchin embryos. In this study, we showed that Golgi/RE units were scattered throughout the cytoplasm at early developmental stages but gathered to form a "Golgi ring" surrounding the centric REs at the blastula stage. This change in the cell-wide arrangement of Golgi/RE units coincided with a dramatic change in microtubule organization from a randomly oriented cortical pattern to radial arrays under the apical plasma membrane. A single gigantic Golgi apparatus surrounding centric RE is clearly associated with the center of the radial microtubule arrays. Furthermore, we found that in some animal species belonging to different clades, Golgi stacks lack lateral connections but are likely centralized by microtubule motors. These results suggest that Golgi centralization depends on the organization of the microtubule array in addition to the lateral linking between Golgi stacks.Key words: Golgi stack, recycling endosome, Golgi-ribbon, microtubule, cilium, sea urchin, ascidian.

The sarcomere is the contractile unit of striated muscle and is composed of actin and myosin filaments. There is increasing evidence to support that actin assembly mediated by Fhod3, a member of the formin family of proteins, is critical for sarcomere formation and maintenance in cardiac muscle. Fhod3, which is abundantly expressed in the heart, localizes to the center of sarcomeres and contributes to the regulation of the cardiac function, as evidenced by the fact that mutations in Fhod3 cause cardiomyopathy. However, the role of Fhod3 in skeletal muscle, another type of striated muscle, is unclear. We herein show that Fhod3 is expressed in the tongue at both mRNA and protein levels, although in smaller amounts than in the heart. To determine the physiological role of Fhod3 expressed in the tongue, we generated embryos lacking Fhod3 in the tongue. The tongue tissue of the Fhod3-depleted embryos did not show any significant structural defects, suggesting that Fhod3 is dispensable for normal development of the mouse tongue. Unexpectedly, the immunostaining analysis revealed the absence of specific sarcomeric signals for Fhod3 in the wild-type tongue when compared to the Fhod3-depleted tongue as a negative control, despite the use of antibodies that had previously been validated by immunostaining of heart tissues. Taken together, although Fhod3 protein is expressed at a significant level in the tongue, Fhod3 in the tongue does not appear to exhibit the same sarcomeric pattern as observed in the heart, suggesting a different role for Fhod3 in the tongue muscles.Key words: actin, formin, sarcomere, striated muscle.

Anhydrobiosis, a phenomenon in which organisms survive extreme dehydration by entering a reversible ametabolic state, is a remarkable example of survival strategies. This study focuses on anhydrobiosis in tardigrades, which are known for their resilience to severe environmental conditions. Tardigrades utilize several protective mechanisms against desiccation, notably the constitutive expression of cytoplasmic abundant heat soluble (CAHS) proteins in Ramazzottius varieornatus. These proteins share similarities in their amphiphatic alpha helices with late embryogenesis abundant (LEA) proteins, but differ significantly in their amino acid sequences. In this study, we further explored the functionality of CAHS proteins by analyzing their role in aggregation and tolerance to hyperosmotic stress in mammalian cells. Using live cell imaging, we examined the subcellular localization of several CAHS and LEA proteins in response to hyperosmotic stress. The expression of CAHS1, CAHS3, and CAHS8 tended to enhance the resilience to the hyperosmotic conditions. These findings not only deepen our understanding of the molecular mechanisms of anhydrobiosis but also highlight the potential of CAHS proteins as cryoprotectants.Key words: anhydrobiosis, tardigrades, live imaging, disordered proteins, desiccation tolerance.

Collagen is the most abundant protein in the extracellular matrix of animals, and 28 types of collagen have been reported in humans. We previously analyzed the endoplasmic reticulum (ER)-to-Golgi transport of fibril-forming type III collagen (Hirata et al., 2022) and network-forming type IV collagen (Matsui et al., 2020), both of which have long collagenous triple-helical regions. To understand the ER-to-Golgi trafficking of various types of collagens, we analyzed the transport of short-chain type X collagen in this study. We fused cysteine-free GFP to the N-telopeptide region of procollagen X (GFP-COL10A1), as employed in our previous analysis of procollagens III and IV, and analyzed its transport by live-cell imaging. Procollagen X was transported to the Golgi apparatus via vesicular and tubular carriers containing ERGIC53 and RAB1B, similar to those used for procollagen III. Carriers containing procollagen X probably used the same transport processes as those containing conventional cargoes such as α-antitrypsin. SAR1, TANGO1, SLY1/SCFD1, and BET3/TRAPPC3 were required for trafficking of procollagen X, which are different from the factors required for trafficking of procollagens III (SAR1, TANGO1, and CUL3) and IV (SAR1 and SLY1/SCFD1). These findings reveal that accommodation of various types of collagens with different shapes into carriers may require fine-tuning of the ER-to-Golgi transport machinery.Key words: collagen, GFP-procollagen X, ER-to-Golgi trafficking, export from ER, TANGO1.

The liver is a complex organ with a highly organized structure in which tight junctions (TJs) play an important role in maintaining their function by regulating barrier properties and cellular polarity. Dysfunction of TJs is associated with liver diseases, including progressive familial intrahepatic cholestasis (PFIC). In this study, we investigated the molecular alterations in a liver-specific ZO-1 and ZO-2 double-knockout (DKO) mouse model, which exhibits features resembling those of PFIC4 patients with mutations in the ZO-2 gene. RNA-seq analysis revealed the upregulation of genes involved in the oxidative stress response, xenobiotic metabolism, and cholesterol metabolism in DKO livers. Conversely, the expression of genes regulated by HNF4α was lower in DKO livers than in the wild-type controls. Furthermore, age-associated analysis elucidated the timing and progression of these pathway changes as well as alterations in molecules related to TJs and apical polarity. Our research uncovered previously unknown implications of ZO-1 and ZO-2 in liver physiology and provides new insights into the molecular pathogenesis of PFIC4 and other tight junction-related liver diseases. These findings contribute to a better understanding of the complex mechanisms underlying liver function and dysfunction and may lead to the development of novel therapeutic strategies for liver diseases associated with tight junction impairment.Key words: tight junctions, ZO-1/ZO-2 knockout mouse, liver, transcriptome analysis, molecular pathological progression.

The Golgi apparatus, a crucial organelle involved in protein processing, including glycosylation, exhibits complex sub-structures, i.e., cis-, medial, and trans-cisternae. This study investigated the distribution of glycosyltransferases within the Golgi apparatus of mammalian cells via 3D super-resolution imaging. Focusing on human glycosyltransferases involved in N-glycan modification, we found that even enzymes presumed to coexist in the same Golgi compartment exhibit nuanced variations in localization. By artificially making their N-terminal regions [composed of a cytoplasmic, transmembrane, and stem segment (CTS)] identical, it was possible to enhance the degree of their colocalization, suggesting the decisive role of this region in determining the sub-Golgi localization of enzymes. Ultimately, this study reveals the molecular codes within CTS regions as key determinants of glycosyltransferase localization, providing insights into precise control over the positioning of glycosyltransferases, and consequently, the interactions between glycosyltransferases and substrate glycoproteins as cargoes in the secretory pathway. This study advances our understanding of Golgi organization and opens avenues for programming the glycosylation of proteins for clinical applications.Key words: Golgi apparatus, glycosyltransferase, 3D super-resolution imaging, N-glycosylation.

Although quantitative analysis of biological images demands precise extraction of specific organelles or cells, it remains challenging in broad-field grayscale images, where traditional thresholding methods have been hampered due to complex image features. Nevertheless, rapidly growing artificial intelligence technology is overcoming obstacles. We previously reported the fine-tuned apodized phase-contrast microscopy system to capture high-resolution, label-free images of organelle dynamics in unstained living cells (Shimasaki, K. et al. (2024). Cell Struct. Funct., 49: 21-29). We here showed machine learning-based segmentation models for subcellular targeted objects in phase-contrast images using fluorescent markers as origins of ground truth masks. This method enables accurate segmentation of organelles in high-resolution phase-contrast images, providing a practical framework for studying cellular dynamics in unstained living cells.Key words: label-free imaging, organelle dynamics, apodized phase contrast, deep learning-based segmentation.

In metazoans, the nuclear envelope (NE) disassembles during the prophase and reassembles around segregated chromatids during the telophase. The process of NE formation has been extensively studied using live-cell imaging. At the early step of NE reassembly in human cells, specific pattern-like localization of inner nuclear membrane (INM) proteins, connected to the nuclear pore complex (NPC), was observed in the so-called "core" region and "noncore" region on telophase chromosomes, which corresponded to the "pore-free" region and the "pore-rich" region, respectively, in the early G1 interphase nucleus. We refer to these phenomena as NE subdomain formation. To biochemically investigate this process, we aimed to develop an in vitro NE reconstitution system using digitonin-permeabilized semi-intact mitotic human cells coexpressing two INM proteins, emerin and lamin B receptor, which were labeled with fluorescent proteins. The targeting and accumulation of INM proteins to chromosomes before and after anaphase onset in semi-intact cells were observed using time-lapse imaging. Our in vitro NE reconstitution system recapitulated the formation of the NE subdomain, as in living cells, although chromosome segregation and cytokinesis were not observed. This in vitro NE reconstitution required the addition of a mitotic cytosolic fraction supplemented with a cyclin-dependent kinase inhibitor and energy sources. The cytoplasmic soluble factor(s) dependency of INM protein targeting differed among the segregation states of chromosomes. Furthermore, the NE reconstituted on segregated chromosomes exhibited active nucleocytoplasmic transport competency. These results indicate that the chromosome status changes after anaphase onset for recruiting NPC components.

Cell biologists have long sought the ability to observe intracellular structures in living cells without labels. This study presents procedures to adjust a commercially available apodized phase-contrast (APC) microscopy system for better visualizing the dynamic behaviors of various subcellular organelles in living cells. By harnessing the versatility of this technique to capture sequential images, we could observe morphological changes in cellular geometry after virus infection in real time without probes or invasive staining. The tune-up APC microscopy system is a highly efficient platform for simultaneously observing the dynamic behaviors of diverse subcellular structures with exceptional resolution.

The ribosome is a molecular machine essential for protein synthesis, which is composed of approximately 80 different ribosomal proteins (Rps). Studies in yeast and cell culture systems have revealed that the intracellular level of Rps is finely regulated by negative feedback mechanisms or ubiquitin-proteasome system, which prevents over- or under-abundance of Rps in the cell. However, in vivo evidence for the homeostatic regulation of intracellular Rp levels has been poor. Here, using Drosophila genetics, we show that intracellular Rp levels are regulated by proteasomal degradation of excess Rps that are not incorporated into the ribosome. By establishing an EGFP-fused Rp gene system that can monitor endogenously expressed Rp levels, we found that endogenously expressed EGFP-RpS20 or -RpL5 is eliminated from the cell when RpS20 or RpL5 is exogenously expressed. Notably, the level of endogenously expressed Hsp83, a housekeeping gene, was not affected by exogenous expression of Hsp83, suggesting that the strict negative regulation of excess protein is specific for intracellular Rps. Further analyses revealed that the maintenance of cellular Rp levels is not regulated at the transcriptional level but by proteasomal degradation of excess free Rps as a protein quality control mechanism. Our observations provide not only the in vivo evidence for the homeostatic regulation of Rp levels but also a novel genetic strategy to study in vivo regulation of intracellular Rp levels and its role in tissue homeostasis via cell competition.Key words: ribosomal protein, proteasomal degradation, Drosophila.

Gaucher disease (GD) is a recessively inherited lysosomal storage disorder characterized by a deficiency of lysosomal glucocerebrosidase (GBA1). This deficiency results in the accumulation of its substrate, glucosylceramide (GlcCer), within lysosomes. Here, we investigated lysosomal abnormalities in fibroblasts derived from patients with GD. It is noteworthy that the cellular distribution of lysosomes and lysosomal proteolytic activity remained largely unaffected in GD fibroblasts. However, we found that lysosomal membranes of GD fibroblasts were susceptible to damage when exposed to a lysosomotropic agent. Moreover, the susceptibility of lysosomal membranes to a lysosomotropic agent could be partly restored by exogenous expression of wild-type GBA1. Here, we report that the lysosomal membrane integrity is altered in GD fibroblasts, but lysosomal distribution and proteolytic activity is not significantly altered.Key words: glucosylceramide, lysosome, Gaucher disease, lysosomotropic agent.

Calcium transients drive cells to discharge prostaglandin E (PGE). We visualized PGE-induced protein kinase A (PKA) activation and quantitated PGE secreted from a single cell by combining fluorescence microscopy and a simulation model. For this purpose, we first prepared PGE-producer cells that express either an optogenetic or a chemogenetic calcium channel stimulator: OptoSTIM1 or Gq-DREADD, respectively. Second, we prepared reporter cells expressing the Gs-coupled PGE reporter EP2 and the PKA biosensor Booster-PKA, which is based on the principle of Förster resonance energy transfer (FRET). Upon the stimulation-induced triggering of calcium transients, a single producer cell discharges PGE to stimulate PKA in the surrounding reporter cells. Due to the flow of the medium, the PKA-activated area exhibited a comet-like smear when HeLa cells were used. In contrast, radial PKA activation was observed when confluent MDCK cells were used, indicating that PGE diffusion was restricted to the basolateral space. By fitting the radius of the PKA-activated area to a simulation model based on simple diffusion, we estimated that a single HeLa cell secretes 0.25 fmol PGE upon a single calcium transient to activate PKA in more than 1000 neighboring cells. This model also predicts that the PGE discharge rate is comparable to the diffusion rate. Thus, our method quantitatively envisions that a single calcium transient affects more than 1000 neighboring cells via PGE.Key words: prostaglandin E, imaging, intercellular communication, biosensor, quantification.

Directional cell rearrangement is a critical process underlying correct tissue deformation during morphogenesis. Although the involvement of F-actin regulation in cell rearrangement has been established, the role and regulation of actin binding proteins (ABPs) in this process are not well understood. In this study, we investigated the function of Coronin-1, a WD-repeat actin-binding protein, in controlling directional cell rearrangement in the Drosophila pupal wing. Transgenic flies expressing Coronin-1-EGFP were generated using CRISPR-Cas9. We observed that Coronin-1 localizes at the reconnecting junction during cell rearrangement, which is dependent on actin interacting protein 1 (AIP1) and cofilin, actin disassemblers and known regulators of wing cell rearrangement. Loss of Coronin-1 function reduces cell rearrangement directionality and hexagonal cell fraction. These results suggest that Coronin-1 promotes directional cell rearrangement via its interaction with AIP1 and cofilin, highlighting the role of ABPs in the complex process of morphogenesis.Key words: morphogenesis, cell rearrangement, actin binding proteins (ABPs).

Secretory pathway proteins are cotranslationally translocated into the endoplasmic reticulum (ER) of metazoan cells through the protein channel, translocon. Given that there are far fewer translocons than ribosomes in a cell, it is essential that secretory protein-translating ribosomes only occupy translocons transiently. Therefore, if translocons are obstructed by ribosomes stalled or slowed in translational elongation, it possibly results in deleterious consequences to cellular function. Hence, we investigated how translocon clogging by stalled ribosomes affects mammalian cells. First, we constructed ER-destined translational arrest proteins (ER-TAP) as an artificial protein that clogged the translocon in the ER membrane. Here, we show that the translocon clogging by ER-TAP expression activates triage of signal sequences (SS) in which secretory pathway proteins harboring highly efficient SS are preferentially translocated into the ER lumen. Interestingly, the translocon obstructed status specifically activates inositol requiring enzyme 1α (IRE1α) but not protein kinase R-like ER kinase (PERK). Given that the IRE1α-XBP1 pathway mainly induces the translocon components, our discovery implies that lowered availability of translocon activates IRE1α, which induces translocon itself. This results in rebalance between protein influx into the ER and the cellular translocation capacity.Key words: endoplasmic reticulum, translocation capacity, translocon clogging, IRE1, signal sequence.

Osteoclasts play a crucial role in bone homeostasis by forming resorption pits on bone surfaces, resulting in bone resorption. The osteoclast expression of Rab38 protein is highly induced during differentiation from macrophages. Here we generated mice with double knockout (DKO) of Rab38 and its paralogue, Rab32, to investigate the roles of these proteins in osteoclasts. Bone marrow-derived macrophages from Rab32/38 DKO mice differentiated normally into osteoclasts in vitro. However, DKO osteoclasts showed reduced bone resorption activity. These osteoclasts also demonstrated defective secretion of tartrate-resistant acid phosphatase and cathepsin K into culture medium. Furthermore, the plasma membrane localization of a3, an osteoclast-specific a subunit of V-ATPase, was abrogated in DKO mice, substantiating the reduced resorption activity. In vivo, Rab32- and Rab38-positive cells were attached to the bone surface. Eight-week-old DKO mice showed significantly thickened trabecular bones in micro-CT and histomorphometry analysis, as well as reduced serum levels of cross-linked C-telopeptide of type I collagen, indicating diminished bone resorption in vivo. In DKO male mice over 10 weeks of age, hyperostosis appeared at the talofibular syndesmosis, the distal junction of the tibia and fibula. Furthermore, middle-aged mice (10 to 12 months of age) exhibited kyphosis, which is not usually observed in wild-type male mice until around 24 months of age. These results indicate that Rab32 and Rab38 contribute to osteoclast function by supporting intracellular traffic, thereby maintaining normal bone homeostasis.Key words: Rab32, Rab38, osteoclast, lysosome-related organelle, secretory lysosome.

Small extracellular vesicles (sEVs) are largely classified into two types, plasma-membrane derived sEVs and endomembrane-derived sEVs. The latter type (referred to as exosomes herein) is originated from late endosomes or multivesicular bodies (MVBs). In order to release exosomes extracellularly, MVBs must fuse with the plasma membrane, not with lysosomes. In contrast to the mechanism responsible for MVB-lysosome fusion, the mechanism underlying the MVB-plasma membrane fusion is poorly understood. Here, we systematically analyze soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) family proteins and identify VAMP5 as an MVB-localized SNARE protein required for exosome release. Depletion of VAMP5 in HeLa cells impairs exosome release. Mechanistically, VAMP5 mediates exosome release by interacting with SNAP47 and plasma membrane SNARE Syntaxin 1 (STX1) or STX4 to release exosomes. VAMP5 is also found to mediate asymmetric exosome release from polarized Madin-Darby canine kidney (MDCK) epithelial cells through interaction with the distinct sets of Q-SNAREs, suggesting that VAMP5 is a general exosome regulator in both polarized cells and non-polarized cells.Key words: exosome, small extracellular vesicle (sEV), multivesicular body, SNARE, VAMP5.

Invadopodia are protrusive structures that mediate the extracellular matrix (ECM) degradation required for tumor invasion and metastasis. Rho small GTPases regulate invadopodia formation, but the molecular mechanisms of how Rho small GTPase activities are regulated at the invadopodia remain unclear. Here we have identified FilGAP, a GTPase-activating protein (GAP) for Rac1, as a negative regulator of invadopodia formation in tumor cells. Depletion of FilGAP in breast cancer cells increased ECM degradation and conversely, overexpression of FilGAP decreased it. FilGAP depletion promoted the formation of invadopodia with ECM degradation. In addition, FilGAP depletion and Rac1 overexpression increased the emergence of invadopodia induced by epidermal growth factor, whereas FilGAP overexpression suppressed it. Overexpression of GAP-deficient FilGAP mutant enhanced invadopodia emergence as well as FilGAP depletion. The pleckstrin-homology (PH) domain of FilGAP binds phosphatidylinositol 3,4-bisphosphate [PI(3,4)P], which is distributed on membranes of the invadopodia. FilGAP localized to invadopodia in breast cancer cells on the ECM, but FilGAP mutant lacking PI(3,4)P-binding showed low localization. Similarly, the decrease of PI(3,4)P production reduced the FilGAP localization. Our results suggest that FilGAP localizes to invadopodia through its PH domain binding to PI(3,4)P and down-regulates invadopodia formation by inactivating Rac1, inhibiting ECM degradation in invasive tumor cells.Key words: invadopodia, breast carcinoma, Rac1, FilGAP, PI(3,4)P.

Ciliary outer-arm dynein (OAD) consists of heavy chains (HCs), intermediate chains (ICs), and light chains (LCs), of which HCs are the motor proteins that produce force. Studies using the green alga Chlamydomonas have revealed that ICs and LCs form a complex (IC/LC tower) at the base of the OAD tail and play a crucial role in anchoring OAD to specific sites on the microtubule. In this study, we isolated a novel slow-swimming Chlamydomonas mutant deficient in the IC2 protein. This mutation, E279K, is in the third of the seven WD repeat domains. No apparent abnormality was observed in electron microscope observations of axonemes or in SDS-PAGE analyses of dynein subunits. To explore the reason for the lowered motility in this mutant, in vitro microtubule sliding experiments were performed, which revealed that the motor activity of the mutant OAD was lowered. In particular, a large difference was observed between wild type (WT) and the mutant in the microtubule sliding velocity in microtubule bundles formed with the addition of OAD: ~35.3 μm/sec (WT) and ~4.3 μm/sec (mutant). From this and other results, we propose that IC2 in an OAD interacts with the β HC of the adjacent OAD, and that an OAD-OAD interaction is important for efficient beating of cilia and flagella.Key words: cilia, axoneme, dynein heavy chain, cooperativity.

In eukaryotic motile cells, the active Ras (Ras-GTP)-enriched domain is generated in an asymmetric manner on the cell membrane through the excitable dynamics of an intracellular signaling network. This asymmetric Ras signaling regulates pseudopod formation for both spontaneous random migration and chemoattractant-induced directional migration. While membrane lipids, such as sphingomyelin and phosphatidylserine, contribute to Ras signaling in various cell types, whether they are involved in the Ras excitability for cell motility is unknown. Here we report that functional Ras excitability requires the normal metabolism of sphingomyelin for efficient cell motility and chemotaxis. The pharmacological blockade of sphingomyelin metabolism by an acid-sphingomyelinase inhibitor, fendiline, and other inhibitors suppressed the excitable generation of the stable Ras-GTP-enriched domain. The suppressed excitability failed to invoke enough basal motility to achieve directed migration under shallow chemoattractant gradients. The fendiline-induced defects in Ras excitability, motility and stimulation-elicited directionality were due to an accumulation of sphingomyelin on the membrane, which could be recovered by exogenous sphingomyelinase or phosphatidylserine without changing the expression of Ras. These results indicate a novel regulatory mechanism of the excitable system by membrane lipids, in which sphingomyelin metabolism provides a membrane environment to ensure Ras excitation for efficient cellular motility and chemotaxis.Key words: cell polarity, cell migration, Ras, excitability, sphingomyelin.

We cloned and characterized two new coral fluorescent proteins: h2-3 and 1-41. h2-3 formed an obligate dimeric complex and exhibited bright green fluorescence. On the other hand, 1-41 formed a highly multimeric complex and exhibited dim red fluorescence. We engineered 1-41 into AzaleaB5, a practically useful red-emitting fluorescent protein for cellular labeling applications. We fused h2-3 and AzaleaB5 to the ubiquitination domains of human Geminin and Cdt1, respectively, to generate a new color variant of Fucci (Fluorescent Ubiquitination-based Cell-Cycle Indicator): Fucci5. We found Fucci5 provided more reliable nuclear labeling for monitoring cell-cycle progression than the 1 and 2 generations that used mAG/mKO2 and mVenus/mCherry, respectively.Key words: fluorescent protein, cell cycle, time-lapse imaging, flow cytometry.

We have recently demonstrated that a LIM domain protein, cysteine and glycine-rich protein 2 (CSRP2 [CRP2]), plays a vital role in the functional expression of myofibroblasts and cancer-associated fibroblasts. CRP2 binds directly to myocardin-related transcription factors (MRTF [MRTF-A or MRTF-B]) and serum response factor (SRF) to stabilize the MRTF/SRF/CArG-box complex, leading to the expression of smooth muscle cell (SMC) genes such as α-smooth muscle actin (α-SMA) and collagens. These are the marker genes for myofibroblasts. Here, we show that the adhesion of cultured human skin fibroblasts (HSFs) to collagen reduces the myofibroblastic features. HSF adhesion to collagen suppresses the expression of CRP2 and CSRP2-binding protein (CSRP2BP [CRP2BP]) and reduces the expression of SMC genes. Although CRP2BP is known as an epigenetic factor, we find that CRP2BP also acts as an adaptor protein to enhance the function of CRP2 mentioned above. This CRP2BP function does not depend on its histone acetyltransferase activity. We also addressed the molecular mechanism of the reduced myofibroblastic features of HSFs on collagen. HSF adhesion to collagen inhibits the p38MAPK-mediated pathway, and reducing the p38MAPK activity decreases the expression of CRP2 and SMC genes. Thus, the activation of p38MAPK is critical for the myofibroblastic features. We also show evidence that CRP2 plays a role in the myofibroblastic transition of retinal pigment epithelial cells (RPEs). Like HSFs, such phenotypic modulation of RPEs depends on the p38MAPK pathway.Key words: CRP2, p38MAPK, MRTF, myofibroblasts, retinal pigment epithelial cells.