Myo5b Transports Fibronectin-Containing Vesicles and Facilitates FN1 Secretion from Human Pleural Mesothelial Cells

Pleural mesothelial cells (PMCs) play a central role in the progression of pleural fibrosis. As pleural injury progresses to fibrosis, PMCs transition to mesenchymal myofibroblast via mesothelial mesenchymal transition (MesoMT), and produce extracellular matrix (ECM) proteins including collagen and fibronectin (FN1). FN1 plays an important role in ECM maturation and facilitates ECM-myofibroblast interaction, thus facilitating fibrosis. However, the mechanism of FN1 secretion is poorly understood. We report here that myosin 5b (Myo5b) plays a critical role in the transportation and secretion of FN1 from human pleural mesothelial cells (HPMCs). TGF-β significantly increased the expression and secretion of FN1 from HPMCs and facilitates the close association of Myo5B with FN1 and Rab11b. Moreover, Myo5b directly binds to GTP bound Rab11b (Rab11b-GTP) but not GDP bound Rab11b. Myo5b or Rab11b knockdown via siRNA significantly attenuated the secretion of FN1 without changing FN1 expression. TGF-β also induced Rab11b-GTP formation, and Rab11b-GTP but not Rab11b-GDP significantly activated the actin-activated ATPase activity of Myo5B. Live cell imaging revealed that Myo5b- and FN1-containing vesicles continuously moved together in a single direction. These results support that Myo5b and Rab11b play an important role in FN1 transportation and secretion from HPMCs, and consequently may contribute to the development of pleural fibrosis.


Introduction
Pleural mesothelial cells (PMCs) form a cellular monolayer that covers the chest wall and the lung forming the parietal and visceral pleural, respectively. They have multiple intercellular adherens junctions and focal adhesions that anchor the cells to the extracellular membrane. However, PMCs change their phenotype from an epithelioid phenotype to more profibrotic phenotype in response to cytokines released during pleural inflammation during the development of fibrosis [1]. During transition, mesothelial cells begin expressing α-smooth muscle actin (α-SMA) and demonstrate cytoskeletal reorganization and phenotypic changes to become myofibroblast. PMC derived myofibroblasts also demonstrate the expression and secretion of extracellular matrix (ECM) proteins such as collagen 1 (Col-1), which is directly involved in pleural fibrosis. As described above, another key ECM protein that is supposed to be involved in the development of pleural fibrosis is fibronectin, yet the role of PMCs in the production and deposition of fibronectin in the progression of pleural fibrosis is poorly understood.
Fibrotic tissue is characterized by the excessive accumulation of cells and extracellular matrix proteins. Col-1 plays a central role in the development of fibrosis. Col-1 fibrils show disorganized structure due to cross-linking in fibrosis with increased collagen crosslinking enzymes such as lysyl oxidase (LOX) [2,3]. Another important ECM protein during the development of fibrosis is fibronectin (FN1). FN1 mediates the interaction between We and others have shown that profibrotic stimuli such as TGF-β induce distinct phenotypic changes of human pleural mesothelial cells (HPMCs). Because of the similarity to epithelial-mesenchymal transition (EMT) this process has been termed mesothelial mesenchymal transition (MesoMT) [21][22][23][24][25]. During MesoMT, there is a significant increase in α-smooth muscle actin (αSMA) expression and the expression of extracellular matrix (ECM) proteins, such as collagen-1 (Col-1), by HPMCs [26,27]. Similar findings were observed in vivo. However, FN1 expression in HPMCs and in the pleura during the development of pleural fibrosis has not been characterized. To address this gap in knowledge, we first examined the deposition of FN1 in the pleura of mice with carbon black bleomycin (CBB) induced pleural fibrosis. Immunofluorescent analysis of αSMA, a marker of MesoMT, showed a marked increase in αSMA expression in the thickened pleural rind of CBB injured mice compared to saline-treated control mice ( Figure 1A). FN1 and (Col-1) expression were likewise enhanced in the visceral pleura of CBB injured mice ( Figure 1B). FN1 deposition was also observed in the thickened pleura of mice having S. pneumoniae-induced empyema ( Figure 1C). Further, FN1 deposition was colocalized with the (Col-1) in the thickened pleural rind of S. pneumoniae injured mice. matrix (ECM) proteins, such as collagen-1 (Col-1), by HPMCs [26,27]. Similar findings were observed in vivo. However, FN1 expression in HPMCs and in the pleura during the development of pleural fibrosis has not been characterized. To address this gap in knowledge, we first examined the deposition of FN1 in the pleura of mice with carbon black bleomycin (CBB) induced pleural fibrosis. Immunofluorescent analysis of αSMA, a marker of MesoMT, showed a marked increase in αSMA expression in the thickened pleural rind of CBB injured mice compared to saline-treated control mice ( Figure 1A). FN1 and (Col-1) expression were likewise enhanced in the visceral pleura of CBB injured mice (Figure 1B). FN1 deposition was also observed in the thickened pleura of mice having S. pneumoniae-induced empyema ( Figure 1C). Further, FN1 deposition was colocalized with the (Col-1) in the thickened pleural rind of S. pneumoniae injured mice. were stained for α-smooth muscle actin (α-SMA; green) and nuclei (red). α-SMA were localized at the pleura in CBB-treated mice. Representative image of n = 3 tissues. (B) Lung tissues sections from saline and CBB-treated mice were stained for FN1 (red), Collagen 1 (Col-1; cyan) and nuclei (blue). FN1 and Col-1 were colocalized at the pleura in CBB-treated mice. Representative image of n = 5 tissues. (C) Lung tissue sections from saline and empyema -treated mice were stained for FN1 (red), Col-1 (cyan) and nuclei (blue). FN1 and Col-1 were colocalized at the pleura in empyema lung tissues. Representative image of n = 5 tissues. White arrows indicate pleura. Scale bars; 100 μm. Carbon black bleomycin (CBB) and empyema. (A) Lung tissues sections from saline and CBB-treated mice were stained for α-smooth muscle actin (α-SMA; green) and nuclei (red). α-SMA were localized at the pleura in CBB-treated mice. Representative image of n = 3 tissues. (B) Lung tissues sections from saline and CBB-treated mice were stained for FN1 (red), Collagen 1 (Col-1; cyan) and nuclei (blue). FN1 and Col-1 were colocalized at the pleura in CBB-treated mice. Representative image of n = 5 tissues. (C) Lung tissue sections from saline and empyema -treated mice were stained for FN1 (red), Col-1 (cyan) and nuclei (blue). FN1 and Col-1 were colocalized at the pleura in empyema lung tissues. Representative image of n = 5 tissues. White arrows indicate pleura. Scale bars; 100 µm.

TGF-β Increases Expression and Secretion of FN1 by HPMCs
We and others have shown that MesoMT (mesothelial mesenchymal transition) of pleural mesothelial cells contributes to the enhanced deposition of collagen in injured pleural tissues [21,[28][29][30]. Therefore, we next asked whether MesoMT contributes to FN1 deposition in pleural tissues. To address this question, we treated primary human pleural mesothelial cells (HPMCs) with TGF-β to induce MesoMT. RNA was then isolated from these cells and analyzed by qPCR. As shown in Figure 2A, TGF-β significantly (p < 0.001) up-regulated FN1 mRNA expression, as well as αSMA, in HPMCs. TGF-β treated lysates and conditioned media were next examined for the expression and secretion of FN1. The qPCR results were confirmed by Western blot analysis that showed significantly increased FN1 protein expression with TGF-β stimulation ( Figure 2B,C). We also found that FN1 was significantly increased in the cell culture supernatant of TGF-β treated HPMCs compared to controls ( Figure 2B,C). It should be noted that the increase in the expression and secretion of FN1(EDA) isoform that has been shown to be related to myofibroblast activation during fibrosis [7] is more prominent (Figure 2B,C). These results suggest that TGF-β stimulation not only increases mRNA and protein expression of FN1, but also enhances FN1 secretion from HPMCs. Further, these data suggest that HPMCs directly contribute to the deposition of FN1 in injured pleural tissues.

TGF-β Increases Expression and Secretion of FN1 by HPMCs
We and others have shown that MesoMT (mesothelial mesenchymal transition) of pleural mesothelial cells contributes to the enhanced deposition of collagen in injured pleural tissues [21,[28][29][30]. Therefore, we next asked whether MesoMT contributes to FN1 deposition in pleural tissues. To address this question, we treated primary human pleural mesothelial cells (HPMCs) with TGF-β to induce MesoMT. RNA was then isolated from these cells and analyzed by qPCR. As shown in Figure 2A, TGF-β significantly (p < 0.001) up-regulated FN1 mRNA expression, as well as αSMA, in HPMCs. TGF-β treated lysates and conditioned media were next examined for the expression and secretion of FN1. The qPCR results were confirmed by Western blot analysis that showed significantly increased FN1 protein expression with TGF-β stimulation ( Figure 2B,C). We also found that FN1 was significantly increased in the cell culture supernatant of TGF-β treated HPMCs compared to controls ( Figure 2B,C). It should be noted that the increase in the expression and secretion of FN1(EDA) isoform that has been shown to be related to myofibroblast activation during fibrosis [7] is more prominent (Figure 2B,C). These results suggest that TGFβ stimulation not only increases mRNA and protein expression of FN1, but also enhances FN1 secretion from HPMCs. Further, these data suggest that HPMCs directly contribute to the deposition of FN1 in injured pleural tissues.  Figure S1. (C) Statistical representation of Western blot analysis. n = 3, *** p < 0.001 vs. PBS. FN1 protein expression levels were analyzed by Western blotting in both cell lysate and culture medium. Culture media were changed at 48 h after TGF-β stimulation and the culture media for Western blotting were collected after 24 h. Cell lysate were prepared after the collection of the culture medium. The red spots in bar graphs display the distribution of data points.

Myo5b Co-Localizes with FN1 in HPMCs
Because TGF-β increased the expression and secretion of FN1 from HPMCs, we hypothesized that the directional transportation system of FN1-containing vesicles in HPMC is likewise activated by TGF-β stimulation. Figure 3A shows co-localization of FN1 with actin structures in HPMCs. Co-localization analysis using confocal imaging of FN1 with F-actin revealed significant co-localization of FN1 along the line of filamentous structure of F-actin ( Figure 3A). The result suggests that FN1-containing vesicles are transported along with actin filaments presumably by means of actin-based motor protein, myosin. Myosin constitutes a large superfamily, and certain classes of myosin family members are suitable motors for cargo transportation because they can associate with the cargo molecules for a prolonged time, and these myosin motors are often called "processive myosin". Class V myosin is the best known processive myosin that can transport the associated cargo molecules for long distances. Among the class V myosin, myosin Va and myosin Vb PBS. FN1 protein expression levels were analyzed by Western blotting in both cell lysate and culture medium. Culture media were changed at 48 h after TGF-β stimulation and the culture media for Western blotting were collected after 24 h. Cell lysate were prepared after the collection of the culture medium. The red spots in bar graphs display the distribution of data points.

Myo5b Co-Localizes with FN1 in HPMCs
Because TGF-β increased the expression and secretion of FN1 from HPMCs, we hypothesized that the directional transportation system of FN1-containing vesicles in HPMC is likewise activated by TGF-β stimulation. Figure 3A shows co-localization of FN1 with actin structures in HPMCs. Co-localization analysis using confocal imaging of FN1 with F-actin revealed significant co-localization of FN1 along the line of filamentous structure of F-actin ( Figure 3A). The result suggests that FN1-containing vesicles are transported along with actin filaments presumably by means of actin-based motor protein, myosin. Myosin constitutes a large superfamily, and certain classes of myosin family members are suitable motors for cargo transportation because they can associate with the cargo molecules for a prolonged time, and these myosin motors are often called "processive myosin". Class V myosin is the best known processive myosin that can transport the associated cargo molecules for long distances. Among the class V myosin, myosin Va and myosin Vb are processive myosin [19,31,32]. As shown in Figure 3A, a notable portion of endogenous Myo5b co-localized with FN1-containing vesicles. Further, the colocalized signals were found on actin filaments. Conversely, Myo5a did not show appreciable colocalization with FN1 ( Figure 3C). are processive myosin [19,31,32]. As shown in Figure 3A, a notable portion of endogenous Myo5b co-localized with FN1-containing vesicles. Further, the colocalized signals were found on actin filaments. Conversely, Myo5a did not show appreciable colocalization with FN1 ( Figure 3C).  To further study the close association of Myo5b and FN1 induced by TGF-β stimulation, we next performed a proximity ligation assay. This assay can detect close association between the two proteins within 50 nm distance [33,34]. As shown in Figure 4A,B, the proximity ligation signal between Myo5b and FN1 was significantly (p < 0.001) increased by TGF-β stimulation compared to PBS controls. Moreover, the signals were found on actin filaments, indicated by arrow heads in Figure 4A (right). These results suggest that TGF-β stimulation induces the association of Myo5b with the FN1-containing vesicles. To further study the close association of Myo5b and FN1 induced by TGF-β stimulation, we next performed a proximity ligation assay. This assay can detect close association between the two proteins within 50 nm distance [33,34]. As shown in Figure 4A,B, the proximity ligation signal between Myo5b and FN1 was significantly (p < 0.001) increased by TGF-β stimulation compared to PBS controls. Moreover, the signals were found on actin filaments, indicated by arrow heads in Figure 4A (right). These results suggest that TGF-β stimulation induces the association of Myo5b with the FN1-containing vesicles. Quantitative analysis of intensity of the signals (mean ± SD), *** p < 0.001. Proximity-ligation assay revealed close association of Myo5b and FN1 that is induced by TGF-β stimulation. The red spots in bar graph displays the distribution of data points.

Myo5b Gene Silencing Attenuates Secretion of FN1 from HPMCs
Since Myo5b associates with FN1-containing vesicles upon TGF-β stimulation, it is plausible that Myo5b is involved in FN1 transportation and secretion. We next determined the role of Myo5b in the transportation/secretion of FN1 in HPMCs. We downregulated Myo5b expression with targeting siRNA and examined the effect of gene silencing of Myo5b on FN1 secretion. Quantitative PCR analysis revealed that Myo5b siRNA significantly reduced the Myo5b mRNA level both in the presence and absence of TGF-β (KD efficiency; 95.5% before TGF-β, 96.5% after TGF-β), suggesting the Myo5b siRNA effectively diminished transcription of myosin 5B ( Figure 5A). Myo5b protein expression was likewise significantly diminished by the siRNA ( Figure 5B). We next examined the (B) Quantitative analysis of intensity of the signals (mean ± SD), *** p < 0.001. Proximity-ligation assay revealed close association of Myo5b and FN1 that is induced by TGF-β stimulation. The red spots in bar graph displays the distribution of data points.

Myo5b Gene Silencing Attenuates Secretion of FN1 from HPMCs
Since Myo5b associates with FN1-containing vesicles upon TGF-β stimulation, it is plausible that Myo5b is involved in FN1 transportation and secretion. We next determined the role of Myo5b in the transportation/secretion of FN1 in HPMCs. We down-regulated Myo5b expression with targeting siRNA and examined the effect of gene silencing of Myo5b on FN1 secretion. Quantitative PCR analysis revealed that Myo5b siRNA significantly reduced the Myo5b mRNA level both in the presence and absence of TGF-β (KD efficiency; 95.5% before TGF-β, 96.5% after TGF-β), suggesting the Myo5b siRNA effectively diminished transcription of myosin 5B ( Figure 5A). Myo5b protein expression was likewise significantly diminished by the siRNA ( Figure 5B). We next examined the effect of Myo5b downregulation on FN1 secretion. FN1 in culture supernatant was significantly reduced by the Myo5b knockdown ( Figure 5C,D). Since siRNA for Myo5a and Myo5b did not affect the mRNA level of FN1 ( Figure 5A), the result suggests that gene silencing of Myo5b diminished the secretion process of FN1.
On the other hand, Myo5a KD did not show the significant changes in FN in culture supernatant ( Figure 5C,D), while it effectively diminished the mRNA ( Figure 5A) and protein expression of Myo5b ( Figure 5B). These results suggest that Myo5b, but not Myo5a, is important for the secretion of FN1-containing vesicles, presumably transportation of the vesicles containing FN1 to facilitate secretion.
To address this notion, we first examined the effect of gene silencing of Rab11 on FN1 secretion. Rab11b siRNA significantly reduced Rab11b mRNA in the presence and absence of TGF-β ( Figure 6A, KD efficiency; 94.3 % before, 93.6% after TGF-β), whereas FN1 mRNA were unchanged by the Rab11b siRNA ( Figure 6A). Likewise, Rab11a siRNA significantly reduced Rab11a mRNA before and after TGF-β stimulation (KD efficiency; 94.4 % before, 92.7% after TGF-β) without changing FN1 mRNA level ( Figure 6A). Western blot analysis of total cell lysates revealed that Rab11b specific siRNA effectively diminished the protein expression of Rab11b. Likewise, Rab11a specific siRNA effectively diminished the protein expression of Rab11a ( Figure 6B). It should be noted that FN1 expression was not affected by Rab11b siRNA or Rab11a siRNA.
Using these siRNAs, we next examined the effect of Rab11b gene silencing on FN1 secretion. While FN1 protein in cell lysates were not significantly altered by the Rab11b gene silencing, FN1 in culture supernatant was significantly reduced by the Rab11b siRNA ( Figure 6C,D). Conversely, Rab11a knock-down did not result in the significant decrease in the FN1 in the culture supernatants ( Figure 6C,D). These results suggest that Rab11b, but not Rab11a, plays a role in the secretion of FN1-containing vesicles, presumably by bridging Myo5b and FN1-containing vesicles.
To further study the effect of TGF-β on the close association of Myo5b and Rab11b, we again performed a proximity ligation assay that can detect close association between the two proteins within 50 nm distance. As shown in Figure 8A, the proximity ligation signal between Myo5b and Rab11b was significantly increased by TGF-β stimulation compared to PBS controls. These results suggest that TGF-β stimulation induces the association of Myo5b with Rab11b-containing vesicles. The proximity ligation signal between FN1 and Rab11b was also significantly increased after TGF-β stimulation ( Figure 8B). Our results suggest that Myo5b resides on FN1-containing vesicles through the binding to Rab11b after TGF-β treatment. This implies that Rab11b is required for Myo5b to transport FN1-containing vesicles. effect of Myo5b downregulation on FN1 secretion. FN1 in culture supernatant was significantly reduced by the Myo5b knockdown ( Figure 5C,D). Since siRNA for Myo5a and Myo5b did not affect the mRNA level of FN1 ( Figure 5A), the result suggests that gene silencing of Myo5b diminished the secretion process of FN1.  % before, 92.7% after TGF-β) without changing FN1 mRNA level (F blot analysis of total cell lysates revealed that Rab11b specific siRNA ished the protein expression of Rab11b. Likewise, Rab11a specific si minished the protein expression of Rab11a ( Figure 6B). It should be pression was not affected by Rab11b siRNA or Rab11a siRNA.

TGF-β Induces Association of Myo5b with Rab11b and FN1-Containing Vesicles in HPMCs
We next determined whether TGF-β facilitates the association of Myo5b to FN1-containing vesicles via Rab11b for transportation of the vesicles. We first examined whether FN1-containing vesicles colocalize with Rab11b and Myo5b. HPMCs were co-infected with vectors expressing GFP-Myo5b, mCherry-FN1, and V5-Rab11b (Figure 7). Infected cells were examined for the localization of these proteins in HPMCs after TGF-β stimulation. V5-Rab11b showed notable colocalization with GFP-Myo5b and mCherry-FN1 (Figure 7). To further study the effect of TGF-β on the close association of Myo5b and Rab11b, we again performed a proximity ligation assay that can detect close association between the two proteins within 50 nm distance. As shown in Figure 8A, the proximity ligation

Activation of Rab11b Induces Association with Myo5b and Facilitates FN1 Secretion
We next asked if the activation of Rab11b facilitates the association of Myo5b with FN1-containing vesicles. Further, if TGF-β induces the activation of Rab11b, thus promoting the association of Myo5b with FN1-containing vesicles for transportation and secretion. Figure 9 shows the colocalization between GFP-Myo5b and V5-Rab11b wild type, V5-Rab11b S25V (dominant negative mutant), and V5-Rab11b S20V (active mutant) in the presence of TGF-β, respectively. The active mutant of Rab11b (S20V) showed punctate localization that was highly colocalized with GFP-Myo5b ( Figure 9). On the other hand, staining of the inactive mutant of Rab11b (S25V) was rather diffuse. Myo5b also showed diffuse localization in the cells expressing Rab11b (S25V). These results suggest that Myo5b associate with vesicles via interaction with the active form of Rab11b. V5-Rab11b wild type also showed notable colocalization with Myo5b (Figure 9 Top panel). This suggests that WT Rab11b is partially activated in the cells after TGF-β stimulation. signal between Myo5b and Rab11b was significantly increased by TGF-β stimulation compared to PBS controls. These results suggest that TGF-β stimulation induces the association of Myo5b with Rab11b-containing vesicles. The proximity ligation signal between FN1 and Rab11b was also significantly increased after TGF-β stimulation ( Figure 8B). Our results suggest that Myo5b resides on FN1-containing vesicles through the binding to Rab11b after TGF-β treatment. This implies that Rab11b is required for Myo5b to transport FN1-containing vesicles.

Activation of Rab11b Induces Association with Myo5b and Facilitates FN1 Secretion
We next asked if the activation of Rab11b facilitates the association of Myo5b with FN1-containing vesicles. Further, if TGF-β induces the activation of Rab11b, thus tate localization that was highly colocalized with GFP-Myo5b (Figure 9). On the other hand, staining of the inactive mutant of Rab11b (S25V) was rather diffuse. Myo5b also showed diffuse localization in the cells expressing Rab11b (S25V). These results suggest that Myo5b associate with vesicles via interaction with the active form of Rab11b. V5-Rab11b wild type also showed notable colocalization with Myo5b (Figure 9 Top panel). This suggests that WT Rab11b is partially activated in the cells after TGF-β stimulation. To further confirm the effect of Rab11b activation on the binding to Myo5b, we performed a Proximity Ligation Assay. V5-Rab11b variants were expressed in HPMCs along with GFP-Myo5b and subjected to Proximity Ligation assay. HPMCs expressing similar levels of the three V5-Rab11b variants were subjected to the analysis ( Figure 10A). Figure  10B shows the proximity ligation signal between V5-Rab11b variants and GFP-Myo5b in the presence of TGF-β. The signal intensity was significantly lower for the dominant negative mutant (S25V) compared to the active mutant (S20V). On the other hand, signal intensity of the wild type (WT) was significantly higher than the negative mutant (S25N) but lower than the active mutant (V5-Rab11b S20V) ( Figure 10B,C). The result suggests that Rab11b wild type is partially activated in TGF-β treated cells, thus associating with Myo5b. The results also suggest that Rab11b associates with FN1-containing vesicles when it forms an active conformation. To test this hypothesis, the role of Rab11b activation on Myo5b binding was examined ( Figure 11). To further confirm the effect of Rab11b activation on the binding to Myo5b, we performed a Proximity Ligation Assay. V5-Rab11b variants were expressed in HPMCs along with GFP-Myo5b and subjected to Proximity Ligation assay. HPMCs expressing similar levels of the three V5-Rab11b variants were subjected to the analysis ( Figure 10A). Figure 10B shows the proximity ligation signal between V5-Rab11b variants and GFP-Myo5b in the presence of TGF-β. The signal intensity was significantly lower for the dominant negative mutant (S25V) compared to the active mutant (S20V). On the other hand, signal intensity of the wild type (WT) was significantly higher than the negative mutant (S25N) but lower than the active mutant (V5-Rab11b S20V) ( Figure 10B,C). The result suggests that Rab11b wild type is partially activated in TGF-β treated cells, thus associating with Myo5b. The results also suggest that Rab11b associates with FN1-containing vesicles when it forms an active conformation. To test this hypothesis, the role of Rab11b activation on Myo5b binding was examined ( Figure 11).   We purified Rab11b and the tail domain of Myo5b, a cargo binding domain (see Materials and Methods) and performed the direct binding assay. The isolated Rab11b was incubated with either GTPγS or GDPβS to produce the active and inactive form, respectively. The mixtures were then subjected to GST-pull-down assay with GST-Myo5b tail. As shown in Figure 11A, GTPγS markedly facilitated the binding of Rab11b to Myo5b tail, indicating that the activation of Rab11b is critical for the binding of Rab11b to Myo5b. The result was consistent with the imaging analysis shown in Figures 9 and 10, and supports our hypothesis that Myo5b binds to Rab11b, when activated. Next, we determined whether TGF-β activates Rab11b in HPMCs. To monitor the activation of Rab11b, we used the Rab11 binding domain of Rab11FIP3 (FIP3-RBD) as a probe [43,44]. Figure 11B shows We purified Rab11b and the tail domain of Myo5b, a cargo binding domain (see Materials and Methods) and performed the direct binding assay. The isolated Rab11b was incubated with either GTPγS or GDPβS to produce the active and inactive form, respectively. The mixtures were then subjected to GST-pull-down assay with GST-Myo5b tail. As shown in Figure 11A, GTPγS markedly facilitated the binding of Rab11b to Myo5b tail, indicating that the activation of Rab11b is critical for the binding of Rab11b to Myo5b. The result was consistent with the imaging analysis shown in Figures 9 and 10, and supports our hypothesis that Myo5b binds to Rab11b, when activated. Next, we determined whether TGF-β activates Rab11b in HPMCs. To monitor the activation of Rab11b, we used the Rab11 binding domain of Rab11FIP3 (FIP3-RBD) as a probe [43,44]. Figure 11B shows the effect of GTPγS and GDPβS on the binding of Rab11b to FIP-RBD. GTPγS markedly enhanced the binding of Rab11b to GST-FIP3-RBD. Using FIP3-RBD as a probe, we next examined the effect of TGF-β stimulation on the activation of Rab11b in HPMCs. Flag-Rab11b was added to the lysates of HPMCs treated with or without TGF-β, and subjected to GST-pull-down assay ( Figure 12C). TGF-β stimulated HPMCs showed a notable increase in the binding with GST-FIP3-RBD compared to PBS treated controls. The result suggests that TGF-β activates Rab11b in HPMCs to form Rab11b-GTP, thus facilitating the binding to Myo5b. the effect of GTPγS and GDPβS on the binding of Rab11b to FIP-RBD. GTPγS markedly enhanced the binding of Rab11b to GST-FIP3-RBD. Using FIP3-RBD as a probe, we next examined the effect of TGF-β stimulation on the activation of Rab11b in HPMCs. Flag-Rab11b was added to the lysates of HPMCs treated with or without TGF-β, and subjected to GST-pull-down assay ( Figure 12C). TGF-β stimulated HPMCs showed a notable increase in the binding with GST-FIP3-RBD compared to PBS treated controls. The result suggests that TGF-β activates Rab11b in HPMCs to form Rab11b-GTP, thus facilitating the binding to Myo5b. Figure 12. Effect of Rab11b on actin activated ATPase activity of Myo5b. Rab11b and GDPβS or GTPγS were pre-incubated at 37 °C for 30 min, and the ATPase activity of Myo5b with 20 μM Factin in 1 mM EGTA at 37 °C was measured in the presence or absence of 15 μM Rab11b as described in "Materials and Methods" (n = 3, mean ± SD). *** p < 0.001. The red spots in bar graph displays the distribution of data points.

Rab11 Activates the Motor Function of Myo5b
Our results suggest that Myo5b binds to the active form of Rab11b, i.e., Rab11b-GTP, and facilitates the binding to FN1-containing vesicles that reside with Rab11b.
We next asked whether the active Rab11b binding to Myo5b promotes the motor function of Myo5b. To address this notion, we examined the effect of Rab11b on the motor activity of Myo5b. Full-length Myo5b was isolated and the effect of Rab11b on the motor function was determined by measuring the actin activated ATPase activity of Myo5b. Rab11b with GTPγS highly activated the actin-activated ATPase activity, while Rab11b-GDPβS maginally activated the ATPase activity ( Figure 12). Since the actin-activated ATPase activity is coupled with the cross-bridge cycling activity, i.e., motor cycling [45], the result indicated that the active form of Rab11b activates the motor function of Myo5b in addition to facilitating the association of Myo5b to the FN1-containing vesicles.

Direct Visualization of Movement of FN by Myo5b in Living Cells
To obtain conclusive evidence that Myo5b transports FN1-containing vesicles, we performed live cell imaging of the movement of FN1-containing vesicles with Myo5b. HPMCs were co-transduced with mCherry-FN1(EDA) expressing viral vector and GFP-Myo5b expressing viral vector. GFP-Myo5b showed punctate localization in HPMCs with TGF-β stimulation ( Figure 13). To analyze Myo5b dependent movement of FN1-containing vesicles, we monitored time projection images. The images clearly demonstrated the directional movement of FN1-containing vesicles in live HPMCs stimulated by TGF-β ( Figure 13A). The trajectory images of mCherry-FN1 and GFP-Myo5b, clearly showed the co-movement of mCherry-FN1 and GFP-Myo5b ( Figure 13B,C and Supplementary Video

Rab11 Activates the Motor Function of Myo5b
Our results suggest that Myo5b binds to the active form of Rab11b, i.e., Rab11b-GTP, and facilitates the binding to FN1-containing vesicles that reside with Rab11b.
We next asked whether the active Rab11b binding to Myo5b promotes the motor function of Myo5b. To address this notion, we examined the effect of Rab11b on the motor activity of Myo5b. Full-length Myo5b was isolated and the effect of Rab11b on the motor function was determined by measuring the actin activated ATPase activity of Myo5b. Rab11b with GTPγS highly activated the actin-activated ATPase activity, while Rab11b-GDPβS maginally activated the ATPase activity ( Figure 12). Since the actin-activated ATPase activity is coupled with the cross-bridge cycling activity, i.e., motor cycling [45], the result indicated that the active form of Rab11b activates the motor function of Myo5b in addition to facilitating the association of Myo5b to the FN1-containing vesicles.

Direct Visualization of Movement of FN by Myo5b in Living Cells
To obtain conclusive evidence that Myo5b transports FN1-containing vesicles, we performed live cell imaging of the movement of FN1-containing vesicles with Myo5b. HPMCs were co-transduced with mCherry-FN1(EDA) expressing viral vector and GFP-Myo5b expressing viral vector. GFP-Myo5b showed punctate localization in HPMCs with TGF-β stimulation ( Figure 13). To analyze Myo5b dependent movement of FN1-containing vesicles, we monitored time projection images. The images clearly demonstrated the directional movement of FN1-containing vesicles in live HPMCs stimulated by TGF-β ( Figure 13A). The trajectory images of mCherry-FN1 and GFP-Myo5b, clearly showed the co-movement of mCherry-FN1 and GFP-Myo5b ( Figure 13B,C and Supplementary Video S1). We next performed Kymograph analysis of the movement that represents the time course of the directional movement of GFP-Myo5b and mCherry-FN1(EDA). The results show that mCherry-FN1(EDA) and GFP-Myo5b continuously move together without dissociation from the track ( Figure 13D). The Myo5b/FN1 complex moved with a velocity of 0.26 µm/sec, stopped, and in the same direction towards cell peripheries ( Figure 13). S1). We next performed Kymograph analysis of the movement that represents the time course of the directional movement of GFP-Myo5b and mCherry-FN1(EDA). The results show that mCherry-FN1(EDA) and GFP-Myo5b continuously move together without dissociation from the track ( Figure 13D). The Myo5b/FN1 complex moved with a velocity of 0.26 μm/sec, stopped, and in the same direction towards cell peripheries ( Figure 13).

Discussion
It has been known that FN1, especially the FN1 EDA variant is highly up-regulated during differentiation of myofibroblasts, and plays an important role in myofibroblast activation and differentiation [11,13]. Moreover, FN1 plays a critical role in ECM assembly [46,47], and collagen network organization. FN1 plays an essential role in the interaction between ECM and myofibroblast, which can influence TGF-β signaling and phenotypic changes of myofibroblasts, yet the mechanism of FN1 secretion from myofibroblasts is poorly understood. Further, as a component of the neomatrix, FN1 abundance may impact resolution of pleural loculations. As such the identification of processes that may regulate the expression and secretion of neomatrix components could affect the effectiveness of therapeutics which promote dissolution of pleural loculations [48]. In the present study, we attempted to clarify the mechanism of the transportation and secretion of FN1 from myofibroblasts that is differentiated from mesothelial cells through mesothelial-mesenchymal transition (MesoMT) [24,49].
The present study is the first to identify Myo5b as a motor protein that transports FN1-containing vesicles in myofibroblasts. Supporting this conclusion, we obtained the following findings: (1)  Myosin constitutes a large superfamily of more than 35 classes and 10 classes are found in vertebrates [17]. Myosin family members can be classified into two groups based upon their nature as a motor protein. One is the motor suitable to produce large force, such as myosin II. This type of myosin spends the majority of the cross-bridge cycle time dissociated from actin and only a short time associated with actin to produce power stroke to move actin [18]. This nature is suitable for producing a large force since many myosin molecules can interact with single actin filaments without interfering with each other. However, this group of myosins is not suitable for vesicular transportation since it cannot move continuously on actin filaments without dissociation. The other group of myosin spends a majority of cross-bridge cycling time associated with actin. This type of myosin, often called "processive" myosin, is a suitable motor for specific cargo transportation such as intracellular vesicles [18]. Class V myosin is typical processive myosin, therefore it is consistent that Myo5b, a member of class V myosin processively transports FN1 vesicles.
It has been shown that the motor activity of Myo5a, a member of class 5 myosin, is regulated through the intramolecular interaction between the globular tail domain and the motor domain [50]. In the inhibited conformation, Myo5a forms a folded conformation, in which the tail domain associates with the motor domain. The binding of the globular tail domain to the motor domain interferes with the cross-bridge cycle movement, thus functioning as an intramolecular inhibitor [20,51]. This tail-dependent inhibited conformation is altered to an active extended conformation when Myo5a binds to a specific binding protein, melanophilin, to the globular tail that releases this inhibition, thus activating the motor activity of Myo5a [20]. A similar scenario of the regulation of Myo5b was reported that the binding of Rab11a to the globular tail can in part attenuate the inhibition [52]. In the present study, we found that Rab11b co-localizes with FN1 and Myo5b in HPMCs. Proximity ligation assay revealed that TGF-β significantly increased the proximity ligation signal intensity between Rab11b and Myo5b indicating a close association between Myo5b and Rab11b (Figure 8). Moreover, the active mutant of Rab11b, i.e., Rab11b S20V but not the inactive mutant of Rab11b (Rab11b S25N) showed colocalization with Myo5b (Figure 7). These results suggest that Myo5b associates with Rab11b and TGF-β stimulation promotes the binding of these proteins (Figures 7-10). It was suggested using a yeast two-hybrid assay that Myo5b tail domain may interact with Rab11b S20V (active mutant), but not Rab11b S25N (dominant negative mutant) [53]. The present results are consistent with that obtained by a yeast two-hybrid assay. Supporting this view, we found that TGF-β stimulation induces the formation of the active form of Rab11b ( Figure 11). Further, Rab11b gene silencing attenuated the TGF-β induced secretion of FN1 ( Figure 6). These results suggest that Myo5b binds to Rab11b, which plays an important role in the transportation and secretion of FN1 in HPMCs.
Quite interestingly, when Myo5b is overexpressed all three proteins, i.e., FN1, Rab11b and Myo5b, showed high colocalization ( Figure 6). The result suggests that Myo5b strengthens the association of Rab11b with FN1 vesicles. In other words, Myo5b is not simply a binding partner of Rab11b, but it may function as a regulator of Rab11b to facilitate association of FN1 vesicles. We also found that the active mutant of Rab11b showed punctate localization that well colocalize with Myo5b while the dominant negative mutant was diffuse (Figure 7). These results are consistent with the above finding and suggest that Myo5b facilitates association of Rab11b with FN1, presumably due to stabilization of Rab11b-GTP form. This view was supported by the finding that Rab11b-GTPγS but not Rab11b-GDPβS directly binds to the Tail domain of Myo5b ( Figure 11).
We found that Rab11b gene silencing significantly diminished the secretion of FN1 from HPMCs. The results are consistent with an earlier report that Rab11b down-regulation decreases FN1 secretion from cultured primary arterial endothelial cells [54]. On the other hand, gene silencing of Rab11a, isoform of the Rab11 family, did not significantly attenuate the secretion of FN1 from HPMCs. Although Rab11a and Rab11b share high amino acid homology (~90%), several reports suggested the distinct function of the two isoforms. Rab11a and Rab11b were localized at the different vesicle compartments in MDCK and gastric parietal cells [55]. Rab11b, but not Rab11a regulated cystic fibrosis transmembrane conductance regulator (CFTR) recycling to the apical membrane although both Rab11a and Rab11b were localized in the same vesicles containing CFTR [56]. The present findings are consistent with these earlier reports and support the distinct function of Rab11a and Rab11b. Based upon these findings, we concluded that Rab11b, but not Rab11a plays an important role in the transportation of FN1-containing vesicles in HMPCs.
It was previously reported that Rab11a partially activated the actin-activated ATPase activity of Myo5b in vitro, presumably due to the disruption of intra-molecular inhibitor function of the globular tail domain [52]. In the present study, we found that Rab11b-GTP binds to the tail domain of Myo5b, and significantly activates the actin-activated ATPase activity of Myo5b that is closely correlated with the movement activity of Myo5b. Importantly, the Rab11b induced activation of Myo5b is dependent on the activation of Rab11b, i.e., formation of Rab11b-GTP form, and TGF-β facilitates the activation of Rab11b in HPMCs. Since the active form of Rab11b directly binds to the globular tail domain of Myo5b, it is likely that Rab11b binding to the tail domain interferes with the intra-molecular interaction between the motor domain and the globular tail domain. Above results suggest that Rab11b activates Myo5b based transportation of FN1-containing vesicles by two folds, i.e., facilitating the association of Myo5b motor with the FN1-containing vesicles and the activation of the motor activity thus activating the movement of the vesicles.
Present study identified that Myo5b is an important motor for FN1 transportation and secretion, thus actin-based transportation system plays an important role in FN1 transportation and secretion. Proposed mechanism of Myo5b driven transportation of FN1-containing vesicles based upon the present findings is as follows. TGF-β stimulation activates Rab11b to form a GTP bound form. The activated Rab11b-GTP recruits Myo5b to the FN1-containing vesicles and activates the motor activity of Myo5b. Activated Myo5b continuously move the FN1-containing vesicles along with the actin filaments towards cell periphery, thus facilitating the secretion of FN from the cells. Future studies will involve the characterization of pleural fibrosis progression in mice with a targeted deletion of Myo5b. Floxed myo5b mice [57] will cross with calb2 cre-mice to generate mice with mesothelial cell specific loss of Myo5b, as we previously reported [49]. These studies will determine the effectiveness of Myo5b targeting on disease progression.
Our results indicate that Myo5b transports FN1 and facilitates secretion from transitioned HPMCs. It has been postulated that the directional transportation of intracellular vesicles is driven by both microtubule and microfilament transportation systems [58]. For instance, both kinesin and Myo5a are involved in the transportation and secretion of melanosomes from melanocytes [59]. For FN1 transportation, it is likely that microtubule motors such as kinesin family members contribute to such a movement. To date, nothing is known about possible microtubule based motors involved in FN1 transportation, and further study is required to understand entire transportation systems for FN1 secretion.

Mouse Disease Models
All experiments involving animals were approved by the Institutional Animal Care and Use Committee at the University of Texas Health Science Center at Tyler. All experiments relating to animals were performed in accordance with relevant guidelines and regulations [IACUC protocol numbers: 648 (Date of approval, 24 June 2019) and 689 (Date of approval, 17 March 2021)]. Wild-type C57BL/6j mice were intrapleurally treated with saline, carbon black/bleomycin (CBB) for 14 days or S. pneumoniae for 7d as previously described [21,60].

HPMC Isolation and Culture
Permission to collect and use HPMCs was approved by the Institutional Human Subjects Review Board of the University of Texas Health Science Center at Tyler [Protocol code, 2020-039 (Date of approval, 14 September 2020)]. All experiments regarding human subjects were performed in accordance with relevant guidelines and regulations. Cells were isolated from pleural fluids collected from patients with congestive heart failure or post coronary bypass pleural effusions [61] and maintained on dishes with CellBIND surface (Corning, New York, NY, USA) using LHC-8 culture medium (Thermo Fisher Scientific, Waltham, MA, USA) or BEGM (without retinoic acid and epinephrine, Lonza, Basel, Switzerland) containing 3% fetal bovine serum (Thermo Fisher Scientific), 2% antibiotic-antimycotic (Thermo Fisher Scientific), and 1% L-glutamine (Thermo Fisher Scientific) in a humidified incubator at 37 • C and 5% CO 2 /95% air.

Cellular Treatment
Cells were incubated in serum-free medium (SFM) of RPMI 1640 (HyClone, Logan, UT, USA) supplemented with GlutaMAX (Thermo Fisher Scientific) for 8 h prior to treatment with recombinant human TGF-β (5 ng/mL, R&D Systems, Minneapolis, MN, USA). Cells were then allowed to incubate for 24 h (Quantitative PCR analysis) or 48 h (Western blotting and immunostaining analysis) at 37 • C and 5 % CO 2 /95 % air.

Antibodies
The following antibodies were used for immunocytochemistry and Western blot analysis: anti-

Histochemistry and Immunofluorescence Staining of Mice Tissues
Tissues sections were first deparaffinized and subjected to antigen retrieval using a citrate buffer at 95 • C for 20 min as previously described [29]. Immunostaining was performed by using antibodies of Col-1 (1310-01, SouthernBiotech), α-SMA (17H19L35, Thermo Fisher Scientific) and FN1 antibody (MAB1918, R&D Systems). Briefly, mouse tissue sections were blocked using a proprietary blocking solution from a M.O.M. kit (Vector Laboratories, Burlingame, CA, USA). Primary antibodies were then incubated overnight at 4 • C in kit diluent. First antibodies were visualized with Alexa Fluor 488, 568 and 647 secondary antibodies (Life Technologies, Carlsbad, CA, USA), and nuclei were stained with Hoechst 33342 (Thermo Fisher Scientific). Tissues were mounted onto slides with Fluoro Gel with DABCO (Electron Microscopy Sciences, Hatfield, PA, USA). Tissue staining images were taken by Leica TCS SP8 systems (Leica Microsystems, Wetzlar, Germany).

Western Blotting
Conditioned media (CM) were collected, and cells were washed with PBS and lysed with 200 µL PBS containing 1% NP40, 20 mM Tris-HCl (PH 7.5), 137 mM NaCl, 1 mM Na 2 VO 4 , 1 mM EDTA supplemented with Halt Protease Inhibitor Cocktail (Thermo Fisher Scientific). Cell lysates were then incubated on ice for 30 min, and debris was removed by centrifuging. Protein concentrations of cell lysates were determined using Pierce BCA assay kit (Thermo Fisher Scientific). Immunoblots were then imaged by Molecular Imager ChemiDoc XRS+ (Bio-Rad, Hercules, CA, USA).

Immunofluorescence Staining
Cells were washed with PBS, placed in fixation solution (4 % formaldehyde, 2 mM MgCl 2 , and 1 mM EGTA in PBS), washed with PBS, and permeabilized with 0.05% Triton X-100 in PBS for 10 min. Cells were then washed and blocked with 5% BSA for 1 h. First antibodies were diluted with 5% BSA, then applied to cells and incubated overnight and visualized with Alexa Fluor 488, 568 and 647 secondary antibodies (Thermo Fisher Scientific), and nuclei were stained with Hoechst 33342 (Thermo Fisher Scientific). Proximity ligation assay was performed according to the Olink Bioscience protocol using Duolink In situ Red starter kit (Sigma Aldrich). After incubation with primary antibodies as described above, cells were washed and incubated with PLA probes (anti-mouse MINUS and antirabbit PLUS) for 1 h at 37 • C. Subsequent ligation and detection were performed according to the manufacturer's protocol. Cells were mounted onto slides with Fluoro Gel with DABCO (Electron Microscopy SciencesFluorescence images were taken by Leica TCS SP8 systems (Leica Microsystems) for confocal microscopy.

Live Cell Imaging
GFP-Myo5a and mCherry-FN1 was cloned in BacMam pCMV-DEST Vector using the Gateway cloning system. For GFP-Myo5a, GFP tag was cloned to N-terminal of Myo5a. For mCherry-FN1, mCherry-tag was cloned between FN1-III domains 3 and 4. The sequence at the insertion site was (FN1-III 3).TTGTGGRMVSK. (mCherry). ELYKGGRPRSD. (FN1-III 4) [the mCherry sequence is underlined; the Not I restriction site added three extra amino acids (GGR) at each end of the mCherry] [62]. The recombinant baculoviruses expressing GFP-Myo5a and mCherry-FN1 were produced using ViraPower BacMam Expression System (Thermo Fisher Scientific) according to manufacture's protocol. Live cell imaging was performed by using DeltaVision OMX (GE Healthcare Life Sciences, Chicago, IL, USA) in 48 h after transfection.
Full length of Myo5b heavy chain and full length Rab11b cDNAs were transferred to pFastbac HT (Thermo Fisher Scientific) that is modified to express FLAG Tag at the N-terminal side of cDNA of interest. Approximately 2 × 10 9 Sf9 cells (Thermo Fisher Scientific,) were infected with Myo5b heavy chain plus calmodulin light chain or Rab11b expressing baculoviruses. The cells were harvested at 72 h after infection.
For Rab11b, the Sf9 cell pellet was homogenized in PBS, 1 mM DTT, and 10 µg/mL leupeptin. The supernatant was cleared by ultracentrifugation, mixed with Ni-NTA resins and the tube containing the suspension was rotated at 4 • C for 1 h. After two washes with 0.3 M NaCl and 10 mM Imidazole-HCl (pH7.5), the resins were packed to a column, and the protein was eluted with 0.3 M NaCl, 200 mM Imidazole-HCl (pH7.5) and 10% sucrose. Purified protein was concentrated with a Vivaspin 6 (10 kDa MWCO, Sartorius). All purified proteins were snap-frozen in liquid nitrogen and stored at −80 • C.

GST-Pull down Assay
The purified flag-Rab11b was pretreated in buffer (20 mM Tris-HCl PH 7.5, 150 mM NaCl, 1 mM MgCl 2 , 1 mM EGTA, 1 mM DTT) in the presence of 5-fold excess of GTPγS, or GDPβS for 30 min at 37 • C. In other assays flag-Rab11b was added to the lysates of cells treated with or without TGF-β. Thirty microliters of 10 µM Rab11b and 2.5 µM Myo5b Tail in binding buffer (20 mM Tris-HCl PH 7.5, 150 mM NaCl, 1 mM MgCl 2 , 1 mM EGTA, 1 mM DTT, 0.05 % Tween, 1% BSA) in the presence of 20 µM GTPγS, GDPβS, or absence of the any analogs were mixed with 5 µL of Glutathione Sepharose 4B (GE Healthcare Life Sciences) and incubated with rotation at 4 • C for 1 h. The beads were collected by brief centrifugation and washed 3 times with 300 ul of buffer (20 mM Tris-HCl PH 7.5, 150 mM NaCl, 1 mM MgCl 2 , 1 mM EGTA, 1 mM DTT, 0.05 % Tween). The bound proteins were eluted with 25 µL of 20 mM Glutathione in Tris-HCl (PH8.0), 100 mM NaCl and 1 mM DTT. The eluted proteins were subjected to SDS-PAGE. GST-Myo5b Tail was visualized by Coomassie Brilliant Blue staining. Flag-Rab11b were analyzed by Western blot using flag antibody.

Conclusions
Present findings indicate that Myo5b and Rab11b play a key role in the transportation and secretion of FN1 from HPMCs. We conclude that TGF-β stimulation activates Rab11b to form Rab11b-GTP, which facilitates the association of Rab11b to FN1-containing vesicles as well as the binding of Ran11b to the tail domain of Myo5b. The bound Rab11b-GTP activates the motor activity of Myo5b, and the activated Myo5b continuously moves FN1containing vesicles to cell peripheries and facilitates FN1 secretion from the cells. The present findings support that Myo5b and Rab11b may contribute to the development of pleural fibrosis.