3

    Synergistic induction of CTGF by cytochalasin D and TGFβ-1 in primary human renal epithelial cells: Role of transcriptional regulators MKL1, YAP/TAZ and Smad2/3 Susanne M¨uhlich, Margot Rehm, Astrid Ebenau, Margarete GoppeltStruebe PII: DOI: Reference:

S0898-6568(16)30238-8 doi: 10.1016/j.cellsig.2016.10.002 CLS 8773

To appear in:

Cellular Signalling

Received date: Revised date: Accepted date:

4 May 2016 26 September 2016 6 October 2016

Please cite this article as: Susanne M¨ uhlich, Margot Rehm, Astrid Ebenau, Margarete Goppelt-Struebe, Synergistic induction of CTGF by cytochalasin D and TGFβ-1 in primary human renal epithelial cells: Role of transcriptional regulators MKL1, YAP/TAZ and Smad2/3, Cellular Signalling (2016), doi: 10.1016/j.cellsig.2016.10.002

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Synergistic induction of CTGF by cytochalasin D and TGF-1 in primary human renal

IP

T

epithelial cells: Role of transcriptional regulators MKL1, YAP/TAZ and Smad2/3

1

Walther Straub Institute of Pharmacology and Toxicology, Ludwig-Maximilians-University,

Goethestrasse 33, D-80336 München, Germany 2

SC R

Susanne Mühlich1, Margot Rehm2, Astrid Ebenau2, Margarete Goppelt-Struebe2§

Department of Nephrology and Hypertension, Friedrich-Alexander-Universität Erlangen-

§

CE P

TE

D

MA

NU

Nürnberg, Loschgestrasse 8, D-91054 Erlangen, Germany

AC

Address correspondence to:

Margarete Goppelt-Struebe, PhD, Friedrich-Alexander-Universität Erlangen-Nürnberg Department of Nephrology and Hypertension Loschgestrasse 8, D-91054 Erlangen, Germany Phone: ++49-9131-8539201; Fax: ++49-9131-8539202; Email: [email protected]

1

ACCEPTED MANUSCRIPT Abstract

Changes in cell morphology that involve alterations of the actin cytoskeleton are a hallmark of diseased renal tubular epithelial cells. While the impact of actin remodeling on gene

T

expression has been analyzed in many model systems based on cell lines, this study

IP

investigated human primary tubular epithelial cells isolated from healthy parts of tumor

SC R

nephrectomies.

Latrunculin B (LatB) and cytochalasin D (CytoD) were used to modulate G-actin levels in a receptor-independent manner. Both compounds (at 0.5 µM) profoundly altered F-actin structures in a Rho kinase-dependent manner, but only CytoD strongly induced the pro-

NU

fibrotic factor CTGF (connective tissue growth factor). CTGF induction was dependent on YAP as shown by transient downregulation experiments. However, CytoD did not alter the

MA

nuclear localization of either YAP or TAZ, whereas LatB reduced nuclear levels particularly of TAZ. CytoD modified MKL1, a coactivator of serum response factor (SRF) regulating CTGF induction, and promoted its nuclear localization.

D

TGF-1 is one of the major factors involved in tubulointerstitial disease and an inducer of CTGF. Preincubation with CytoD but not LatB synergistically enhanced the TGF-1-

TE

stimulated synthesis of CTGF, both in cells cultured on plastic dishes as well as in polarized epithelial cells. CytoD had no direct effect on the phosphorylation of Smad2/3, but facilitated

CE P

their phosphorylation and thus activation by TGF-1. Our present findings provide evidence that morphological alterations have a strong impact on cellular signaling of one of the major pro-fibrotic factors, TGF-1. However, our data also

AC

indicate that changes in cell morphology per se cannot predict those interactions which are critically dependent on molecular fine tuning of actin reorganization.

Key Words: MKL1, YAP/TAZ, CTGF, primary tubular epithelial cells; TGF

Abbreviations: CTGF, connective tissue growth factor, CCN2; CytoD, cytochalasin D; Gactin, globular actin; F-actin, filamentous actin; hPTEC, human primary tubular epithelial cells; LatB, latrunculin B; MKL1, Megakaryoblastic Leukemia 1; MYPT, myosin phosphatase targeting subunit; TAZ, transcriptional coactivator with PDZbinding motif; SRF, serum response factor; TGFtransforming growth factor beta; YAP, Yes-associated protein;

2

ACCEPTED MANUSCRIPT 1. Introduction

Acute and chronic kidney injuries are associated with structural damage of tubular epithelial cells. Depending on the severity of the disease, tubular injury may be more focal or rather

T

extended leading to progressive functional and potentially irreversible renal failure [1].

IP

Tubular cells are embedded in a network of interacting cells communicating by multiple

SC R

soluble mediators [2]. By activating neighboring interstitial cells, damaged tubular cells may contribute to tubulointerstitial fibrosis, a hallmark of end stage renal disease. Examples are crystal-induced renal failure by uranyl acetate [3], selective damage of proximal tubules by diphtheria toxin [4] or enforced cell cycle arrest [5]. Dilation of tubules implies extensive

as well as regenerating epithelial cells.

NU

reorganization of cell morphology, as do the mesenchymal alterations observed in damaged

MA

Morphological alterations of tubular cells are accompanied and mediated by changes in Factin (filamentous actin) structures, primarily regulated by the activity of GTPases of the Rho family [6]. Reorganization of the F-actin cytoskeleton has been linked to gene expression implicating signaling via RhoA and MKL1 (Megakaryoblastic Leukemia 1, also named MRTF-

D

A, MAL, BSAC). The transcriptional coactivator MKL1 has the exquisite capability to

TE

transduce rearrangements in the actin cytoskeleton into gene expression via the transcription factor SRF (serum response factor). In the repressed state, monomeric actin binds in a 5:1

CE P

complex to the N-terminal region of MKL1 [7]. Actin polymerization in response to Rho GTPase signaling causes dissociation of the G-actin (globular actin)-MKL1 complex and translocation of MKL1 to the nucleus, where it binds to and activates SRF [8]. Complex formation of MKL1 and SRF leads to the induction of a subset of SRF sensitive genes

AC

involved in motile, contractile and muscle-specific functions [9]. Downregulation of SRF activity is facilitated by nuclear export of actin-bound, phosphorylated MKL1 [10, 11]. The link between actin rearrangements and gene expression has been studied most intensively in mesenchymal cells, notably fibroblasts and vascular smooth muscle cells, but is also relevant in epithelial cells. We have shown earlier that activation of RhoA by lysophosphatidic acid (LPA) leads to a MKL1-dependent induction of the pro-fibrotic factor connective tissue growth factor (CTGF, also named CCN2) in proximal tubular cells [12]. Using LPA as stimulus implicates expression of appropriate LPA receptors, which have been described in proximal tubules [13], in line with in vitro studies, where proximal but not distal tubular epithelial cells reacted to LPA with an increase in CTGF expression [14]. Modulation of RhoA and Rac-1 activities was used to show a link between structural changes of F-actin and CTGF expression in a cell line (HKC-8) derived from human proximal tubular epithelial cells [12].

3

ACCEPTED MANUSCRIPT Multiple enzyme activities are involved in the regulation of F-actin polymerization, structuring and depolymerization [15]. Pharmacologically, F-actin structures can be modulated by drugs, namely latrunculin B (LatB) and cytochalasin D (CytoD), which differentially interact with Gactin and F-actin leading to the dissociation of F-actin. LatB binds to G-actin and thus rapidly

T

sequesters the monomeric form [16, 17]. CytoD, by contrast, primarily binds to the barbed

IP

end of F-actin fibers, prevents their constantly ongoing assembly and disassembly and

SC R

increases the cellular content of G-actin [18]. CytoD thus competes with the binding of the capping protein to the barbed end and indirectly also affects interactions with other proteins involved in anchoring F-actin [19]. Both compounds also differentially affect complex formation of MKL1 and G-actin [9]: LatB inhibits the release of MKL1 from G-actin, whereas

NU

CytoD disrupts the MKL1 - G-actin complex, leading to nuclear localization of MKL1 and activation of SRF, as shown in fibroblasts [20]. In this model, CytoD is assumed to directly

MA

bind monomeric actin [8]. However, the details of the interaction of CytoD, MKL1 and G-actin need further analysis.

Regulation of gene transcription by actin reorganization is not restricted to MKL1-SRF signaling. More recently, transcription factors YAP (Yes-associated protein) and TAZ

D

(transcriptional coactivator with PDZ-binding motif) have been shown to be regulated by cell

TE

geometry and the stiffness of the actin cytoskeleton in mesenchymal cells [21, 22]. In addition to mechanical input YAP/TAZ are regulated by input from multiple activators of the

CE P

Hippo signaling pathway [23, 24]. As CTGF is one of the target proteins regulated by YAP/TAZ (e.g. [25, 26]), we sought to compare the role of YAP/TAZ signaling to that of MKL1 in the regulation of CTGF in epithelial cells, exposed to disruption of the cytoskeleton. Tubular epithelial cell lines of different species and different origin are available for in vitro

AC

studies. In our experiments we used freshly isolated human primary tubular epithelial cell cultures (hPTEC) obtained from healthy parts of tumor nephrectomies to work with nontransformed cells obtained from different patients and originating from different parts of the tubular system. In these cells we investigated the impact of changes in cell architecture on CTGF expression together with an analysis of MKL1 and YAP-TAZ signaling, as well as TGF-1 (transforming growth factor beta)-mediated SMAD2/3 activation. We show that disruption of F-actin fibers by CytoD leads to enhanced nuclear localization of the coactivator MKL1 but also increases phosphorylation of regulatory Smads 2/3 by TGF-1. These effects were not seen when F-actin fiber formation was impaired by LatB, indicating that the impact of tubular epithelial cell injury on interstitial fibrosis is largely dependent on the subsequent modulation of intracellular signaling pathways.

4

ACCEPTED MANUSCRIPT 2. Materials and Methods 2.1 Materials - DMEM/Ham’s F12 medium was purchased from Biochrom, DMEM medium and Hank´s BSS from Sigma, insulin-transferrin-selenium supplement from Gibco, fetal calf

T

serum (FCS) from PAN Biotech (Aidenbach, Germany), triiodothyronine, hydrocortisone and

IP

lysophosphatidic acid (LPA) from Sigma-Aldrich, and epidermal growth factor from

SC R

PeproTech. TGFβ-1 was obtained from tebu-bio, Rho kinase inhibitor Y27632, MEK kinase inhibitor U0126, TGF- RI kinase inhibitor IV (SB431542), LatB and CytoD were obtained from Calbiochem.

NU

2.2 Cell culture - Human primary tubular epithelial cells were isolated from renal cortical tissues collected from healthy parts of tumor nephrectomies essentially as described

MA

previously [27]. Isolation of human cells from healthy parts of tumor nephrectomies was approved by the local ethics committee (Reference number 3755, Ethik-Kommission der Medizinischen Fakultät der Friedrich-Alexander Universität Erlangen-Nürnberg). Written

D

informed consent was obtained from all participants involved in this study. Part of the tissue was fixed with PFA and used for immunolocalization of YAP/TAZ. In brief, sections were

TE

treated with citrate buffer, exposed to micro wave treatment, and blocked with horse serum. As primary antibodies, mouse anti-YAP/TAZ (SC-101199, Santa Cruz) and rabbit anti-YAP

CE P

(#4912, Cell Signaling Technologies), and as secondary antibodies, PromoFluor Fluor antimouse or anti-rabbit (Promokine) were used. For experiments, hPTEC were seeded in medium containing 2.5 % FCS to facilitate cell attachment. After 24 h, medium was replaced with FCS-free epithelial cell selective medium

AC

(DMEM/Ham’s F12 medium containing 2 mM  L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, insulin-transferrin-selenium supplement, 10 ng/ml epidermal growth factor, 36 ng/ml hydrocortisone and 4 pg/ml triiodothyronine). For part of the experiments proximal and distal cells were separated by their differential adherence to cell culture plastic. Trypsinization for 3 min resulted in a culture enriched in proximal cells (about 60 % Ncadherin positive cells), while the remaining cells were over 90 % E-cadherin positive and thus represented cells of distal tubular origin. Data obtained with proximal or distal cells were combined when no significant difference was detected between cell populations. Polarized cells were cultured on permeable transwell inserts (Millicell PCF, Millipore) as described previously [14]. For scratch assays, cells were cultured for 10 days on transparent cell culture inserts (PET membranes, Becton Dickinson). Monolayers were wounded with a pipet tip and the cells were allowed to migrate into the open space for 6 h. When CTGF secretion from polarized cells was analyzed, heparin (100 µg/ml) was added in the upper and

5

ACCEPTED MANUSCRIPT lower compartment to prevent CTGF sticking to the transwells. In all experiments, hPTEC at passages 1-3 were used. 2.3 Cell fractionation - Separation of nuclear and cytosolic fractions was performed

T

essentially as described before [28, 29], by lysing the cells in buffer containing 10 mM

IP

Hepes, pH 7.9, 0.1 mM EDTA, 10 mM KCl, 1 mM DTT, 0.66 % NP-40 and protease

SC R

inhibitors. After centrifugation the washed pellet was resuspended in buffer containing 10 mM Hepes, pH 7.9, 0.1 mM EDTA, 400 mM NaCl, 1 mM DTT and protease inhibitors. After centrifugation the supernatant contained soluble nuclear proteins. Proteins detected in

NU

nuclear fractions by Western blotting were related to lamin A/C as nuclear marker protein.

2.4 Western blot analysis - Cells were lyzed in buffer containing 50 mM HEPES pH 7.4,

MA

150 mM NaCl, 1 % Triton X-100, 1 mM EDTA, 10 % glycerol, 2 mM sodium vanadate and protease inhibitors complete EDTA-free (Roche Diagnostics, Mannheim, Germany) or in phosphate-buffered saline containing 5 % SDS plus inhibitors to detect phosphorylated proteins. hPTEC secret CTGF which can be recovered from the cell culture supernatants. To

D

detect secreted CTGF, equal volumes of cell culture supernatants (200 µl) were precipitated

TE

with ethanol or acetone. Western blot analyses were performed essentially as described before as described before [30] using the following antibodies: goat anti-CTGF (SC-14939),

CE P

mouse anti-vinculin (SC-5573), mouse anti-YAP-TAZ (SC-101199), mouse anti-RhoA (SC418), and peroxidase-conjugated donkey anti-goat IgG (SC-2020) from Santa Cruz; rabbit anti-phosphoMYPT (Thr853) (#4563), and rabbit anti-lamin A/C (#2032) from Cell Signaling Technologies, mouse anti-tubulin (T0198) from Sigma-Aldrich, and sheep anti-mouse IgG

AC

(NA931V) and donkey anti-rabbit IgG (NA934V) from Amersham Biosciences. A rabbit antiMKL1 serum was generated by injecting rabbits with glutathione S-transferase (GST)-MKL1 (amino acids 601 to 931) as described previously [11]. To ensure equal loading and blotting, blots were redetected with an antibody directed against vinculin, tubulin or lamin A/C. The immunoreactive bands were quantified using the luminescent image analyzer (LAS-1000 Image Analyzer, Fujifilm, Berlin, Germany) and AIDA 4.15 image analyzer software (Raytest, Berlin, Germany), or the Odyssey infrared imaging system (Li-Cor, Biosciences). All data are presented as mean + SD of n different experiments with cells obtained from at least 3 different donors. To summarize data obtained from different cell cultures, relative band intensities were normalized as indicated in the legends.

2.5 Immunocytochemistry - Cells were fixed with paraformaldehyde (3.5 % in PBS) for 10 min and afterwards permeabilized with 0.5 % Triton X-100 in PBS for 10 min. After washing

6

ACCEPTED MANUSCRIPT three times with PBS, cells were blocked in 1 % BSA in PBS for 1 h at room temperature and washed once. Primary antibodies were mouse anti-E-cadherin (ab1416, Abcam), rabbit anti-N-cadherin (SC-7939), goat anti-MKL1 (SC-21558, Santa Cruz) and mouse anti-YAP/TAZ (SC-101199,

T

Santa Cruz). Secondary antibodies (1:500, PromoFluor Fluor anti-mouse or anti-rabbit)

IP

were from PromoKine. F-actin was stained with PromoFluor 488 or 555 phalloidin from

SC R

PromoKine, nuclei were visualized with Hoechst (Sigma-Aldrich).

After mounting, slides were viewed using a Nikon Eclipse 80i fluorescence microscope and digital images recorded by Visitron Systems 7.4 Slider camera (Diagnostic Instruments, Puchheim, Germany) using Spot Advanced software (Diagnostic Instruments) or Keyence

NU

fluorescence microscope BZ9000.Three dimensional images were evaluated by

MA

epifluorescence microscopy including Apotome technique (Zeiss, Göttingen, Germany).

2.6 siRNA Transfection – siRNA transfections were performed essentially as described previously [31]. To down-regulate MKL or YAP in primary cells, cells were transfected 3 h

D

after seeding using Saint Red (Synvolux) according to the manufacturer’s instructions.

TE

Silencing of YAP (EHU113021, Sigma, consisting of 4 different siRNAs directed against human YAP) was about 80 % (80 + 6 %, means + SD of 6 experiments). Silencing of YAP did not significantly alter the expression of TAZ [29]. Silencing of MKL1 has been described

CE P

in [32], (si 5’ GAA UGU GCU ACA GUU GAA A and 5'-GUG UCU UGG UGU AGU GUA A). Silencing of RhoA was over 75 %, si: 5’GAC AUG CUU GCU CAU AGU C [12] and

control.

AC

Hs_RHOA_7 HP validated (SI02654267, Qiagen). siRNA directed against GFP was used as

2.7 Statistical analysis – All graphs are presented as means + SD. To compare multiple conditions, statistical significance was calculated by one-way ANOVA with appropriate post hoc tests, or one sample t-test using GraphPad software. A value of p< 0.05 was considered to indicate significance.

3. Results 3.1 Activation of Rho kinases contributes to LatB- and CytoD-mediated changes in Factin structures and CTGF secretion. Disruption of F-actin fibers by LatB (0.5 µM) or CytoD (0.5 µM) in hPTEC induced formation of F-actin aggregates (arrows in Fig. 1A). At these concentrations, cell-cell contacts became loose but the cells did not detach (for higher magnification see Fig. S1A). Changes in cell 7

ACCEPTED MANUSCRIPT architecture were concentration-dependent in a narrow range: while 0.1 µM LatB or CytoD mostly affected intracellular F-actin fibers with little alterations of the cortical F-actin, barely any cell-cells contacts were left when the concentrations of LatB or CytoD were increased to 1 µM (Fig. S1A). Long F-actin containing extensions formed between separated neighboring

T

cells (Fig. S1A, 1 µM, higher magnification). Since cell-cell contacts are essential for

IP

epithelial cells (arrow heads in Fig. S1A), 0.5 µM LatB and CytoD were used throughout this

SC R

study.

Loosening of cell contacts was associated with rearrangement of cadherins which also clustered in and co-localized with F-actin aggregates (Fig S1B). Activation of Rho kinase signaling was involved in the formation of F-actin clusters as shown by the profound effect of

NU

the Rho kinase inhibitor Y27632, which almost completely prevented formation of F-actin clusters (Fig. 1A). In line with these findings, the down-stream target of Rho kinases, MYPT

MA

(myosin phosphatase targeting subunit) was phosphorylated upon LatB and CytoD treatment with no significant difference in proximal and distal cell preparations (Fig. 1B). Of interest, kinetics differed between LatB and CytoD, with LatB being active within 30 min whereas a comparable activation of MYPT by CytoD was observed only after about 2 h (Fig S2). These

D

observations were in accordance with the different molecular mechanisms leading to the

TE

inhibition of actin polymerization, with LatB being a direct binding partner of G-actin and CytoD binding to actin filaments and thereby blocking polymerization of actin.

CE P

Next we analyzed CTGF secretion which was more pronounced in cultures of proximal than of distal cells (Fig. 1C, control prox vs control dist). A minor increase was observed in LatBtreated pre-activated proximal cells, whereas a robust increase in CTGF secretion was observed in distal cells (Fig. 1C). Even though the changes in F-actin structures were less

AC

pronounced by CytoD compared to LatB, CTGF secretion was strongly enhanced (Fig. 1C). This indicated that the rapid binding of G-actin by LatB and the activation of Rho kinase signaling contributed but were not sufficient for a strong CTGF induction. Inhibition of Rho kinase by Y27632 profoundly interfered with CytoD-mediated CTGF induction (Fig. 1D). Ongoing CTGF synthesis can also be detected in the cellular homogenates (Fig. 1D, cCTGF). The reduction by Rho kinase inhibition indicated an effect at the transcriptional level rather than impairment of secretion by Y27632. 3.2 YAP or TAZ are not translocated to the nuclear compartment by CytoD or LatB in hPTEC YAP/TAZ are essential transcription factors in the regulation of CTGF expression. Their nuclear or cytoplasmic localization has been shown to be dependent on cell density and cell structure. Nuclear and cytosolic localization of YAP and TAZ was quantified by Western blotting in respective extracts obtained from subconfluent (sc) and confluent (c) hPTEC (Fig. 8

ACCEPTED MANUSCRIPT 2A). The proportion of nuclear YAP/TAZ was calculated considering the different volumes used for extraction of cytosolic and nuclear fractions. The proportion of nuclear YAP and TAZ was 2fold higher in subconfluent (sc) compared to confluent (c) cells (Fig. 2A). In confluent cells less than 20% YAP or TAZ were observed in the nuclear compartment. However,

T

compared to cell lines, primary tubular cells showed a much more variable expression

IP

pattern of YAP/TAZ, as shown by immunofluorescence. Proximal cells in culture,

SC R

characterized by N-cadherin, tend to form cell clusters (Fig. 2B, arrows, [31]), which showed far less nuclear YAP/TAZ than spread cells. Differences in YAP/TAZ distribution was also observed in distal tubular hPTEC (negative for N-cadherin) when cultured in plastic dishes (Fig. 2B). Cell density-dependent localization became strikingly obvious when polarized distal

NU

tubular epithelial cells were compared to migrating cells after wounding of the monolayer (Fig. 2C). Whereas dense polarized cells showed dominant cytosolic localization of

MA

YAP/TAZ, there was a strong nuclear localization detectable in the migrating, non-polarized cells.

Nuclear as well as cytoplasmic localization of YAP/TAZ were detected in tubular cells by immunocytochemistry in sections obtained from the tissue used for isolation of hPTEC. Only

D

selected tubules showed strictly cytoplasmic localization of YAP/TAZ, whereas in most

TE

tubular cells, YAP/TAZ also localized to the nucleus, albeit with varying intensity (Fig. 2D) Quantification of subcellular distribution by Western blotting showed that LatB reduced

CE P

nuclear localization of TAZ and to a lesser extent of YAP. In proximal cells, LatB seemed to affect TAZ synthesis or degradation as both, cytosolic and nuclear TAZ were reduced (Fig. 2E). The partial reduction of nuclear YAP and/or TAZ was not sufficient to decrease CTGF expression in LatB-treated cells (Fig. 1C), most likely due to the activation of other signaling

AC

pathways including Rho kinase signaling. CytoD did not significantly increase the nuclear localization of YAP or TAZ in proximal or in distal hPTEC (Fig. 2E). As a control we used LPA as stimulus which enhanced YAP and TAZ in the nuclear fraction (Fig S3). Based on these results, the strong induction of CTGF by CytoD did not seem to be mediated by an increase in nuclear YAP/TAZ. Almost complete downregulation of YAP by siRNA (< 80 %) was necessary to interfere with upregulation of CTGF by CytoD (Fig. 2F). This indicated that a certain concentration of nuclear YAP was essential for CTGF induction in the context of other transcription factors. 3.3 Nuclear localization of modified MKL1 is increased in CytoD-treated cells The SRF coactivator MKL1 has been shown to be involved in CTGF induction and is sensitive to changes in G- and F-actin levels. Therefore, its localization and functional role in hPTEC were investigated. In confluent proximal and distal hPTEC MKL1 appeared to be primarily cytoplasmic when analyzed by immunofluorescence (Fig. 3A). The nucleus was not 9

ACCEPTED MANUSCRIPT spared indicative of partial nuclear localization of MKL1 in cultured cells (Fig. 3A). This was confirmed by Western blotting of nuclear and cytosolic fractions (Fig. 3B). Considering the volumes of extraction solutions, about 40 % of MKL1 were detected in the nuclear compartment. LatB contracted the cells impairing the analysis of MKL1 localization by

T

immunocytochemistry (Fig. 3A). The quantitative analysis of nuclear and cytosolic fractions

IP

by Western blotting showed a decrease of nuclear MKL1 in LatB-treated cells (Fig. 3B).

SC R

CytoD, by contrast, increased nuclear MKL1 in proximal and distal tubular cells, confirmed by Western blot analysis (Fig. 3 A/B). Treatment of the cells induced a shift in the molecular weight of MKL1, in the cytosolic as well as in the nuclear fraction. It was observed as early as 30 min of CytoD treatment (Fig. 3C). It was observed as early as 30 min of Cyto D treatment

NU

and was reminiscent of the shift caused by ERK-dependent phosphorylation of MKL1 [11]. However, the modification of MKL1 by CytoD was insensitive to MEK inhibition by UO126

MA

(Fig. 3C), which excluded phosphorylation by ERK. Furthermore, the modification was independent of RhoA activity as shown by down-regulation of RhoA by two different siRNAs (Fig. 3C and Fig. S4). In line with these results, inhibition of Rho kinases by Y27632 (10 µM, Fig. 3C) did not affect the CytoD-induced modification of MKL1. Lysophosphatidic acid (LPA)

D

was included in this experiment as it is a known activator of RhoA signaling. Thus far, the

TE

chemical nature of the modification has not been identified and the functional impact on MKL1 target genes such as CTGF remains to be investigated. Transient downregulation of

CE P

MKL1 reduced but did not abolish CytoD-induced CTGF expression indicating involvement of other transcriptional (co)activators in CytoD-mediated regulation of CTGF (Fig. 3D).

3.4 CytoD enhances TGF-1-mediated induction of CTGF in polarized hPTEC

AC

Tubular epithelial cells in vivo are polarized with a distinct apical and basolateral side. Therefore, we generated distal polarized hPTEC in permeable cell culture inserts [14, 31]. The polarized phenotype was not distorted by treatment with LatB or CytoD, even though the F-actin cytoskeleton was rearranged (Fig. 4A). As observed in 2D cultures, F-actin fibers were reorganized into clusters. Rho kinase signaling seemed to be involved as determined by phosphorylation of MYPT, particularly in LatB-mediated rearrangements (Fig. 4B). Apart from the change in morphology there was no significant difference between control cells and CytoD-treated cells in terms of YAP/TAZ localization, whereas LatB shifted YAP/TAZ to the cytosolic compartment with more pronounced staining of the nuclei (Fig. S5, nuclei in blue). In some control cells, mostly at the boundaries of cell clusters, MKL1 showed nuclear localization in control cells (arrow in Fig. 4C), while in most cells, MKL1 appeared localized to the cytosol and the nucleus in polarized cells. Relocation of MKL1 from the nuclear compartment to the cytosolic compartment was evident upon treatment with LatB (Fig. 4C). In line with the strong shift of MKL1 and YAP/TAZ to the cytosol, there was very little 10

ACCEPTED MANUSCRIPT secretion of CTGF upon treatment with LatB (Fig. 4D: 2h, Fig. 4E: 24 h). CytoD, by contrast, stimulated MKL1 nuclear localization (Fig. 4C) correlating with secretion of CTGF to the apical side (Fig. 4D/E). CTGF is a very adhesive protein and therefore, heparin was added to the media to prevent CTGF from binding to the inserts. Under these conditions, secretion of

T

CTGF from TGF-1-stimulated cells was detected primarily in the basolateral compartment

IP

(Fig. 4E). Minor amounts of CTGF were detectable after 6 h, whereas robust secretion was

SC R

observed after 24 h (Fig. 4E). Enhancement of the TGF-1-mediated basolateral secretion was detected when cells were preincubated with CytoD, whereas the apical secretion induced primarily by CytoD was not altered (Fig. 4E).

NU

3.5 CytoD facilitates phosphorylation of Smad2/3 by TGF-1

To get an insight into the molecular mechanism of the interaction between CytoD and TGF-

MA

1, we also analyzed cells cultured under 2D conditions. Within 6 h, TGF-1 induced a significant increase of CTGF secretion from proximal or distal hPTEC (combined in the graph Fig. 5A). TGF-1-induced secretion of CTGF was not modulated by LatB, whereas pre-

D

incubation of the cells with CytoD induced a more than additive induction of CTGF (additive:

TE

2.0 + 0.8, combined: 3.3 + 1.6 fold induction, n=7, p<0.05, two sided t-test). Regulatory Smads, Smad2 and Smad3, are major mediators of TGF signaling. Smad2 and Smad3 were detectable in both, the nuclear and the cytosolic compartment (data not shown).

CE P

However, only cells stimulated with TGF-1 showed nuclear phosphorylated Smad2 and Smad3 (Fig. 5B). By itself, CytoD had no effect on Smad2/3 phosphorylation even at concentrations as high as 2.5 µM (Fig. S6A). However, preincubation with CytoD significantly

AC

increased TGF-1-mediated phosphorylation and thus activation of Smad2 and Smad3 (Fig. 5B). Both, phosphorylation of Smads by TGF-1 as well the increased stimulation in the presence of CytoD were dependent on TGF receptor activity as shown by complete abrogation in the presence of SB431542, an inhibitor of TGF receptor kinases ALK4 and ALK5 (Fig. S6B). Higher concentrations of CytoD, which more strongly affected F-actin structures and cell morphology, did not further increase Smad2/3 activation (Fig. S6A). 4. Discussion Tissue mechanics are translated into activation or repression of transcription factors and their coactivators [33]. In a previous study we have shown that the small GTPases RhoA and Rac1 play a role a modulators of the actin cytoskeleton and subsequent regulation of gene expression in proximal tubular epithelial cells, HKC-8 cells and primary cells [12]. In this study we used two well established drugs, CytoD and LatB, to disturb the actin cytoskeleton and concomitantly cell morphology by directly interfering with actin polymerization. 11

ACCEPTED MANUSCRIPT Concentrations of CytoD (0.5 µM) and LatB (0.5 µM) were chosen based on morphological evidence to achieve disruption of F-actin fibers within 1 h without detachment of the cells from the substratum and loss of cell-cell contacts. Much higher concentrations of both compounds, most notably CytoD, are often reported in the literature. As CytoD binds to the

T

barbed end of F-actin, it interferes with other anchoring proteins [19]. These direct effects as

SC R

depending on the concentration of the drug and the cell type.

IP

well as the subsequent reorganization of cell structures will differentially affect cell signaling

Both drugs induced the formation of actin aggregates. These have been analyzed in detail in MCF-7 cells [34]. Sequestration of signaling molecules such as active MAP kinases or members of adhesion complexes such as paxillin was observed in those cells. Formation of

NU

actin aggregates may thus contribute to the resolution of dense epithelial structures as observed in hPTEC.

MA

Even though both drugs bind directly to actin, albeit at different sites as outlined in the introduction, rearrangement of F-actin led to an activation of Rho kinases. This activation was necessary to form actin aggregates and more general for the cytoskeletal reorganization as shown by the Rho kinase inhibitor Y27632, which did not prevent but modulate drug-

D

induced changes in cell structure.

TE

The renal tubules are lined by epithelial cells which differ in their functional and structural properties. Therefore we had expected much more pronounced differences between cells

CE P

obtained from distal parts of the tubules and those obtained from proximal parts. However, only minor differences were observed such as TAZ degradation upon treatment with LatB in proximal tubular cells, which was not observed in distal cell preparations. This indicates that in spite of all their differences, morphological alterations per se lead to comparable

signals.

AC

responses in both cell types which may then be modulated by additional receptor-mediated

Even though the cytoskeletal alterations upon incubation with LatB occurred more rapidly and were stronger than those observed upon treatment with CytoD, induction of CTGF expression was much more pronounced with CytoD. These results were reminiscent of earlier studies in renal fibroblasts which did not respond with increased CTGF synthesis to LatB but only to CytoD [35]. Diverse signaling pathways have been linked to the translation of mechanical signals into gene expression [6, 36]. In terms of CTGF expression we concentrated on MKL1-SRF signaling and the role of YAP/TAZ coactivators, both of which have been implicated in mechanotransduction and regulation of CTGF expression. Localization of these transcription factors was differently affected by LatB and CytoD. LatB reduced nuclear localization of YAP/TAZ and MKL1, whereas CytoD increased nuclear localization of MKL1 with no effect on YAP/TAZ. The differential effects of LatB and CytoD on the cellular signaling may not only be attributable to their alterations of F-actin structures but 12

ACCEPTED MANUSCRIPT may involve interactions with other factors such as e.g. thymosin beta 4 which modulates MKL1 – actin binding [37, 38], or angiomotins, which regulate YAP localization in an F-actindependent manner [39]. Cell density has been shown in various cell types to be a modulator of YAP/TAZ localization.

T

This was confirmed in our study in primary tubular epithelial cells. Even in confluent cell

IP

monolayers, variation of subcellular localization of YAP/TAZ was observed. Whether this was

SC R

exclusively due to differences in cell density or whether it was also attributable to subtypes of tubular cells remains to be investigated. However, exclusion from the nucleus as reported in some cell lines or upon comparison of single cells with cells in monolayers was not detectable in 2D-cultures [22, 40]. Most strikingly, polarized cells were almost void of nuclear

NU

YAP/TAZ, which was prominent in cohorts of migrating hPTEC. Interestingly, a heterogeneous distribution of YAP/TAZ was also observed in tissue sections of human

MA

kidney. Only certain tubular structures showed exclusion of YAP/TAZ from the nucleus. The nature of these tubular segments needs to be further analyzed. CytoD barely affected YAP or TAZ localization, whereas LatB attenuated the nuclear content of YAP/TAZ, and also reduced cytosolic TAZ indicative of increased proteasomal

D

degradation. Reduced levels of total TAZ were also reported in human trabecular meshwork

TE

cells at concentrations of 2 µM LatB [41]. Interestingly, that effect was dependent on the stiffness of the substratum the cells were grown on and only observed in cells plated on soft

CE P

hydrogel. This indicates that regulation of TAZ degradation is not dependent on LatB alone but further influenced by additional mechanical factors. Such differences may also contribute to the differences between proximal and distal cells, with distal cells being much more adherent and less motile than proximal cells. Induction of CTGF by CytoD was thus not

AC

triggered by YAP/TAZ translocation, but still dependent on the presence of YAP/TAZ in the nucleus, as shown by transient downregulation of YAP by siRNA. MKL1, a transcriptional coactivator of SRF in the context of certain DNA binding motives, is regulated at multiple levels, binding partners having been identified in the cytosol as well as the nucleus [8]. G-actin plays a crucial role in MKL1 regulation and a decrease in G-actin upon actin polymerization leads to increased nuclear MKL1 localization and SRF activation in many systems. MKL1 binds to G-actin directly and is retained by G-actin in the cytosol [42], because actin monomers occupying the N-terminal RPEL-domains of MKL1 block access of importins to MKL1’s nuclear localization sequence [43, 44]. Furthermore, binding of G-actin to phosphorylated MKL1 is required for efficient nuclear export of MKL1 [10, 11]. Consistent with this, disruption of the MKL1 - G-actin complex by CytoD increased MKL1 nuclear localization, correlating with a strong increase in CTGF secretion. CytoD-induced CTGF secretion was reduced, but not abrogated upon MKL1 depletion, indicating other transcriptional activators being involved in CytoD-mediated regulation of CTGF. 13

ACCEPTED MANUSCRIPT In this study we observed a stable modification of MKL1 upon treatment of the cells with CytoD but not with LatB or other modulators of the cytoskeleton such as e.g. colchicine. The shift in gel mobility was reminiscent of ERK-dependent MKL1 phosphorylation [11]. However, the CytoD-induced modification was independent on ERK activity as shown by the MEK

T

inhibitor UO126. In a very recent publication, Panayiotou et al. described CytoD-dependent

IP

phosphorylation of overexpressed mouse MKL1 in fibroblasts, which was related to RhoA

SC R

activation [45]. By contrast, phosphorylation of the endogenous MKL1 in hPTEC: proved to be independent of RhoA signaling as shown by down-regulation of RhoA. Furthermore, there was no effect of the Rho kinase inhibitor Y27632. While the chemical nature of the modification remains to be defined, it is intriguing to speculate it may influence the interaction

NU

of MKL1 with SRF or other nuclear factors leading to the strong induction of CTGF. Within the nucleus, actin polymers were shown to contribute to MKL1 activation [46]. We

MA

recently described FilaminA as new MKL1 binding partner that mediates an association between F-actin and MKL1 required for MKL1/SRF activity [47]. Whether Filamin A also plays a role in CytoD-induced modulation of MKL1 localization and activation in tubular epithelial cells warrants further investigation.

D

TGF-1 is one of the factors associated with renal fibrosis and the induction of pro-fibrotic

TE

proteins such as CTGF. In vitro, induction of CTGF expression by TGF-1 was densitydependent with higher secretion from subconfluent cells [27], which may relate to the higher

CE P

nuclear content of YAP/TAZ in those cells serving as coactivators of regulatory Smads. Preincubation of hPTEC with CytoD synergistically increased TGF-1-mediated induction of CTGF. While 2 h of treatment with CytoD were sufficient to induce CTGF, induction by TGF-

AC

1 occurred slower and was detectable only after 4-6 h and more prominent after an overnight incubation. Preincubation with CytoD did not alter the kinetics of CTGF induction and no synergistic effect was observed at early time points. This indicated that CytoD enforced TGF-1 signaling, which was confirmed by increased phosphorylation of Smad2/3 in the presence of CytoD, which was strictly dependent on TGF-1 signaling. In polarized cells, rapid secretion of CTGF to the apical compartment was observed upon treatment with CytoD which was not altered by treatment of the cells with TGF-1 from the basolateral side. As shown earlier, Smad2/3 are activated upon application of TGF-1 from the basolateral side [14]. Preincubation with CytoD promoted this signaling pathway as evidenced by the increase of CTGF secretion in the basolateral compartment. Preincubation with CytoD increased phosphorylation and nuclear translocation of Smad2 and Smad3. As phosphorylation of Smads occurs in the cytoplasm upon TGF receptor activation, modulation of TGF-1 signaling by CytoD is assumed to occur in the cytoplasm and is most likely independent of MKL1 nuclear translocation. It was interesting to note that CytoD at 0.5 µM was sufficient to increase Smad phosphorylation with no significant increase 14

ACCEPTED MANUSCRIPT at higher concentrations of CytoD. This suggested that displacement of capping proteins and other actin anchoring proteins was relevant for CytoD effects rather than the subsequent rearrangement of F-actin fibers which were more prominent at higher CytoD concentrations. Scaffolding of actin-binding proteins also seems to be involved in the modulation of ion

T

channel activity observed upon treatment of epithelial cells with cytochalasins [48]. The

IP

molecular mechanisms which lead to the increased phosphorylation of Smad2/3 by TGF-1

SC R

need further investigation and may add to the complexity of interactions between alterations of the cytoskeleton and signal transduction by plasma membrane receptors.

5. Conclusions

NU

It was the aim of our study to investigate the impact of alterations of the cytoskeleton on profibrotic signaling. We provide evidence for multiple levels of interaction, at the level of

MA

transcriptional activation as well in the regulation of cytosolic signaling. Most notable, our data with human primary tubular epithelial cells show that the outcome of reorganization of Factin fibers cannot be predicted without careful analysis of the underlying mechanisms.

D

Translated to the in vivo situation this means that, depending on the trigger, tubular injury will

CE P

TE

differentially contribute to interstitial injury.

Acknowledgements

We are grateful to B. Wullich and his team, Department of Urology, University of ErlangenNürnberg, for providing us with kidney tissue. This study was supported by the Deutsche

AC

Forschungsgemeinschaft (MU2737/2-2 to SM) and by the Department of Nephrology and Hypertension, University of Erlangen-Nürnberg.

15

ACCEPTED MANUSCRIPT References

AC

CE P

TE

D

MA

NU

SC R

IP

T

[1] M.A. Venkatachalam, K.A. Griffin, R. Lan, H. Geng, P. Saikumar, A.K. Bidani, Acute kidney injury: a springboard for progression in chronic kidney disease, Am J Physiol Renal Physiol, 298 (2010) F1078-1094. [2] M.A. Venkatachalam, J.M. Weinberg, W. Kriz, A.K. Bidani, Failed Tubule Recovery, AKICKD Transition, and Kidney Disease Progression, J Am Soc Nephrol, 26 (2015) 1765-1776. [3] Y. Fujigaki, Y. Muranaka, D. Sun, T. Goto, H. Zhou, M. Sakakima, H. Fukasawa, K. Yonemura, T. Yamamoto, A. Hishida, Transient myofibroblast differentiation of interstitial fibroblastic cells relevant to tubular dilatation in uranyl acetate-induced acute renal failure in rats, Virchows Arch, 446 (2005) 164-176. [4] I. Grgic, G. Campanholle, V. Bijol, C. Wang, V.S. Sabbisetti, T. Ichimura, B.D. Humphreys, J.V. Bonventre, Targeted proximal tubule injury triggers interstitial fibrosis and glomerulosclerosis, Kidney Int, 82 (2012) 172-183. [5] L. Yang, T.Y. Besschetnova, C.R. Brooks, J.V. Shah, J.V. Bonventre, Epithelial cell cycle arrest in G2/M mediates kidney fibrosis after injury, Nat Med, 16 (2010) 535-543 [6] E.K. Rajakyla, M.K. Vartiainen, Rho, nuclear actin, and actin-binding proteins in the regulation of transcription and gene expression, Small GTPases, 5 (2014) e27539. [7] S. Mouilleron, C.A. Langer, S. Guettler, N.Q. McDonald, R. Treisman, Structure of a pentavalent G-actin*MRTF-A complex reveals how G-actin controls nucleocytoplasmic shuttling of a transcriptional coactivator, Sci Signal, 4 (2011) ra40. [8] G. Posern, R. Treisman, Actin' together: serum response factor, its cofactors and the link to signal transduction, Trends Cell Biol, 16 (2006) 588-596. [9] E.N. Olson, A. Nordheim, Linking actin dynamics and gene transcription to drive cellular motile functions, Nat Rev Mol Cell Biol, 11 (2010) 353-365. [10] M.K. Vartiainen, S. Guettler, B. Larijani, R. Treisman, Nuclear actin regulates dynamic subcellular localization and activity of the SRF cofactor MAL, Science, 316 (2007) 17491752. [11] S. Muehlich, R. Wang, S.M. Lee, T.C. Lewis, C. Dai, R. Prywes, Serum-induced phosphorylation of the serum response factor coactivator MKL1 by the extracellular signalregulated kinase 1/2 pathway inhibits its nuclear localization, Mol Cell Biol, 28 (2008) 63026313. [12] K. Giehl, C. Keller, S. Muehlich, M. Goppelt-Struebe, Actin-mediated gene expression depends on RhoA and Rac1 signaling in proximal tubular epithelial cells, PLoS ONE, 10 (2015) e0121589. [13] J.P. Pradere, J. Gonzalez, J. Klein, P. Valet, S. Gres, D. Salant, J.L. Bascands, J.S. Saulnier-Blache, J.P. Schanstra, Lysophosphatidic acid and renal fibrosis, Biochim Biophys Acta, 1781 (2008) 582-587. [14] J. Zuehlke, A. Ebenau, B. Krueger, M. Goppelt-Struebe, Vectorial secretion of CTGF as a cell-type specific response to LPA and TGF-beta in human tubular epithelial cells, Cell Commun Signal, 10 (2012) 25. [15] E.M. De La Cruz, M.L. Gardel, Actin Mechanics and Fragmentation, J Biol Chem, 290 (2015) 17137-17144. [16] L. Lanzetti, P.P. Di Fiore, G. Scita, Pathways linking endocytosis and actin cytoskeleton in mammalian cells, Exp Cell Res, 271 (2001) 45-56. [17] M. Coue, S.L. Brenner, I. Spector, E.D. Korn, Inhibition of actin polymerization by latrunculin A, FEBS Lett., 213 (1987) 316-318. [18] G. Kustermans, J. Piette, S. Legrand-Poels, Actin-targeting natural compounds as tools to study the role of actin cytoskeleton in signal transduction, Biochem Pharmacol, 76 (2008) 1310-1322. [19] T. Wakatsuki, B. Schwab, N.C. Thompson, E.L. Elson, Effects of cytochalasin D and latrunculin B on mechanical properties of cells, J Cell Sci, 114 (2001) 1025-1036. [20] F. Miralles, G. Posern, A.I. Zaromytidou, R. Treisman, Actin dynamics control SRF activity by regulation of its coactivator MAL, Cell, 113 (2003) 329-342.

16

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

[21] S. Dupont, L. Morsut, M. Aragona, E. Enzo, S. Giulitti, M. Cordenonsi, F. Zanconato, J. Le Digabel, M. Forcato, S. Bicciato, N. Elvassore, S. Piccolo, Role of YAP/TAZ in mechanotransduction, Nature, 474 (2011) 179-183. [22] S. Dupont, Role of YAP/TAZ in cell-matrix adhesion-mediated signalling and mechanotransduction, Exp Cell Res, (2015) 42-53. [23] F. Liu, D. Lagares, K.M. Choi, L. Stopfer, A. Marinkovic, V. Vrbanac, C.K. Probst, S.E. Hiemer, T.H. Sisson, J.C. Horowitz, I.O. Rosas, L.E. Fredenburgh, C. Feghali-Bostwick, X. Varelas, A.M. Tager, D.J. Tschumperlin, Mechanosignaling through YAP and TAZ drives fibroblast activation and fibrosis, Am J Physiol Lung Cell Mol Physiol, 308 (2015) L344-357. [24] B.M. Gumbiner, N.G. Kim, The Hippo-YAP signaling pathway and contact inhibition of growth, J Cell Sci, 127 (2014) 709-717. [25] T. Shimomura, N. Miyamura, S. Hata, R. Miura, J. Hirayama, H. Nishina, The PDZbinding motif of Yes-associated protein is required for its co-activation of TEAD-mediated CTGF transcription and oncogenic cell transforming activity, Biochem Biophys Res Commun, 443 (2014) 917-923. [26] B. Zhao, X. Ye, J. Yu, L. Li, W. Li, S. Li, J.D. Lin, C.Y. Wang, A.M. Chinnaiyan, Z.C. Lai, K.L. Guan, TEAD mediates YAP-dependent gene induction and growth control, Genes Dev, 22 (2008) 1962-1971. [27] S. Kroening, E. Neubauer, B. Wullich, J. Aten, M. Goppelt-Struebe, Characterization of connective tissue growth factor expression in primary cultures of human tubular epithelial cells: modulation by hypoxia, Am J Physiol Renal Physiol, 298 (2010) F796-806. [28] J. Samarin, J. Wessel, I. Cicha, S. Kroening, C. Warnecke, M. Goppelt-Struebe, FoxO proteins mediate hypoxic induction of connective tissue growth factor in endothelial cells, J Biol Chem, 285 (2010) 4328-4336. [29] F. Preisser, K. Giehl, M. Rehm, M. Goppelt-Struebe, Inhibitors of oxygen sensing prolyl hydroxylases regulate nuclear localization of the transcription factors Smad2 and YAP/TAZ involved in CTGF synthesis, Biochim Biophys Acta, 1863 (2016) 2027-2036. [30] S. Kroening, E. Neubauer, J. Wessel, M. Wiesener, M. Goppelt-Struebe, Hypoxia interferes with connective tissue growth factor (CTGF) gene expression in human proximal tubular cell lines, Nephrol Dial Transplant, 24 (2009) 3319-3325. [31] C. Keller, S. Kroening, J. Zuehlke, F. Kunath, B. Krueger, M. Goppelt-Struebe, Distinct mesenchymal alterations in N-cadherin and e-cadherin positive primary renal epithelial cells, PLoS ONE, 7 (2012) e43584. [32] V. Hampl, C. Martin, A. Aigner, S. Hoebel, S. Singer, N. Frank, A. Sarikas, O. Ebert, R. Prywes, T. Gudermann, S. Muehlich, Depletion of the transcriptional coactivators megakaryoblastic leukaemia 1 and 2 abolishes hepatocellular carcinoma xenograft growth by inducing oncogene-induced senescence, EMBO Mol Med, 5 (2013) 1367-1382. [33] P.A. Janmey, R.G. Wells, R.K. Assoian, C.A. McCulloch, From tissue mechanics to transcription factors, Differentiation, 86 (2013) 112-120. [34] K. Mortensen, L.I. Larsson, Effects of cytochalasin D on the actin cytoskeleton: association of neoformed actin aggregates with proteins involved in signaling and endocytosis, Cell Mol Life Sci, 60 (2003) 1007-1012. [35] C. Ott, D. Iwanciw, A. Graness, K. Giehl, M. Goppelt-Struebe, Modulation of the expression of connective tissue growth factor by alterations of the cytoskeleton, J Biol Chem, 278 (2003) 44305-44311. [36] G. Halder, S. Dupont, S. Piccolo, Transduction of mechanical and cytoskeletal cues by YAP and TAZ, Nat Rev Mol Cell Biol, 13 (2012) 591-600. [37] E.G. Yarmola, T. Somasundaram, T.A. Boring, I. Spector, M.R. Bubb, Actin-latrunculin A structure and function. Differential modulation of actin-binding protein function by latrunculin A, J Biol Chem, 275 (2000) 28120-28127. [38] P. Muller, A. Langenbach, A. Kaminski, J. Rychly, Modulating the actin cytoskeleton affects mechanically induced signal transduction and differentiation in mesenchymal stem cells, PLoS ONE, 8 (2013) e71283. [39] S. Mana-Capelli, M. Paramasivam, S. Dutta, D. McCollum, Angiomotins link F-actin architecture to Hippo pathway signaling, Mol Biol Cell, 25 (2014) 1676-1685. 17

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

MA

NU

SC R

IP

T

[40] B. Zhao, X. Wei, W. Li, R.S. Udan, Q. Yang, J. Kim, J. Xie, T. Ikenoue, J. Yu, L. Li, P. Zheng, K. Ye, A. Chinnaiyan, G. Halder, Z.C. Lai, K.L. Guan, Inactivation of YAP oncoprotein by the Hippo pathway is involved in cell contact inhibition and tissue growth control, Genes Dev, 21 (2007) 2747-2761. [41] S.M. Thomasy, J.T. Morgan, J.A. Wood, C.J. Murphy, P. Russell, Substratum stiffness and latrunculin B modulate the gene expression of the mechanotransducers YAP and TAZ in human trabecular meshwork cells, Exp Eye Res, 113 (2013) 66-73. [42] G. Posern, A. Sotiropoulos, R. Treisman, Mutant actins demonstrate a role for unpolymerized actin in control of transcription by serum response factor, Mol.Biol.Cell, 13 (2002) 4167-4178. [43] R. Pawlowski, E.K. Rajakyla, M.K. Vartiainen, R. Treisman, An actin-regulated importin alpha/beta-dependent extended bipartite NLS directs nuclear import of MRTF-A, EMBO J, 29 (2010) 3448-3458. [44] E.K. Rajakyla, T. Viita, S. Kyheroinen, G. Huet, R. Treisman, M.K. Vartiainen, RNA export factor Ddx19 is required for nuclear import of the SRF coactivator MKL1, Nature communications, 6 (2015) 5978. [45] R. Panayiotou, F. Miralles, R. Pawlowski, J. Diring, H.R. Flynn, M. Skehel, R. Treisman, Phosphorylation acts positively and negatively to regulate MRTF-A subcellular localisation and activity, eLife, 5 (2016) 15460. [46] C. Baarlink, H. Wang, R. Grosse, Nuclear actin network assembly by formins regulates the SRF coactivator MAL, Science, 340 (2013) 864-867. [47] P. Kircher, C. Hermanns, M. Nossek, M.K. Drexler, R. Grosse, M. Fischer, A. Sarikas, J. Penkava, T. Lewis, R. Prywes, T. Gudermann, S. Muehlich, Filamin A interacts with the coactivator MKL1 to promote the activity of the transcription factor SRF and cell migration, Sci Signal, 8 (2015) ra112. [48] C. Mazzochi, D.J. Benos, P.R. Smith, Interaction of epithelial ion channels with the actinbased cytoskeleton, Am J Physiol Renal Physiol, 291 (2006) F1113-1122.

18

ACCEPTED MANUSCRIPT Legends to the figures

Fig. 1: Activation of Rho kinases contributes to LatB- and CytoD-mediated changes in F-actin structures and CTGF secretion.

T

A: Confluent hPTEC were preincubated with Y27632 (10 µM) for 30 min and then treated

IP

with LatB (0.5 µM) or CytoD (0.5 µM) for 2 h. F-actin was visualized by rhodamine-phalloidin.

SC R

Scale bar: 40 µm.

B: pMYPT was detected in cellular homogenates of cells treated as in A for 2 h. LatB (LA), CytoD (CD). The graph combines data obtained in n=7 experiments with proximal and distal cells, respectively. **p<0.01, ***p<0.001 compared to control cells.

NU

C: CTGF was detected in cell cultures supernatants of proximal and distal hPTEC. Samples were detected on one blot which had to be rearranged (dotted lines). The graph summarizes

MA

data obtained with n=5 prox hPTEC and n=3 distal hPTEC preparations. In each experiment the release obtained with LatB-treated cells was set to 1. **p<0.01, ***p<0.001 compared to control cells.

D: CTGF was detected in cell culture supernatants (secreted CTGF, sCTGF) and in the

D

homogenates (cellular CTGF, cCTGF) of hPTEC treated as in A. Data obtained with proximal

TE

and distal cells were combined. Secretion in control cells, LatB- and CytoD-treated cells was

CE P

set to 1. **p<0.01 compared to the respective supernatant in the absence of Y27632, n=3.

Fig. 2: Nuclear localization of YAP/TAZ is dependent on cell density, but not increased upon CytoD treatment

A: YAP and TAZ were detected in cytosolic and nuclear fractions of confluent (c) and

AC

subconfluent (sc) cells. Expression of YAP and TAZ was corrected for buffer volumes used in subcellular fractionation. The percentage of nuclear YAP/TAZ in confluent cells was set to 1 in each experiment, n=3, ** p<0.01, one-sided t-test. In dense cells, 15 + 3 % YAP and 19 + 4 % TAZ were localized to the nucleus (n= 4 preparations). B: Confluent hPTEC were stained with N-cadherin to detect cells of proximal origin. YAP/TAZ localization was analyzed by indirect immunocytochemistry. The arrow indicates dense proximal cells. Scale bar: 20 µm. C: Polarized distal hPTEC were wounded with a pipet tip. Cells migrated into the open space for 6 h. YAP/TAZ were detected by immunocytochemistry. The image is representative of 3 independent experiments with comparable results. D: YAP/TAZ were detected tissue section of human renal cortex by immunocytochemistry. The image shown is representative for tissue obtained from three different donors. The antibody used (SC -101199) detected YAP and TAZ. A comparable localization was also observed with a YAP-specific antibody (#4912, Cell Signaling). Scale bar: 30 µm. 19

ACCEPTED MANUSCRIPT E: Nuclear and cytosolic fractions of proximal and distal hPTEC preparations were analyzed for YAP and TAZ expression as indicated. Stimulation with LatB (La) and CytoD (CD) was for 2 h. In each set, expression in control cells was set to 1. n=4-5 experiments, **p<0.01. F: hPTEC were transfected with siYAP or siGFP for 48 h and then stimulated with LatB or

T

CytoD for 2 h. Samples were detected on one blot which had to be rearranged (dotted line).

IP

The graph summarizes the data of n=3 experiments. Secretion in LatB-treated cells was set

SC R

to 1 in each experiment. p<0.001, siYAP vs siGFP-treated cells, ANOVA with Bonferroni post hoc correction.

Fig. 3 Modified MKL1 is translocated to the nuclear compartment upon CytoD

NU

treatment in hPTEC

A: MKL1 and E-cadherin (E-Cad) were detected in confluent hPTEC treated with LatB or

MA

CytoD for 2 h. Scale bar: 20 µm.

B: MKL1 was detected in the nuclear and cytosolic fraction of cells treated with LatB (La) or CytoD (CD) for 2 h. The graph summarizes data obtained with n=4 nuclear preparations. Nuclear MKL1 in control cells was set to 1 in each experiment, *p<0.05.

D

C: Upper panel: Cells were pre-treated with Y27632 (10 µM) and U0126 (1 µM) for 30 min

TE

and then stimulated with CytoD for the times indicated. Data are representative of two independent experiments. Lower panel: RhoA was transiently downregulated by siRNA.

CE P

Cells were stimulated with CytoD for 2 h (CD). For comparison, LPA (10 µM) was included in the experiment. siGFP- and siRhoA-treated samples were detected on one blot that had to be rearranged.

D: hPTEC were treated with 2 different siRNAs directed against MKL1 for 48 h and then

AC

stimulated for 2 h with CytoD (CD). Samples were detected on one blot which had to be rearranged (dotted line). MKL1 was detected in the cellular homogenates with vinculin as loading control. Secreted CTGF was related to cellular protein as surrogate marker of cells numbers. The graph summarizes n=5 experiments, data with both siRNAs being combined. CTGF secretion from CytoD-stimulated cells was set to 1 in each experiment. ***p<0.001.

Fig. 4: Synergistic basolateral secretion of CTGF by CytoD and TGF-1 in polarized hPTEC A: Distal hPTEC were polarized in transwell inserts for 10 days and then treated with CytoD (0.5 µM) or LatB (0.5 µM) for 2 h from the apical side. F-actin is shown along the x-z axis, and in the x-y level along the middle level of the nuclei. Scale bar: 10 µm. B: pMYPT was detected in cellular homogenates of cells treated as in A for 2 h. pMYPT in control cells was set to 1 in each experiment, n=3, **p<0.01.

20

ACCEPTED MANUSCRIPT C: MKL1 was detected in polarized cells by immunocytochemistry. The arrows indicate nuclear MKL1 in control cells. Scale bar: 10 µm. D: CTGF was precipitated from the upper compartment of polarized cells (apical secretion). Data are means + SD of 6 experiments. In each experiment, release from CytoD-treated

T

cells was set to 1. ***p<0.001, ANOVA with Tukey multiple comparison test; **p<0.01, two-

IP

sided paired t-test, compared to control cells.

SC R

E: Polarized hPTEC were treated with CytoD or LatB for one hour from the apical side and then stimulated with TGF-1 (5 ng/ml) from the basolateral side for 6 h. CTGF was detected in the apical and basolateral compartment, respectively. The blot is representative of 3

NU

experiments.

Fig. 5: CytoD facilitates activation of Smad2/3 by TGF-1

MA

A: Proximal and/or distal cells were pretreated with LatB (La) and CytoD (CD) for 1 h and then stimulated with TGF-1 (2 ng/ml) for 6 h. CTGF was detected in the cell culture supernatant by Western blotting. The graph summarizes the data of n=7 experiments; data

D

obtained with proximal and distal cells were combined. Secretion of CTGF from CytoDtreated cells was set to 1 in each experiment. #p<0.05 compared to control cells. ***p<0.01

TE

as indicated.

B: pSmad 2/3 was detected in nuclear fractions of cells pretreated with CytoD for 1 h and

CE P

then stimulated with TGF-1 (2 ng/ml) for 2 h. The graph summarizes data obtained in 3-4 experiments. Expression of pSmad in samples stimulated with TGF-1 plus CytoD was set to 1 in each experiment and compared to samples stimulated with only TGF-1. **p<0.01, one

AC

sample t-test.

21

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

22

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

23

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

24

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

25

AC

CE P

TE

D

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

26

MA

NU

SC R

IP

T

ACCEPTED MANUSCRIPT

AC

CE P

TE

D

Graphical abstract

27

ACCEPTED MANUSCRIPT Highlights

CE P

TE

D

MA

NU

SC R

IP

T

Both, Latrunculin B and Cytochalasin D reorganize F-actin into globular structures. Only Cytochalasin D leads to a strong induction of CTGF. Cytochalasin D modifies MKL1 and increases its nuclear localization YAP/TAZ are essential for CTGF synthesis, but not affected by Cytochalasin D. Cytochalasin D promotes TGF-1-mediated phosphorylation of Smad2/3.

AC

    

28