PI3K as Mediator of Apoptosis and Contractile Dysfunction in TGFβ1-Stimulated Cardiomyocytes

Simple Summary TGFβ1 is a growth factor that plays a major role in the remodeling process of the heart by inducing cardiomyocytes dysfunction and apoptosis, as well as fibrosis, thereby restricting heart function. TGFβ1 mediates its effect via the TGFβ receptor I (ALK5) and the activation of SMAD transcription factors. But, TGFβ1 is also known as activator of phosphoinositide-3-kinase (PI3K) via the non-SMAD signaling pathway. The aim of this study was to investigate whether PI3K is also involved in TGFβ1–induced cardiomyocytes apoptosis and contractile dysfunction. Pharmacological inhibition of PI3K with Ly294002 reduced TGFβ-induced apoptosis and reduced cell shortening. Inhibition of the PI3Kγ isoform also abolished the TGFβ effect on apoptosis and cell shortening. These data support a role for a PI3K and ALK5/SMAD pathway in TGFβ1-induced apoptosis and impaired cell shortening, which in part appears to be PI3Kγ-dependent. Abstract Background: TGFβ1 is a growth factor that plays a major role in the remodeling process of the heart by inducing cardiomyocyte dysfunction and apoptosis, as well as fibrosis thereby restricting heart function. TGFβ1 mediates its effect via the TGFβ receptor I (ALK5) and the activation of SMAD transcription factors, but TGFβ1 is also known as activator of phosphoinositide-3-kinase (PI3K) via the non-SMAD signaling pathway. The aim of this study was to investigate whether PI3K is also involved in TGFβ1–induced cardiomyocytes apoptosis and contractile dysfunction. Methods and Results: Incubation of isolated ventricular cardiomyocytes with TGFβ1 resulted in impaired contractile function. Pre-incubation of cells with the PI3K inhibitor Ly294002 or the ALK5 inhibitor SB431542 attenuated the decreased cell shortening in TGFβ1–stimulated cells. Additionally, TGFβ-induced apoptosis was significantly reduced by the PI3K inhibitor Ly294002. Administration of a PI3Kγ-specific inhibitor AS605240 abolished the TGFβ effect on apoptosis and cell shortening. This was also confirmed in cardiomyocytes from PI3Kγ KO mice. Induction of SMAD binding activity and the TGFβ target gene collagen 1 could be blocked by the PI3K inhibitor Ly294002, but not by the specific PI3Kγ inhibitor AS605240. Conclusions: TGFβ1-induced SMAD activation, cardiomyocyte apoptosis, and impaired cell shortening are mediated via both, the ALK5 receptor and PI3K, in adult cardiomyocytes. PI3Kγ specifically contributes to apoptosis induction and impairment of contractile function independent of SMAD signaling.


Introduction
The transition of compensated hypertrophy to heart failure is accompanied by an increased expression of transforming growth factor β (TGFβ) [1,2] that contributes to

Isolation of Murine Cardiomyocytes
Mice were anaesthetized by an inhalation of 5% isoflurane followed by the euthanasia with the cervical dislocation. The hearts were extracted and rinsed with 4 • C cold 0.9% NaCl. Thereafter, the hearts were digested with retrograde perfusion in a Langendorff apparatus for 25 min containing collagenase and calcium-free buffer (in mmol/L: 10 glucose monohydrate D + , 25 HEPES, 2.5 KCl, 1.2 KH 2 PO 4 , 1.2 MgSO 4 × 7 H 2 O, 110 NaCl) with a pH of 7.4 at 37 • C. Subsequently, the tissue was minced and incubated for another five minutes in the recirculating buffer. The suspension was filtered and the cardiomyocytes were separated from the other cells by centrifugation. To reconstruct the physiological calcium concentration the material was resuspended with an step-wise increase of calcium up to 1 mM followed by resuspension in the culture medium (in mmol/L: 2.5 CaCl 2 -dihydrate, 5 glucose, 10 HEPES, 4.7 KCl, 1.2 KH 2 PO 4 , 0.8 MgSO 4 , 118 NaCl and 1.9 Na-pyruvate). Finally, the cardiomyocytes were plated on laminin-coated (5 µg/mL) culture dishes and incubated at 37 • C (5.5% CO 2 , 95% humidity). After 1 h, the cells were washed with fresh culture medium and could be then stimulated. The cardiomyocytes were treated with TGFβ 1 . Times of stimulation are given for the respective experiments in the figure legends. TGFβ 1 concentrations will be given in the figure legends. Untreated samples served as a control.

Isolation of Rat Cardiomyocytes
The isolation of ventricular cardiomyocytes from 200 to 250 g male Wistar rats were performed similar to the isolation of murine cardiomyocytes. The cells were suspended in culture medium and subsequently plated onto preincubated culture dishes (4% fetal calf serum in medium 199, overnight). Two hours later, the cells were washed twice with fresh CCT medium (modified medium 199 including Earle's salts, 2 mM l-carnitine, 5 mmol/L taurine, 100,000 IU/L penicillin, 100 mg/L streptomycin and 10 µmol/L cytosineβ-d-arabinofuranoside (pH 7.4). On average, about 90% of the cultures were quiescent rod-shaped cells. Subsequent to the cell washing, they were treated with 1 ng TGFβ 1 . Untreated samples served as controls.

Determination of Cell Function
Cell contraction behavior was analysed at room temperature by using a cell-edgedetection system. The cultured cardiomyocytes were stimulated by using two AgCl electrodes with biphasic electrical stimuli constituted out of two equal but opposite rectangular 50 V stimuli of 0.5 ms duration. Each cell was stimulated with a voltage of 2 Hz and was measured four times. To calculate the cell shortening of a given cell, the mean of these measurements was used. In addition, cell lengths could be analysed by using a line camera, recording at 500 Hz. The resulting data are presented as cell shortening normalized to diastolic cell length (dL/L (%)).

Electrophoretic Mobility Shift Assay (EMSA)
For the extraction of nuclear extracts, isolated cardiomyocytes were homogenized in swelling buffer (10 mM Tris-HCl, pH 7.9, 1 mM MgCl 2 , 10 mM KCl, 1 mM DTT). Thereafter, the homogenates were placed on ice for one hour and nuclei were pelleted by centrifugation at 900 rpm for ten minutes. Supernatants were used for Western blots. Pellets were homogenized in 10 mM Tris-HCl, pH 7.9, 300 mM sucrose, 1mM DTT, 1.5 mM MgCl 2 , 0.3% triton X-100 and another time centrifuged as above. These pellets were resuspended in storage buffer (10 mM HEPES, pH 7.5, 50 mM KCl, 300 mM NaCl, 1 mM EDTA, 1 mM DTT, 1 mM PMSF, 20% glycerol) on ice for 30 min and centrifuged at 13,000 rpm, 4 • C, for five minutes. For electrophoretic mobility shift assays (EMSAs) the resulting supernatants were taken. Complementary sequences of SBE (SMAD binding elements) oligonucleotides were hybridized. 5 -GTACAT TGTCAGTCTAGACATACT-3 was the sequence for binding of SMAD transcription factors [3]. In the presence of Cy3-dCTP, oligonucleotides were incubated with terminal transferase. Using gel filtration. Unincorporated nucleotides were Biology 2021, 10, 670 4 of 13 parted from fluorescence-labeled oligonucleotides. Ten µL nuclear extracts were incubated with labeled oligonucleotides in the presence of one µg poly (dIdC) at 30 • C for 30 min. The samples were run on 4% native polyacrylamide gels. Complete gels were exposed on fluorescence imager (BioRad, Hercules, CA, USA).

Real-Time RT-PCR
According to the instructions by the manufacturer, total RNA of cardiomyocytes was extracted with Trizol (Invitrogen, Hercules, CA, USA). Reverse transcription was done for one hour at 37 • C in a final volume of 10 µL using 1 µg of RNA, eight units of RNasin, 1 mol/L dNTPs, 100 ng of oligo (dT) 15 and 60 units of Moloney murine leukemia virus reverse transcriptase. Aliquots (1.5 µL) of the synthesized cDNA were used for PCR in a final volume of 10 µL including primer pairs at 1.5 µmol/L, 1.5 mol/L MgCl 2 , 0.4 mol/L dNTPs, and 1 unit of Taq polymerase. For each tested gene, annealing temperature, and the number of cycles resulting in a linear amplification range were tested. RT-PCR was run in an automated thermal cycler. SYBR Green fluorescence was used for quantification and detected with the Bio-Rad detection system (Bio-Rad). The calculations of the results were done according to the 2 −∆∆Ct methods as described [36]. Determination of the "most stable" housekeeping genes in real-time RT-PCR was performed by using repeated pair-wise correlation analysis of GAPDH 18sRNA, B2M, and HPRT [37] as listed in Table 1.

Caspase 3/7 Assay
The kit "Caspase-Glo 3/7" (Promega, Mannheim, Germany) was used for the measurement of caspase activity as recommended by the company. In brief, cardiomyocytes were stimulated with TGFβ 1 for six hours. Thereafter, a caspase-3/7 pro-luminescent substrate was added to the cells. This led to cell lysis and caspase cleavage of the substrate.
Using a spectrophotometer the generation of a "glow-type" luminescent signal could be measured. The values obtained were normalised to the amount of protein per well.

Statistical Analysis
The SPSS programme (IBM SPSS Statistics 27, IBM Deutschland GmbH, Ehningen, Germany) was used for statistical data analysis. Results are expressed as mean ± standard error of the mean (SEM) as indicated in the figure legends. A one-way ANOVA was applied followed by Student-Newman-Keuls post hoc test, and the Mann-Whitney-U test to further evaluate differences between two means. Differences among multiple groups were evaluated using a two-way analysis of variance (ANOVA) for the two factors genotype (WT versus PI3Kγ knock-out) and treatment (with/without TGFβ 1 ), followed by Tukey's post hoc test. Values of p < 0.05 were marked to be statistically significant.

TGFβ 1 -Induced Apoptosis Depends on PI3K
To study the impact of PI3K on TGFβ 1 -induced apoptosis, adult rat cardiomyocytes were treated with TGFβ 1 and the PI3K inhibitor Ly294002. Increase of caspase 3/7 activation by TGFβ 1 was reduced by the administration of Ly294002 ( Figure 1).

Caspase 3/7 Assay
The kit "Caspase-Glo 3/7" (Promega, Mannheim, Germany) was used for the measurement of caspase activity as recommended by the company. In brief, cardiomyocytes were stimulated with TGFβ1 for six hours. Thereafter, a caspase-3/7 pro-luminescent substrate was added to the cells. This led to cell lysis and caspase cleavage of the substrate. Using a spectrophotometer the generation of a "glow-type" luminescent signal could be measured. The values obtained were normalised to the amount of protein per well.

Statistical Analysis
The SPSS programme (IBM SPSS Statistics 27, IBM Deutschland GmbH, Ehningen, Germany) was used for statistical data analysis. Results are expressed as mean ± standard error of the mean (SEM) as indicated in the figure legends. A one-way ANOVA was applied followed by Student-Newman-Keuls post hoc test, and the Mann-Whitney-U test to further evaluate differences between two means. Differences among multiple groups were evaluated using a two-way analysis of variance (ANOVA) for the two factors genotype (WT versus PI3Kγ knock-out) and treatment (with/without TGFβ1), followed by Tukey's post hoc test. Values of p < 0.05 were marked to be statistically significant.

TGFβ1-Induced Apoptosis Depends On PI3K
To study the impact of PI3K on TGFβ1-induced apoptosis, adult rat cardiomyocytes were treated with TGFβ1 and the PI3K inhibitor Ly294002. Increase of caspase 3/7 activation by TGFβ1 was reduced by the administration of Ly294002 ( Figure 1).

Figure 1.
TGFβ1-induced apoptosis was determined with and without addition of the PI3K inhibitor Ly294002 (10 µM) by caspase 3/7 activation after six hours stimulation with 1 ng/mL TGFβ1. Data are shown as means ± SEM of 11 different rat cardiomyocytes preparations. Significance is indicated by * p < 0.05.

Involvement of PI3K in TGFβ1-Dependent SMAD Binding Activity and TGFβ Target Genes
TGFβ1-induced apoptosis is mediated via SMAD transcription factors. To evaluate whether PI3K is involved in enhanced SMAD binding activity under TGFβ stimulation, electromobility shift assays were performed. Increase in SMAD binding activity could be abolished by inhibition of PI3K with Ly294002 ( Figure 2).

Involvement of PI3K in TGFβ 1 -Dependent SMAD Binding Activity and TGFβ Target Genes
TGFβ 1 -induced apoptosis is mediated via SMAD transcription factors. To evaluate whether PI3K is involved in enhanced SMAD binding activity under TGFβ stimulation, electromobility shift assays were performed. Increase in SMAD binding activity could be abolished by inhibition of PI3K with Ly294002 ( Figure 2). Analysis of two TGFβ target genes revealed an increase in expression upon TGFβ stimulation. Whereas SMAD7 expression was not affected, collagen 1 mRNA expression was down-regulated by the PI3K inhibitor Ly294002 (Figure 3).  Analysis of two TGFβ target genes revealed an increase in expression upon TGFβ stimulation. Whereas SMAD7 expression was not affected, collagen 1 mRNA expression was down-regulated by the PI3K inhibitor Ly294002 (Figure 3).

PI3K Affected TGFβ1-Dependent Depression of Contractile Function
Isolated cardiomyocytes of adult rats responded with a decrease in cell shortening upon TGFβ1 stimulation. Pre-incubation of the cells with Ly294002 abolished this effect ( Figure 4A). In addition, pre-incubation of cardiomyocytes with TGFβ receptor I (ALK5) inhibitor SB431542, similarly inhibited the impairment of cardiomyocytes contractile function by TGF β ( Figure 4B) demonstrating that both PI3K and SMADs mediate contractile dysfunction of cardiomyocytes under TGFβ stimulation.

PI3K Affected TGFβ 1 -Dependent Depression of Contractile Function
Isolated cardiomyocytes of adult rats responded with a decrease in cell shortening upon TGFβ 1 stimulation. Pre-incubation of the cells with Ly294002 abolished this effect ( Figure 4A). In addition, pre-incubation of cardiomyocytes with TGFβ receptor I (ALK5) inhibitor SB431542, similarly inhibited the impairment of cardiomyocytes contractile function by TGFβ ( Figure 4B) demonstrating that both PI3K and SMADs mediate contractile dysfunction of cardiomyocytes under TGFβ stimulation.
Analysis of two TGFβ target genes revealed an increase in expression upon TGFβ stimulation. Whereas SMAD7 expression was not affected, collagen 1 mRNA expression was down-regulated by the PI3K inhibitor Ly294002 (Figure 3).

PI3K Affected TGFβ1-Dependent Depression of Contractile Function
Isolated cardiomyocytes of adult rats responded with a decrease in cell shortening upon TGFβ1 stimulation. Pre-incubation of the cells with Ly294002 abolished this effect ( Figure 4A). In addition, pre-incubation of cardiomyocytes with TGFβ receptor I (ALK5) inhibitor SB431542, similarly inhibited the impairment of cardiomyocytes contractile function by TGF β ( Figure 4B) demonstrating that both PI3K and SMADs mediate contractile dysfunction of cardiomyocytes under TGFβ stimulation.

PI3Kγ Is Involved in TGFβ 1 -Dependent Depression of Contractile Function and in TGFβ-Dependent Apoptosis Induction
Since PI3Kγ is highly expressed in cardiomyocytes, we now analyzed, if this specific isoform contributes to TGFβ-induced contractile dysfunction. Pharmacological inhibition of PI3Kγ by AS605240 did not affect baseline cardiomyocytes shortening but led to a reduction of TGFβ 1 -dependent depression of cardiomyocytes shortening ( Figure 5A). This was validated in cardiomyocytes from PI3Kγ KO mice. Whereas WT cardiomyocytes responded to TGFβ stimulation with a decrease in cell shortening, PI3Kγ cardiomyocytes were not affected by TGFβ ( Figure 5B). This demonstrates that PI3Kγ is involved in the impairment of contractile function under TGFβ stimulation.
of PI3Kγ by AS605240 did not affect baseline cardiomyocytes shortening but led to a reduction of TGFβ1-dependent depression of cardiomyocytes shortening ( Figure 5A). This was validated in cardiomyocytes from PI3Kγ KO mice. Whereas WT cardiomyocytes responded to TGFβ stimulation with a decrease in cell shortening, PI3Kγ cardiomyocytes were not affected by TGFβ ( Figure 5B). This demonstrates that PI3Kγ is involved in the impairment of contractile function under TGFβ stimulation. Since PI3Kγ influences cardiomyocyte function, we wanted to know whether this isoform also has an influence on apoptosis. Pharmacological inhibition of PI3Kγ by AS605240 blocked TGFβ1-induced caspase 3/7 activation ( Figure 6). Since PI3Kγ influences cardiomyocyte function, we wanted to know whether this isoform also has an influence on apoptosis. Pharmacological inhibition of PI3Kγ by AS605240 blocked TGFβ 1 -induced caspase 3/7 activation ( Figure 6). sponded to TGFβ stimulation with a decrease in cell shortening, PI3Kγ cardiomyocytes were not affected by TGFβ ( Figure 5B). This demonstrates that PI3Kγ is involved in the impairment of contractile function under TGFβ stimulation. Since PI3Kγ influences cardiomyocyte function, we wanted to know whether this isoform also has an influence on apoptosis. Pharmacological inhibition of PI3Kγ by AS605240 blocked TGFβ1-induced caspase 3/7 activation ( Figure 6). Figure 6. Cardiomyocytes of adult rats were stimulated with 1 ng/mL TGFβ 1 for 6 h, with and without addition of PI3Kγ inhibitor AS605240 (1 µM). Quantitative determination of caspase 3/7 activation (n = 14-16 independent cardiomyocytes preparations). Data are shown as means ± SEM. Significance is indicated by *** p < 0.001.

PI3Kγ Action Is Independent of SMAD Activation
Further, we analysed whether the influence of the TGFβ 1 -dependent effects of PI3Kγ were linked to the canonical TGFβ signalling pathway. SMAD activation demonstrated by phosphorylation of SMAD2 revealed no inhibition of TGFβ-induced SMAD2 activation by the use of the PI3Kγ-specific inhibitor AS605240. (Figure 7A,B). Addition of the PI3Kγ inhibitor AS605240 did not block SMAD binding activity ( Figure 7C). Also, expression of the TGFβ target genes SMAD7 and collagen 1 were not affected by inhibition of PI3Kγ ( Figure 7D,E).
Further, we analysed whether the influence of the TGFβ1-dependent effects of PI3Kγ were linked to the canonical TGFβ signalling pathway. SMAD activation demonstrated by phosphorylation of SMAD2 revealed no inhibition of TGFβ-induced SMAD2 activation by the use of the PI3Kγ-specific inhibitor AS605240. (Figures 7A,B). Addition of the PI3Kγ inhibitor AS605240 did not block SMAD binding activity ( Figure 7C). Also, expression of the TGFβ target genes SMAD7 and collagen 1 were not affected by inhibition of PI3Kγ ( Figures 7D,E).

Discussion
We have identified PI3K as a further signalling molecule involved in TGFβ-mediated decrease in cell shortening as well as apoptosis in cardiomyocytes. That TGFβ is capable of decreasing cell shortening was demonstrated in adult rat cardiomyocytes stimulated with AngII [6], which induces activation and release of TGFβ [38]. Addition of neutralizing TGFβ antibodies into the culture medium of AngII-stimulated cardiomyocytes improved their function [6]. Banerjee et al. [39] also demonstrated that TGFβ decreased contractile function in embryonic cardiomyocytes of mice in a SMAD-dependent manner.

Discussion
We have identified PI3K as a further signalling molecule involved in TGFβ-mediated decrease in cell shortening as well as apoptosis in cardiomyocytes. That TGFβ is capable of decreasing cell shortening was demonstrated in adult rat cardiomyocytes stimulated with AngII [6], which induces activation and release of TGFβ [38]. Addition of neutralizing TGFβ antibodies into the culture medium of AngII-stimulated cardiomyocytes improved their function [6]. Banerjee et al. [39] also demonstrated that TGFβ decreased contractile function in embryonic cardiomyocytes of mice in a SMAD-dependent manner. Isolated cardiomyocytes of TGFβ-overexpressing mice had an attenuated cell shortening [7], which was accompanied by decreased mitochondrial energy metabolism due to increased expression of mitochondrial uncoupling proteins (UCP). However, none of the above studies proposed an involvement of PI3K in TGFβ-mediated contractile dysfunction.
Both beneficial and detrimental roles in the regulation of cardiomyocyte contractility have already been shown for PI3K. Enhancement of PI3Kα or reduction of PI3Kγ as a therapeutic approach in heart disease has been suggested [40]. Accordingly, overexpression of cardiac-specific PI3Kα resulted in increased contractility in mice [41], whereas cardiac-specific ablation of p110α-the catalytic subunit of PI3Kα-in adult mice caused significant reductions in voltage-dependent L-type Ca 2+ channel density, Ca 2+ transients, and cardiomyocyte contraction that resulted in compromised cardiac contractility [42]. In contrast, targeted deletion of PI3Kγ significantly increased contractility [35,43] and enhanced cardiac function as assessed by increased fractional shortening [27,35]. Additionally, mice expressing a kinase-dead PI3Kγ (PI3K KD ) displayed protection from maladaptive remodeling and cardiac dysfunction up to 16 weeks after aortic constriction [44]. Also in the present study, supporting previous findings, PI3Kγ KO improved cardiomyocytes shortening under basal conditions.
In contrast to the PI3Kγ KO cardiomyocytes, pharmacological PI3Kγ inhibition itself had no effect on the cardiomyocytes function of WT mice. However, the inhibition of PI3Kγ with AS605240 could reverse TGFβ-induced depression of cell shortening as seen also in PI3Kγ KO cardiomyocytes. Cardiomyocytes from PI3Kγ KO mice already have higher cell shortening due to cAMP elevation than WT mice already at baseline [34], and therefore TGFβ 1 probably could not influence contractility in PI3Kγ KO cardiomyocytes. Our findings on differences in contractility depending on the complete KO of PI3Kγ or inhibition of PI3Kγ activity by AS605240 has been seen comparing PI3Kγ −/− to PI3Kγ KD mice. In contrast to PI3Kγ KO mice, the loss of kinase activity of PI3Kγ KD did not lead to a change in contractility [34].
The TGFβ and also the PI3K signalling pathway are important for the organism to control various cell responses, such as apoptosis. Many studies suggest that these two pathways may counteract each other, especially in cell survival where PI3K has a protective role via activation of AKT (for a review see Kandel and Hay [45]. Although PI3K signaling usually improves survival in combination with SMAD signaling this effect is reversed. The decisive factor seems to be the extent to which SMAD or AKT is in the predominance [46,47]. There is also evidence that both pathways work in the same direction, for example, for progression of tumour cells [48]. In addition to SMAD activation [3], we now demonstrate that PI3K is also necessary for TGFβ-induced apoptosis. This is supported by Bakin et al. [49], who showed an interaction between PI3K and TGFβ. In epithelial cells, inhibition of PI3K led to blockage of TGFβ-induced SMAD2 activation, which thereby regulated signaling molecules of TGFβ-induced apoptosis. In addition, Conery et al. [47] identified a crosstalk between the TGFβ and PI3K pathways in different cell lines that led to apoptosis. This study showed that with an increase in the amount of SMAD3 translocating into the nucleus, apoptosis induction was enhanced. We also could decrease SMAD translocation by inhibition of PI3K demonstrated by EMSA. This is in line with other studies showing PI3K/SMAD signaling being harmful for the heart [50,51].
Recently, Aki et al. [52] elucidated the crucial role of PI3K for TGFβ signaling in vascular endothelial cells. PI3Kα knockdown nearly completely abolished TGFβ 1 -induced phosphorylation and nuclear translocation of SMAD2/3-the canonical transcription factors downstream of TGFβ. Our study now identified PI3K as another down-stream signaling molecule of TGFβ-induced apoptosis in cardiomyocytes with PI3Kγ being the important isoform. Other studies confirm that PI3K modulates the SMAD signalling pathway and thus apoptosis. In the present study we did not see an influence of PI3Kγ on SMADS, although PI3Kγ did influence apoptosis. This suggests that PI3Kγ acts independently of SMADs.
SMAD transcription factors are the canonical mediators of TGFβ signaling, where SMAD2 and SMAD3 are phosphorylated upon receptor activation and SMAD7 acts as an inhibitory SMAD. Whereas SMAD7 transcription could be abolished by SB431542 [3], the inhibition of PI3K had no effect on SMAD7 transcription. Our findings fit quite well with the studies by Edlund et al. [53], here pretreatment with the PI3K inhibitor LY294002 did not abolish the Smad7-induced activation of Cdc42 in human prostate cancer cell line (PC-3U). They stated that Smad7 has probably different roles in addition to its inhibitory function in TGFβ signaling which suggests that there is a parallel signaling pathway. In contrast to SMAD7, the TGFβ-induced upregulation of collagen 1 expression could be blocked by inhibition of PI3K. This could also be measured in human mesangial cells, where the PI3K inhibitor LY294002 decreased TGFβ-stimulated Col1 mRNA expression by blocking induction of COL1A2 promoter activity [54]. This supports the notion that the TGFβ-induced increase in collagen 1 expression is PI3K-dependent. However, in the present study, PI3Kγ had no effect on collagen 1 expression, which was also supported by the fact that neither SMAD translocation nor SMAD2 phosphorylation could be inhibited.
TGFβ-induced upregulation of collagen-1 expression may be mediated via another PI3K isoform that has not yet been identified.

Conclusions
Our data support a role for a PI3K and ALK5/SMAD pathway in TGFβ 1 -induced apoptosis and impaired cell shortening. More specifically, the influence on apoptosis and contractile dysfunction of TGFβ 1 stimulation in part appears to be PI3Kγ-dependent.