Bachelor thesis
Microtubule remodeling in the heart as a possible
cause for dysfunction of Nav1.5
Effects of taxol and parthenolide on dynamic instability of
microtubules associated with altered sodium currents
ABSTRACT
Dysfunction of the main cardiac sodium channel Nav1.5 frequently leads to conduction diseases and sudden cardiac death. Besides the well-studied causes of Nav1.5 dysfunction, such as loss-of-function and gain-of-loss-of-function mutations of SCN5A that alter the biophysical properties of Nav1.5, recent research has focused on the role of microtubule remodeling. Previous work observed alterations in sodium currents in neonatal rat cardiomyocytes and HEK293 cells following treatment with the microtubule stabilizing agent taxol and the microtubule destabilizing agent parthenolide (PTL). Taxol and PTL are well known anticancer drugs. These measurements were obtained using patchclamp techniques, but this data was not supported by molecular investigation of microtubule remodeling. In this thesis report the possible effects of microtubule remodeling on cardiac sodium channel (Nav1.5) function will be studied by investigating microtubule remodeling mediated by taxol and PTL in three cell models during the conditions in which the altered sodium currents were measured. Detyrosinated--tubulin is an indicator for microtubule remodeling, because taxol should increase the amount of detyrosination of -tubulin, whilst PTL should have the opposite effect. These effects are well studied in human HeLa cells and mouse fibroblasts but are not described in HEK293 cells, human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) or mouse cardiomyocytes during the conditions in which the functional measurements were conducted. Therefore, in three models (HEK293 cells, hiPSC-CMs and mouse cardiomyocytes), cells were treated with taxol and PTL in the same conditions as the cells that were used for the functional measurements and immunolabeled for -tubulin and detyrosinated--tubulin. -Tubulin functions as an indicator for the normal structure of the microtubule network and is used to quantify the amount of detyrosination by calculating a mean ratio of the fluorescence intensity of the detyrosinated and normal microtubule network. Stainings from this study showed that the amount and density of detyrosinated--tubulin in HEK293 cells and hiPSC-CMs were increased upon taxol treatment. Treatment with PTL in HEK293 cells and hiPSC-CMs resulted in a disturbed structure of the detyrosinated microtubule network, instead of the expected decrease in amount and density of the detyrosinated--tubulin. It can be concluded that taxol mediates microtubule remodeling and that PTL affects the detyrosinated microtubule structure. These data partly complement the functional data from previous work that demonstrate a decreased sodium current upon taxol treatment and an increased sodium current upon PTL treatment. It can therefore be concluded that microtubule remodeling may affect the function of Nav1.5.
INTRODUCTION... 4
MATERIALS & METHODS... 6
RESULTS... 8
CONCLUSION & DISCUSSION... 16
ACKNOWLEDGEMENTS... 19
LITERATURE LIST... 20
INTRODUCTION
A number of rare cardiac pathologies, such as long QT syndrome type 3 (LQT3), Brugada syndrome, (progressive) conduction disease, sick sinus syndrome, atrial standstill, atrial fibrillation, dilated cardiomyopathy and sudden infant death (SIDS) are related to SCN5A mutations that have been described in patients with these pathologies (Shy et al., 2013). SCN5A is the human gene encoding the -subunit of the major cardiac sodium channel Nav1.5. The initiation and propagation of action potentials throughout the myocardium is triggered by the fast upstroke of the cardiac action potential that originates from the influx of sodium ions through cardiac voltage-gated sodium channels such as Nav1.5. Since Nav1.5 plays an important role in the excitability of myocardial cells and the proper conduction of electrical signals within the heart, dysfunction of Nav1.5 leads to conduction abnormalities and severe conduction diseases such as arrhythmogenesis and sudden cardiac death (Remme, 2013).
Over the past two decades, several genetic, electrophysiological and molecular mechanisms underlying cardiac sodium channel dysfunction have been elucidated with the help of studies in patients with inherited arrhythmia syndromes (Remme, 2013). Arrhythmias caused by genetic mutations that change biophysical properties of ion channels are the most common. However, disruptions at the transcriptional, translational and post-translational levels of ion channel production are also able to disturb proper functional expression of ion channels (Mohler, 2009). Studies have shown recently that mutations in SCN5A are related to trafficking defects such as endoplasmic reticulum (ER) exit defects. These mutations cause retention and/or degradation of the -subunit of Nav1.5 from the ER (Balse & Boycott, 2017). As a consequence, these channels are degraded in the ERAD system (endoplasmic reticulum-associated degradation). Besides the mutations in SCN5A that alter the biophysical properties of the sodium channel and the possibility of ER exit defects due to SCN5A mutations, there is some evidence that points towards the hypothesis that microtubule remodeling may lead to dysfunctional sodium channels without the involvement of mutations that affect the function of Nav1.5.
The first indication for a role of microtubule remodeling on the dysfunction of Nav1.5 came from the observation that taxol users reported a more than average occurrence of arrhythmias (Arbuck et al., 1993). Taxol is an anti-cancer drug that is able to stabilize microtubules. Casini et
al. (2010) showed that rat cardiomyocytes treated with taxol had a decreased INa-current in comparison to the control situation, as shown in Figure 1. This finding indicates that the cardiac microtubule network is required for normal Nav1.5 function. Microtubules are the main transporting structures that move Nav1.5 from the Golgi to the lateral membrane and intercalated disks of cardiomyocytes (Shy et al., 2013). Microtubule filaments are also known as the highways of cells. They perform a wide range of cellular processes, such as the capture, transport and spatial organization of cargos and organelles, changes in cell shape, cell division and cell motility (Naghavi & Walsh, 2017). Microtubules consist of heterodimers which are formed by -tubulin and -tubulin. Microtubules undergo constant cycles of catastrophe and rescue events, or fast periods of polymerization and depolymerization respectively, which can be referred to as dynamic instability (Lacroix et al., 2016). This type of behavior is regulated by a variety of proteins such as adapter proteins and by posttranslational modifications (Westermann
& Weber, 2003).
Detyrosination of microtubules is an example of a reversible posttranslational modification of the -tubulin subunit, which is able to regulate the dynamic behavior of microtubules. Normally a tyrosine residue forms the end of a microtubule and enables dynamic behavior. However, the tyrosine residue can be proteolytically removed by carboxypeptidase resulting in detyrosinated microtubules, which are more rigid (Westermann & Weber, 2003). The total amount of detyrosinated -tubulin varies between cell types and a limited fraction (<10%) is present in the free tubulin pool (Gundersen et al., 1987). Two types of microtubules can be distinguished:
stable (more detyrosinated) and dynamic (less detyrosinated) microtubules. Several compounds are known to have an effect on microtubule stability. Taxol (paclitaxel) is a taxane and was originally identified from plants of the genus Taxus. Nowadays, taxol is primarily used as an anticancer chemotherapeutic drug (Long, 1994). The working mechanism of taxol is based on the increase of detyrosinated--tubulin (Schiff et al., 1980; Arnal & Wade, 1995), however how taxol manages this is so far unknown. On the other side there is parthenolide (PTL), also an anti-cancer drug with anti-inflammatory properties, derived from the “Feverfew” plant (Tanacetum
parthenium L., astercaea family). PTL is known as a sesquiterpene lactone (Fonrose et al., 2017)
and this compound inhibits the enzyme carboxypeptidase. Normally, this enzyme converts -tubulin into detyrosinated---tubulin. The inhibition of this enzyme by PTL causes less detyrosinated microtubules and therefore enhances the dynamic behavior of microtubules
(Mathema, 2012).
Casini et al. (2010) showed that pre-incubation of neonatal rat cardiomyoctes with the anti-cancer drug taxol led to reduced Nav1.5 expression on the sarcolemma and to a decreased sodium current (INa+) as is stated before and can be seen in Figure 1. Unpublished data from S.
Casini demonstrates an increased (INa+) current upon PTL treated HEK293 cells stably expressing
Nav1.5. These results suggest that dynamic instability of microtubules might play a significant role in the (dys)function of Nav1.5. However, the exact role of microtubules in this process remains unclear. Therefore, in this thesis report the possible effects of microtubule remodeling on sodium channel (Nav1.5) function is further investigated. It is hypothesized that altered sodium current mediated by taxol and PTL may be due to microtubule (de)stabilization mediated by these pharmacological compounds.
To test this hypothesis, immunocytochemistry stainings have been performed on HEK293 cells that stably express Nav1.5, hiPSC-CMs and mouse cardiomyocytes which are treated with taxol and PTL to increase and decrease microtubule detyrosination respectively. The results will be compared with functional measurements that previously showed a decreased sodium current upon taxol treatment of HEK293 cells, hiPSC-CMs and mouse cardiomyocytes and ongoing studies that show an increased sodium current upon PTL treatment of HEK293 cells (Figure 2). This comparison will confirm or reject the hypothesis that the altered sodium currents mediated by taxol and PTL are due to microtubule remodeling mediated by these compounds. It is
and after exposure of PTL. Note. Adapted from unpublished data by Casini, S.
MATERIALS & METHODS
All the experiments in this study conform with the Guide for the Care and Use of Laboratory
Animals published by the US National Institutes of Health (NIH Publication No. 85-2, revised
1996) and was approved by the institutional animal experiments committee.
HEK293 Nav1.5 culture
Human embryonic kidney (HEK293) cells that stably express the -subunit of Nav1.5 were cultured from frozen stocks at 37°C in Gilbco’s minimum essential medium consisting of 4.5g/l D-glucose and 0.11g/L sodium pyruvate, supplemented with nonessential amino acid solution L-glutamine, fetal bovine solution (FBS), P/S and zeocine in a 5% CO2 incubator at 37°C.
This cell line was routinely split once a week with a split ratio of 1:10, when ~80% confluency was reached. Before incubation with taxol and PTL, cells were split and adhered to glass coverslips coated with gelatin and cultured for 1 or 2 days in the previous described medium in a 5% CO2 incubator at 37°C. To study the possible effects of microtubule remodeling on disrupted
Nav1.5 function, HEK293 cells were incubated with 5M or 10M PTL (Sigma Co.), 1M taxol (Sigma Co.) or DMSO (1:1300) for 2 h at 37°C. DMSO functions as a control group, since PTL and taxol are dissolved in DMSO to enhance solubility in the aqueous medium. After incubation with these pharmacological compounds cells were fixed using methanol for 5 min at -20°C and used for immunocytochemistry. The fixation method needed to be optimized to obtain optimal results. Several common fixation methods are routinely used, such as 4% paraformaldehyde (PFA) and methanol. Which method is preferred depends on both cell type and the subcellular compartment that is desired to be visualized. Both 4% PFA and methanol are used for microtubule fixation, but it is argued that fixation with 4% PFA leads to altered microtubules since PFA not immediately fixes the microtubules and modifies the microtubules during the fixation process. Therefore, it is recommended to use methanol for fixing microtubules, because opposed to PFA, methanol fixes microtubules immediately.
Human induced pluripotent stem cell-derived cardiomyocyte culture
Human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) were obtained from the company Ncardia. They were defrosted and incubated with 5M or 10M PTL, 1M taxol or DMSO for 2 h at 37°C. After incubation with these pharmacological compounds cells were fixed using methanol for 5 min at -20°C and used for immunocytochemistry.
Mouse cardiomyocytes
Left ventricular mouse cardiomyocytes were isolated from hearts of adult male and female C57Bl6 mice and kept in solution as described previously (Rivaud et al., 2018). After 2 h incubation with 5M or 10M PTL, 1M taxol or DMSO at 37°C, cells were fixed using methanol for 5 min at -20°C and used for immunocytochemistry.
Immunocytochemistry
Cells were permeabilized at room temperature for 8 min with 0.1% Triton X-100 in PBS. 4% horse serum was added for 1 h to prevent non-specific staining. Primary antibodies (mouse monoclonal antibody against -tubulin and rabbit polyclonal antibody against detyrosinated-tubulin, Santa Cruz sc-5286 and Millipore AB3201 respectively, 1:250) were diluted in 4% horse serum and added to the coverslips for an overnight incubation at 4°C. Secondary antibodies (donkey-anti-mouse against -tubulin and donkey-anti-rabbit against detyrosinated-tubulin, Life Technologies A10037 and Life Technologies A21206 respectively, 1:250) were diluted in 4% horse serum and incubated at room temperature for 1 h. Coverslips were mounted on glass
slides using Mowiol (Sigma 81381) and imaged. The first stainings were imaged using a Leica DM5500 Q confocal microscope and it became evident that it is challenging to visualize the -tubulin and detyrosinated---tubulin network using this type of microscope. This microscope generates images of specific sections of the sample and the light is emitted at one point (Fig. S3). This technique enables the reconstruction of 3D images of the microtubule network by generating z-stacks. However, the generation of these images is very time-consuming, and the resolution is not sufficient to visualize a clear microtubule network. Therefore, this microscope was replaced by the Leica DM6000 widefield microscope. Figure S3 shows that the whole specimen of interest is exposed to the light source with widefield microscopy, instead of only a small volume of the specimen with confocal microscopy. The structure of the microtubule network is clearer when z-stacks are generated using the widefield microscope (Fig. S4) and so it is recommended to make use of the widefield microscope to visualize microtubules.
Quantification using ImageJ
Mean fluorescence intensity of -tubulin and detyrosinated--tubulin was measured in a preset area of the cell in all the cell models using ImageJ. For the HEK293 and hiPSC-CMs cell models, twenty cells were measured per condition. For the mouse cardiomyocyte cell model, 4 cells were measured per condition. A mean ratio of the amount of detyrosination was calculated by dividing the fluorescence intensity of detyrosinated--tubulin by the fluorescence intensity of -tubulin.
Statistical analysis
The mean ratios were used for statistical analysis. Comparisons between groups in HEK293 cells and hiPSC-CMs were performed with two-way analysis of variance (ANOVA). Two-way ANOVA with repetitive measurements followed by a Tukey HSD test for post hoc analysis was performed where appropriate. To compare the mean ratios during control conditions and after incubation with PTL in mouse cardiomyocytes, unpaired Student’s t-test was used. P<0.05 was considered
RESULTS
Optimization primary antibody concentration
The concentrations of the primary antibodies needed to be optimized before they could be used for staining. Since detyrosination occurs on the -tubulin subunit, it is desirable to stain -tubulin as a marker of the microtubule network, because there is competition between the antibody for detyrosinated -tubulin and the antibody for -tubulin. Several test stainings in mouse cardiomyocytes (Fig. S2) show that a concentration of 1:200 is optimal for the -tubulin antibody. Unfortunately, there were no sufficient amounts of this antibody to complete all stainings. Therefore, in follow up stainings an antibody for -tubulin was used, although this antibody is not preferred due to previous described competition with the antibody for
detyrosinated--tubulin. Figure S1 demonstrates that a concentration of 1:250 of the -tubulin antibody in HEK293 cells results in the most optimal images of the microtubule network. Furthermore, a concentration of 1:200 for detyrosinated--tubulin, also named glu-tubulin, shows the clearest structure of the microtubules in cardiomyocytes (Fig. S2), whilst a concentration of 1:250 is preferred in HEK293 cells (Fig. S1).
Taxol increases the amount and density of detyrosinated--tubulin in HEK293 cells, while PTL seems to disrupt the detyrosination network
Figure 3 demonstrates the amount and density of the detyrosinated microtubule network (glu-tubulin; green) and the global microtubule network (-(glu-tubulin; red) in HEK293 cells treated with taxol, PTL or DMSO (control). Treatment of the cells with taxol under the same conditions as used for the sodium current measurements (Fig. 1), resulted in a strong increase in the amount and density of the detyrosinated microtubules (Fig. 3: TAXOL (glu-tub)). Immunolabeling of tubulin shows an extensive microtubule network (Fig. 3; control). -Tubulin also seems to be increased in specific areas after taxol treatment (Fig. 3; TAXOL). Although these observations indicate that taxol affects the -tubulin network, this was not observed in all areas and cells. Clearly, detyrosinated--tubulin density was strongly increased
after taxol treatment.
To investigate whether PTL decreases the amount of detyrosinated--tubulin reported by several papers (Fonrose et al., 2017; Kerr et al., 2015), HEK293 cells were treated with two different concentrations, 5M or 10M of PTL under the same conditions as used for the sodium current measurements (Fig. 2). It is striking that the network of -tubulin is disrupted in both PTL conditions (Fig. 3; PTL (5M) and PTL (10M)). This effect could be visualized in all the HEK293 cells treated with PTL in multiple stainings. Figure 3 (PTL (5M) and PTL (10M)) also demonstrates that the amount and density of detyrosinated--tubulin upon 5M and 10M PTL treatment could not be measured, because it appears that PTL disrupts the detyrosination network. Furthermore, it became evident from figure 3 (PTL (5M) and PTL (10M)) that the
cell size is decreased upon PTL treatment.
Figure 4 shows the results from the quantification of the fraction of detyrosinated microtubules, (measured as the ratio of glu-tubulin versus -tubulin) in HEK293 cells upon taxol, 5M and 10M PTL treatment. An ANOVA test was performed to compare mean ratios of the amount of detyrosination between the control and treated conditions. From table S2 it can be concluded that the mean ratios differed significantly between the conditions and therefore a post hoc Tukey HSD test was performed. It becomes clear from figure 4 that taxol (3,04 ± 0,27, Tukey HSD post-hoc, p<0.05, see table 1 and S1) increases the amount of detyrosination significantly in comparison to the control condition (1,87 ± 0,10). Figure 4 also clearly shows that 5M PTL (3,78 ± 0,13, Tukey HSD post-hoc, p=<0,05)and 10M PTL treatment (3,64 ± 0,10, Tukey HSD post-hoc, p<0,05) increase the amount of detyrosination significantly compared to the control condition (1,87 ± 0,10), apparently even more than taxol treatment.
Figure 3: Detyrosinated -tubulin is increased following taxol treatment and the detyrosination network is disrupted following 5M and 10M PTL treatment in HEK293 cells
HEK293 cells stably expressing Nav1.5 were treated with taxol, 5M or 10M PTL or DMSO (control) before fixation, and immunostained for detyrosinated--tubulin (green) and -tubulin (red). A clear increase in the amount and density of microtubule detyrosination is observed following taxol treatment. Furthermore, the detyrosinated microtubule structure completely disappeared upon both PTL treatments (5M and 10M). It is also striking that the cell size seems to be decreased upon PTL treatment.
Figure 4: The amount of detyrosination is increased upon taxol and PTL treatment in HEK293 cells The fluorescence intensity of detyrosinated--tubulin and -tubulin of twenty cells (n=20) per condition was measured to calculate the mean ratio of the amount of detyrosination. The amount of detyrosination is significantly increased upon taxol (3,04 ± 0,27, Tukey HSD post-hoc, p<0.05, see table 1 and S1), 5M PTL (3,78 ± 0,13, Tukey HSD post-hoc, p=<0,05) and 10M PTL (3,64 ± 0,10, Tukey HSD post-hoc, p<0,05) treatment in comparison to the control condition (1,87 ± 0,10). 5M PTL treated cells show the highest amount of detyrosination.
The mean difference, standard error (Std. Error) and significance (Sig.) are shown for the control condition compared to the taxol, 5M PTL and 10M PTL treated condition. The amount of detyrosination differed significantly between the control condition and the taxol and PTL treated conditions. This table is adapted from table S3 that shows the original post-hoc Tukey HSD results.
(I) Control
condition (J) Treated condition Mean Difference (I-J) Std. Error Sig.
DMSO PTL 5M -1,91371* 0,23318 0
PTL 10M -1,77418* 0,23318 0
Taxol -1,17087* 0,23318 0
Taxol increases the amount and density of detyrosinated--tubulin in hiPSC-CMs, while PTL seems to disrupt the detyrosination network
Figure 5 demonstrates the amount and density of the detyrosinated microtubule network (glu-tubulin; green) and the global microtubule network (-(glu-tubulin; red) in hiPSC-CMs treated with taxol, PTL or DMSO (control). Treatment of the cells with taxol under the same conditions as used for the sodium current measurements (Fig. 1), resulted in a strong increase in the amount and density of the detyrosinated microtubules (Figure 5: TAXOL (glu-tub)). Figure 5 (-tubulin) reveals that -tubulin is unaffected by taxol since equal amounts of -tubulin are visualized in both the taxol treated condition and the control condition.
Figure 5 (PTL (5M) and PTL (10M)) shows that -tubulin is in general not altered between the PTL treated and the control condition. However, some smaller cells in both PTL conditions have a higher intensity of -tubulin in comparison to the cells of the control condition. Figure 5 also demonstrates that the detyrosinated--tubulin structure is disrupted due to PTL treatment (both concentrations). Treatment with 10M PTL in particular seems to increase the
immunofluorescence intensity of detyrosinated--tubulin. It was also evident from multiple stainings that the cell size was dramatically decreased after 5M and 10M PTL treatment. Figure 6 shows the quantification of the amount of detyrosination in hiPSC-CMs cells upon taxol, 5M PTL and 10M PTL treatment. An ANOVA test was performed to compare mean ratios of the amount of detyrosination between the control and treated conditions. Table 2 shows that these means did not differ significantly between the conditions. It becomes clear from figure 6 and table 2 that taxol treated cells (2,42 ± 0,23, see table S4) seem to have a higher amount of detyrosination than the control cells (1,88 ± 0,32), however these results were not statistically significant. Figure 6 also seems to demonstrate a higher amount of detyrosination upon 10M PTL treatment (2,7 ± 0,35) in comparison to the control condition (1,88 ± 0,32), but this increase was not statistically significant. Furthermore, 5M PTL treated cells (1,74 ± 0,27) show a slight decrease of the amount of detyrosination in comparison to the control cells (1,88 ± 0,32) that was again not significant.
Figure 5: Detyrosinated -tubulin is increased following taxol treatment and the detyrosination network is disrupted following 5M and 10M PTL treatment in hiPSC-CMs
hiPSC-CMs were treated with taxol, 5M or 10M PTL or DMSO (control) before fixation, and immunostained for detyrosinated--tubulin (green) and -tubulin (red). The amount and density of detyrosinated--tubulin are increased upon taxol treatment. The detyrosinated microtubule structure is disrupted following 5M and 10M PTL treatment in comparison to the control condition. 10M PTL treated cells show a strongly decreased cell size.
Figure 6: The amount of detyrosination is increased following taxol and 10M PTL treatment and slightly decreased upon 5M PTL treatment in hiPSC-CMs
The fluorescence intensity of detyrosinated--tubulin and -tubulin of twenty cells (n=20) per condition was measured to calculate the mean ratio of the amount of detyrosination. The amount of detyrosination is not
significantly increased following taxol treatment (2,42 ± 0,23, see table 2 and S4). This amount is also not significantly decreased upon 5M PTL treatment (1,74 ± 0,27) in comparison to the control condition (1,88 ± 0,32). The amount of detyrosination of the -tubulin network seems to be increased upon 10M PTL treatment (2,7 ± 0,35) in comparison to the control condition (1,88 ± 0,32), but this increase is not significant.
Table 2: ANOVA test results show that there is no significant difference of the mean ratio of the amount of detyrosination between the control, taxol and PTL treated conditions in hiPSC-CMs. The sum of squares, degrees of freedom (df), mean square, test statistic (F) and the significance (Sig.) are shown. There is no significant difference between the mean ratios of the investigated conditions.
Sum of Squares df Mean Square F Sig.
Between
Groups 12,693 3 4,231 2,345 0,079
Within Groups 138,952 77 1,805
Effect of PTL treatment in mouse cardiomyocytes
Figure 7 demonstrates the amount and density of the detyrosinated microtubule network (glu-tubulin; green) and the global microtubule network (-(glu-tubulin; red) in mouse cardiomyocytes treated with 5M PTL or DMSO (control). A mesh-like microtubule network in control mouse cardiomyocytes is revealed by immunolabeling -tubulin (Fig. 7; control and PTL (5M)), which is consistent with previous reports (Nishimura et al., 2006; Ghosh et al., 2012). Figure 7 (control and PTL (5M)) visualizes that the immunofluorescence signal for detyrosinated--tubulin is slightly decreased following PTL treatment. However, it must be stated that the characteristic microtubule structure in this condition is not completely visible. Another striking observation that was observed in multiple stainings is that the general size of the cardiomyocytes treated with PTL is reduced in comparison to those of the control condition. The intensity of the nucleus staining is also diminished upon PTL treatment (Fig. 7; PTL (5M)).
Figure 8 shows the quantification of the amount of detyrosination in mouse cardiomyocytes upon 5M PTL treatment. It becomes clear from this figure and table S6 that PTL treatment significantly decreases the amount of detyrosination (0,39 ± 0,02, see table S5) in comparison to the control condition (0,83 ± 0,19), t(7)= 4,49, p<0,05.
Figure 7: Detyrosinated--tubulin is decreased following 5M PTL treatment Mouse cardiomyocytes were treated with 5M PTL or DMSO (control) before fixation, and immunostained for detyrosinated--tubulin (green) and -tubulin (red). Detyrosinated--tubulin is slightly decreased upon PTL
Figure 8: The amount of detyrosination is steeply decreased after treatment with PTL in mouse cardiomyocytes The fluorescence intensity of detyrosinated--tubulin and -tubulin of 4 cells (n=4) per condition was measured to calculate the mean ratio of the amount of detyrosination. The amount of detyrosination upon 5M PTL treatment (0,83 ± 0,19, see table S5) decreased significantly in comparison to the control condition (0,39 ± 0,02), t(7)= 4,49, p<0,05.
-CONCLUSION & DISCUSSION
The aim of this study was to investigate whether microtubule remodeling mediated by taxol and PTL could be observed in HEK293 cells, hiPSC-CMs and mouse cardiomyocytes under the conditions where the altered sodium currents were measured. Therefore, the amount and density of detyrosinated--tubulin and -tubulin in these three models was visualized using immunocytochemistry. It became evident from the stainings in HEK293 cells and hiPSC-CMs that taxol increases the amount of detyrosinated microtubules, while the data suggests but cannot conclude that PTL disrupts the structure of the detyrosinated microtubules. Although
immunocytochemistry is not an absolute quantitative method, it can be argued that the effect of taxol is more substantial and clear than the effect of PTL. These results can be influenced by a variety of factors that are discussed in more detail in the following section.
The microtubule network is responsible for the transport of Nav1.5 from the Golgi apparatus to specific subcellular compartments such as the lateral membrane and the intercalated disks. Changes in the microtubule network, mimicked by the interventions with taxol and PTL, could influence the (dys)function of Nav1.5. The results of the current study confirm previous findings from Schiff and colleagues (1980) that taxol is a microtubule stabilizing agent and therefore shows that microtubule remodeling might affect (dys)function of Nav1.5. However, this study does not support the hypothesis that PTL decreases the amount and density of detyrosinated microtubules. Therefore, these data partly complement the functional data that demonstrate a decreased sodium current upon taxol treatment and an increased sodium current upon PTL treatment (Fig. 1 and Fig. 2). A possible explanation for the observed effect of PTL in the
stainings might be that the baseline levels of detyrosinated microtubules in healthy cells are very low. PTL treatment of healthy cells would then result in a small or no effect on microtubules. It is hypothesized that the amount of detyrosination in diseased hearts is significantly higher in comparison to healthy hearts. Thus, PTL treated diseased heart tissue would probably show a more clear decrease in the amount and density of detyrosination.
The results obtained from this research could be influenced by a variety of factors that are described per cell type. Firstly, due to suboptimal quality of the HEK293 cells, microtubule structure in these cells was not as characteristic as could be seen in hiPSC-CMs. Suboptimal cell quality was mainly due to challenging conditions during the fixation of the cells. Methanol must be ice cold (-20°C for at least a couple of hours before fixation) to fix HEK293 cells properly. However, in the experiments in this study methanol was held on ice for 30 minutes before fixation, because at that moment it was thought that this method would be sufficient to fix cells adequately. Consequently, the majority of HEK293 cells was washed off during fixation. Due to suboptimal fixation, methanol might have influenced the quality of the cells, since they appeared to be round after fixation, while they looked healthy before fixation. The microtubule structure may also have been affected, and this explains that the microtubules in HEK293 cells are not as characteristic as those observed in the hiPSC-CMs. In further studies it is recommended to keep methanol at -20°C for at least a couple of hours before fixation.
Secondly, in mouse cardiomyocytes, the characteristic structure of the detyrosinated
qualitative method instead of a quantitative one. The main reason for using this method was its ability to visualize microtubules and determine the morphology of the cells. From the stainings in HEK293 cells and hiPSC-CMs it became evident that taxol mediates microtubule remodeling and that PTL affects the structure of the detyrosination network. Despite these findings, it was desirable to quantify the results. In this study, quantification was performed by manually measuring the fluorescence intensity in a preset area of twenty cells per condition in HEK293 cells and hiPSC-CMs and four cells per condition in mouse cardiomyocytes. A down side of this method is that the fluorescence intensity showed variation in cells of the same condition. Depending on the area selected in these cells different results were obtained. The results from this method are therefore biased and less reliable than for instance quantification using Western blotting. In further research it is therefore recommended to use Western blotting.
There are several potential mechanisms that explain how taxol is able to decrease the sodium current in cells. Kinesin-1 is the motor protein that traffics Nav1.5 from the Golgi apparatus to the intercalated disks and T-tubules in cardiomyocytes via the microtubule network (Steele & Fedida, 2014). Taxol stabilizes microtubules by increasing the amount of detyrosinated -tubulin, which also increases the affinity of kinesin-1 to bind to microtubules (Reuther et al., 2016). As a consequence of the increased amount of kinesin-1 on the stabilized microtubules, it is logical to expect more trafficking of Nav1.5 to the intercalated disks and lateral membrane. Consequently, the sodium current should be increased at the membrane due to a higher amount and density of Nav1.5. However, the results from Casini et al. (2010) showed an opposite effect on the sodium current, due to other effects of taxol. It appears that taxol influences the function of the plus end binding protein EB1 (Morrison et al., 1998). In a normal situation EB1 interacts with the adapter protein Ankyrin-G that is linked to connexin 43 to deliver Nav1.5 to the
intercalated disks. But, since the function of EB1 is disrupted by taxol, so is the delivery process, resulting in a decreased amount and density of Nav1.5. This is the main reason for the decreased sodium current. This mechanism reveals the complexity of microtubule remodeling and all the processes that need to be taken into account to study the role of microtubule remodeling on the function of Nav1.5.
Although the results do not show clear microtubule remodeling mediated by PTL, it is thought that PTL has an effect on the microtubule structure. As mentioned before, the concept of low baseline levels of detyrosination in healthy cells could explain why there is no clear effect visible in the stainings. Nevertheless, the functional measurements (unpublished data from Casini et al., 2010) show a very clear effect of PTL on the sodium current. Several mechanisms could underlie the altered sodium currents upon PTL treatment. Since PTL inhibits the activity of
carboxypeptidase and enhances the dynamic instability of microtubules, it is thought that kinesin-1 binds with a lower affinity to destabilized microtubules. As a result, there should be less transport of Nav1.5 and subsequently a lower sodium current at the membrane.
Controversially, unpublished data from Casini et al. (2010) demonstrate that PTL mediates a higher sodium current in both HEK293 cells and hiPSC-CMs. It is unknown if any other effects of PTL on microtubule binding partners, as is the case for taxol, could be responsible for the observations. Another option that explains the observed results for PTL is described below. As aforementioned, it is posed by Mohler (2009) that the microtubule network could possibly direct the localization of Nav1.5 to specific membrane domains such as the intercalated disks and the lateral membrane. It could very well be that PTL has an effect on this process. If further research proves that PTL has a rescuing effect in research models with a diminished sodium current, future studies could look into the influence of PTL on the process of Nav1.5 localization. It is hypothesized that PTL could enhance trafficking of Nav1.5 to a greater or lesser extent to the intercalated disks and the lateral membrane.
As was previously stated in the introduction, so far only biophysical alterations of Nav1.5 due to loss-of-function and gain-of-function mutations in SCN5A were studied and proposed as causes for Nav1.5 dysfunction. This study provides unique evidence that microtubule remodeling might also affect the (dys)function of Nav1.5. If further research is able to elucidate the exact
mechanisms underlying Nav1.5 (dys)function due to microtubule remodeling, then new therapeutics could be developed to treat pathologies caused by dysfunction of Nav1.5.
Especially the increased sodium currents due to PTL treatment are promising since this indicates that PTL might be able to rescue Nav1.5 function in models with a diminished sodium current. Since PTL influences the trafficking of Nav1.5, the beneficial effects of this drug likely only apply to situations in which the trafficking of the sodium channel is disrupted. PTL will for example not have a rescuing effect when the expression of Nav1.5 is reduced. Another point that needs to be mentioned is that PTL could also enhance the trafficking of dysfunctional sodium channels (for instance caused by a mutation). It would be ideal to find a compound that only affects the trafficking of the ‘correct’ sodium channels but investigating this is very challenging taking into account that the functions and dynamicity of microtubules are common processes in the cell. It is therefore likely that the microtubule modulating effects of taxol and PTL not only affect Nav1.5, but probably also influence other ion channels. It is as yet unknown if different sodium channels have a distinct sensitivity for microtubule remodeling. Although the search for suitable drugs will be challenging, it is promising that microtubule remodeling could be a new target for medicines and therapies. This will broaden the therapeutic options for patients suffering from electrical conduction diseases characterized by dysfunction of Nav1.5.
In this study, the first pieces of evidence are provided that microtubule remodeling is involved in the dysfunction of Nav1.5. However, the amount and density of detyrosinated--tubulin are solely investigated in healthy cell models after taxol and PTL interventions. Further research should study the effects of these pharmacological compounds in diseased heart tissue to
elucidate if microtubule remodeling is affected in diseased conditions with consequent effects on sodium channel function.
Taken together, this data suggests that microtubule remodeling may cause dysfunction of Nav1.5. The microtubule network therefore functions as a new target for therapies, providing new therapeutic options for patients suffering from pathologies marked by dysfunction of Nav1.5.
ACKNOWLEDGEMENTS
First and foremost, I would like to express my appreciation to C.A. Remme for giving me the chance to do my internship and thesis project at the experimental cardiology department of the AMC. I would also like to thank C.A. Remme for her supervision, guidance and enthusiastic encouragement during my internship. I would like to extend my thanks to V.M. Portero for his daily supervision, which was very enthusiastic and educative. I would also like to show gratitude to the technicians of the laboratory of the experimental cardiology department, especially S.C.M. van Amersfoorth, for their help in performing this study.
LITERATURE LIST
- Arbuck, S. G., Strauss, H., Rowinsky, E., Christian, M., Suffness, M., Adams, J., ... & Gibbs, H. (1993). A reassessment of cardiac toxicity associated with Taxol. Journal of the National
Cancer Institute. Monographs, (15), 117-130.
- Arnal, I., & Wade, R. H. (1995). How does taxol stabilize microtubules?. Current
biology, 5(8), 900-908.
- Balse, E., & Boycott, H. E. (2017). Ion Channel Trafficking: Control of Ion Channel Density as a Target for Arrhythmias?. Frontiers in physiology, 8, 808.
- Casini, S., Tan, H. L., Demirayak, I., Remme, C. A., Amin, A. S., Scicluna, B. P., ... & Veldkamp, M. W. (2010). Tubulin polymerization modifies cardiac sodium channel expression and gating. Cardiovascular research, 85(4), 691-700.
- Fonrose, X., Ausseil, F., Soleilhac, E., Masson, V., David, B., Pouny, I., ... & Lafanechere, L. (2007). Parthenolide inhibits tubulin carboxypeptidase activity. Cancer research, 67(7), 3371-3378.
- Ghosh, D. K., Dasgupta, D., & Guha, A. (2012). Models, regulations, and functions of microtubule severing by katanin. ISRN molecular biology, 2012.
- Gundersen, G. G., Khawaja, S., & Bulinski, J. C. (1987). Postpolymerization detyrosination of alpha-tubulin: a mechanism for subcellular differentiation of microtubules. The Journal of Cell Biology, 105(1), 251-264.
- Kerr, J. P., Robison, P., Shi, G., Bogush, A. I., Kempema, A. M., Hexum, J. K., ... & Prosser, B. L. (2015). Detyrosinated microtubules modulate mechanotransduction in heart and skeletal muscle. Nature communications, 6, 8526.
- Lacroix, B., Ryan, J., Dumont, J., Maddox, P. S., & Maddox, A. S. (2016). Identification of microtubule growth deceleration and its regulation by conserved and novel
proteins. Molecular biology of the cell, 27(9), 1479-1487.
- Long, H. J. (1994, April). Paclitaxel (Taxol): a novel anticancer chemotherapeutic drug. In Mayo Clinic Proceedings (Vol. 69, No. 4, pp. 341-345). Elsevier.
- Mathema, V. B., Koh, Y. S., Thakuri, B. C., & Sillanpää, M. (2012). Parthenolide, a sesquiterpene lactone, expresses multiple anti-cancer and anti-inflammatory activities. Inflammation, 35(2), 560-565.
- Mohler, P. J. (2009). Sodium channel traffic on the cardiac microtubule highway.
- Morrison, E. E., Wardleworth, B. N., Askham, J. M., Markham, A. F., & Meredith, D. M. (1998). EB1, a protein which interacts with the APC tumour suppressor, is associated with the microtubule cytoskeleton throughout the cell cycle. Oncogene, 17(26), 3471.
- Naghavi, M. H., & Walsh, D. (2017). Microtubule regulation and function during virus infection. Journal of virology, 91(16), e00538-17.
- Nishimura, S., Nagai, S., Katoh, M., Yamashita, H., Saeki, Y., Okada, J. I., ... & Sugiura, S. (2006). Microtubules modulate the stiffness of cardiomyocytes against shear stress. Circulation Research, 98(1), 81-87.
- Reuther, C., Diego, A. L., & Diez, S. (2016). Kinesin-1 motors can increase the lifetime of taxol-stabilized microtubules. Nature nanotechnology, 11(11), 914.
- Rivaud, M. R., Baartscheer, A., Verkerk, A. O., Beekman, L., Rajamani, S., Belardinelli, L., ... & Remme, C. A. (2018). Enhanced late sodium current underlies pro-arrhythmic
- Westermann, S., & Weber, K. (2003). Post-translational modifications regulate microtubule function. Nature Reviews Molecular Cell Biology, 4(12), 938.
SUPPLEMENTAL INFORMATION
Supplemental Inventoryo Figure S1: Optimization anti--tubulin and anti-glu-tubulin concentration in HEK293 Nav1.5 cells
o Figure S2: Optimization anti--tubulin and anti-glu-tubulin concentration in mouse cardiomyocytes
o Figure S3: Illumination differences between confocal and widefield microscopy
o Figure S4: Detyrosinated--tubulin in hiPSC-CMs with confocal microscopy and widefield microscopy
o Statistical analysis
o Table S1: Descriptives of the mean ratio of the amount of detyrosination in HEK293 cells treated with DMSO (control), taxol and PTL.
o Table S2: ANOVA test results show that there is a significant difference of the mean ratio of the amount of detyrosination between the control, taxol and PTL treated conditions in HEK293 cells.
o Table S3: Post-hoc Tukey HSD results of the amount of detyrosination in HEK293 cells treated with DMSO (control), taxol and PTL.
o Table S4: Descriptives of the mean ratio of the amount of detyrosination in hiPSC-CMs treated with DMSO (control), taxol and PTL.
o Table S5: Descriptives of the mean ratio of the amount of detyrosination in mouse cardiomyocytes treated with DMSO (control), taxol and PTL.
o Table S6: Unpaired Student’s t-test shows that there is a significant difference between the mean ratios of the amount of detyrosination between the 5M PTL treated cardiomyocytes and the DMSO (control) condition.
o Experimental procedures
o Protocol 1: HEK293 Nav1.5 cell culture
Figure S1
Figure S1: Optimization anti--tubulin and anti-glu-tubulin concentration in HEK293 Nav1.5 cells
Upper images showing -tubulin by different concentrations (1:100, 1:200 and 1:250) of the antibody for -tubulin. Lower images showing glu-tubulin by different concentrations (1:50, 1:200 and 1:250) of the antibody for glu-tubulin.
Figure S2
Figure S2: Optimization anti--tubulin and anti-glu-tubulin concentration in mouse cardiomyocytes
Upper images showing -tubulin by different concentrations (1:200, 1:500 and 1:1000) of the antibody for -tubulin. Lower images showing glu-tubulin by different concentrations (1:50, 1:200 and 1:500) of the antibody for glu-tubulin.
Figure S3
Figure S3: Illumination differences between confocal and widefield microscopy
Confocal microscopes emit their light at a specific small volume of the specimen of interest because the light is focused through a pinhole for illumination, while the whole specimen volume of interest is exposed to light with widefield microscopy.
Figure S4
Figure S4: Detyrosinated--tubulin in hiPSC-CMs with confocal microscopy and widefield microscopy
The detyrosinated--tubulin network is shown using confocal and widefield microscopy. The structure is clearer when the widefield microscope is used.
STATISTICAL ANALYSIS
Table S1: Descriptives of the mean ratio of the amount of detyrosination in HEK293 cells treated with DMSO (control), taxol and PTL. The amount of cells measured per condition (N), mean, standard deviation (Std. Deviation), standard error (Std. Error), 95% confidence interval for mean and minimum and maximum values of the mean are shown.
N Mean Std. Deviatio
n
Std. Error 95% Confidence Interval for Mean
Minimu m
Maximu m Lower Bound Upper Bound
Control 20 1,866 4 ,43814 ,09797 1,6614 2,0715 ,97 2,93 PTL 5M 20 3,780 1 ,57842 ,12934 3,5094 4,0508 2,76 4,92 PTL 10M 20 3,640 6 ,44037 ,09847 3,4345 3,8467 2,86 4,44 Taxol 20 3,037 3 1,20597 ,26966 2,4729 3,6017 1,63 6,00 Total 80 3,081 1 1,04883 ,11726 2,8477 3,3145 ,97 6,00
Table S2: ANOVA test results show that there is a significant difference of the mean ratio of the amount of detyrosination between the control, taxol and PTL treated conditions in HEK293 cells. The sum of squares, degrees of freedom (df), mean square, test statistic (F) and the significance (Sig.) are shown. There is a significant difference between the mean ratios of the investigated conditions.
Sum of Squares df Mean Square F Sig.
Between Groups 45,581 3 15,194 27,945 ,000
Within Groups 41,322 76 ,544
Table S3: Post-hoc Tukey HSD results of the amount of detyrosination in HEK293 cells treated with DMSO (control), taxol and PTL. The mean difference, standard error (Std. Error), significance (Sig.) and 95% confidence interval are shown for the control condition compared to the taxol, 5M PTL and 10M PTL treated condition. The amount of detyrosination differed significantly between the control condition and the taxol and PTL treated conditions.
Tukey HSD
(I) staining (J) staining Mean Difference (I-J)
Std. Error Sig. 95% Confidence Interval Lower Bound Upper Bound DMSO PTL 5M -1,91371* ,23318 ,000 -2,5262 -1,3012 PTL 10M -1,77418* ,23318 ,000 -2,3867 -1,1617 Taxol -1,17087* ,23318 ,000 -1,7834 -,5584 PTL 5M DMSO 1,91371* ,23318 ,000 1,3012 2,5262 PTL10 M ,13953 ,23318 ,932 -,4730 ,7520 Taxol ,74284* ,23318 ,011 ,1303 1,3553 PTL 10M DMSO 1,77418* ,23318 ,000 1,1617 2,3867 PTL 5M -,13953 ,23318 ,932 -,7520 ,4730 Taxol ,60331 ,23318 ,055 -,0092 1,2158 Taxol DMSO 1,17087* ,23318 ,000 ,5584 1,7834 PTL 5M -,74284* ,23318 ,011 -1,3553 -,1303 PTL 10M -,60331 ,23318 ,055 -1,2158 ,0092
*. The mean difference is significant at the 0.05 level.
Table S4: Descriptives of the mean ratio of the amount of detyrosination in hiPSC-CMs treated with DMSO (control), taxol and PTL. The amount of cells measured per condition (N), mean, standard deviation (Std. Deviation), standard error (Std. Error), 95% confidence interval and minimum and maximum values of the mean are shown.
Conditions N Mean Std.
Deviation Std. Error 95% Confidence interval
Minimum Maximum
Lower Bound Upper Bound
Taxol 20 2,4175434 75 1,050938732 0,234997045 1,925689008 2,909397942 1,166012785 4,286092376 PTL 5M 20 1,7396210 13 1,205436559 0,269543809 1,175459338 2,303782689 0,709163837 5,819325068 Control 20 1,8804320 36 1,391110424 0,319142619 1,209938274 2,550925798 0,326131438 4,954811715 PTL 10M 20 2,7030704 57 1,626048977 0,346674807 1,982120729 3,424020185 0,837064066 6,79212357 Total 80 2,2017166 48 1,37679391 0,152977101 1,897282515 2,506150781 0,326131438 6,79212357
Table S5: Descriptives of the mean ratio of the amount of detyrosination in mouse cardiomyocytes treated with DMSO (control), taxol and PTL. The amount of cells measured per condition (N), mean, standard deviation (Std. Deviation) and standard error mean (Std. Error Mean) are shown.
Conditions N Mean Std. Deviation Std. Error Mean DMSO 4 ,8322 ,19347 ,08652 PTL 5M 4 ,3890 ,02310 ,01155
Table S6: Unpaired Student’s t-test shows that there is a significant difference between the mean ratios of the amount of detyrosination between the 5M PTL treated cardiomyocytes and the DMSO (control) condition. Levene’s test for equality of variances was performed and the test statistic (F) and significance (Sig.) are shown. The result from this test is not significant. The t-test for equality of means was performed and the test statistic (t), degrees of freedom (df), significance (Sig. (2-tailed)), mean difference, standard error difference (Std. Error Difference) and the 95% confidence interval of the difference are shown. The mean ratio differed significantly between the control and the PTL treated condition.
Levene’s test for equality of variances T-test for equality of means
F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference
95% Confidence Interval of the Difference
Lower Upper
Mean ratio
Equal variances assumed
6,508 ,038 4,493 7 ,003 ,44316 ,09863 ,20994 ,67639
EXPERIMENTAL PROCEDURES
PROTOCOL 1 HEK293 NAV1.5 PASSAGE 1. Remove media from cells
2. Rinse cells with 5ml PBS
3. Add 0.5ml of trypsin and incubate 2min at 37°C
4. Add 4.5ml of complete media
5. Pipet up and down to resuspend the cells in the media
6. Divide cells 1:10 into new flasks, adding fresh media for a total volume of 5ml in each
PROTOCOL 2 IMMUNOFLUORESCENCE IN CELLS (APPIE) 1. Wash 3 times PBS
2. Permeabilize 8 min. at RT with 0.1% Triton X-100 in PBS (100l per 15mm coverslips) 3. Wash 3 times PBS
4. Preincubate cells with 4% normal horse serum (NHS) (in PBS) for 1 hour at RT (or O/N at 4°C)
5. Incubate with first antibody in 4% NHS for 1 hour at RT (or O/N at 4°C) 6. Wash 4 times 5 min. with PBS-Tween 0.05%
7. Incubate with secondary antibody in 4% NHS for 1 hour at RT (or O/N at 4°C) 8. Wash 5 times 5 min. with PBS-Tween 0.05%
9. For nuclear staining incubate 5 min DAPI 1:1000 in PBS, 5 minutes incubation 10. Wash shortly in miliQ
11. Embed coverslips in Mowiol (1-2l) 12. Wait 2 hours at 4°C for polymerization
