The neonatal splice variant of Nav1.5 potentiates in vitro invasive behaviour of MDA-MB-231 human breast cancer cells

Cell culture

MDA-MB-231 and PC-3M cells were cultured as described previously [6, 27]. Cells were seeded into Falcon tissue culture dishes (Becton-Dickinson, Oxford, UK) and incubated at 37 °C, 5 % CO₂ and 100 % relative humidity.

RNA interference

Two different siRNAs targeting unique nucleotides within the 5′-DI:S3 alternatively spliced exon of nNav1.5 were designed using an online algorithm (Dharmacon, Lafayette, CO), as follows:

siNESO1 target sequence: 5′-UUUGUCGGCUCUUCGAACU-3′

siNESO2 target sequence: 5′-GAGUCCUGAGAGCUCUAAA-3′

Their target specificity was ensured by nucleotide BLAST screening optimised for short sequences [34]. All siRNA duplexes were synthesised by Dharmacon and resuspended and stored according to the manufacturer’s instructions. Cells (70 % confluent in 6-well plates) were transfected for 4 h with siRNA (300 pmol/well) using Oligofectamine (Invitrogen, Paisley, UK), according to the manufacturer’s instructions. mRNA, protein level and functional activity were assayed 3-13 days post-transfection, and compared with mock or control siRNA-treated cells (siControl Non-Targeting siRNA 2). Transfection efficiency was assessed independently using a positive control siRNA (siControl Lamin A/C).

Real-time PCR

Extraction of total RNA, synthesis of cDNA and real-time PCR were performed as described previously [15, 27], with NADH/cytochrome b5 reductase (Cytb5R) measured as a normalising gene. The following primer pairs and annealing temperatures were used:

1. Cytb5R: 5′-TATACACCCATCTCCAGCGA-3′ and 5′-CATCTCCTCATTCACGAAGC-3′; annealing temperature, 60 °C [35].

2. Adult Nav1.5: 5′-CATCCTCACCAACTGCGTGT-3′ and 5′-ACATTGCCCAGGTCCACAAA-3′; annealing temperature, 60 °C.

3. Neonatal Nav1.5: 5′-CATCCTCACCAACTGCGTGT-3′ and 5′-CCTAGTTTTTCTGATACA-3′; annealing temperature, 58 °C.

4. Nav1.7: 5′-TATGACCATGAATAACCCGC-3′ and 5′-TCAGGTTTCCCATGAACAGC-3′; annealing temperature, 59 °C [35].

5. Lamin A/C: 5′-CTGCGGCGGGTGGATGCTGAGAAC-3′ and 5′-CCACGGCTGCGCTGCGAGGTAGG-3′; annealing temperature, 67 °C.

The threshold amplification cycles were determined using the Opticon Monitor 2 software (MJ Research, Waltham, MA) and then analysed by the 2^−ΔΔCt method [36].

Western blotting

Equivalent amounts of protein lysate (40 μg/well) were resolved by SDS polyacrylamide gel electrophoresis, as described previously [7, 32]. Three primary antibodies were used, at given dilutions, as follows:

1. Pan-VGSC antibody (1 μg/ml; Upstate);

2. NESO-pAb antibody (1 μl/ml) [32]; and

3. Anti-actinin antibody (1 μl/ml; Sigma, Dorset, UK).

The secondary antibodies were peroxidase-conjugated swine anti-rabbit for 1 and 2, and goat anti-mouse for 3 (Dako, Glostrup, Denmark). Blots were developed with ECL (Amersham, Little Chalfont, Buckinghamshire, UK) and visualised by exposure to Super RX100NF film (Fujifilm, London, UK). Densitometric analysis was performed using the Image-Pro Plus software (Media Cybernetics). Signal intensity was normalised to anti-actinin antibody as a loading control/reference, for at least three separate treatments. For each antibody, linearity of signal intensity with respect to increasing protein loading in the range 20-80 μg was ensured using a standard dilution of MDA-MB-231 protein extract.

Migration, proliferation and invasion assays

Cells (2 × 10⁵ /ml and 7.5 × 10⁴ /ml, respectively) were plated onto 12 μm-pore Transwell migration filters in 12-well plates (Corning, NY), or Matrigel-coated invasion chambers in 24-well plates (Becton-Dickinson) according to the manufacturers’ instructions. Cells were incubated with or without TTX (10 μM), and/or NESO-pAb (0.01-20 μg/ml) in a 1-10 % FBS chemotactic gradient. The number of cells migrating over 7 h or invading over 24 h was determined using the thiazolyl blue tetrazolium bromide (MTT) assay [17]. Results were compiled as the mean of at least four repeats. A polyclonal rabbit anti-laminin antibody (10 μl/ml; Serotec, Oxford, UK) and a rabbit IgG fraction from healthy non-immunised rabbits (1 μg/ml; Dako) were used as controls for the NESO-pAb experiments. Proliferation was determined using the MTT assay [17], in at least five different experiments.

Immunocytochemistry and confocal microscopy

Immunocytochemistry was performed as described previously [6, 32], with the following modifications: Cells were labelled with FITC-conjugated concanavalin A (Sigma) as a plasma membrane marker. The secondary antibody was Alexa567-conjugated goat anti-rabbit IgG (Dako).

Samples were viewed using a Leica DM IRBE microscope (with a X100 objective) with a confocal laser scanner (Leica TCS-NT with Ar/Kr laser). FITC and/or Alexa567 were excited with the 488 nm and 568 nm laser lines, respectively. The images (512 × 512 pixels) were obtained simultaneously from two channels using a confocal pinhole of 226.98 μm (Airy 1).

Digital image analysis

Densitometric analysis was performed using the LCS Lite software (Leica), as follows:

1. Cell surface VGSC expression was determined using the “freeform line profile” function drawn around the cell surface, determined by concanavalin A staining. Measurements were taken from 30 cells per condition, for three repeat treatments.

2. Protein distribution was determined using the “straight line profile” function drawn across the cytoplasm avoiding the nucleus, as described previously [37-39]. Signal intensity in plasma membrane region, set to cover 1.5 μm inward from the edge of concanavalin A staining, was compared with cytoplasmic signal intensity within the central 30 % of the line profile. Measurements were taken from ≥ 6 cells (randomly chosen) per condition, for three repeat treatments.

3. For analysing nNav1.5 staining in migrated vs. non-migrated cells, the “histogram profile” function was used to obtain measurements from the whole cell. Measurements were taken from ≥ 15 cells (chosen randomly) per condition, for three repeat treatments.

Electrophysiology

Whole-cell patch clamp recordings were performed 13 days after transfection, when patch pipettes could be successfully sealed upon cell membranes. The procedures used were mostly as described previously [40]. Voltage-activated membrane currents were recorded using an Axopatch 1D patch clamp amplifier (Axon Instruments, Wokingham, UK) compensating for series resistance by ~80 %. Currents were digitised using a Digidata 1200 interface (Axon Instruments), low-pass filtered at 5 kHz, sampled at 50 kHz, and analysed using pClamp 6 software (Axon Instruments). Linear components of leak were subtracted using the leak subtraction facility on the amplifier and/or using the pClamp software.

Two voltage-clamp protocols were used (holding potential = −100 mV), as follows:

1. Basic current-voltage (I-V) protocol: Cells were depolarised to test potentials within the range −70 to +70 mV in 5 mV steps. The test pulse duration was 60 ms; the interpulse duration was 2 s.

2. Steady-state inactivation protocol: Prepulses were in the range −130 to −10 mV (10 mV steps) for durations of 1 s. A test pulse of −10 mV was immediately applied for 80 ms; the interpulse duration was 2 s.

Recordings were obtained from a minimum of 20 cells per condition, from at least three repeat treatments. Data from individual dishes were combined to provide an overall mean and standard error (SEM). Conductance-voltage relationships and curve fitting were performed as described previously [41]. TTX dose-repsonse data were fitted using Excel (Micosoft) and Origin (OriginLab, Northampton, MA) software to a double Langmuir adsoption isotherm:

where IC₅₀S and IC₅₀R are the concentrations of TTX for 50 % VGSC blockage, for the TTX-sensitive and TTX-resistant components, respectively; and P_TTXs is the proportion of current that is TTX-sensitive [42].

Data analysis

All quantitative data are presented as means ± standard errors, unless stated otherwise. Pairwise statistical significance was determined with Student’s t test, or Mann-Whitney rank sum test, as appropriate. Multiple comparisons were made using ANOVA followed by Newman-Keuls post-hoc analysis. Real-time PCR data were analysed using the 2^−ΔΔCT method [36]. Results were considered significant at P < 0.05 (*).

siRNA significantly reduced the nNav1.5 mRNA level

MDA-MB-231 cells were transfected with a positive control siRNA targeting lamin A/C to confirm an efficient transfection protocol. After 3 days, the siRNA significantly reduced the lamin A/C mRNA level by 91 %, compared to mock-transfected cells (P < 0.01; n = 3). Next, cells were transfected independently with one of two siRNAs (siNESO1 and 2) targeting nNav1.5. As both siRNAs produced very similar results, data hereon are presented from either. In siRNA-transfected cells, the nNav1.5 mRNA level was reduced by 82 % after 3 days (P < 0.05; n = 3), increasing to 91 % after 5 days (P < 0.001; n = 3), compared to mock-transfected cells (Figure 1A). In contrast, after 5 days, the mRNA levels of adult Nav1.5 (aNav1.5) and Nav1.7 were unaffected (P = 0.73; n = 3 for both; Figure 1B). After 9 days and 13 days, the reduction in nNav1.5 mRNA level was less marked but still significant, reaching 74 % (P < 0.05; n = 3), and 65 % (P < 0.05; n = 3), respectively (Figure 1A), compared to cells transfected with non-targeting control siRNA (siControl).

siRNA reduced the VGSC protein level

Western blot with pan-VGSC and NESO-pAb antibodies revealed smaller reductions in the total VGSC α-subunit protein level of siRNA-transfected cells. After 3 days, protein was reduced by 24 % (P = 0.23; n = 6), and 23 % (P < 0.05; n = 3), for pan-VGSC and NESO-pAb, respectively, compared to mock-transfected cells (Figure 2A, B). Similarly, after 5 days, protein was reduced by 19 % (P = 0.06; n = 4), and 22 % (P = 0.17; n = 3), for pan-VGSC and NESO-pAb, respectively. The reduction in protein level persisted such that after 9 days, there was a reduction of 19 % (P < 0.05; n = 3), and 21 % (P = 0.19; n = 3), and after 13 days, protein was reduced by 26 % (P < 0.01; n = 3), and 25 % (P = 0.07; n = 3), for pan-VGSC and NESO-pAb, respectively, compared to cells transfected with siControl (Figure 2A, B).

Consistent with the Western blot data, analysis of confocal images of cells labelled with NESO-pAb revealed that after 13 days, siRNA transfection had significantly reduced the level of nNav1.5 protein at the plasma membrane by 36 % (P < 0.001; n = 90; Figure 2C, D). However, the subcellular (plasma membrane vs. internal/cytoplasmic) distribution of nNav1.5 protein along cellular cross-sections was unchanged by the siRNA treatment. The level of nNav1.5 immunoreactivity in a 1.5 μm section measured inward from the cell edge was assumed to represent the ‘plasma membrane’ fraction; the middle 30 % of the cross-section was assumed to represent the ‘internal’ cytoplasmic fraction [39]. The relative levels of nNav1.5 immunoreactivity for both fractions (expressed as a percentage of the total) were the same in control and siRNA-treated cells (P = 0.74 and P = 0.77, respectively; n = 20 for each; Figure 2E).

In conclusion, the reduction in the total VGSC protein by the siRNA treatment was much smaller and slower than the decrease in the nNav1.5 mRNA level. There was also a reduction in the level of VGSC at the plasma membrane. However, the balance of the subcellular distribution/(re)cycling of protein was not affected.

siRNA reduced VGSC functional activity

The effect of siRNA on VGSC functional activity was assessed using whole-cell patch clamp electrophysiology, at the same stage, when recordings became possible. Thus, after 13 days, transfection with siRNA significantly reduced peak VGSC current density by 33 %, from 20.6 ± 2.5 pA/pF to 13.7 ± 1.6 pA/pF, compared to siControl (P < 0.05; n > 20 cells of each; Figure 3A-C; Table I). Treatment with siRNA had no effect on activation voltage, voltage for current peak, voltage dependence of activation or steady-state inactivation (Figure 4A; Table I). There was a slight enlargement of the window current in the siRNA-treated cells (Figure 4A inset).

The time to peak of VGSC current was significantly slower in the voltage range −20 to 15 mV in cells transfected with siRNA (Figure 4B). However, fast (τ_f) and slow (τ_s) time constants of inactivation at 0 mV, derived from fitting current inactivations to a double exponential function, were unaffected by the siRNA treatment (Table I). Thus, the siRNA treatment slowed the activation kinetics, consistent with a possible change in the functional VGSC isoform expression profile.

The dose-dependent blockage of VGSC current in MDA-MB-231 cells by TTX was best fitted to a double Langmuir adsorption isotherm, with TTX-sensitive, and TTX-resistant components. The sensitivity of VGSC currents to low concentrations of TTX (< 2 μM) was increased in cells transfected with siRNA, compared to siControl-transfected cells (Figure 4C). The siRNA significantly increased sensitivity to 200 nM TTX, from 14.0 ± 3.9 % to 26.7 ± 4.1 % (P < 0.05; n = 6 cells of each; Table I). We concluded that by downregulating the level of TTX-resistant nNav1.5 with siRNA, the relative proportion of the other TTX-sensitive VGSC isoform(s) was increased. The latter was likely to involve Nav1.7 [6]

Migrated cells expressed more VGSC protein than non-migrated cells

Confocal immunocytochemistry with NESO-pAb was used to compare the nNav1.5 protein expression in plasma membranes of cells that did/did not migrate through the Transwell filters (Figure 5A, B). Cells that migrated through the filter had 2.2-fold more plasma membrane nNav1.5 protein immunoreactivity compared with cells that did not migrate (P < 0.001; n > 45; Figure 5C).

siRNA and NESO-pAb both suppressed VGSC-dependent migration and invasion

These assays were performed also after 13 days, when the status of the VGSC signalling was known. Thus, transfection with siRNA significantly reduced the number of cells migrating through 12 μm-pore Transwell filters filter by 43 % compared to siControl-transfected cells (P < 0.05; n = 5; Figure 6A). When TTX was applied while the siControl-transfected cells were migrating, their migration was reduced by 53 % (P < 0.05; n = 4), as reported earlier (FRASER). Importantly, TTX applied to siRNA-transfected cells during the assay had no further effect on their migration (P = 0.78; n = 4). Thus, the VGSC-sensitive component of migration was completely suppressed by siRNA targeting nNav1.5.

NESO-pAb, applied during Transwell assays, reduced migration in a dose-dependent manner, reaching significance, 38 % reduction at 1 μg/ml (P < 0.001; n = 7; Figure 6B). NESO-pAb (1 μg/ml) also reduced Matrigel invasion of MDA-MB-231 cells by 49 % compared to control (P < 0.001; n = 7), but had no effect on human metastatic PCa PC-3M cells, in which functional Nav1.7 is dominant [35] (J. K. J. Diss, unpublished observation). IgG control and an unrelated anti-laminin antibody had no effect on invasiveness of MDA-MB-231 or PC-3M cells (Figure 6C). TTX (10 μM) reduced invasion of MDA-MB-231 cells by 50 % (P = 0.001; n = 7). Importantly, co-application of NESO-pAb (1 μg/ml) and TTX (10 μM) also reduced cell invasion by 47 % (P < 0.01; n = 7), but had no further effect compared to either NESO-pAb or TTX alone (P = 0.87 and 0.58, respectively; n = 7 for each; Figure 6C). None of the antibody treatments had any effect on proliferation of MDA-MB-231 or PC-3M cells (data not shown).

Effects of RNAi on nNav1.5 mRNA and protein expression

In this study, the unique DI:S3 5′-splice region at exon 6 of nNav1.5 was targeted with siRNA. Although RNAi has previously been used to target ion channels [43-45], including VGSC α-subunit [46, 47], this is the first report where siRNA has been used to specifically target one particular splice variant of a VGSC isoform. There are several potential technical problems associated with the RNAi technique, including off-target effects and poor mRNA/protein-level downregulation [48-50]. In this study, the effects of RNAi were monitored over a 13-day time-course at a hierarchy of levels: mRNA, protein (level and distribution) and functioning; basically the same effects were observed using the two different siRNAs. Although this duration would seem longer than most siRNA experiments, it was necessary (i) to perform the patch clamp recordings to ascertain the level of VGSC signalling, and (ii) to enable correlation of the siRNA effects at the various levels in a systematic approach. The siRNA caused ~90 % reduction of nNav1.5 mRNA after 5 days, beyond which, the level started to partially recover, reaching ~70 % reduction by 13 days. A similar pattern of mRNA silencing has been reported previously [51]. Nevertheless, both adult Nav1.5, and Nav1.7 mRNA levels were not affected by this treatment, confirming the specificity of the siRNA for nNav1.5.

The reduction of nNav1.5 protein level, detected by Western blot using NESO-pAb, was much smaller, becoming significant only after 13 days. A pan-VGSC antibody gave similar results. Confocal immunocytochemistry with NESO-pAb [32] showed that nNav1.5 was present at the cell surface, in agreement with previous observations [6], and that the siRNA reduced the level of nNav1.5 protein expressed at the plasma membrane, without changing the balance of its (re)cycling. At present, the dynamics of VGSC mRNA/protein expression in cancer cells are not known, and this could be compounded by activity-dependent (auto)regulation of VGSCs [39]. The apparent disparity between the time course and extent of target mRNA and protein-level reduction may, in part, be due to large stores of intracellular protein and/or slow turnover rates [52]. Such disparities have been seen before in siRNA experiments. For example, siRNA targeting vitamin K epoxide reductase gene reduced mRNA level after 3 days, but protein was not reduced until after 11 days [51]. Also, in human BCa cells, siRNA-induced protein downregulation persisted for at least 7 days [53]. Given that VGSC protein half-life is 18-26 h [54, 55] and that there was intense intracellular immunoreactivity with NESO-pAb, the slower and smaller reduction in nNav1.5 protein could be due to it being present in large quantity in internal stores and cycling to the plasma membrane persisting.

Involvement of nNav1.5 in the functional VGSC activity of MDA-MD-231 cells

The siRNA treatment reduced the VGSC functional activity in MDA-MB-231 cells by 33 %, in agreement with the molecular data, suggesting a decrease in the availability of VGSC protein at the plasma membrane. Importantly, the siRNA was found also to increase the sensitivity of cells to TTX concentrations of < 2 μM. This implies that the proportion of TTX-sensitive functional VGSCs was increased after the siRNA treatment. This TTX-sensitive component is likely to be Nav1.7, which we have previously shown to constitute ~20 % of the VGSC α-subunit mRNA in MDA-MB-231 cells [6]. This is consistent with the dose-response data (Figure 4C) showing an increase in the level of the TTX-sensitive component up to 20-25 % of total current. On the other hand, the slower time to peak recorded in siRNA-treated cells is unlikely to be due to an increased proportion of Nav1.7, because TTX-sensitive VGSCs tend to have faster kinetics than TTX-resistant VGSCs [56, 57]. This raises the possibility that downregulation of nNav1.5 may have induced secondary, knock-on effects, e.g. alteration of the relative profile of β-subunits, which can have a marked influence on VGSC functional properties [58-60]. Further work is required to determine the functional consequences of the β-subunits and Nav1.7 expression in MDA-MB-231 cells.

A role for nNav1.5 in potentiating metastatic activity

MDA-MB-231 cells were shown to express significantly more nNav1.5 protein at the plasma membrane than weakly metastatic MCF-7 cells [6]. In addition, migrated MDA-MB-231 cells were found to express more plasma membrane nNav1.5 protein than non-migrated cells. Together, these data suggest a correlation between surface expression of nNav1.5 and invasiveness in BCa, as in vivo [6].

In MDA-MB-231 cells, TTX (10 μM) treatment has been shown previously to suppress a variety of in vitro behaviours associated with the metastatic cascade, including galvanotaxis, endocytic membrane activity, migration and invasion, consistent with the involvement of TTX-resistant VGSC activity [6, 22]. The siRNA treatment reduced migration by 43 %, and this was not further suppressed by addition of TTX during the migration assay. Thus, the siRNA almost completely eliminated the VGSC-dependent component of the cells’ invasiveness. Furthermore, NESO-pAb, which can directly inhibit nNav1.5 functional activity [32], was found to reduce both the migration and invasion of MDA-MB-231 cells to a similar extent, without affecting their proliferation, when applied during the assays. Importantly, co-application of TTX did not further reduce the cells’ invasiveness below that of NESO-pAb alone. In addition, NESO-pAb had no effect on the invasion of PC-3M PCa cells, which, like the parent PC-3 cells (34), express primarily Nav1.7 [35] (J. K. J. Diss, unpublished observations), confirming further the specificity of NESO-pAb for nNav1.5. Furthermore, an anti-laminin antibody had no effect on the migration of MDA-MB-231 cells, consistent with the effect of NESO-pAb being specifically upon the nNav1.5 channel. Together, these data imply that nNav1.5 alone is primarily responsible for the VGSC-dependent potentiation of invasiveness in MDA-MB-231 cells. In both approaches, one interesting characteristic was apparent: blocking VGSC activity by only ~30 % was sufficient to eliminate its potentiating effect upon migration or invasion. This would imply that the relationship between VGSC activity and migration/invasion enhancement is rather steep.

Clinical implications

This work has important clinical implications, whereby reduction of nNav1.5 activity, e.g. generally with non-cytotoxic VGSC-blocking drugs [28], could result in suppression of metastatic BCa. More specifically, given that nNav1.5 expression is an oncofetal phenomenon [61] and is not present in normal Nav1.5-expressing adult tissues [6, 32], it appears to have the hallmarks of an ideal target for specific anticancer treatment utilising siRNA [62-64], and/or antibody [65].