b-Subunits of voltage-gated sodium channels in human ... - Nature

Prostate Cancer and Prostatic Diseases (2008) 11, 325–333 & 2008 Nature Publishing Group All rights reserved 1365-7852/08 $30.00 www.nature.com/pcan

ORIGINAL ARTICLE

b-Subunits of voltage-gated sodium channels in human prostate cancer: quantitative in vitro and in vivo analyses of mRNA expression JKJ Diss1,2, SP Fraser2, MM Walker3, A Patel3,4, DS Latchman1 and MBA Djamgoz2 1

Medical Molecular Biology Unit, Institute of Child Health, University College London, London, UK; 2Division of Cell and Molecular Biology, Neuroscience Solutions to Cancer Research Group, Imperial College London, London, UK; 3Department of Histopathology, Imperial College School of Medicine, St Mary’s Hospital, Norfolk Place, London, UK and 4Department of Urology, St Mary’s Hospital, Imperial College School of Medicine, London, UK

We previously identified high levels of Nav1.7 voltage-gated sodium channel a-subunit (VGSCa) mRNA and protein in human prostate cancer cells and tissues. Here, we investigated auxillary b-subunit (VGSCbs) expression. In vitro, the combined expression of all four VGSCbs was significantly (B4.5-fold) higher in strongly compared to weakly metastatic cells. This was mainly due to increased b1-expression, which was under androgenic control. In vivo, b1–b4 mRNAs were detectable and their expression in CaP vs non-CaP tissues generally reflected the in vitro levels in relation to metastatic potential. The possible role(s) of VGSCbs (VGSCa-associated and VGSCaindependent) in prostate cancer are discussed. Prostate Cancer and Prostatic Diseases (2008) 11, 325–333; doi:10.1038/sj.pcan.4501012; published online 25 September 2007

Keywords: sodium channel; a-subunit; b-subunit; real-time PCR; metastasis; androgen

Introduction Functional expression of voltage-gated sodium channels (VGSCs), determined by patch-clamp electrophysiological recording, was found to distinguish strongly and weakly metastatic human/rat prostate cancer (CaP) cells: PC-3/MAT-LyLu and LNCaP/AT-2, respectively.1,2 Importantly, blocking VGSC activity in CaP cells with the highly specific VGSC-blocker tetrodotoxin suppressed in vitro cellular behaviours integral to metastasis, especially directional motility and Matrigel invasion.1–4 Thus, VGSCs could actively contribute to the metastatic process. The VGSC a-subunit (VGSCa) expressed at the highest levels and most likely to be responsible for ion pore formation was Nav1.7.5 mRNA levels of this VGSCa were also markedly upregulated in CaP in vivo, increasing with Gleason grade.6 In fact, receiver operator curve analysis suggested that Nav1.7 could be an effective in vivo marker for CaP.6 VGSCas are often, but not always, physically associated with one or more smaller, transmembrane auxillary b-subunit proteins (VGSCbs). Four different VGSCb genes have been identified in the genomes of higher Correspondence: Dr JKJ Diss, Medical Molecular Biology Unit, Institute of Child Health, University College London, 30 Guilford Street, London WC1N 1EH, UK. E-mail: [email protected] Received 17 July 2007; accepted 3 August 2007; published online 25 September 2007

vertebrates.7–10 At least one of these (b1) can be expressed in different isoforms, generated by alternative splicing and intron-retention mechanisms.11–13 VGSCbs can affect VGSC functioning via multiple mechanisms, including modification of voltage-dependent gating, activation and inactivation.7,14–17 These subunits can also increase the amplitude of VGSC currents by (i) modulating the intracellular trafficking of VGSCa protein (thereby incorporating more VGSCs into the cell’s plasma membrane) and (ii) increasing their cell surface stability via association with cytoskeletal molecules such as ankyrin.7,8,18,19 VGSCbs are unique among voltage-gated ion channel auxillary subunits in that they can also function independently of pore-forming VGSCas, in cell adhesion, cell migration and in neurite outgrowth.20–23 Indeed, it has been proposed that the VGSCa-independent functions of VGSCbs may be more important than their role in modulating VGSC activity.19 VGSCbs participate in both homophilic cell–cell interaction (for example, resulting in ankyrin recruitment at contact points22), and heterophilic adhesion with contactin and neurofascin,23,24 but can also associate with extracellular matrix molecules (for example, tenascin-C and tenascin-R) to induce cellular repulsion.20,25 VGSCb-expression has not previously been investigated in CaP cells or tissues despite (i) adhesion/ migration being of fundamental importance to metastasis26 and (ii) co-transfection studies indicating that Nav1.7 can associate with VGSCbs.27,28 Here, we used real-time reverse transcriptase (RT)-PCR techniques to

Sodium channel b-subunits in prostate cancer JKJ Diss et al

326

investigate expression of VGSCb1-b4 mRNAs since antibodies to all four VGSCbs were not commercially available. Thus, VGSCb mRNA-expression in human prostatic cell lines of a range of metastatic potentials was determined and compared with expression in vivo, using biopsy samples taken from CaP patients and agematched non-CaP controls (in a quantitative approach similar to that which we used before to identify Nav1.7 as a potential marker for CaP6). The main aims were to determine (i) which VGSCb(s) were expressed and (ii) whether VGSCb mRNA-expression levels were altered in the disease state.

Materials and methods Tissue culture Three human CaP cell lines of increasing metastatic potential were used: LNCaP, PC-3 and PC-3M, maintained in RPMI-1640 growth medium containing 4 mM L-glutamine, 1.5 g l1 sodium bicarbonate, 1 mM Na pyruvate, 4.5 g l1 glucose and 10 mM HEPES. A fourth cell line, PNT2-C2,29 derived from ‘normal’ prostatic epithelial cells immortalized with SV40 was grown in RPMI-1640 medium containing only 4 mM L-glutamine. For all four cell lines, growth medium was supplemented with 10% fetal bovine serum and cultures were maintained in a 37 1C incubator with 5% CO2. Cells were plated in 10 cm dishes at six serial densities (ranging from 3.2 105 to 2.0 107 cells per dish) and harvested after 2 days. Androgen treatment PC-3 or LNCaP cells (6.25 105) were plated in normal growth medium in 10 cm dishes. After 2 days, the medium was removed, cells washed twice with phosphate-buffered saline and then re-fed with normal growth medium supplemented with 10% charcoalstripped fetal bovine serum containing either (i) 10 nM (5a, 17b)-17-hydroxy-androstan-3-one (dihydrotestosterone (DHT), the predominant androgen found in human prostate) in 0.01% ethanol; (ii) 100 nM DHT in 0.01% ethanol or (iii) 0.01% ethanol only. RNA was extracted from the cells after 1, 2 and 3 days of treatment. Prostate biopsy samples Twenty prostate tissue samples, consisting of five benign prostatic hyperplasia with no evident CaP (following thorough histopathological examination) and 15 CaP, were obtained from patients undergoing needle-biopsies or radical prostatectomy for biopsy-proven adenocarcinoma at St Mary’s Hospital, London, with local ethics approval. The patients varied in age from 47 to 83 years. Those with diagnosed CaP had summed Gleason gradings ranging from 4 to 8. All specimens were snapfrozen in liquid nitrogen in the operating room and stored at 80 1C until used for RNA extraction. RNA extraction and cDNA synthesis For cell lines, RNA was extracted using Trizol (Invitrogen, Paisley, UK), according to the manufacturer’s instructions, and then subjected to DNase I treatment (Amersham Pharmacia Biotech, Little Chalfont, UK). For tissues, RNA was extracted using the StrataPrep Total Prostate Cancer and Prostatic Diseases

RNA Miniprep kit (Stratagene, Cambridge, UK), as described previously.6 Approximate RNA quality was verified visually by gel electrophoresis and spectrophotometrically from 260/280 nm absorbance ratios. One microgram of total RNA, as determined by spectrophotometric measurement at 260 nm, was then used as the substrate for single-stranded cDNA synthesis using Superscript II reverse-transcriptase (Invitrogen). Reactions were primed with 250 ng random hexamer mix and performed in a 20 ml final volume. For real-time PCR experiments, this was subsequently made up to 100 ml with H2O, once the reverse-transcription process was completed.

Fixed endpoint PCR and VGSCb cloning Specific PCRs were performed on each of two extracts of the cell lines utilizing the primer pairs listed in Table 1. These PCRs amplified nucleotides corresponding to 58, 48, 54 and 67% respectively, of the coding regions of the VGSCbs. Amplified products spanned at least one conserved intron position. Reactions were performed on 2.5 ml of the synthesized cDNA, using 200 mM of each dNTP (Amersham Pharmacia Biotech), 1 unit of Taq DNA polymerase (Amersham Pharmacia Biotech), 1 Taq buffer and 0.5 mM of each primer, in a final volume of 20 ml, as follows: 94 1C for 5 min; 1 U enzyme added; 94 1C for 1 min; 60–64 1C for 1 min; 72 1C for 1 min, with the main section repeated 35–45 times (depending on the reaction). In each PCR run, a blank reaction without added cDNA was also performed to control for crosscontamination from other sources. All products were analysed by gel electrophoresis on 0.8% agarose gels and those derived from LNCaP and PC-3 cells were gel purified, ligated into the pGEM T-vector (Promega pGEM T-vector System) and used to transform E. coli (pMosBlue, Amersham Pharmacia Biotech). Plasmid DNA was recovered from bacterial cell cultures following blue/white, X-gal/b-galactosidase selection using a modified version of the Vistra Labstation 625 miniprep procedure (Vistra DNA Systems, Amersham Pharmacia Biotech) and resultant clones were sequenced using Amersham Thermo Sequenase fluorescent cycle sequencing kits and the Vistra DNA 725 automated sequencer, as described previously.5 Real-time PCRs Real-time PCR utilizing SYBR I Green technology was performed using the DNA Engine Opticon system (MJ Research, Waltham, MA, USA) to determine VGSCb-expression levels in the four cell lines. VGSCb primer pairs used were as described above. Three ‘control’ genes known to be expressed at similar levels in all four cell lines (b-actin, b2-microglobulin and NADH-cytochrome b5 reductase)5,6,30 were amplified from all cDNAs and used as ‘RNA load’ normalizing genes, controlling for possible variables including sample-to-sample differences in RNA input, RNA quality and reverse-transcription efficiency that could contribute to inter-sample expression differences. Reactions were performed simultaneously on each sample in triplicate, in a final volume of 20 ml, containing 5 ml cDNA, 500 nM of each specific primer and 1 QuantiTect SYBR Green PCR mix (Qiagen, Crawley, UK). Amplification was via


327

Table 1 Primer pairs and ATs used in PCRs on cell lines and tissues Target b1 b2 b3 b4 b-Actin b2-Microglobulin NADH-cytochrome b5 reductase Cytokeratin-8 PSA Nav1.7 IL-6

Primer sequences 0

0

5 -AGAAGGGCACTGAGGAGTTT-3 50 -GCAGCGATCTTCTTGTAGCA-30 50 -GAGATGTTCCTCCAGTTCCG-30 50 -TGACCACCATCAGCACCAAG-30 50 -CTGGCTTCTCTCGTGCTTAT-30 50 -TCAAACTCCCGGGACACATT-30 50 -TAACCCTGTCGCTGGAGGTG-30 50 -TGAGGATGAGGAGCCCGATG-30 50 -AGCCTCGCCTTTGCCGA-30 50 -CTGGTGCCTGGGGCG-30 50 -TGCTGTCTCCATGTTTGATGTATCT-30 50 -TCTCTGCTCCCCACCTCTAAGT-30 50 -TATACACCCATCTCCAGCGA-30 50 -CATCTCCTCATTCACGAAGC-30 50 -GCGGCGCACAAAGACTGAGA-30 50 -ACTTGGCGTTGGCATCCTTA-30 50 -TGCCCACTGCATCAGGAACA-30 50 -GTCCAGCGTCCAGCACAC-30 50 -TATGACCATGAATAACCCGC-30 50 -TCAGGTTTCCCATGAACAGC-30 50 -GGTACATCCTCGACGGCATCT-30 50 -GTGCCTCTTTGCTGCTTTCAC-30

AT (1C)

AS (nt)

60

379

62

310

60

353

64

459

67

167

60

80

60

492

62

158

60

420

60

389

62

81

Abbreviations: AS, amplicon size; ATs, annealing temperatures; PSA, prostate-specific antigen.

an initial denaturation at 95 1C for 15 min to activate the HotStar Taq, with subsequent three-step cycling of 95 1C for 30 s, 59–67 1C (depending on the primer pair) for 30 s and 72 1C for 30 s. A final melt curve was also carried out from 65 to 95 1C with 0.3 1C steps and gel electrophoresis performed to verify product composition. Additionally, blank control reactions without added cDNA were carried out in each PCR run. Initially, PCRs for the four VGSCb subtypes were carried out separately, permitting comparison of the relative-expression levels across the four cell lines. Subsequently, the relative proportion of each VGSCb subtype comprising the total VGSCbexpression in the cells was determined by performing a PCR run (60 1C annealing temperature) on single reactions, each containing cDNA derived from the cell lines when grown at the different cell densities. Finally, the relative levels of b1 and Nav1.7 mRNAs in PC-3 and PC3M cells were compared (60 1C annealing temperature), using the Nav1.7 primers shown in Table 1. Experiments with prostate tissues were performed as above except that b-actin, b2-microglobulin and cytokeratin-831 were utilized as normalizing genes.

Data analysis For real-time PCRs, the threshold amplification cycle (Ct) representing the first discernable amplification cycle above which, the product fluorescence was above the background noise was determined for each reaction using the Opticon software. Subsequently, these values were analysed using the 2DDCt method6,32 to determine target-expression levels (expressed as means7s.d.). This analysis method was valid since all VGSCb subtype PCRs worked with similar efficiencies, as determined by reactions performed on serial dilutions of cDNA derived from LNCaP, human umbilical vein endothelial cells and RH-1 neuroblastoma cells (data not shown). The VGSCbexpression in each sample was normalized relative to the sample-matched expression level of all three normalizing

genes. Differences in VGSCb-expression levels between strongly and weakly metastatic cell lines were tested for statistical significance using analysis of variance followed by Bonferroni post hoc tests and P-values o0.05 were considered significant. In vitro and in vivo mRNAexpression differences for (a) androgen-treated vs nontreated CaP cells, (b) CaP vs non-CaP samples and (c) low-grade vs high-grade CaPs, were tested for statistical significance by comparing normalized means and medians using the non-parametric Mann–Whitney U-test (a and b) and the Kruskal–Wallis test (c).33

Results Expression of VGSCb subunits in vitro Fixed endpoint PCRs showed that all four known VGSCbs were expressed in each of the prostate epithelial cell lines studied (Figure 1a). Two VGSCbs (b3 and b4) were present in more than one splice form, with RT-PCRs amplifying products from an exon-skipped form expressed at lower levels than the full-length transcript (Figure 1a). These isoforms, if translated, would code for highly truncated VGSCbs lacking the transmembrane domain with presently unknown function, if any. Using PC-3M cells for relative measurements, real-time RT-PCR determined that the total VGSCb level in strongly metastatic cells was on average 4.5-fold higher than in non/weakly metastatic cells: 20.578.6% (combined PNT2-C2 and LNCaP mean) vs 92.4710.8% (combined PC-3 and PC-3M mean) (Po0.05; Figure 1b). These data were essentially the same using all three normalizing genes (not shown). The four VGSCb subtypes were expressed at markedly different levels within each cell line (Figure 1c). b1 was predominant in every case, typically constituting 490% of the total VGSCb mRNA expression. b3 (3–7%) and b4 (o2%) were present at lower levels, and b2-expression was notably very low in all four cell lines studied. Prostate Cancer and Prostatic Diseases


328

∗

100

1

3

2

4

2

3

1

2

3

∆ 1

3

189 nt

∆

379 nt

310 nt

3

353 nt

3

2

4

2

3

4

Total relative VGSC mRNA expression

PN LN PC PM B

4

459 nt

80

60

40

20

305 nt 0 PNT2

Proportional expression of VGSCs

PC-3

PC-3M

∗

∗

120

LNCaP

∗

1

100

2 ∗

80

3 4

60

40 ∗ 20

0 PNT2

LNCaP

PC-3

PC-3M

10 8 6 4 2 0

140

140

120

120 mRNA expression (% PC-3 1 levels)

VGSC mRNA expression (fold. units)

12

(% LNCaP)

PNT2 LNCaP PC-3 PC-3M

14

4 mRNA expression

Figure 1 VGSCb-expression in prostatic epithelial cell lines. (a) Detectable expression of b1–b4 mRNAs by PCR in four cell lines of increasing metastatic potential PNT2-C2 (‘PN’), LNCaP (‘LN’), PC-3 (‘PC’) and PC-3M (‘PM’). No amplification was evident from the blank controls (‘B’). (b) Histogram showing total VGSCb-expression in the four prostatic epithelial cell lines, normalized with b2 microglobulin. Total mRNA levels were significantly higher in combined strongly metastatic compared to weakly/non-metastatic cells (denoted by an asterisk). Similar results were obtained with b-actin and NADH-cytochrome b5 reductase normalizations. (c) Histogram showing the relative proportion of each subtype in the VGSCb transcriptome of the four prostatic epithelial cell lines. b1 was the predominant VGSCb in all cell lines. b4 constituted a larger fraction of the VGSCb transcriptome in LNCaP compared to the other cell lines.

100 80 60 40

2

3

4

80 60 40 20

20 0

1

100

0 PNT2 LNCaP PC-3 PC-3M

1 3 Nav1.7 PC-3

1 3 Nav1.7 PC-3M

1 3 Nav1.7 LNCaP

Figure 2 Quantitative differences in VGSCb subtypes of CaP cells in vitro. (a) Plot showing the expression levels of each VGSCb subtype relative to b4 levels in PC-3 cells. (b) Histogram displaying b4-expression in the four cell types, relative to levels in LNCaP cells. (c) Plot showing b1 and b3 mRNA-expression relative to Nav1.7 levels in PC-3, PC-3M and LNCaP cells. b1 and b3 mRNA levels were significantly different from those of Nav1.7 in all three cell lines. All data shown in (a)–(c) were normalized with b2 microglobulin. Similar results were obtained with b-actin and NADH-cytochrome b5 reductase normalizations. Each bar denotes mean7s.e. Asterisks denote statistical significance (Po0.05).

Increased total VGSCb-expression in strongly metastatic cell lines was mainly a result of increased b1expression (Figure 2a). On average, b1-expression was 5-fold higher in strongly vs weakly metastatic cells: Prostate Cancer and Prostatic Diseases

1.770.4-fold units (PNT2-C2 and LNCaP mean) vs 8.572.9-fold units (PC-3 and PC-3M mean) (Po0.01; Figure 2a). b2- and b3-expression was generally similar in all cells. On the other hand, b4 levels, although very


levels was apparent following DHT application at 24 and 48 h time points but this effect was lost after 72 h of treatment. Surprisingly, PC-3 cells also responded to DHT treatment with significant (two- to threefold) timedependent increases in b1 and Nav1.7 mRNA expression (Figure 3b).

low, were reversed, being significantly greater in LNCaP and PNT2-C2 cells compared to the strongly metastatic cells. Indeed, PC-3 cells expressed only B1% of the b4 level in LNCaP cells (Po0.05; Figure 2b). The observed cell line differences in VGSCb levels were maintained at comparable cell densities over an B100-fold range (data not shown). In summary, in CaP cells of strong metastatic potential, expression of b1 was significantly greater and b4 (present at a much lower level) significantly lower, while b2/b3 remained unchanged, compared to cells of weak metastatic potential.

VGSCb mRNA expression in vivo mRNAs of the four VGSCbs were readily detected in all human prostate biopsy tissues tested, as found in vitro. However, (i) the derived Ct values for the VGSCb subtypes were similar, indicating that expression levels of b1–b4 were comparable in the CaP/non-CaP biopsy samples, unlike the cell lines, and (ii) the medianexpression levels were not significantly different between non-CaP and CaP tissues (Figures 4a–d). In contrast, Nav1.7 and IL-6, known to be expressed at greatly increased levels in CaP vs non-CaP tissues,6,35 showed 19- and 42-fold median increases, respectively, in CaP vs non-CaP tissues and these differences were significant (Po0.05 for both; Figures 4e and f). These data were essentially the same when normalized with respect to all three normalizing genes. VGSCb-expression was also analysed within the CaP samples in relation to Gleason score divided into high (sum X6) and low (sumo6) grades.4 This analysis also did not reveal a significant difference in VGSCb-expression in high-grade vs low-grade CaP samples. A different approach, plotting in vivo vs in vitro fold-changes in b1–b4 and Nav1.7 mRNA-expression levels revealed a significant positive linear correlation, close to the origin (Figure 5). This analysis of fold differences suggested that VGSCbexpression was likely to be altered in CaP in vivo in line with the in vitro changes.

Comparison with VGSCa-expression in vitro We next analysed VGSCb mRNAs with respect to the expression of the predominant VGSCa in metastatic CaP cells, Nav1.7.6 A very similar profile was obtained for PC3 and PC-3M cells (Figure 2c). Levels of the predominant b1 were 6- to 8-fold higher than Nav1.7 in strongly metastatic cells (PC-3 and PC-3M cells, respectively; Po0.05 for both; Figure 2c). In contrast, b3, the second most abundant VGSCb in these cells, was expressed four- to sixfold lower than Nav1.7. Thus, assuming similar translation efficiencies, strongly metastatic cells could contain significant amounts of b1 not associated with Nav1.7. It would follow from the 5-fold difference in b1 mRNA levels between strongly and weakly metastatic cells that even PNT2-C2 and LNCaP cells would possess sufficient b1 to associate with all the Nav1.7 in PC-3 and PC-3M cells (Figure 2c). Effect of androgen treatment on b1 mRNA expression Since strongly and weakly metastatic cells also differ in their androgen sensitivity, we investigated whether stimulating the androgen receptor (AR) with DHT might reduce VGSCb-expression. Application of 10 or 100 nM DHT to androgen-deprived AR-positive LNCaP cells over a 3-day time period increased mRNA levels of the prostate-specific antigen gene, used here as a positive control, by up to 15-fold in a time-dependent manner (Figure 3a), as reported previously.34 Importantly, DHT treatment significantly decreased b1-expression in LNCaP cells (Figure 3b). A 1.6-fold reduction in b1 22

The main results of the present study are as follows: (1) mRNAs of all four VGSCbs were expressed in human prostate epithelial cell lines, overall at much higher levels in strongly vs weakly metastatic cells. (2) b1 was predominant and its expression level was much higher 5

18 16 14 12 10 8

∗

β1 LNCaP β1 PC-3

4

∗ ∗

6 ∗

4 2

∗

0

mRNA expression (vs control)

PSA mRNA expression (vs control)

Discussion

∗

20

329

Nav1.7PC-3 ∗ ∗

3 2 1

∗ ∗

∗

∗

0 Con 10nM 100nM DAY 1

Con 10nM 100nM DAY 2

Con10nM 100nM DAY 3




Figure 3 Effect of androgen treatment on b1 mRNA-expression in vitro. Plots showing (a) the expression level of the androgen-responsive prostate-specific antigen gene in LNCaP cells, and (b) the expression of b1 in LNCaP (black bars) and PC-3 (white bars) cells and Nav1.7 in PC-3 cells (grey bars), treated with 10 and 100 nM dihydrotestosterone for 1, 2 and 3 days relative to levels in untreated cells. All data shown were normalized with both NADH-cytochrome b5 reductase and b2 microglobulin. Each bar denotes mean7s.e. Asterisks denote statistical significance (Po0.05). Prostate Cancer and Prostatic Diseases


330

4.0

3.5

3.5

2.5 2 1.5 1

1.05 n=15

1.00 n=5

2 mRNA expression

1 mRNA expression

3

3.0 2.5 2.0 1.5 1.0

1.00 n=5

0.5

0.64 n=15

0.5 0.0

0 Non-CaP

CaP

Non-CaP

CaP

2

3.5

1.8

2.5 2 1.5 1

1.00

0.99 n=15

n=5

4 mRNA expression

3 mRNA expression

3

0.5

1.6 1.4 1.2 1 0.8

1.00 n=5

0.6

0.54

0.4

n=15

0.2 0

0 Non-CaP

CaP

Non-CaP

160

CaP

120

100

60

40 n=15 19.8

20 n=5 0

80

60

40 n=13 19.9

20 n=5

1.0 Non-CaP

IL-6 mRNA expression

Nav1.7 mRNA expression

100 80

0 CaP

1.0 Non-CaP

CaP

Figure 4 VGSCb mRNA-expression in vivo in human non-CaP and CaP tissues. (a–f) Plots of b1 (a), b2 (b), b3 (c), b4 (d), Nav1.7 (e) and IL-6 (f) mRNA-expression levels in non-CaP vs CaP patients, normalized to b2-microglobulin. Medians are shown as horizontal lines with the value, and the number of cases, given on the side. All amplicons were electrophoresed to verify identity (not shown). b3 and b4 PCRs amplified exon-skipped (D) isoforms (as found in vitro) in addition to full-length amplicons, but at much lower expression levels.

than the predominant VGSCa, Nav1.7. Accordingly, b1 would probably have a functional role both in association with, and independently of, Nav1.7. (3) Treatment with an androgen analogue decreased b1-expression in androgen-sensitive CaP cells, but increased b1 levels in Prostate Cancer and Prostatic Diseases

androgen-insensitive CaP cells apparently via AR and non-AR mechanisms, respectively. (4) In vivo, all four VGSCbs were readily detectable. Although no significant differences were evident in CaP and non-CaP tissues, median fold-changes in vivo (non-CaP vs CaP) correlated


25 ♦

20 R2 = 0.76

15 10

magnitude of the in vivo change for b1 and b4 is consistent with the B50-fold ‘dilution’ of the in vitro change observed previously for Nav1.7. Together, these analyses would suggest that our in vitro and in vivo findings show considerable consistency. Further work exclusively on the epithelial/CaP cells of micro-dissected biopsy tissues would clarify the in vivo situation further.

331

5 -2♦

♦

-1

♦ 1

2

3

-5 -10 Figure 5 VGSC in vitro and in vivo mRNA-expression differences are correlated. Plot of the logarithmic in vitro fold change against the in vivo fold change for b1, b3, b4 and Nav1.7. In vitro fold changes represent mean expression differences between strongly and weakly metastatic cells, with the sign indicating whether expression was greater in strongly (positive) or weakly (negative) metastatic cells. In vivo fold changes similarly represent median expression differences between non-CaP and CaP tissues (positive if greater in CaP). A regression line was fitted with a regression coefficient (r2 value) of 0.76.

well with corresponding in vitro changes (between strongly and weakly metastatic cells).

VGSCa/b-expression in CaP VGSCa mRNA levels were shown previously to be significantly upregulated with increased metastatic potential in CaP with 420-fold higher overall expression in strongly compared to weakly metastatic human CaP cell lines, including an B1000-fold elevation in the expression of a particular VGSCa subtype, Nav1.7.5 However, this B1000-fold in vitro elevation translated to a much smaller, but still measurable B20-fold significant increase in Nav1.7 mRNA in CaP vs non-CaP in vivo.6 In the present study, we found that VGSCb mRNA levels were much less altered with cellular metastatic potential in vitro. Thus, VGSCb mRNA levels overall were only B4.5fold higher in strongly metastatic (PC-3 and PC-3M) compared to weakly metastatic (PNT2-C2 and LNCaP) cells, the difference being almost entirely due to the upregulation of b1. No statistically significant difference was evident in the expression of any of the four VGSCbs between CaP and non-CaP in vivo. This may have been due to the nature of the whole biopsy tissue chunks, with epithelial cells (from which CaP cells arise) only constituting a small fraction of the overall content. In such a situation, any epithelial cell-specific VGSC-expression changes with CaP in vivo would be hard to detect, especially if the non-epithelial cell types present in prostate (for example, smooth muscle cells, peripheral nerves, fibroblasts) also expressed significant levels of VGSCs, which is very likely.36–39 Indeed, although not statistically significant, median fold-changes in vivo (non-CaP vs CaP) did appear to correlate very well with corresponding in vitro changes (between strongly and weakly metastatic cells), as shown in Figure 5. In particular, (a) the direction of the change is the same for b1 (increased with disease progression in vitro and in vivo) and b4 (decreased with disease progression), and (b) the

VGSCa and VGSCb interactions in CaP cells The predominant VGSC a- and b-subunits in CaP cells, Nav1.7. and b1, are also coexpressed in numerous other cell types, such as Schwann cells.40 Importantly, when coexpressed in Xenopus oocytes, Nav1.7 and b1 were found to functionally associate in a 1:1 ratio, with the resultant VGSC currents exhibiting more rapid inactivation, a small but significant (B5 mV) hyperpolarizing shift in steady-state activation and inactivation, and accelerated recovery from fast inactivation, compared to currents resulting from Nav1.7 a-subunit expression alone.41,42 In the present study, we found that Nav1.7 mRNA levels were much lower than b1, even in strongly metastatic cells (Figure 2c). Indeed, assuming similar translation efficiencies, there is more than required b1-expression in CaP cells (regardless of metastatic potential) to associate in a 1:1 ratio with Nav1.7. This may explain why there is no need for a large increase in b1-expression (concomitant with the large increase in VGSCa-expression) in strongly metastatic CaP cells, and would also be consistent with overexpression of VGSCa alone in LNCaP cells being sufficient to increase significantly their in vitro invasiveness.43 Despite the absence of a large increase in VGSCbexpression, these subunits might still have an important role in VGSC functioning in CaP, and thus in VGSCmediated potentiation of metastatic cell behaviours. Indeed, reducing b1-expression or perturbing its association with Nav1.7 could alter key channel properties such as activation/inactivation and recovery from fast inactivation that would result in less VGSC activity and inhibition of VGSC-potentiated metastatic cell behaviour. Furthermore, a reduction in VGSCb-expression would also likely cause a reduction in the amount of functional (cell surface) sodium channels via intracellular VGSCa protein trafficking and reduced cell surface stability mechanisms. It is also noteworthy that VGSCb association can significantly alter the efficacy of VGSC-blocking drugs. For example, blockage of Nav1.3 by lidocaine was reduced in a concentration-dependent manner by coexpression with b1 and to a lesser extent with b3.44 Thus, reduction of VGSCb-expression, and b1 in particular, could be a successful strategy for reducing VGSC activity (and hence metastatic cell behaviour) via several mechanisms. VGSCa-independent roles for VGSCbs in CaP cells If VGSCbs are more abundant than VGSCas in CaP cells, they will either be ‘inactive’ in a large intracellular pool awaiting VGSCa upregulation, or, perhaps more likely, may be functioning independently of VGSCas. Important VGSCa-independent roles of VGSCbs in cell adhesion and migration, via interactions with extracellular matrix and cell adhesion molecules, have been documented.19–25 Such effects are known to also be of fundamental importance to the metastatic process.26 Prostate Cancer and Prostatic Diseases


332

More recently, a novel role for VGSCbs in p53-dependent apoptosis has been suggested.45 Two p53-response elements were located in the b3 gene and b3 was found to be markedly upregulated by (a) p53 overexpression and (b) DNA damage in p53-expressing cell lines. Furthermore, overexpression of b3 with p53 in glioma and osteosarcoma cells reduced their colony-forming ability. Since PC-3 and PC-3M cell lines carry mutations of the p53 gene that prevent its normal expression,46 the functional role of b3 may be different in strongly vs weakly/non-metastatic cells.

Androgen regulation of VGSCbs in CaP cells Differences in b1 levels in the different CaP cell lines raised the possibility that its expression in these cells is under the control of androgen. Indeed, treatment of androgen-deprived, AR-positive LNCaP cells with DHT revealed an inhibitory influence of androgen on b1expression, consistent with lower basal levels of b1 mRNA compared with AR-negative cells. Furthermore, and surprisingly, androgen treatment of PC-3 cells increased b1-expression. Since these cells do not possess functional AR this effect must be mediated by a non-AR mechanism. Thus, b1-expression in CaP cells is under androgen control in both AR-positive and negative cells but in opposite directions. Importantly, Nav1.7-expression was also affected (increased) by androgen in ARnegative cells, consistent with the proposed role of Nav1.7 in CaP metastasis and the classic role of androgen in CaP progression.47 The effects of androgen on ARnegative cells may involve cross-talk with growth factor (for example EGF) signalling pathways47 also known to regulate VGSC expression and activity.48

Conclusion In the present study, we found that auxillary b-subunits of VGSCs are expressed in CaP cells in vitro and in vivo. VGSCb-expression was significantly higher in strongly compared to weakly metastatic cells in vitro, mainly due to increased b1-expression. However, this increase was significantly less than that previously observed for the predominant VGSCa, Nav1.7. Median VGSCb-expression levels generally reflected in vitro changes, but did not individually reach statistical significance. Even without strikingly altered expression, VGSCbs are likely to have dual importance in CaP. First, VGSCbs, especially b1, can modulate VGSC functioning and, thus, VGSC-mediated potentiation of metastasis. Second, independently of VGSCas, VGSCbs can have an important role(s) in modulating cell adhesion, an integral part of the metastatic cascade. Accordingly, reducing-expression of the predominant VGSCb (b1) or perturbing its association with Nav1.7 could be a successful strategy for reducing metastatic cell behaviour.

Acknowledgements This study was supported by the Pro Cancer Research Fund (PCRF). We thank Dr Rory Curtis (Elixir Pharmaceuticals) for his comments on the manuscript and Petros Prostate Cancer and Prostatic Diseases

Andrikopoulos for his assistance with the real-time PCR experiments.

References 1 Grimes JA, Fraser SP, Stephens GJ, Downing JE, Laniado ME, Foster CS et al. Differential expression of voltage-activated Na+ currents in two prostatic tumour cell lines: contribution to invasiveness in vitro. FEBS Letts 1995; 369: 290–294. 2 Laniado ME, Lalani EN, Fraser SP, Grimes JA, Bhangal G, Djamgoz MB et al. Expression and functional analysis of voltageactivated Na+ channels in human prostate cancer cell lines and their contribution to invasiveness in vitro. Am J Path 1997; 150: 1213–1221. 3 Djamgoz MB, Mycielska M, Madeja Z, Fraser SP, Korohoda W. Directional movement of rat prostate cancer cells in directcurrent electric field: involvement of voltage gated Na+ channel activity. J Cell Sci 2001; 114: 2697–2705. 4 Fraser SP, Salvador V, Manning EA, Mizal J, Altun S, Raza M et al. Contribution of functional voltage-gated Na+ channel expression to cell behaviors involved in the metastatic cascade in rat prostate cancer: I. Lateral motility. J Cell Physiol 2003; 195: 479–487. 5 Diss JK, Archer SN, Hirano J, Fraser SP, Djamgoz MB. Expression profiles of voltage-gated Na+ channel a-subunit genes in rat and human prostate cancer cell lines. Prostate 2001; 48: 165–178. 6 Diss JK, Stewart D, Pani F, Foster CS, Walker MM, Patel A et al. A potential novel marker for human prostate cancer: voltage-gated sodium channel expression in vivo. Prostate Cancer Prostate Dis 2005; 8: 266–273. 7 Isom LL, De Jongh KS, Patton DE, Reber BF, Offord J, Charbonneau H et al. Primary structure and functional expression of the beta 1 subunit of the rat brain sodium channel. Science 1992; 256: 839–842. 8 Isom LL, Ragsdale DS, De Jongh KS, Westenbroek RE, Reber BF, Scheuer T et al. Structure and function of the beta 2 subunit of brain sodium channels, a transmembrane glycoprotein with a CAM motif. Cell 1995; 83: 433–442. 9 Morgan K, Stevens EB, Shah B, Cox PJ, Dixon AK, Lee K et al. beta 3: an additional auxiliary subunit of the voltage-sensitive sodium channel that modulates channel gating with distinct kinetics. Proc Nat Acad Sci USA 2000; 97: 2308–2313. 10 Yu FH, Westenbroek RE, Silos-Santiago I, McCormick KA, Lawson D, Ge P et al. Sodium channel beta4, a new disulfidelinked auxiliary subunit with similarity to beta2. J Neurosci 2003; 23: 7577–7585. 11 Oh Y, Waxman SG. Beta 1 subunit mRNA of the rat brain Na+ channel is expressed in glial cells. Proc Natl Acad Sci USA 1994; 91: 9985–9989. 12 Kazen-Gillespie KA, Ragsdale DS, D’Andrea MR, Mattei LN, Rogers KE, Isom LL. Cloning, localization, and functional expression of sodium channel beta1A subunits. J Biol Chem 2000; 275: 1079–1088. 13 Qin N, D’Andrea MR, Lubin ML, Shafaee N, Codd EE, Correa AM. Molecular cloning and functional expression of the human sodium channel beta1B subunit, a novel splicing variant of the beta1 subunit. Eur J Biochem 2003; 270: 4762–4770. 14 Fahmi AI, Patel M, Stevens EB, Fowden AL, John JE, Lee K et al. The sodium channel beta-subunit SCN3b modulates the kinetics of SCN5a and is expressed heterogeneously in sheep heart. J Physiol 2001; 537: 693–700. 15 Qu Y, Curtis R, Lawson D, Gilbride K, Ge P, Di Stefano PS et al. Differential modulation of sodium channel gating and persistent sodium currents by the beta1, beta2, and beta3 subunits. Mol Cell Neurosci 2001; 18: 570–580. 16 Meadows LS, Chen YS, Powell AJ, Clare JJ, Ragsdale DS. Functional modulation of human brain Nav1.3 sodium channels,


17

18

19

20

21

22

23

24

25

26 27

28

29

30

31

expressed in mammalian cells, by auxiliary beta 1, beta 2 and beta 3 subunits. Neuroscience 2002; 114: 745–753. Vijayaragavan K, Powell AJ, Kinghorn IJ, Chahine M. Role of auxiliary beta1-, beta2-, and beta3-subunits and their interaction with Na(v)1.8 voltage-gated sodium channel. Biochem Biophys Res Commun 2004; 319: 531–540. Chen C, Bharucha V, Chen Y, Westenbroek RE, Brown A, Malhotra JD et al. Reduced sodium channel density, altered voltage dependence of inactivation, and increased susceptibility to seizures in mice lacking sodium channel beta 2-subunits. Proc Nat Acad Sci USA 2002; 99: 17072–17077. Malhotra JD, Koopmann MC, Kazen-Gillespie KA, Fetman N, Hortsch M, Isom LL. Structural requirements for interaction of sodium channel beta 1 subunits with ankyrin. J Biol Chem 2002; 277: 26681–26688. Xiao ZC, Ragsdale DS, Malhotra JD, Mattei LN, Braun PE, Schachner M et al. Tenascin-R is a functional modulator of sodium channel beta subunits. J Biol Chem 1999; 274: 26511–26517. Malhotra JD, Kazen-Gillespie K, Hortsch M, Isom LL. Sodium channel beta subunits mediate homophilic cell adhesion and recruit ankyrin to points of cell-cell contact. J Biol Chem 2000; 275: 11383–11388. Davis TH, Chen C, Isom LL. Sodium channel beta 1 subunits promote neurite outgrowth in cerebellar granule neurons. J Biol Chem 2004; 279: 51424–51432. Kazarinova-Noyes K, Malhotra JD, McEwan DP, Mattei LN, Berglund EO, Ranscht B et al. Contactin associates with Na+ channels and increases their functional expression. J Neurosci 2001; 21: 7517–7525. Ratcliffe CF, Westenbroek RE, Curtis R, Catterall WA. Sodium channel beta1 and beta3 subunits associate with neurofascin through their extracellular immunoglobulin-like domain. J Cell Biol 2001; 154: 427–434. Srinivasan J, Schachner M, Catterall WA. Interaction of voltagegated sodium channels with the extracellular matrix molecules tenascin-C and tenascin-R. Proc Nat Acad Sci USA 1998; 95: 15753–15757. Okegawa T, Li Y, Pong RC, Hsieh JT. Cell adhesion proteins as tumor suppressors. J Urol 2002; 167: 1836–1843. Johnson D, Montpetit ML, Stocker PJ, Bennett ES. The sialic acid component of the beta1 subunit modulates voltage-gated sodium channel function. J Biol Chem 2004; 279: 44303–44310. Chevrier P, Vijayaragavan K, Chahine M. Differential modulation of Nav1.7 and Nav1.8 peripheral nerve sodium channels by the local anesthetic lidocaine. Br J Pharmacol 2004; 142: 576–584. Berthon P, Dimitrov T, Stower M, Cussenot O, Maitland NJ. A microdissection approach to detect molecular markers during progression of prostate cancer. Br J Cancer 1995; 72: 946–951. Moriyama N, Yamaguchi T, Takeuchi T, Sakamoto E, Ueki T, Tsujimoto G et al. Semi-quantitative evaluation of a1A-adrenoceptor subtype mRNA in human hypertrophied and nonhypertrophied prostates: regional comparison. Life Sci 1999; 64: 201–210. Hudson DL, Guy AT, Fry P, O’Hare MJ, Watt FM, Masters JR. Epithelial cell differentiation pathways in the human prostate:

32

33

34

35

36

37

38 39

40

41

42

43

44

45

46

47 48

identification of intermediate phenotypes by keratin expression. J Hist Cytochem 2001; 49: 271–278. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods 2001; 25: 402–408. De Kok JB, Verhaegh GW, Roelofs RW, Hessels D, Kiemeney LA, Aalders TW et al. DD3(PCA3), a very sensitive and specific marker to detect prostate tumors. Cancer Res 2002; 62: 2695–2698. DePrimo SE, Diehn M, Nelson JB, Reiter RE, Matese J, Fero M et al. Transcriptional programs activated by exposure of human prostate cancer cells to androgen. Genome Biol 2002; 3: RESEARCH0032.1–0032.12. Corcoran NM, Costello AJ. Interleukin-6: minor player or starring role in the development of hormone-refractory prostate cancer? BJU Int 2003; 91: 545–553. Jo T, Nagata T, Iida H, Imuta H, Iwasawa K, Ma J et al. Voltagegated sodium channel expressed in cultured human smooth muscle cells: involvement of SCN9A. FEBS Lett 2004; 567: 339–343. Gosling M, Harley SL, Turner RJ, Carey N, Powell JT. Human saphenous vein endothelial cells express a tetrodotoxinresistant, voltage-gated sodium current. J Biol Chem 1998; 273: 21084–21090. Estacion M. Characterization of ion channels seen in subconfluent human dermal fibroblasts. J Physiol 1991; 436: 579–601. Fraser SP, Diss JK, Lloyd LJ, Pani F, Chioni AM, George AJ et al. T-lymphocyte invasiveness: control by voltage-gated Na+ channel activity. FEBS Lett 2004; 569: 191–194. Hinson AW, Gu XQ, Dib-Hajj S, Black JA, Waxman SG. Schwann cells modulate sodium channel expression in spinal sensory neurons in vitro. Glia 1997; 21: 339–349. Shcherbatko A, Ono F, Mandel G, Brehm P. Voltage-dependent sodium channel function is regulated through membrane mechanics. Biophys J 1999; 77: 1945–1959. Vijayaragavan K, O’Leary ME, Chahine M. Gating properties of Na(v)1.7 and Na(v)1.8 peripheral nerve sodium channels. J Neurosci 2001; 21: 7909–7918. Bennett ES, Smith BA, Harper JM. Voltage-gated Na+ channels confer invasive properties to human prostate cancer cells. Pflugers Arch 2004; 447: 908–914. Lenkowski PW, Shah BS, Dinn AE, Lee K, Patel MK. Lidocaine block of neonatal Nav1.3 is differentially modulated by co-expression of beta1 and beta3 subunits. Eur J Pharmacol 2003; 467: 23–30. Adachi K, Toyota M, Sasaki Y, Yamashita T, Ishida S, Ohe-Toyota M et al. Identification of SCN3B as a novel p53-inducible proapoptotic gene. Oncogene 2004; 23: 7791–7798. Van Bokhoven A, Varella-Garcia M, Korch C, Johannes WU, Smith EE, Miller HL et al. Molecular characterization of human prostate carcinoma cell lines. Prostate 2003; 57: 205–225. Culig Z, Klocker H, Bartsch G, Hobisch A. Androgen receptors in prostate cancer. Endocr Relat Cancer 2002; 9: 155–170. Wada A. Roles of voltage-dependent sodium channels in neuronal development, pain, and neurodegeneration. J Pharmacol Sci 2006; 102: 253–268.

333

Prostate Cancer and Prostatic Diseases