Valproic acid inhibits adhesion of vincristine- and cisplatin-resistant neuroblastoma tumour cells to endothelium

Drug resistance to chemotherapy is often associated with increased malignancy in neuroblastoma (NB). In pursuit of alternative treatments for chemoresistant tumour cells, we tested the response of multidrug-resistant SKNSH and of vincristine (VCR)-, doxorubicin (DOX)-, or cisplatin (CDDP)-resistant UKF-NB-2, UKF-NB-3 or UKF-NB-6 NB tumour cell lines to valproic acid (VPA), a differentiation inducer currently in clinical trials. Drug resistance caused elevated NB adhesion (UKF-NB-2VCR, UKF-NB-2DOX, UKF-NB-2CDDP, UKF-NB-3VCR, UKF-NB-3CDDP, UKF-NB-6VCR, UKF-NB-6CDDP) to an endothelial cell monolayer, accompanied by downregulation of the adhesion receptor neural cell adhesion molecule (NCAM). Based on the UKF-NB-3 model, N-myc proteins were enhanced in UKF-NB-3VCR and UKF-NB-3CDDP, compared to the drug naïve controls. p73 was diminished, whereas the p73 isoform deltaNp73 was upregulated in UKF-NB-3VCR and UKF-NB-3CDDP. Valproic acid blocked adhesion of UKF-NB-3VCR and UKF-NB-3CDDP, but not of UKF-NB-3DOX, and induced the upregulation of NCAM surface expression, NCAM protein content and NCAM coding mRNA. Valproic acid diminished N-myc and enhanced p73 protein level, coupled with downregulation of deltaNp73 in UKF-NB-3VCR and UKF-NB-3CDDP. Valproic acid also reverted enhanced adhesion properties of drug-resistant UKF-NB-2, UKF-NB-6 and SKNSH cells, and therefore may provide an alternative approach to the treatment of drug-resistant NB by blocking invasive processes.

Multiple-agent chemotherapy is the conventional therapy for patients with advanced stages of neuroblastoma (NB) and disseminated NB. However, drug resistance arises in the majority of stage IV and relapsed NB, often leading to treatment failure (Keshelava et al, 1998). Development of novel antitumoural strategies is therefore highly desired to overcome resistance mechanisms and to prevent tumour progression. Molecules modulating cellular function have been identified in the majority of tumours and their manipulation might be the key to decreasing malignancy.
Histone deacetylases (HDAC) represent one of the most important intracellular targets, as these molecules modulate a wide variety of cellular functions. Abnormal histone acetylation status can result in undesirable phenotypic changes, including developmental disorders and cancer. Indeed, aberrant histone acetylation may be an aetiological factor in several types of cancer by derepressing gene transcription. Hence, HDAC inhibitors may be useful for cancer prevention, due to their ability to 'reactivate' the expression of epigenetically silenced genes, including those involved in differentiation, invasion and metastasis. Most notably, recent data indicate that HDAC inhibition may be successful in treating refractory or relapsing tumours after conventional chemotherapy. Histone deacetylase inhibition has been demon-strated to block cell growth of drug-resistant small-cell lung cancer lines (Tsurutani et al, 2003), abrogate resistance in breast cancer cells (Hirokawa et al, 2005) and induce apoptosis in drug-resistant ovarian cancer cells (Sonnemann et al, 2006), myeloma cells (Maiso et al, 2006) and hepatoma cell lines (Pathil et al, 2006).
The branched-chain fatty acid valproic acid (VPA) has been shown to possess HDAC inhibitory properties and to affect the growth and survival of tumour cells in vitro and in vivo (Cinatl et al, 1996;Blaheta et al, 2005). This is highly relevant since VPA is an established drug in the long-term therapy of epilepsy. It can be applied orally, negative side effects are rare and it demonstrates expedient pharmacokinetic properties. To clearly assess whether VPA might be of benefit in treating relapsed NB, we evaluated the potential of therapeutic VPA concentration to block the interaction of drug-resistant NB cells with vascular endothelium. The experiments were based on an in vitro model of acquired drug resistance (Kotchetkov et al, 2003(Kotchetkov et al, , 2005, closely resembling progressive NB disease through long-term treatment of NB cell lines with vincristine (VCR), cisplatin (CDDP) or doxorubicin (DOX) to establish the resistant tumour cell sublines UKF-NB-3 VCR , UKF-NB-3 CDDP and UKF-NB-3 DOX . A co-culture binding assay allowed the analysis of NB cells that adhered to an endothelial cell monolayer. Since surface receptors are strongly involved in tumour invasion and data have indicated that changes in neural cell adhesion molecule (NCAM, CD56) expression play an essential part in the progression of NB, we investigated NCAM expression (Blaheta et al, 2002), and the NCAM regulating proteins N-myc, p73 and deltaNp73 (Blaheta et al, 2004) under the influence of VPA.

VPA treatment
Parental NB tumour cells and their drug-resistant sublines were treated with VPA (gift from GL Pharma GmbH, Lannach, Austria) at a final concentration of 1 mM for 3 or 5 days. Tumour cell adhesion, NCAM, p73, deltaNp73 and N-myc expression were then measured in VPA-treated cells. Results were compared to untreated controls. Viability of tumour cells in presence of VPA was assessed by propidium iodide dsDNA intercalation.

Tumour cell adhesion
Human endothelial cells were transferred to six-well multiplates (Falcon Primaria; Becton Dickinson, Heidelberg, Germany) in complete HUVEC medium. When confluency was reached, 0.5 Â 10 6 parental NB tumour cells or their drug-resistant sublines (VPA treated vs non-treated) were carefully added to the HUVEC monolayer for 60 min. Subsequently, non-adherent tumour cells were washed off using warmed (371C) M199. The adherent cells were fixed with 1% glutaraldehyde and counted in five different fields (5 Â 0.25 mm 2 ) using a phase contrast microscope (20 Â objective) to calculate the mean cellular adhesion rate.

Cell proliferation
Proliferative activity of NB tumour cells and HUVEC was estimated by the PicoGreen assay as described elsewhere (Blaheta et al, 1998). Briefly, at several time points after plating the cells in six-well multiplates, culture medium was removed and cells were digested with papain (0.125 mg protein ml À1 ) for 20 h at 601C. Thereafter, the fluorescent dye PicoGreen (MoBiTec, Goettingen, Germany), which shows high specificity for dsDNA, was added (1 : 200 dilution) for 10 min at 201C. Fluorescence intensity was determined using a computer-controlled fluorescence reader (Cytofluor 2300 plate scanner; Millipore, Eschborn, Germany) at lex ¼ 485 nm and lem ¼ 530 nm.

Evaluation of NCAM surface expression
Neuroblastoma cells were disaggregated mechanically, washed in blocking solution (PBS, 0.5% BSA) and then incubated for 60 min at 41C with an FITC-conjugated monoclonal antibody anti-CD56 which detects the NCAM 120, 140 and 180 kDa isoform (clone 16.2). Neural cell adhesion molecule expression of NB cells was then measured using a FACscan (Becton Dickinson; FL-1 H (log) channel histogram analysis; 1 Â 10 4 cells/scan) and expressed as mean fluorescence units (MFU). A mouse IgG1-FITC was used as the isotype control.
To explore NCAM localisation, tumour cells were transferred to round cover slips, which were placed in a 24-well multiplate. Upon reaching confluency, cell cultures were washed two times with PBS (with Ca 2 þ and Mg 2 þ ) and then fixed in cold (À201C) methanol/ acetone (60/40 v/v). Subsequently, cells were washed again with PBS (without Ca 2 þ and Mg 2 þ ), and later once with blocking buffer (0.5% BSA in PBS without Ca 2 þ and Mg 2 þ ). After removing the washing buffer, cells were incubated for 60 min with FITCconjugated anti-NCAM monoclonal antibody. To prevent photobleaching of the fluorescent dye, cover glasses with stained cells were taken out of the wells and the residual liquid was removed. The cells were then embedded in an antifade reagent/mounting medium mixture (ProLongt Antifade Kit, MoBiTec) and mounted on slides. The slides were viewed using a confocal laser-scanning microscope (LSM 10; Zeiss, Jena, Germany) with a plan-neofluar Â 100/1.3 oil immersion objective.

Western blot analysis
Total NCAM content in NB cells was evaluated by Western blot analysis: tumour cell lysates were applied to a 7% polyacrylamide gel and electrophoresed for 90 min at 60 V. The protein was then transferred to nitrocellulose membranes. After blocking, the membranes were incubated overnight with the anti-NCAM antibody (dilution 1 : 1000). HRP-conjugated goat anti-mouse IgG (Upstate Biotechnology, Lake Placid, NY, USA; dilution 1 : 5000) served as the secondary antibody. The membrane was briefly incubated with ECL detection reagent (ECLt, Amersham/GE Healthcare, München, Germany) to visualise the proteins and exposed to an x-ray film (Hyperfilmt ECt, Amersham).

Semi-quantitative reverse transcription/polymerase chain reaction
Total RNA was extracted and purified with Trizol reagent according to the manufacturer's instructions and treated with RNase-free DNase. The NCAM primer sequences were as follows: for NCAM-180: 5 0 CGAGGCTGCCTCCGTCAGCACC 3 0 and 5 0 CCGG ATCCATCATGCTTTGCTCTCG 3 0 ; for NCAM-140: 5 0 GAACCTG ATCAAGCAGGATGACGG 3 0 and 5 0 CCGGATCCATCATGCTTTGC TCTCG 3 0 (Kleinschmidt-DeMasters et al, 1999). Internal controls for the reverse transcription/polymerase chain reaction (RT-PCR) reaction was performed by running parallel reaction mixtures with the housekeeping gene GAPDH: 5 0 ATCTTCCAGGAGCGAGATCC 3 0 and 5 0 ACCACTGACACGTTGGCAGT 3 0 . Ribonucleic acid (1 -10 mg) was reverse transcribed and the resulting cDNA directly added to the PCR. Amplification reactions (20 ml) were performed in the presence of 1/10 (2 ml) of the cDNA reaction, with an initial incubation step at 941C for 1 min. Cycling conditions consisted of denaturation at 941C for 1 min, annealing at 551C for 1 min and extension at 721C for 1 min over a total of 30 cycles. The reaction was completed by another incubation step at 721C for 10 min. The PCR products were subjected to electrophoresis in 2% agarose gel and visualised by ethidium bromide.

Statistics
All experiments were performed 3 -6 times. Statistical significance was investigated by the Wilcoxon -Mann -Whitney U-test. Differences were considered statistically significant at Po0.05.

VPA downregulates cell adhesion of CDDP-and VCRresistant UKF-NB-3 tumour cells
Adhesion of UKF-NB-3, UKF-NB-3 CDDP , UKF-NB-3 VCR or UKF-NB-3 DOX was quantified 60 min after plating the cells on to an endothelial cell monolayer ( Figure 1). Nearly 200 parental (drugsensitive) UKF-NB-3 cells mm À2 were attached to HUVEC during this time (SD interassay o50%, SD intraassay o10%). The amount of adherent cells increased fourfold when UKF-NB-3 became resistant to CDDP or VCR. Doxorubicin resistance did not induce any effects on tumour cell binding to HUVEC. Table 1 provides a profile of cross-resistance among the sublines examined.
The application of 1 mM VPA to UKF-NB-3 CDDP or UKF-NB-3 VCR significantly blocked the cellular adhesion process (Figure 1). A 5-day incubation period evoked stronger effects than a 3-day incubation period. Notably, treatment of UKF-NB-3 CDDP with VPA completely reverted the elevated adhesion behaviour induced by drug resistance. Valproic acid also acted on the parental cell lines, as evidenced by a significant downregulation of the number of adherent UKF-NB-3. The 60-min adhesion rate was reduced by 59.4718.6% (n ¼ 6).
The PicoGreen assay did not reveal any proliferative activity during the experiment, which rules out the possibility that adhesion differences between drug resistant, VPA-treated and control NB cells may be caused by different cell growth capacity.  (VPA5). Neuroblastoma cells were added at a density of 0.5 Â 10 6 cells/well to HUVEC monolayers for 60 min. Nonadherent tumour cells were washed off in each sample, the remaining cells were fixed and counted in five different fields (5 Â 0.25 mm 2 ) using a phase contrast microscope. Adhesion capacity is depicted as tumour cell adhesion mm À2 (mean7s.d.; n ¼ 6). Reduction of NCAM surface expression was observed on UKF-NB-3 CDDP and UKF-NB-3 VCR , when compared to the parental control cell line. However, no changes were seen with NCAM expression on UKF-NB-3 DOX , compared to the controls. Treatment of CDDP-or VCR-resistant cell lines with VPA led to a significant increase in NCAM, which exceeded the control values after a 5-day incubation period.

VPA upregulates NCAM surface expression
Valproic acid also acted on the parental cell lines, as evidenced by a significant upregulation of NCAM by þ 56.9722.8% (n ¼ 5).

VPA enhances NCAM protein content
Similar to the modifications of the NCAM surface expression level, UKF-NB-3 CDDP and UKF-NB-3 VCR were characterised by a strong reduction of NCAM proteins when compared to the parental UKF-NB-3 control cells (Figure 3). No differences were seen between UKF-NB-3 and UKF-NB-3 DOX .
When UKF-NB-3 CDDP or UKF-NB-3 VCR were treated with VPA for 3 or 5 days, NCAM protein content became upregulated, partially exceeding the control values. Applying VPA to UKF-NB-3 DOX did not induce any alterations in NCAM, independent of the incubation time.

VPA modifies NCAM mRNA expression
Assessment of NCAM mRNA showed distinct expression of mRNA encoding the 140 kDa isoform in UKF-NB-3 control cells, which however became down-modulated in UKF-NB-3 CDDP or UKF-NB-3 VCR (Figure 4). Only slight differences were visualised between UKF-NB-3 and UKF-NB-3 DOX . The presence of VPA was accompanied by elevated 140 kDa mRNA levels in UKF-NB-3 CDDP or UKF-NB-3 VCR . This effect was not seen in UKF-NB-3 DOX , irrespective if VPA was applied for 3 or 5 days.
VPA alterates p73, deltaNp73 and N-myc expression p73, deltaNp73 and N-myc have been identified to trigger NCAM expression (Blaheta et al, 2004). They were therefore used as biomarkers to further explore the influence of VPA on NCAMtriggered NB adhesion. p73 was detected in UKF-NB-3, expression of which was significantly reduced in UKF-NB-3 CDDP and UKF-NB-3 VCR . Doxorubicin resistance was not accompanied by a distinct p73 downregulation (Table 2). When VPA was added to the cell cultures, downregulation of p73 seen in UKF-NB-3 CDDP and UKF-NB-3 VCR was reverted and the protein became upregulated after 3 and 5 days. deltaNp73 became enhanced in UKF-NB-3 CDDP and UKF-NB-3 VCR compared to the drug naïve UKF-NB-3 controls. Application of VPA led to a significant reduction of deltaNp73 in CDDP-or VCR-resistant cell lines.
N-myc proteins were enhanced in UKF-NB-3 CDDP and UKF-NB-3 VCR , but not in UKF-NB-3 DOX , compared to the UKF-NB-3 control cell line ( Figure 5). Moderate N-myc downregulation was induced by  VPA in UKF-NB-3 VCR . N-myc became nearly undetectable in UKF-NB-3 CDDP after a 5-day treatment with VPA. The distinct influence of VPA on UKF-NB-3 CDDP or UKF-NB-3 VCR was also proven by flow cytometry, which revealed left shifting of the N-myc-specific fluorescence intensity, indicating N-myc loss. N-myc histogram analysis concentrates on isotype controls and specific fluorescence of untreated vs VPA-treated (5-day treatment) UKF-NB-3 CDDP , UKF-NB-3 VCR and UKF-NB-3 DOX . The complete experimental data set (MFU7s.d.; n ¼ 6) is given below the histogram.

VPA acts on further NB tumour cell lines
To strengthen the relevance of our findings, further NB tumour cell lines were included in the study. Analysis of UKF-NB-2 and     drug-resistant sublines indicated a significant increase in UKF-NB-2 CDDP , UKF-NB-2 VCR and UKF-NB-2 DOX cell adhesion, compared to the parental controls. Adhesion correlated inversely with the NCAM surface level (Table 3). A similar phenomenon was observed in CDDP-(moderate) and VCR-resistant UKF-NB-6 cell lines. A 3-day VPA treatment blocked cell binding to HUVEC in UKF-NB-2 and their drug-resistant sublines, and upregulated NCAM expression. Valproic acid also acted on UKF-NB-6, UKF-NB-6 CDDP , UKF-NB-6 VCR and SKNSH. However, although VPA (partially) reverted the increased adhesion phenotype and restored NCAM levels, it did not re-sensitise the drug-resistant NB cells to VCR, DOX or CDDP (data not shown). N-myc was significantly diminished by VPA in UKF-NB-2, UKF-NB-2 CDDP , UKF-NB-6 and UKF-NB-6 CDDP , compared to the controls. However, no differences were seen between VPA-treated and non-treated VCR or DOX-resistant tumour cell lines.
To better interprete our data, antiproliferative effects of VPA on the chemosensitive vs chemoresistant NB cells were analysed in final experiments. Table 4 document growth inhibition by this compound, whereas higher VPA concentrations were necessary to reach 50% reduction in chemotherapy-resistant cell lines compared to the parental controls.

DISCUSSION
Based on this cell culture model, we have demonstrated that drugresistant NB cancer cells develop an increased malignant phenotype as evidenced by enhanced adhesion to vascular endothelial cells, accompanied by significant downregulation of the adhesion receptor NCAM. Valproic acid reverted this process by downregulating cell adhesion and upregulating NCAM expression. Considering clinical utility, it is encouraging that VPA was active in NB cell lines resistant to existing chemotherapies, since overcoming resistance to anticancer agents is a major challenge in the development of novel antitumour protocols.
In fact, several cases have been documented showing antineoplastic effects of VPA in patients with relapsed tumours. When VPA was given as maintenance therapy for childhood malignant glioma after postoperative combined chemotherapy and irradiation, about 10% of these patients were maintained in continuous complete remission and an equal number of patients showed at least partial responses ). An additional pediatric patient with glioblastoma multiforme responded to VPA after showing progressive disease shortly after having received combined chemotherapy and irradiation as well as topotecane (Witt et al, 2004). Another pediatric patient suffering from a relapsed supratentorial primitive neuroectodermal tumour while receiving chemotherapy (CCNU, VCR and cisplatinum) after total resection and irradiation showed conspicuous signs of glial differentiation induction and a nonmalignant morphology on histological examination. This patient had received VPA for epilepsy treatment for a period of several months before the tumour recurred .
In vitro, VPA has been shown to inhibit proliferation in acute myeloid leukaemia cells expressing P-glycoprotein (P-gp) and MDR-associated protein 1 (Tang et al, 2004), and to increase sensitivity towards apoptosis in hepatoma cells resistant to epirubicin (Schuchmann et al, 2006). Although the underlying mode of action has not been explored in these studies, the data clearly indicate that VPA may alter the malignant behaviour of tumours that do not respond to chemotherapy. With particular emphasis on NB, VPA significantly prevented the interaction between tumour cells and endothelium. This finding is important, because binding of single cancer cells to the vessel wall represents the first step in the haematogenous invasion cascade proceeding transendothelial migration and invasion into surrounding tissue. We therefore conclude that VPA may have a direct impact on metastasis formation. In good accordance to this hypothesis, VPA enhanced the NCAM surface level, expression of which is strongly involved in tumour cell adhesion and penetration.
In primitive neuroectodermal tumour cells, an increase in NCAM was paralleled by a significant reduction in cellular motility and adhesion capacity (Owens et al, 1998;Prag et al, 2002). In a rat model, NCAM-transfected glioma tumour cells became less invasive and destructive than control cells with a low NCAM expression level (Edvardsen et al, 1994). Diminished expression of NCAM was also associated with clinically aggressive colon cancers (Sampson-Johannes et al, 1996;Roesler et al, 1997;Huerta et al, 2001), and dissemination of pancreatic b-tumour cells (Perl et al, 1999;Cavallaro et al, 2001). Tezel et al (2001) suggested that NCAM expression in tubular adenocarcinoma of the pancreas has a significant impact on overall patient survival. We recently demonstrated an inverse correlation between NCAM expression and NB cell adhesion, assessed on 11 NB cell lines. In particular, transfection with a cDNA encoding the human NCAM-140 kD Mean standard deviations were as follows: cell adhesion intra-assay o25%, cell adhesion inter-assay o80%. NCAM expression intra-assay o5%, NCAM expression inter-assay o30%. N-myc expression intra-assay o5%, N-myc expression inter-assay o50%.  (Blaheta et al, 2002). It is currently assumed that NCAM, in its function as a homophilic receptor, stabilises the primary tumour or tumour cell aggregates, while circulating in the blood vessels. Reduction of the NCAM expression level might lead to a reduction in cell -cell binding forces, and hence to the release of tumours as single cells. The less NCAM, the more metastatic cells leave the tumour mass, and the more penetration events can take place (Blaheta et al, 2002). Consequently, NCAM upregulation observed in CDDP-and VCRresistant NB tumour cells under VPA might reduce cell transmigration and extravasation processes. There is some evidence from the literature that N-myc downregulates NCAM expression, thus increasing the invasiveness of NB cells. Transfection of the rat NB cell line B104 with an N-myc expression vector resulted in a dramatic reduction in the levels of NCAM polypeptides and mRNAs, and increased metastatic ability (Akeson and Bernards, 1990). Cytomegalovirus-induced acceleration of NB adhesion and transendothelial penetration was evoked by increasing N-myc protein content, accompanied by a diminished NCAM surface level (Blaheta et al, 2004). The present data reveal strong upregulation of N-myc in UKF-NB-3 CDDP , compared to the parental cells, process of which was coupled to NCAM loss and enhanced adhesion capacity. It may therefore be concluded that N-myc plays an important role in NCAM-driven NB adhesion, and that VPA has an impact on N-myc protein expression. Nevertheless, the situation is more complex than initially thought. Our analysis on further NB cell lines indicated that VPA reverts cell adhesion, restored NCAM and suppressed N-myc expression level on UKF-NB-2, UKF-NB-6 and their CDDP-resistant sublines. However, VPA effects on VCR-resistant sublines were not accompanied by N-myc alterations. Furthermore, incubation of multidrug-resistant SKNSH with VPA induced very strong adhesion blockade and NCAM upregulation, although N-myc was detected in non-treated controls just very slightly over threshold values. Based on this, we assume that endogenous Nmyc expression level of NB cells may not correlate with their responsiveness to VPA-induced NCAM upregulation, and VPAinduced loss of N-myc may be limited to CDDP-resistant and drug-sensitive tumour cells.
p73 and deltaNp73, an isoform of p73 lacking the N-terminal transactivation domain, are both suggested to be associated with NCAM expression. In vitro experiments demonstrated that transfected full-length p73 cDNA induces expression of NCAM and downregulation of N-myc in N1E-115 NB cells. Inversely, transfection of dominant-negative p73 abrogated the transactivation of the NCAM promoter (De Laurenzi et al, 2000). Based on UKF-NB-4 tumour cells, we have recently postulated a direct association between tumour progression, upregulation of N-myc and deltaNp73, and downregulation of p73 and NCAM (Blaheta et al, 2004). According to this statement, VPA evoked UKF-NB-3 adhesion blockade was accompanied by diminished N-myc and deltaNp73, and an enhanced p73 and NCAM level. With special respect to the UKF-NB-3 model, VPA may interfere in the N-myc/ deltaNp73 signalling system that causes, as at least one consequence, distinct upregulation of NCAM biosynthesis and receptor processing. Neural cell adhesion molecule processing finally attributes to the lowered invasive capacity of the tumour cells.
To summarise, evidence is presented showing that drugresistant NB cells are sensitive to VPA treatment. Valproic acid distinctly reduced the invasive properties of NB and may therefore be well suited to amend the current treatment protocol with particular emphasis on those tumours that do not respond to chemotherapy. Nevertheless, our data are particularly limited to four NB cell lines. Therefore, the hypothetical possibility that VPA might overcome drug resistance in general needs further investigation. Remarkably, VPA blocked cell adhesion of parental, but not of DOX-resistant UKF-NB-3 or UKF-NB-6 tumour cells in our assay. This implies that DOX may, under certain circumstances, induce resistance to VPA. A similar phenomenon was found using NB cells with 'naturally' arised DOX resistance, SMS-KANR and SMS-KCNR (Reynolds et al, 1986). Tabe et al (2006) demonstrated in this context that the HDAC-inhibitor FK228 induces P-gp expression and prevents growth inhibition and apoptosis in acute promyelocytic leukaemia cells subsequently incubated with DOX. Okada et al (2006) observed resistance development in DOX-resistant clones of osteosarcoma and Ewing's family of tumours after exposure to FK228. Nevertheless, FK228 is chemically different from VPA and, therefore, further experiments are necessary to explore this issue.
Differences between VCR/CDDP-and DOX-resistant cell lines may also point to different resistance mechanisms that are operational in these cell lines. Suppression of MAP kinase (MEK-ERK) signalling has been observed in NB cells with acquired resistance to DOX (Mattingly et al, 2001;Armstrong et al, 2006). Down-modulation of ERK1/2 phosphorylation has further been documented in NB cells with acquired resistance to CDDP or VCR. However, these cell lines were additionally characterised by a distinct Akt activation (Kotchetkov et al, 2005;Servidei et al, 2006). Therefore, although purely speculative, fine-tuned alterations of the ERK and Akt signalling system may be -at least partially -responsible for establishing VCR and CDDP, but not DOX resistance in our NB cell model. However, detailed knowledge of resistance mechanism in individual cancer cells is necessary to allow better prediction of the clinical use of VPA.
Other HDAC inhibitors have also been shown to inhibit tumour growth and to overcome multidrug resistance. Notably, suberoylanilide hydroxamic acid (SAHA) has been demonstrated to act on VCR-resistant leukaemia cell lines (Ruefli et al, 2002), adriamycinresistant breast (Castro-Galache et al, 2003) and paclitaxel-resistant ovarian cancer cells (Sonnemann et al, 2006). SAHA at 5mM significantly diminished UKF-NB-3 CDDP or UKF-NB-3 VCR proliferation and cell adhesion to HUVEC in our own experiments (data not shown). Histone deacetylase inhibitors different from VPA may therefore be considered to become additional options for the treatment of drug-resistant NB. However, detailed studies are necessary to explore their clinical value.