Study on AS and transcription factor mRNA levels in leukaemic patient samples
INTRODUCTION
The anti-tumour activity of L-asparaginase has been known for decades and the drug has been used systematically in the treatment of ALL. Justification of its use as an anti-leukaemic drug was based on the fact that leukaemic cells exhibit reduced AS activity in comparison with normal cells. These particular cells are entirely dependent on the extracellular availability of asparagine and glutamine in the circulatory system to survive. This inability to produce endogenous levels of the essential amino acid asparagine sufficient to maintain survival during treatment with the amino acid-depleting drug has made L-asparaginase a very useful agent in the treatment of leukaemia. Administration of L-asparaginase in patients undergoing chemotherapy will deplete the levels of asparagine from the blood and bone marrow leaving leukaemic cells defenceless against starvation which will lead to cell cycle arrest and eventually apoptosis.
Despite its success story, L-asparaginase treatment is frequently facing the challenge of resistance development in patients. Through the years evidence has suggested that this resistance is the result of elevated AS activity in leukaemic cells. Moreover, in vitro studies have demonstrated a strong correlation between resistance and increase in cellular AS activity, AS mRNA and AS protein levels (Hutson et al, 1997). In addition, use of l-asparaginase has been shown to be affected by other parameters such as severe side effects and hypersensitivity reactions considered to be consequences of inhibition of protein synthesis or immunological reactions. These adverse reactions have been proven to be significant obstacles in the efficient use of L-asparaginase to treat patients suffering from leukaemia.
In order to avoid the development of resistance and the occurrence of adverse reactions to the drug, research is being conducted aiming to elucidate the molecular mechanisms involved in this process. The direction this particular research has taken recently is towards the hypothesis that the use of L-asparaginase could be optimised based on the sensitivity of the individual patient to the drug rather than applying a universal protocol for the use of this agent. In other words, the question raised is whether the efficiency of this drug could be enhanced and the toxicity reduced through a tailor-made treatment according to the profile of the individual patient.
Hypothesis and aim of this study
The work carried out by Leslie et al 2006 on leukaemic patient samples showed variability of AS levels between samples and suggested the probability of differences in in vivo sensitivity of patients during L-asparaginase treament. The most recent evidence from studies carried out on cell lines has proposed a new possible model for the regulation of AS gene expression during nutritional deprivation suggesting the involvement of transcription factors, each one possessing a unique role in inducing or suppressing the expression of the AS gene as a consequence of cellular exposure to nutritional stress.
In addition to furthering the current work and enriching the already existing evidence regarding the transcriptional regulation of AS genes in cell lines it would be of great importance to carry out studies on samples collected from patients suffering from leukaemia and undergoing treatment. Evidence collected from such studies would provide a better understanding of the process of AS gene expression regulation and could provide an insight into the development of resistance to L-asparaginase treatment.
The aim of this particular part of the study was to examine the mRNA levels of AS and the transcription factors of interest in a small cohort of patient samples (n=7) and identify any possible correlation between these levels. The initial hypothesis adopted for this part of the study is that levels of transcription factors involved in the regulation of AS gene expression would exhibit strong correlation with AS mRNA levels.
Determination of mRNA levels in leukaemic patient samples
In order to determine the mRNA levels of AS, ATF4, ATF3 and C/EBP-β in patient samples the real-time RT-PCR assay previously used to determine mRNA levels in MOLT4 and Nalm6 cell lines was also utilised in this part of the study. The patient sample cohort comprised 7 ALL patient samples all at the stage of presentation. Of these 7 samples, 5 samples were from patients suffering from childhood ALL while the remaining 2 were adult patients. The adult samples were peripheral blood samples while the childhood samples consisted of 1 peripheral blood sample and 4 bone marrow samples.
RESULTS
AS and transcription factor mRNA levels in patient samples
Levels of mRNA expression were detectable for all targets and in all 7 patient samples even though in some cases at very low values. Relative AS mRNA levels did not seem to vary significantly with values ranging between rAS = 0.03–0.15 (5-fold).
Levels of ATF4 also appeared not to vary greatly with values ranging between rATF4 = 0.10–0.52 (5-fold). However, measurement of ATF3 and C/EBP-β mRNA levels in these patient samples revealed greater variations with values spanning a wider range between rATF3 = 0.29-3.95 (13-fold) and rC/EBP-β = 0.05-1.49 (30-fold).
Correlations between AS and transcription factor mRNA levels
Since the aim of this particular piece of work included in this study was to investigate possible correlations between AS and transcription factor mRNA levels in patient samples, measurement of mRNA levels in these 7 patient samples was followed by correlation analysis. Pearson’s correlation coefficient was determined for AS levels correlated with the three individual transcription factors. This analysis provided only low to medium correlations. ATF4 exhibited the lowest correlation to AS with r = 0.19. ATF3 and C/EBP-β showed medium correlation to AS with correlation coefficients r = -0.40 and r = -0.30 respectively.
DISCUSSION
Asparagine synthetase levels in patient samples
A large study was previously carried out involving the quantification of AS mRNA levels in primary leukaemic blasts from patients with acute leukaemia. The aim was to investigate the potential for the development of a method for the individualisation of L-asparaginase treatment based on the expression of the AS gene. Although extensive evidence in the literature has suggested cellular AS levels as a major determinant of cell sensitivity to L-asparaginase, there has not been significant work carried out on the measurement of AS levels in primary leukaemic blasts.
Leslie et al (2006) reported that the median rAS value for normal adult peripheral blood mononuclear (PBMN) cells was significantly higher than that obtained for lymphoblasts from children and adults with ALL. In vitro studies have shown that AS mRNA levels parallel AS activity and that cell lines exhibiting high AS activity are resistant to L-asparaginase (Hutson et al, 1997). Hence theoretically normal healthy cells should have higher AS mRNA expression levels than lymphoblasts. Kaspers et al (1991) investigated in vitro drug resistance to L-asparaginase and observed that PBMN cells from healthy individuals were significantly more resistant than lymphoblasts from children with ALL. These findings were supported by Asselin et al (1989), who observed that lymphoblasts from the bone marrow of children suffering from ALL were more sensitive to L-asparaginase than normal PBMN cells. The conclusion derived from the combined findings from these studies is that the resistance of normal PBMN cells to L-asparaginase is most likely due to an elevation in AS gene expression and, hence, cell sensitivity to L-asparaginase treatment is a consequence of low or reduced AS expression. Despite the significant difference in cell sensitivity to L-asparaginase between normal PBMN and lymphoblasts from children with ALL reported by Kaspers et al (1991), the differences in levels of AS mRNA observed by Leslie et al (2006) were only slight. This evidence suggests that a slight change in AS mRNA levels could potentially account for significant differences in L-asparaginase sensitivity. This is consistent with a previous study carried out by Hutson et al (1997) reporting a significant increase in L-asparaginase resistance at a magnitude of 1000-fold following treatment while AS expression levels increased only slightly (7-fold). This evidence highlights an inverse correlation between AS expression levels and L-asparaginase sensitivity in PBMN cells and primary leukaemic blasts.
Despite the evidence pointing towards an inverse correlation between AS mRNA levels and L-asparaginase sensitivity there are studies which conflict with the hypothesis. For instance, in vitro work carried out by Ramakers-van Woerden et al (2003) demonstrated a higher degree of sensitivity for patients with the t(12;21) TEL-AML1 translocation. Stams et al (2003) confirmed the association of L-asparaginase sensitivity with the TEL-AML1 translocation and also observed that TEL-AML1 positive patients exhibited higher median levels of AS mRNA. No correlation between AS mRNA levels and L-asparaginase sensitivity was reported and it was noted that median AS mRNA levels were lower in normal peripheral blood and bone marrow samples than in leukaemic blasts. From the evidence presented in these studies it was concluded that sensitivity of patients with the TEL-AML1 translocation to L-asparaginase is not related to AS gene expression. More recent studies have suggested that measurement of AS protein content might also be a more appropriate marker of cell sensitivity to L-asparaginase rather than AS mRNA (Su et al, 2007).
Leslie et al (2006) observed that relative AS mRNA levels varied largely between patients with ALL and in particular in childhood ALL samples. Variation in these samples reached up to 64-fold difference in mRNA levels. In contrast AS mRNA levels did not seem to vary significantly in cells from healthy volunteers. These findings were very interesting, suggesting that there might be a large difference in sensitivity among ALL patients. Patients with high levels of AS mRNA expression levels or at least equivalent to the levels observed in normal cells might be less sensitive and consequently resistant to L-asparaginase. In these cases leukaemic cells might possess the ability to produce sufficient levels of asparagine in order to counterbalance the nutritional stress inflicted on them through L-asparaginase administration.
Comparison of the AS mRNA levels in presentation samples and samples collected at Day 8 of treatment, at which point two doses of 6000 IU/m2 L-asparaginase had been administered, revealed 10-fold higher levels in Day 8 cells. An explanation given by Leslie et al (2006) for this increase is that cells at Day 8 induce an increase in AS mRNA levels in response to the treatment, attempting to develop a resistance mechanism to L-asparaginase treatment. Leslie et al (2006) investigated the possible correlation of AS mRNA levels with established ALL prognostic factors. However, there were no correlations reported between AS mRNA levels and any prognostic factors such as age, sex or white cell count. However when initial response to treatment was correlated with AS expression levels patients with less than 5% blasts in the bone marrow, and therefore considered to be good responders to treatment, exhibited lower AS levels than poor responders (> 5% blasts). These findings did not reach statistical significance but the observation alone is a clear indicator supporting the hypothesis that sensitivity to L-asparaginase is linked with a favourable response to initial treatment.
In this study, relative AS mRNA levels in the patients’ samples examined were found to be very low, in agreement with previous evidence in the literature (Leslie et al, 2006; Su et al, 2007). There was no significant variation observed as previously reported by Leslie et al (2006) and there were no notable differences in AS levels for the adult vs. child samples and the bone marrow vs. peripheral blood samples.
ATF4 relative mRNA levels also appeared to be low in these samples and there was little variation observed between levels at a magnitude of 5-fold differences similar to the variation observed for AS mRNA levels. ATF4 levels appeared not to correlate strongly to AS mRNA levels (r=0.19, P > 0.05).
In contrast, ATF3 and C/EBP-β mRNA levels appeared to be higher in these samples and exhibited considerable variation at a magnitude of 13 and 30-fold respectively. Both transcription factors showed medium correlation to AS levels (ATF3 r= -0.40, C/EBP-β r= -0.30) but these results did not reach statistical significance (P > 0.05). However, the small number of samples examined (n=7) is a parameter that should be taken into consideration and it does not allow the formulation of solid conclusions about correlations between expression levels at this point. Despite this, the medium correlation observed between AS levels and ATF3, C/EBP-β levels might be an indication of a possible stronger correlation present if a larger number of cases are examined. Such a hypothesis is an exciting prospect since it might suggest an important role for these two transcription factors in AS regulation in patients, similarly to the roles proposed earlier in this study in the context of AS regulation in leukaemic cell lines.
In summary, the aim of this small patient sample study was to initiate an investigation in leukaemic patient samples of the possible involvement of the transcription factors already known to regulate AS gene expression in cell lines and attempt to identify any particular correlation between AS and transcription factor mRNA levels. This investigation did not provide any clear evidence that could suggest a strong correlation between AS levels and transcription factor levels. It would be necessary to expand this investigation into a larger cohort of patients and re-attempt to identify correlations between the expression levels of AS and any of the transcription factors of interest. Also, the absence of evidence regarding the expression of these transcription factors in patients with leukaemia or in a similar context make this investigation a necessity in order to obtain a clearer picture of the potential role of these proteins in AS regulation in patients.
In addition, it is worth noting that all the samples used in this study were samples collected at presentation prior to any treatment initiation. Hence the mRNA levels measured represent basal levels of expression in these patients. It would be extremely interesting to carry out a study investigating transcription factor mRNA levels in patient samples collected at presentation as well as following treatment at Day 8 at which point AS levels have been reported to rise as a consequence of L-asparaginase administration (Leslie et al, 2006). A comparison of this kind could directly link the expression levels of these transcription factors to the AS increase, in case a parallel elevation in transcription factor mRNA levels is observed.
Moreover, since recent evidence has suggested that AS protein levels might be of great importance in determining sensitivity to L-asparaginase, it would be necessary to investigate this aspect in patient samples already available. Measurement of AS protein levels in a large cohort of leukaemic samples for which AS mRNA levels have already been determined could provide valuable evidence through comparison and shed further light into which investigation might be of greater significance as a sensitivity marker to L-asparaginase.