We examined apoptotic cell percentages of Jurkat and two NK-cell lines after treatment with various anti-cancer agents including VCR, MTX, Ara-C, DNR, DXR and l-asparaginase using flow cytometric TUNEL assay (Sugimoto et al, 1998, 2002). At first, we determined the treatment conditions (drug concentration and treatment duration), which caused 30–70% of apoptosis in Jurkat cells. NK-YS and NK-92 cells were extremely resistant to these drugs, and at most 10% of them were apoptotic. On the contrary, 84% of NK-YS and 67% of NK-92 cells underwent apoptosis after 48 h of incubation with 0·01 IU/ml of l-asparaginase, although the same treatment induced apoptosis only in 19% of Jurkat cells. Even at different concentrations and treatment durations, l-asparaginase induced apoptosis rather specifically in those NK-cell lines (Fig 1A). Essentially the same results were obtained in four independent experiments. Among various anti-tumour agents tested here, only l-asparaginase was more effective to the two NK-cell lines than to Jurkat cells. As MTX and Ara-C, whose effects are known to be independent of P-gp status (Pastan & Gottesman, 1994), had little effect on these two NK-cell lines, high expression of P-gp cannot fully explain the refractoriness of NK-cell tumours to various anti-tumour agents. However, l-asparaginase was obviously effective to NK-cell tumour cell lines, regardless of any mechanisms of drug resistance. Figure 1B shows the results of flow cytometric TUNEL assay on NK-YS cells. Treatment with 300 nmol/l of DXR for 24 h caused a small decrease in the G1 peak and small amount of apoptosis in NK-YS cells (b, e). Although treatment with l-asparaginase did not apparently change the cell cycle distribution of NK-YS (Fig 1B-c), most of the non-G1 cells underwent apoptosis (Fig 1B-f). Similar results were obtained in NK-92 cells (data not shown). We next examined the effectiveness of l-asparaginase to NK-cell tumour samples. Even without anti-tumour drug treatment, NK-cell samples often underwent apoptosis spontaneously in in vitro culture conditions. We thus evaluated the anti-tumour effect of l-asparaginase using the corrected apoptotic cell percentage, as indicated in the study design. Figure 2 shows the corrected apoptotic cell percentages of four nasal-type NK-cell lymphoma, two aggressive NK-cell leukaemia, two blastic NK-cell lymphoma, 10 chronic NK lymphocytosis and six ALL samples, treated with 450 nmol/l of DXR or with 10 IU/ml of l-asparaginase for 24 h. The corrected apoptotic cell percentages after DXR treatment were nearly 0% in chronic NK lymphocytosis, <10% in NK-cell tumours and about 40% in ALL. In contrast, l-asparaginase treatment caused corrected apoptotic cell percentages of 20–60% in NK-cell tumour samples except one, and around 10% of apoptosis in chronic NK lymphocytosis and ALL samples. Therefore, our results showed that l-asparaginase was more effective than DXR not only to NK-cell lines but also to patient NK-cell tumour samples. On the contrary, ALL samples were more sensitive to DXR than to l-asparaginase. Chronic NK lymphocytosis was resistant to both agents. The characteristics, diagnosis, treatment regimen(s), survival duration of each patient are shown in Table I. According to the sensitivity to these two agents, ALL, NK-cell tumours, and chronic NK lymphocytosis could be clearly divided into three categories; i.e. DXR-sensitive, l-asparaginase-sensitive and resistant to both agents respectively (Fig 2).
Figure 1. Apoptosis of NK-YS, NK-92 and Jurkat cells treated with various anti-tumour agents including l-asparaginase. (A) Apoptotic cell percentages of NK-YS, NK-92 and Jurkat cells treated with various anti-tumour agents were determined by flow cytometric TUNEL assay. Treatment conditions were as follows: 10 nmol/l of vincristine (VCR) for 24 h, 100 nmol/l of methotrexate (MTX) for 48 h 1 μmol/l of cytosine β-d-arabinofuranose (Ara-C) for 48 h, 120 nmol/l of daunorubicin (DNR) for 24 h, 300 nmol/l of doxorubicin (DXR) for 24 h, 0·01 IU/ml of l-asparaginase for 24 and 48 h, and 0·1 IU/ml of l-asparaginase for 24 h. TUNEL assay was performed in triplicate for all cell lines. The arithmetic means and standard deviations were calculated with three values. NK-YS and NK-92 cells were selectively sensitive to l-asparaginase. (B) Cell cycle positions of apoptotic and non-apoptotic NK-YS cells determined by flow cytometric TUNEL assay. Untreated (a, d), treated with 300 nmol/l of doxorubicin for 24 h (b, e) and 0·01 IU/ml of l-asparaginase for 48 h (c, f). Most NK-YS cells at non-G1 cell cycle positions were apoptotic.
Download figure to PowerPoint
As expression of asparagine synthetase is one known mechanism for resistance to l-asparaginase (Aslanian et al, 2001), we examined asparagine synthetase expression in the samples described above using RQ-PCR and immunostaining. Table II shows the normalised asparagine synthetase mRNA expression value (AS/GAPDH) of each sample. In nasal-type NK-cell lymphoma, asparagine synthetase mRNA levels of two samples were comparatively high (patients 1 and 2), and those of another two samples were low (patients 3 and 4). In this disease entity, asparagine synthetase mRNA levels inversely correlated with in vitro sensitivity to l-asparaginase. Although the sample number was relatively small, asparagine synthetase mRNA levels were high (>0·05) in two aggressive NK-cell leukaemia and two blastic NK-cell lymphoma cases. Expression of asparagine synthetase mRNA was high in almost all samples of chronic NK lymphocytosis. Then, we created a two-dimensional plot of asparagine synthetase mRNA expression and in vitrol-asparaginase sensitivity. Samples belonging to each disease entity occupied a distinct area in this scheme (Fig 3). In spite of comparatively low asparagine synthetase mRNA expression, ALL cells lacked sensitivity to l-asparaginase in vitro. All nine cases of chronic NK lymphocytosis that were resistant to l-asparaginase, had high expression levels of asparagine synthetase mRNA. On the contrary, although l-asparaginase induced considerable apoptosis in two aggressive NK-cell leukaemia and two blastic NK-cell lymphoma samples, asparagine synthetase mRNA expression of these samples was high. As for nasal-type NK-cell lymphoma, there was good correlation between the results of in vitro TUNEL assay and asparagine synthetase mRNA expression (Table II, Fig 3).
Figure 4 shows representative results of asparagine synthetase staining in NK-cell tumour cell lines and patient samples. Compared with the positive control (K562 cells; Fig 4A), NK-92, NK-YS (Fig 4B) and Jurkat cells were all negative for asparagine synthetase staining. As summarised in Table II, asparagine synthetase was detectable in one of four cases for nasal NK-cell lymphoma, two of two cases for aggressive NK-cell leukaemia and two of four cases of ALL. Although asparagine synthetase mRNA levels were relatively high in cases positive for asparagine synthetase immunostaining, high levels of asparagine synthetase mRNA expression did not always warrant the positive staining.
Figure 4. Immunohistochemical detection of asparagine synthetase protein in cell lines and NK-cell tumours. (A) Positive control (K562 cells), (B) negative control (NK-92 cells), (C) nasal-type NK-cell lymphoma (patient 1), (D) aggressive NK-cell leukaemia (patient 5), (E) nasal-type NK-cell lymphoma (patient 2), (F) blastic NK-cell lymphoma (patient 7). Original magnifications are ×400 for A and B, and ×200 for C–F.
Download figure to PowerPoint