Rescue of K562 cells from MDM2-modulated p53-dependent apoptosis by growth factor-induced differentiation

The wild-type human MDM2 protooncogene was tested for its ability to modulate apoptotic activity of the de novo expressed p53 tumor suppressor gene in K562 cells. We also studied the role of some cytokines in this phenomenon. K562, a human myeloid leukemia cell line, does not express p53 at the mRNA or protein level. In this study, we stably transfected K562 with eukaryotic vectors containing either normal p53 cDNA (pC53-SN3) or mutated p53 (143Val-->Ala) cDNA (pC53-SCX3). Transfectants expressing WT p53 or those expressing mutant p53 are called K562 SN and K562 SM respectively. Many leukemic cell lines undergo apoptosis when de novo WT p53 is expressed alone. In contrast, while the resulting clones (K562 SN and K562 SM) expressed p53, they did not undergo apoptosis. However, when treated with MDM2 mRNA antisense (MDM2 AS) oligodeoxynucleotides (ODNs), K562 SN demonstrated apoptotic features at both molecular and morphological levels. No change was observed when the other clones (K562 and K562 SM) were treated with MDM2 AS. Apoptosis induced in this manner was associated with a relatively small increase in intracellular calcium [Ca2+]i. Cells cultured in medium previously supplemented with recombinant human (rh) interleukin (IL)-3 and rh-erythropoietin (Epo) did not undergo apoptosis. Moreover, K562 SN cells were induced to differentiate. This differentiation was evaluated by measuring hemoglobin (Hb) level in cellular extracted proteins and by analyzing erythroid colony number and morphology. High Hb synthesis was obtained when K562 SN cells were cultured with cytokines (IL-3 + Epo) combined with MDM2 AS. Our results are consistent with the hypothesis that the function of the proto-oncogene MDM2 is to provide a 'feedback' mechanism for the p53-dependent pathway of apoptosis that could be shunted toward differentiation.     

In vivo infiltration of mononuclear cells in squamous cell carcinoma of the head and neck correlated with the ability to expand tumour-infiltrating T cells in vitro and with the expression of MHC class I antigens on tumour cells

A series of 18 head and neck squamous cell carcinoma biopsies, 6 primary and 12 recurrent, were investigated for tumour-infiltrating mononuclear cells with monoclonal or polyclonal antibodies. Our results suggest that the number of T cells at the tumour edge in vivo correlates well with their ability to expand in vitro in the presence of high-dose interleukin-2 (2000 U/ml). High MHC class I antigen expression on tumour cells was found to be positively correlated with p53 overexpression, suggesting that p53-derived peptides. wild-type or mutated ones, presented by MHC class I antigens, are potential targets for MHC-restricted cytotoxic T cells in head and neck squamous cell carcinomas. However, lack of correlation between peritumoural T cell infiltration in vivo and T cell expansion in vitro on the one hand, and p53 over-expression on tumour cells, on the other hand, suggests absence of p53-peptide-specific T cells in the patients. Eight out of ten expanded tumour-infiltrating lymphocyte (TIL) cultures showed T-cell-mediated cytotoxicity. ?Promiscuous” cytotoxic T cell activity against the natural-killer-cell-sensitive K562 target cell line was observed in three out of ten TIL expansion cultures.     

cDNA microarray analysis of altered gene expression in Ara-C-treated leukemia cells

The acute lymphoblastic leukemia cell line CCRF-CEM is sensitive to Ara-C and undergoes apoptosis. In contrast, the chronic myelogenous leukemia (CML) cell line K562 is highly resistant to Ara-C, which causes the cells to differentiate into erythrocytes before undergoing apoptosis. We used cDNA microarrays to monitor the alterations in gene expression in these two cell lines under conditions leading to apoptosis or differentiation. Ara-C-treated CCRF-CEM cells were characterized by a cluster of down-regulated chaperone genes, whereas Ara-C-treated K562 cells were characterized by a cluster of up-regulated hemoglobin genes. In K562 cells, Ara-C treatment induced significant down-regulation of the asparagine synthetase gene, which is involved in resistance to L-asparaginase. Sequential treatment with Ara-C and L-asparaginase had a synergistic effect on the inhibition of K562 cell growth, and combination therapy with these two anticancer agents may prove effective in the treatment of CML, which cannot be cured by either drug alone.     

Enhancement of cytotoxicity of artemisinins toward cancer cells by ferrous iron

Iron(II) heme-mediated activation of the peroxide bond of artemisinins is thought to generate the radical oxygen species responsible for their antimalarial activity. We analyzed the role of ferrous iron in the cytotoxicity of artemisinins toward tumor cells. Iron(II)-glycine sulfate (Ferrosanol) and transferrin increased the cytotoxicity of free artesunate, artesunate microencapsulated in maltosyl-beta-cyclodextrin, and artemisinin toward CCRF-CEM leukemia and U373 astrocytoma cells 1.5- to 10.3-fold compared with that of artemisinins applied without iron. Growth inhibition by artesunate and ferrous iron correlated with induction of apoptosis. Cell cycle perturbations by artesunate and ferrous iron were not observed. Treatment of p53 wild-type TK6 and p53 mutated WTK1 lymphoblastic cells showed that mutational status of the tumor suppressor p53 did not influence sensitivity to artesunate. The effect of ferrous iron and transferrin was reversed by monoclonal antibody RVS10 against the transferrin receptor (TfR), which competes with transferrin for binding to TfR. CCRF-CEM and U373 cells expressed TfR in 95 and 48% of the cell population, respectively, whereas TfR expression in peripheral mononuclear blood cells of four healthy donors was confined to 0.4-1.3%. This indicates that artemisinins plus ferrous iron may affect tumor cells more than normal cells. The IC(50) values for a series of eight different artemisinin derivatives in 60 cell lines of the U.S. National Cancer Institute were correlated with the microarray mRNA expression of 12 genes involved in iron uptake and metabolism by Kendall's tau test to identify iron-responsive cellular factors enhancing the activity of artemisinins. This pointed to mitochondrial aconitase and ceruloplasmin (ferroxidase).     

Effects of wild-type p53 expression on the quantity and activity of topoisomerase IIalpha and beta in various human cancer cell lines.

The p53 null HL-60 cell line was transfected with plasmids coding for either the wild-type p53 or mutant p53 gene. The stable expression of wild-type p53 resulted in a significant increase in sensitivity to the topoisomerase II poisons etoposide and doxorubicin, but not to the topoisomerase II inhibitors razoxane and ADR-529. HL-60 cells expressing wild-type p53 demonstrated 8- to 10-fold more VP-16 induced DNA breaks by the alkaline elution assay. The effect of inducible expression of wild-type p53 was also studied in the p53 null erythroblastoid cell line K562 and in the human squamous carcinoma cell line SqCC. The inducible expression of wild-type p53 in the K562 cell line resulted in a 3-fold increase in sensitivity to VP-16. The quantity of topoisomerase IIalpha was not altered by the transfection as determined by immunoblotting, while the amount of the beta isoform was increased 2.5-fold in HL-60 cells. The topo II catalytic activity present in nuclear extracts was measured as the decatenation of kinetoplast DNA, and found to be unaltered by p53 expression. Immunostaining for topoisomerase IIalpha was substantially diminished in both stable and inducible wild-type p53 expressing cells when three different antibodies were used (two polyclonal and one monoclonal). However, the addition of VP-16 resulted in a rapid appearance of nuclear fluorescence for topoisomerase IIalpha. No changes in topoisomerase IIbeta immunostaining were observed. These results suggest that an epitope for topoisomerase IIalpha is concealed in cells expressing wild-type p53 and that a complex between topoisomerase IIalpha and p53 may be disrupted by the addition of antitumor drugs.     

Analysis of DNA content in multidrug-resistant cells by image and flow cytometry

Abstract. Nuclear DNA content was assessed in multidrug-resistant (MDR) cells by image and flow cytometry. Two human MDR cell lines (K562-Dox and CEM-VLB) obtained by in vitro drug selection and overexpressing mdr1 gene were compared to their respective sensitive counterparts (K562 and CCRF-CEM) and to the MDR hamster LR73-R cell line obtained by transfection of mouse mdr1 cDNA. Both cell lines obtained by selection displayed a decreased DNA content, as measured by image cytometry after Feulgen staining, or by flow cytometry after staining with propidium iodide, ethidium bromide, or Hoechst 33342. This decrease was not accompanied by changes in cell cycle phase distribution of cells. Moreover, image cytometry of cells stained after various hydrolysis times in 5 M HCl indicated that MDR cells displayed the same hydrolysis kinetics and sensitivity as drug-sensitive cells with a well-preserved stoichiometry of the Feulgen reaction. LR73-R cells transfected with mdr1 cDNA exhibited only a very limited change in propidium iodide staining as compared with sensitive LR73 cells, suggesting that mdr1 gene overexpression alone could not account for the alterations in DNA content observed in the selected MDR cells.     

Expression of wild-type p53 is not compatible with continued growth of p53-negative tumor cells.

Inactivation of the cellular p53 gene is a common feature of Friend virus-induced murine erythroleukemia cell lines and may represent a necessary step in the progression of this disease. As well, frequent loss or mutation of p53 alleles in diverse human tumors is consistent with the view of p53 as a tumor suppressor gene. To examine the significance of p53 gene inactivation in tumorigenesis, we have attempted to express transfected wild-type p53 in three p53-negative tumor cell lines: murine DP16-1 Friend erythroleukemia cells, human K562 cells, and SKOV-3 cells. We found that aberrant p53 proteins, which differ from wild-type p53 by a single amino acid substitution, were expressed stably in these cells, whereas wild-type p53 expression was not tolerated. The inability of p53-negative tumor cell lines to support long-term expression of wild-type p53 protein is consistent with the view that p53 is a tumor suppressor gene.     

Effects of exogenous wild-type p53 on a human lung carcinoma cell line with endogenous wild-type p53

Several studies have shown that expression of exogenous wild-type p53 is detrimental to the growth of cell lines with absent or mutant p53. In this study, wild-type p53 cDNA expression plasmids were transfected into A549 lung carcinoma cells which had previously been shown by sequencing to contain wild-type p53. When a constitutively expressed wild-type p53 plasmid containing the neomycin resistance gene was transfected into these cells, no G418-resistant colonies contained the exogenous p53 cDNA even though the neomycin resistance gene was integrated. When cells were transfected with a dexamethasone-inducible wild-type p53 cDNA expression plasmid, induction of p53 expression resulted in a decreased growth rate and a decreased proportion of S-phase cells. Continuous treatment with dexamethasone resulted in continued p53 expression for 16 days, but beyond that time expression ceased and could not be reinduced. These data indicated that although the A549 cell line could proliferate in the presence of endogenous wild-type p53 there was a strong selection pressure against continued expression of additional exogenous wild-type p53.     

Fast apoptosis and erythroid differentiation induced by imatinib mesylate in JURL-MK1 cells

We compare the effects of Imatinib mesylate (Glivec) on chronic myeloid leukemia derived cell lines K562 and JURL-MK1. In both cell lines, the cell cycle arrests in G(1)/G(0) phase within 24 h after the addition of 1 microM Imatinib. This is followed by a decrease of Ki-67 expression and the induction of apoptosis. In JURL-MK1 cells, the apoptosis is faster in comparison with K562 cells: the caspase-3 activity reaches the peak value (20 to 30 fold of the control) after about 40 h and the apoptosis proceeds to its culmination point, the DNA fragmentation, within 48 h following 1 microM Imatinib addition. Unlike K562 cells, JURL-MK1 cells possess a probably functional p53 protein inducible by TPA (tetradecanoyl phorbol acetate) or UV-B irradiation. However, no increase in p53 expression was observed in Imatinib-treated JURL-MK1 cells indicating that the difference in the apoptosis rate between the two cell lines is not due to the lack of p53 in K562 cells. Imatinib also triggers erythroid differentiation both in JURL-MK1 and K562 cells. Glycophorin A expression occurred simultaneously with the apoptosis, even at the single cell level. In K562 cells, but not in JURL-MK1 cells, the differentiation process involved increased hemoglobin synthesis. However, during spontaneous evolution of JURL-MK1 cells in culture, the effects produced by Imatinib progressively changed from the fast apoptosis to the more complete erythroid differentiation. We suggest that the apoptosis and the erythroid differentiation are parallel effects of Imatinib and their relative contributions, kinetics and completeness are related to the differentiation stage of the treated cells.     

Norcantharidin-induced post-G(2)/M apoptosis is dependent on wild-type p53 gene

Norcantharidin (NCTD), a synthetic analogue of phosphatase type 2A inhibitors, cantharidin, was shown to have limited effects in treating human and animal tumors. The tumor cell killing mechanisms by norcantharidin, however, remain unclear. In this report, we wished to investigate the mechanisms of norcantharidin-mediated cytotoxicity. Effort was made to investigate whether norcantharidin exerted its cytotoxicity through a p53-dependent or -independent mechanism. RT-2 (wtp53) and U251 (mutant p53) glioblastoma cell lines were exposed to norcantharidin at different dosages. Time-course fluorescent-activated cell sorting (FACS) analysis showed that high doses of norcantharidin arrested the cells at the G(2)/M phase and subsequent post-G(2)/M apoptosis in RT-2 cell line. In comparison, the U251 cell line was found resistant to norcantharidin-induced cytotoxicity. Restoring wild-type p53 gene function in the U251 cell line after adenoviral infections induced tumor cell cytotoxicity after exposure to norcantharidin. These results showed that norcantharidin kills tumor cells efficiently corresponding to their endogenous p53 gene status. The results also showed the feasibility of using adenoviral p53 gene therapy to enhance chemosensitivity of tumor cells to norcantharidin.     

The ability to cleave 28S ribosomal RNA during apoptosis is a cell-type dependent trait unrelated to DNA fragmentation

Discrete cleavages within 28S rRNA divergent domains have previously been found to coincide with DNA fragmentation during apoptosis. Here we show that rRNA and DNA cleavages can occur independently in apoptotic cells, i.e. that the previously observed correlation is likely to be coincidental. In HL-60 cells, apoptosis with massive DNA fragmentation could be induced without any signs of rRNA cleavage. The opposite situation; rRNA cleavage without concomitant internucleosomal DNA fragmentation, was found in okadaic acid-treated Molt-4 cells. Other leukemia cell lines underwent apoptosis either without (K562 and Molt-3) or with (U937) both forms of polynucleotide cleavage. In K562 cells transfected with a temperature-sensitive p53 mutant, internucleosomal DNA fragmentation but not 28S rRNA cleavage was inducible by wild-type p53 expression. The absence of apoptotic rRNA cleavage in some cell types suggests that this phenomenon is tightly regulated and unrelated to DNA fragmentation or a presumed scheme for general macromolecular degradation in apoptotic cells.     

Comparative study of DNA damage, cell cycle and apoptosis in human K562 and CCRF-CEM leukemia cells: role of BCR/ABL in therapeutic resistance

The Philadelphia translocation t(9;22) resulting in the bcr/abl fusion gene is the pathogenic principle of almost 95% of human chronic myelogenous leukemia (CML). Imatinib mesylate (STI571) is a specific inhibitor of the BCR/ABL fusion tyrosine kinase that exhibits potent antileukemic effects in CML. BCR/ABL-positive K562 and -negative CCRF-CEM human leukemia cells were investigated. MTT survival assay and clonogenic test of the cell proliferation ability were used to estimate resistance against idarubicin. DNA damage after cell treatment with the drug at the concentrations from 0.001 to 3 microM with or without STI571 pre-treatment were examined by the alkaline comet assay. We found that the level of DNA damages was lower in K562 cells after STI571 pre-treatment. It is suggested that BCR/ABL activity may promote genomic instability, moreover K562 cells were found to be resistant to the drug treatment. Further, we provided evidence of apoptosis inhibition in BCR/ABL-positive cells using caspase-3 activity colorimetric assay and DAPI nuclear staining for chromatin condensation. We suggest that these processes associated with cell cycle arrest in G2/M checkpoint detected in K562 BCR/ABL-positive compared to CCRF-CEM cells without BCR/ABL expression might promote clone selection resistance to drug treatment.     

Interleukin-2- and mitogen-activated NK-like killer cells from highly purified human peripheral blood T cell (CD3+ N901-) cultures

In this study, highly purified (HP) CD3-positive N901-negative T lymphocytes could be induced to become natural killer (NK)-like in culture in the presence of recombinant interleukin-2 (rIL-2) and phytohemagglutinin (PHA). Thus, purified CD3+ N901- T cells from fresh human peripheral blood were obtained by negative selection using an indirect panning technique. To ensure that T lymphocyte fractions were completely devoid of any detectable NK cells, two additional purification procedures were employed: incubation of post-pan T cells with the NK-cytotoxic lysomotropic agent L-leucinemethylester, and complement-mediated lysis using the NK cell specific NKH1a monoclonal antibody. Purity of CD3+ N901- cells could be confirmed by surface marker analysis, whereby two NK-associated antigens, N901 and H-25, were undetectable, while 94 +/- 1% of cells expressed the CD3 (Leu-4) antigen. On functional analysis, fresh HP CD3+ N901- cells exhibited no cytotoxic activity against the standard NK target K562. When HP NK-depleted T lymphocytes were cultured for 7 days in the presence of rIL-2 (100 U/ml), neither surface antigen expression nor cytotoxic activity against K562 changed significantly. However, significant cytotoxicity against K562 [18 +/- 5% specific lysis at 25:1 effector:target (E/T) ratio] could be induced when HP CD3+ N901- cells were grown for 7 days in the presence of rIL-2 and PHA (0.5% v/v). Concomitantly, antigens N901 and H-25 were found to be coexpressed on a minor proportion (22 +/- 16 and 22 +/- 6%, respectively) of CD3+ (88 +/- 2% on day 7) cells. Four-week long-term culture of HP NK-depleted T cells in the presence of rIL-2 and PHA yielded a continuous increase in cytotoxicity against K562 cells (0 up to 46% specific lysis at 25:1 E/T ratio). Of particular interest was the emergence of cytotoxicity against the NK-resistant Daudi cell target (15 +/- 8% specific lysis at 25:1 E/T ratio on day 21). Expression of antigens N901 and H-25 as well as CD3 remained essentially unchanged in long-term culture. In sorting experiments, the H-25+ cell fraction was significantly enriched for cytotoxicity against K562, when compared to both H-25- and unseparated cell fractions. In summary, our results suggest that a proportion of HP CD3+ N901- T lymphocytes may give rise to cells that exhibit NK-like functional and phenotypic properties.     

Involvement of p53 in gemcitabine mediated cytotoxicity and radiosensitivity in breast cancer cell lines

Gemcitabine (2',2'-difluoro-2'-deoxycytidine; dFdCyd) is one of the anti-metabolites drugs that target DNA replication. We evaluated dFdCyd cytotoxicity and its radiosensitizing ability in human breast cancer cell lines, MCF-7 (wild-type p53) and MDA-MB-231 (mutant-type p53) along with normal mammary epithelial cell line (MCF-12) for comparison. Radiosensitivity and cytotoxicity were measured by the clonogenic survival assays. DNA DSBs was studied by Pulse Field Gel Electrophoresis (PFGE) and cell cycle distribution was analyzed by flow cytometry. MDA-MB-231 cells were the most sensitive to the cytotoxicity of dFdCyd (IC(50) 5 nM) then MCF-7 (IC(50) 10nM), whereas MCF-12 cells were the most resistant to the cytotoxicity of dFdCyd (IC(50) 70 nM). MCF-12 and MCF-7 cell lines did not show any radiosensitization to dFdCyd, whereas the MDA-MB-231 cells showed significantly increased radioresistant to dFdCyd at equimolar concentration (p=0.002) and at IC(50) concentration (p<0.001). The DNA double strand breaks (DSBs) repair showed that dFdCyd neither increases DNA DSBs nor decreases the rate of their repair in MCF-12 and MCF-7 cell lines, while the same treatment in MDA-MB-231 cell line led to decrease the rate of DSBs or increase the rate of DNA repair (p=0.034). Therefore, dFdCyd is a cytotoxic agent, especially in the cancer cells irrespective of having wild-type or mutated p53 protein, but it is not effective as radiosensitizer in the cell lines used in this study. dFdCyd combined with radiation reduces the efficacy of chemo-radiotherapy in p53 mutated cells. Therefore, p53-mutated cancer could be a counter-indication for radiation-gemcitabine combined treatment.     

Cultivation with K562 cells leads to blastogenesis and increased cytotoxicity with changed properties of the active cells when compared to fresh lymphocytes.

Lymphocytes from healthy donors were cultivated with K562, a cell line sensitive to natural cytotoxicity and lacking HLA antigens. Blastogenesis and cytotoxicity were generated in the presence of the serum of the lymphocyte donor. The anti K562 killing effect of the cocultivated populations exceeded considerably that of the fresh lymphocytes. In the majority of cases lymphocytes cultured alone had no cytotoxic activity.We have attempted to characterize the active lymphocyte subset, using the fractionation procedure designed for fresh lymphocytes. Separation of T cells on the basis of SRBC rosetting was not efficient because a high proportion of E rosettes remained in the interface when centrifuged on the Ficoll. Furthermore a high number of E rosetting cells adhered to nylon wool.The cytotoxic potential of various fractions correlated to the presence of blasts and EA positive cells. The least active population was the non-adherent, E rosette sedimented one. The cytotoxicity was not restricted to the sensitizing K562 cell line.     

Establishment of human T cell clones exhibiting natural killer-like activity

We have succeeded in establishing a method to reproducibly immortalize human T cells by oncogene(s) transfection (Alam, 1997). This study was based on our previous discoveries that these immortalized T cell lines contained T cells which showed cytotoxicity against K562 cells in MHC-nonrestricted manner. Then we attempted to obtain human T cell clones exhibiting natural killer-like activity. Here, we tried to establish clones from these immortalized T cell lines by limiting dilution after stimulation with K562 cells, and then obtained 16 T cell clones. Two clones among them maintained their stability and showed vigorous growth phenotype. Thus we selected these two clones for further analysis. One is derived from the T cell line transfected with oncogenes ras and fos, the other is from the T cell line transfected with myc and fos. Both clones were demonstrated to be CD4⁺ T cells, indicating that CD4⁺ T cells were preferably expanded from T cell lines immortalized by oncogene transfection. These two clones showed cytotoxicity against K562 cells, indicating that these two T cell clones still retain a natural killer-like activity of killing target cells of K562 cells in a MHC-nonrestricted manner. The natural killer-like activity of the T cell clones was shown to be stable for more than 2 yr when cultured in the presence of IL-2, indicating that introduction of two oncogenes such as ras/fos or myc/fos resulted in the acquisition of infinite replicative life-span but not in transformational alteration of cellular function. This revised version was published online in August 2006 with corrections to the Cover Date.     

Synthesis and potent antileukemic activities of 10-benzyl-9(10H)-acridinones

A novel series of 10-benzyl-9(10H)-acridinones and 1-benzyl-4-piperidones were synthesized and tested for their in vitro antitumor activities against CCRF-CEM cells. Assay-based antiproliferative activity study using CCRF-CEM cell lines revealed that the acridone group and the substitution pattern on the benzene unit had significant effect on cytotoxicity of this series of compounds, among which 10-(3,5-dimethoxy)benzyl-9(10H)-acridinone (3b) was found to be the most active compound with IC(50) at about 0.7 microM. Compound 3b was also found to have antiproliferative activity against two other human leukemic cell lines K562 and HL60 using the MTT assay. The antitumor effect of 3b is believed to be due to the induction of apoptosis, which is further confirmed by PI (Propidium iodide) staining and Annexin V-FITC/PI staining assay using flow cytometry analysis.     

Human cytosolic and mitochondrial folylpolyglutamate synthetase are electrophoretically distinct. Expression in antifolate-sensitive and -resistant human cell lines

Folylpolyglutamate synthetase (FPGS) activity in CCRF-CEM human leukemia cells was found in the cytosolic ( approximately 67% of total) and mitochondrial ( approximately 22%) fractions. A polyclonal antipeptide antibody (430Ab) to human FPGS specifically recognized distinct immunoreactive bands ( approximately 60 kDa) present in each subcellular fraction. Human cytosolic FPGS (hcFPGS) migrated more rapidly than mitochondrial FPGS (hmFPGS); their estimated difference in molecular mass was 1 kDa. The human K562 acute nonlymphocytic leukemia and the A253 and FaDu head and neck cancer cell lines also expressed the two FPGS isoforms, and the ratio of hcFPGS to hmFPGS protein in each cell line was similar. Since K562 and A253 cells are intrinsically resistant to pulse methotrexate (MTX) exposure relative to CCRF-CEM and FaDu cells, respectively, because of decreased MTX polyglutamate synthesis (despite having similar levels of total FPGS activity expression), these data suggest that the natural difference in drug sensitivity cannot be explained by compartmentalization of FPGS activity. Higher expression of hmFPGS relative to hcFPGS was observed in some sublines of CCRF-CEM with acquired MTX resistance suggesting that differential expression of the hmFPGS isoform may contribute to MTX resistance caused by decreased FPGS activity.