sign in sign up
<<
Home , Antifolate, Antimetabolite, Methotrexate

ANTICANCER RESEARCH 27: 769-774 (2007)

Antifolate Pseudo-resistance Due to Elevated Levels of and in a Commercial Serum Preparation

MARCEL SIMON1, JOHANNES BLATTER2 and CHRISTOF GRANZOW1

1Department Molekulare Toxikologie (G120), Deutsches Krebsforschungszentrum, D-69120 Heidelberg; 2Lilly Deutschland GmbH, Medical Oncology, D-61350 Bad Homburg, Germany

Abstract. Background: Batch variability of sera used for cell It has long been recognized that extracellular culture is of considerable experimental concern. A novel fetal calf allow salvage synthesis that circumvents the serum product, FCS Gold, was claimed to be the first defined fetal inhibitory effects of (3, 4). To eliminate calf serum free of batch variation. Materials and Methods: The extracellular sources in vitro, the use of efficacy of (MTX) and LY231514 (multitargeted thymidine-less media and of dialyzed serum has been antifolate, MTA) in CCRF-CEM cells and KB cells was proposed (4). Alternatively, in test settings requiring compared using media supplemented with FCS Gold or undialyzed serum, it has been suggested to minimize the conventional fetal bovine serum. Results: IC50 values from tests impact of extracellular nucleosides by increasing cell using conventional serum corresponded to published data. FCS numbers, which may lead to initial, rapid depletion of the Gold fully protected the cells from antifolate cytotoxicity. extracellular nucleoside pool (5). Dialysis of FCS Gold restored responsiveness to antifolate drugs. Here, we show the salvage from MTX and MTA action Elevated levels of hypoxanthine and thymidine were present in on two frequently used, drug-sensitive continuous human FCS Gold. They were approximately 10-fold greater than the cell lines by a novel, fractionated and reconstituted fetal calf concentrations required to overcome growth arrest mediated by 2 serum preparation containing elevated levels of thymidine ÌM MTX. Conclusion: FCS Gold or identical products, e.g. FBS and hypoxanthine. Gold, should not be used in studies on antifolate drug action. Materials and Methods Methotrexate (MTX) and LY231514 (multitargeted antifolate, MTA) are drugs inhibiting -requiring Chemicals, reagents and cell culture media. If not specified essential for de novo thymidine and synthesis. otherwise, suppliers and manufacturers mentioned resided in Germany. Aqua ad iniectabilia, HEPES (4-(2-hydroxyethyl)-1- The major MTX target is (1). MTA piperazinethansulfonic acid), trypsin 1:250, bovine serum albumin has been shown to inhibit dihydrofolate reductase, thymidylate fraction V and L-glutamine were products of Braun (Melsungen), synthase and glycinamide ribonucleotide formyltransferase (2). Carl Roth (Karlsruhe), Difco (Detroit, MI, USA), Sigma (Munich) and Biochrom (Berlin), respectively. Hypoxanthine p.a. and thymidine p.a. were purchased from Calbiochem (Schwalbach). Methotrexate (MTX) was obtained as a 5.5 mM pharmaceutical Disclosure: None of the authors have anything to disclose with solution from Lederle Arzneimittel (Münster). LY231514 was respect to the subject matter in the manuscript. provided in powder form by Eli Lilly (Bad Homburg). Minimum essential medium modified for suspension cultures (MEM-S) and Abbreviations: IC50, concentration inhibiting growth by 50 percent; RPMI 1640 medium were prepared, in soluble or powder form, by MEM-S, minimum essential medium for suspension culture; MTA, Biochrom (Berlin). Deviating from the original formulation, multitargeted antifolate; MTX, methotrexate; RPMI 1640, Roswell riboflavin, glutamine and were omitted. Park Memorial Institute medium 1640. Sera, serum handling and serum metabolite contents. Fetal bovine Correspondence to: Prof. Christof Granzow, DKFZ (G120), Im serum (FBS) was obtained from Integro (Zaandam, The Neuenheimer Feld 280, D-69120 Heidelberg, Germany. e-mail: Netherlands). Here, it is designated FBS Integro [cf. (6)]. The fetal [email protected] calf serum preparation FCS Gold (lot Nos A01120-151, A01120- 650 and A01120-711) was obtained from PAA (Coelbe). An Key Words: Antifolate resistance, antifolate salvage, FBS Gold, identical product is being marketed as FBS Gold by the same FCS Gold, fetal bovine serum, KB cells, CCRF-CEM cells. supplier. Sera were shipped in dry ice and stored at –25ÆC. Prior

0250-7005/2007 $2.00+.40 769 ANTICANCER RESEARCH 27: 769-774 (2007) to use, the contents of individual serum flasks were slowly thawed Table I. Proliferation of CCRF-CEM cells in MEM-S supplemented with at 8ÆC and used within 4 weeks. Normally, undialyzed serum was FBS Integro (n=4). used. For some experiments, sera were dialyzed in the dark at 4ÆC for 24 h against 100 volumes of RPMI 1640 (three changes) using Inocula Duration of Cell numbers Cell numbers Doubling Visking type 20/32 dialysis tubing (molecular weight cut-off 14000 (x104 exponential reached while after time Dalton; Carl Roth, Karlsruhe). Dialyzed sera were used immediately. cells/well) growth (h) growing incubation (h) The concentrations of folate, thymidine and hypoxanthine were 0.022 exponentially for 72 h 4 4 ÌM, 0.8 ÌM and 22.8 ÌM in FBS Integro, and 0.025 ÌM, 63 ÌM and (x10 /well) (x10 /well) 75 ÌM in FCS Gold, as determined by Laborärzte Limbach, Schmidt- 4 84 64.20±1.61 45.33±0.29 21.53 Gayk and Coll., Heidelberg, Germany. 8 84 117.30±0.60 90.89±1.81 21.87 16 72 153.30±5.35 153.30±5.35 22.00 Cell lines, preparation of culture media, cell culture and photochemical precautions. CCRF-CEM T-cell cells and KB cells were obtained from the American Type Culture Collection (Bethesda, MD, USA) and grown in MEM-S and RPMI indicated. Concentration gradients of drugs were formed by adding 1640 media, respectively. Powder media were processed according 1% (v/v) of 100-fold concentrated drug solutions. In some to (6). No were used. The final concentrations of experiments, only one very high drug concentration was tested. Cell sodium bicarbonate and HEPES were 9 mM and 3 mM for MEM- suspension was added to a density of 5x104 cells/well for CCRF- S, and 13.5 mM and 4.5 mM for RPMI 1640, respectively. Media CEM cells and of 3x104 cells/well for KB cells. Cell numbers were with 1 mM L-glutamine were completed by adding, under the determined after incubation under standard conditions for 72 exclusive illumination from a 30 W SOX sodium discharge lamp hours. Control wells produced 5x105 to 8x105 CCRF-CEM cells, (Philips, Hamburg), 10% (v/v) of the serum preparation indicated, and 2x105 to 5x105 KB cells. Growth inhibition was expressed as dispensed to brown glass bottles (Schott, Mainz), kept in the dark at percentage of growth compared to untreated controls. If feasible, 8ÆC and used within 10 days. Cells were subcultivated at alternating the concentration resulting in 50% growth inhibition (IC ) was intervals of 3 and 4 days, using respective inocula of 5x105 and 2x105 50 read from percentage of growth vs. log drug molarity graphs. CCRF-CEM suspension cells and of 3x105 and 2x105 trypsinized KB monolayer cells per 10 mL of culture medium. Cultures were Salvage of CCRF-CEM cells from MTX action by thymidine and incubated under air in sealed 25 cm2 plastic tissue culture flasks hypoxanthine. The interference of added thymidine or (Greiner, Frickenhausen) at 36.5ÆC. Cell numbers were determined hypoxanthine with MTX action was studied by identifying the using a Casy I cell analyser and Casystat software (Schärfe System, concentrations of these metabolites required to abolish MTX- Reutlingen). Only CCRF-CEM cells and KB cells with respective mediated growth inhibition of CCRF-CEM cells. MEM-S diameters between 8 Ìm and 30 Ìm and between 10 Ìm and 30 Ìm supplemented with undialyzed FBS Integro was used. Test were considered. To prevent photochemical artifacts, all concentrations of thymidine and hypoxanthine were chosen procedures involving sera, serum-containing media and/or according to (2). The thymidine concentration required was antifolate drugs were performed under flavin-protecting conditions determined by preparing quadruplicate wells containing MTX (2 (6). Tests were performed in 24-well cell culture plates (Greiner, ÌM), hypoxanthine (100 ÌM) and the thymidine concentrations Frickenhausen) using total volumes per well of 1 mL and 2 mL for indicated. The hypoxanthine concentration required was CCRF-CEM cells and KB cells, respectively. determined using MTX (2 ÌM), thymidine (5 ÌM) and the hypoxanthine concentrations indicated. Cell densities were Determination of proliferation kinetics in CCRF-CEM cells. Inocula determined after incubation for 72 h. of 4x104, 8x104 and 16x104 CCRF-CEM cells/well suspended in MEM-S supplemented with FBS Integro were seeded, in duplicate, into seven cell culture plates and incubated under standard Results conditions (36.5ÆC, 2.5% CO2 in humidified air). Every 12 h, cells of one of the plates were counted. Doubling times were read from Kinetics of CCRF-CEM cell proliferation. From inocula of up log cell number vs. time graphs. to 16x104 cells/well, CCRF-CEM cells grew in FBS Integro- supplemented media exponentially for at least 72 h with Test for the influence of serum preparations on cell proliferation. In population doubling times between 21.5 and 22 h (Table I). duplicate tests, inocula of 5x104 CCRF-CEM cells/well and 3x104 KB cells/well were suspended in the appropriate media supplemented with the serum preparation indicated and incubated Effect of different sera on KB . As shown in Table II, as above. Cell counts were performed after 72 h. KB cells grew fastest in media supplemented with FBS Integro. With respect to supporting KB cell growth, minor differences Cytotoxicity assessment. Dilutions of the pharmaceutical MTX existed between the three charges of FCS Gold tested. solution or of frozen aliquots of a 1 mM aqueous MTA stock solution were prepared in aqua ad iniectabilia containing 0.2 mg bovine serum albumin fraction V per mL. After appropriate drug Effects of different sera and of serum dialysis on the growth of CCRF-CEM cells. Numerical growth of CCRF-CEM cells concentration ranges (between IC20 and IC80) had been identified in preliminary experiments, quadruplicate tests were performed was indistinguishable in FBS Integro-supplemented media using the respective media containing the serum preparation and in media supplemented with FCS Gold of lot No.

770 Simon et al: Antifolate Drug Action Blocked by a Commercial Serum Preparation

Table II. Proliferation of KB cells in RPMI 1640 medium supplemented Table IV. IC50 of MTX and MTA in tests using FBS Integro- with various serum preparations (inoculum: 3x104 cells/well, n=1). supplemented media.

Serum Cell numbers after Cell line MTX (nM) n MTA (nM) n incubation for 72 h (x104/well) CCRF-CEM 16.50±1.00 4 27.63±2.97 8 KB 8.25±1.25 6 18.50±2.12 2 FBS Integro 56.30 FCS Gold (Lot No. A01120-151) 52.83 FCS Gold (Lot No. A01120-650) 47.75 FCS Gold (Lot No. A01120-711) 47.65 Table V. Ineffective concentrations of MTX and MTA in tests using FCS Gold-supplemented media.

Table III. Proliferation of CCRF-CEM cells in MEM-S supplemented Cell line MTX (ÌM) n MTA (ÌM) n with various undialyzed and dialyzed serum preparations (inoculum: 5x104 cells/well, n=1). CCRF-CEM 55 1 20 3 KB 55 2 10 3 Cell numbers after incubation for 72 h (x104/well)

Serum Undialyzed Dialyzed Table VI. MTX efficacy on CCRF-CEM cells in tests using MEM-S supplemented with dialyzed preparations of FCS Gold (n=1). FBS Integro 62.72 33.11 FCS Gold (Lot No. A01120-151) 62.32 36.93 Lot No. IC50 (nM) FCS Gold (Lot No. A01120-650) 15.40 37.82 FCS Gold (Lot No. A01120-711) 14.18 38.54 A01120-151 12.00 A01120-650 13.00 A01120-711 12.50

A01120-151. Two other charges of FCS Gold, however, supported growth poorly. Dialysis leveled most of these Table VII. IC50 values (nM) of MTX and MTA in tests using dialyzed differences to an intermediate level (Table III). and undialyzed serum.

MTX MTA Efficacy of MTX and MTA in FBS πntegro-supplemented media. The IC50 values shown in Table IV indicate higher Cell line Undialyzed Dialyzed Undialyzed Dialyzed molar efficacy of MTX and MTA in KB cells, as compared serum serum serum serum to CCRF-CEM cells. CCRF-CEM 16.5 12.5 27.63 16 (14) 14 (6; HS) 4.2 (12) 25 (11) 25.4 (12) Absence of effect of MTX and MTA in FCS Gold- KB 8.25 – 18.5 – supplemented media. In tests using FCS Gold-supplemented 11 (10) – – – media (Table V), MTX concentrations several thousand- fold higher, and MTA concentrations more than 500-fold HS, horse serum; references in brackets; own results using FBS Integro in bold type. higher than the IC50 values of Table IV, failed to inhibit the growth of both KB cells and CCRF-CEM cells.

Restitution of MTX efficacy on CCRF-CEM cells by dialysis CEM cells from MTX-mediated growth inhibition in media of FCS Gold. In tests on CCRF-CEM cells using media supplemented with FBS Integro is illustrated in Figure 1. In supplemented with dialyzed FCS Gold, 12 to 13 nM of the absence of added hypoxanthine, 2 ÌM MTX caused a MTX caused a 50% growth inhibition (Table VI). Results 50% growth inhibition if 5 ÌM thymidine was present. were almost identical in all FCS Gold charges tested. The Addition of hypoxanthine further reduced the MTX effect apparent MTX efficacy was higher than in tests with which was abolished at 10 ÌM (Figure 1A). However, at a undialyzed FBS Integro (cf. Table IV). hypoxanthine concentration of 100 ÌM, 2 ÌM MTX suppressed growth almost completely if no thymidine was Thymidine and hypoxanthine concentrations required to salvage added. Added thymidine decreased MTX efficacy. Salvage CCRF-CEM cells from MTX effect. The salvage of CCRF- was almost complete at 5 ÌM thymidine (Figure 1B).

771 ANTICANCER RESEARCH 27: 769-774 (2007)

cell culture. As a novel approach, FCS Gold was introduced and advertised to be the first defined fetal calf serum without batch variation (10). In the present study on the efficacy of the two antifolate drugs, MTX and MTA, in CCRF-CEM and KB cells, we compared FCS Gold with conventional fetal bovine serum. In media supplemented with conventional serum, both of the cell strains proliferated flawlessly and IC50 values were similar to those of other investigators (2, 11-14). An overview is given in Table VII. Two out of three tested charges of FCS Gold failed to adequately support the growth of CCRF-CEM cells. A third, proliferation-competent charge, simulated antifolate resistance of the cell lines under investigation by protecting both of them from antifolate cytotoxicity. The effectiveness of MTX was restored by dialysis of FCS Gold. Moreover, the hypoxanthine and thymidine concentrations required to abolish the growth arrest mediated by 2 ÌM MTX in CCRF- CEM cells growing in media supplemented with conventional serum were similar to the concentrations of these metabolites present in culture media containing 10% (v/v) FCS Gold. These results indicate that, in tests involving FCS Gold, the antifolate drugs lost their efficacy due to the presence of prohibitive levels of hypoxanthine and thymidine in the fractionated and reconstituted serum preparation. Thus, the use of FCS Gold or identical products, e.g. FBS Gold, should be avoided in studies on antifolate drug action. It is unknown to us whether the Figure 1. Salvage of CCRF-CEM cells from MTX effect by hypoxanthine and thymidine. MEM-S supplemented with conventional fetal bovine elevation of metabolite concentrations observed in FCS serum was used. A, hypoxanthine: Freshly harvested CCRF-CEM cells Gold is restricted to hypoxanthine and thymidine. Since the were added to medium containing MTX, thymidine, and hypoxanthine to supplier provided no detailed information on the make respective final concentrations of 2 ÌM, 5 ÌM, and the values procedures used for preparing FCS Gold, it is unresolved indicated. B, thymidine: Cells were added to medium containing MTX, whether the observed anomalies are related to hypoxanthine, and thymidine to make respective final concentrations of 2 ÌM, 100 ÌM, and the values indicated. Cell densities were determined manufacturing. after incubation for 72 h. Error bars are mean±S.D. of quadruplicates. Acknowledgements

Discussion The authors thank Barbara Liebetrau for excellent technical assistance. Animal serum plays an important role in both cell culture and experimentation in vitro by providing hormones and References other growth factors, transport , adhesion factors, 1 Genestier L, Paillot R, Quemeneur L, Izeradjene K and and stabilizing as well as detoxifying agents. The changeable Revillard JP: Mechanisms of action of methotrexate. nature of serum, being a complex, undefined mixture Immunopharmacol 47: 247-257, 2000. varying for example, from individual to seasonally and with 2 Shih C, Chen VJ, Gossett LS, Gates SB, MacKellar WC, diet, results in batch to batch variability which may alter Habeck LL, Shackelford KA, Mendelsohn LG, Soose DJ, Patel experimental outcomes. Numerous approaches have been VF, Andis SL, Bewley JR, Rayl EA, Moroson BA, Beardsley made to circumvent the disadvantages of serum, e.g. using GP, Kohler W, Ratnam M and Schultz RM: LY231514, a serum from different species (7), treatment before use by pyrrolo[2,3-d]-based antifolate that inhibits multiple folate-requiring enzymes. Res 57: 1116-1123, 1997. filtration, diafiltration, dialysis, heat or fractionation, 3 Kinsella AR, Smith D and Pickard M: Resistance to reduction of serum content (8) or total omission of serum chemotherapeutic : a function of salvage from culture media (9). Still, to date there is no definitive pathway involvement and cellular response to DNA damage. Br solution offering invariant as well as optimal conditions for J Cancer 75: 935-945, 1997.

772 Simon et al: Antifolate Drug Action Blocked by a Commercial Serum Preparation

4 Sobrero AF and Bertino JR: Endogenous thymidine and 11 Taylor EC, Kuhnt D, Shih C, Rinzel SM, Grindey GB, Barredo hypoxanthine are a source of error in evaluating methotrexate J, Jannatipour M and Moran RG: A dideazatetrahydrofolate cytotoxicity by clonogenic assays using undialyzed fetal bovine analogue lacking a chiral center at C-6, N-[4-[2-(2-amino-3,4- serum. Int J Cell Cloning 4: 51-62, 1986. dihydro-4-oxo-7H-pyrrolo[2,3-d]pyrimidin-5-yl)ethyl]benzoyl]- 5 de Anta JM, Real FX and Mayol X: Low tumor cell density L-glutamic acid, is an inhibitor of . J Med environment yields survival advantage of tumor cells exposed to Chem 35: 4450-4454, 1992. MTX in vitro. Biochim Biophys Acta 1721: 98-106, 2005. 12 Schultz RM, Chen VJ, Bewley JR, Roberts EF, Shih C and 6 Heuser M, Kopun M, Rittgen W and Granzow C: Cytotoxicity Dempsey JA: Biological activity of the multitargeted antifolate, determination without photochemical artifacts. Cancer Lett MTA (LY231514), in human cell lines with different resistance 223: 57-66, 2005. mechanisms to antifolate drugs. Semin Oncol 26: 68-73, 1999. 7 Evege EK, Nash CR, Bodziak ML and Jayme DW: Alternative 13 Miwa T, Hitaka T, Akimoto H and Nomura H: Novel serum substitutes for fetal bovine serum. J Cell Biol 101: 489a, pyrrolo[2,3-d]pyrimidine : synthesis and antitumor 1985. activities. J Med Chem 34: 555-560, 1991. 8 Bodziak ML, Grefrath PL, Price PJ and Jayme DW: 14 Gangjee A, Vidwans A, Elzein E, McGuire JJ, Queener SF and Mammalian cell proliferation in low serum medium. In Vitro Kisliuk RL: Synthesis, antifolate, and antitumor activities of Cell Dev Biol 21: 47A (abstract no. 146), 1985. classical and nonclassical 2-amino-4-oxo-5-substituted- 9 Barnes D and Sato G: Methods for growth of cultured cells in pyrrolo[2,3-d]. J Med Chem 44: 1993-2003, 2001. serum-free medium. Anal Biochem 102: 255-270, 1980. 10 Foetal Calf Serum Gold, Product Sheet. PAA Laboratories GmbH, Haidmannweg 9, A-4061 Pasching, Austria. http:// Received January 8, 2007 www.paa.at Accepted February 21, 2007

773