Cancer of the colon and rectum is the second most frequent malignancy in affluent societies. About one million cases occur annually worldwide, and nearly half a million die from it each year (World Health Organization, www.who.int). Adjuvant chemotherapy has been demonstrated to improve, substantially, life expectancy in advanced colorectal cancer, and has increased the likelihood of cure among patients with stage III disease. Although new combination programs are resulting in high response rates and an increase in survival, most colorectal cancers are unresponsive to chemotherapy and unresectable advanced colorectal cancers remain uncurable (Meyerhardt and Mayer, 2005). Genetic and phenotypic heterogeneity appear as key factors to maintain the tumor's ability to survive, grow and metastasize and are also responsible for the failure of otherwise effective therapeutic strategies (Leith and Dexter, 1986; Heppner and Miller, 1998). Colorectal cancer develops as a result of the progressive accumulation of genetic and epigenetic alterations that lead to the transformation of normal colonic epithelium to colon adenocarcinoma (Fearon and Vogelstein, 1990). The loss of genomic stability is a key molecular step in this process and serves to create a permissive environment for the occurrence of alterations in tumor suppressor genes and oncogenes. Alterations in these genes appear to promote colon tumorigenesis by perturbing the function of signaling pathways or by affecting genes that regulate genomic stability (Grady and Markowitz, 2002; Snijders et al., 2003; Vogelstein and Kinzler, 2004).

In sporadic colorectal cancer, different mechanisms of genomic destabilization have been hypothesized, with only one of them firmly demonstrated: microsatellite instability (MSI) (Perucho et al., 1994). This mechanism is well characterized and affects a minority of tumors (about 5–15% of sporadic colorectal cancer). Other mechanisms may also exist, involving a complex pattern of chromosomal aberrations, (Mitelman et al., 1997; Lengauer et al., 1998; Ribas et al., 2003; Rajagopalan and Lengauer, 2004), but the specific events that initiate, direct and enable this instability are poorly understood. Based on the heterogeneous nature of the genetic and epigenetic alterations, together with distinctive morphological and biological features exhibited by tumors, alternative genetic pathways have been proposed in colorectal carcinogenesis (Perucho et al., 1994; Dutrillaux, 1995; Olschwang et al., 1997; Goel et al., 2003; Risques et al., 2003; Frattini et al., 2004).

Methotrexate (MTX) and its polyglutamate forms are potent competitive inhibitors of the dihydrofolate reductase (DHFR) enzyme, which plays a key role in intracellular folate metabolism and is essential for DNA synthesis and cell growth (Chen et al., 1984). MTX is one of the earliest cytotoxic drugs used in cancer therapy, and despite the isolation of multiple other folate antagonists, MTX maintains its significant role as a treatment for different types of cancer (breast, bladder, head and neck cancers, osteogenic sarcoma and leukemias) (Chu et al., 1996). MTX is also notable for its use in inflammatory disease, rheumatoid arthritis and dermatological disorders. The usefulness of treatment with MTX is limited by the development of drug resistance, which may be acquired through different ways, including increased expression of the target gene DHFR (in chromosome 5q14) via gene amplification (Nunberg et al., 1978; Rots et al., 2000; Banerjee et al., 2002), downregulation of reduced folate carrier (RFC) (Matherly, 2001; Bosson, 2003), inefficient polyglutamylation of MTX because of decreased activity of folylpolyglutamate synthase (FPGS) (Rots et al., 1999), upregulation of -glutamate hydrolase (GGH) (Cole et al., 2001) and mutation of DHFR gene resulting in a decreased affinity for MTX (Dicker et al., 1993; Blakley and Sorrentino, 1998). Understanding the molecular mechanisms underlying drug resistance is essential to improve the response rate of current therapies and in the development of new ones. We hypothesize that the chemoresistance capacity and the molecular mechanisms underlying such resistance are conditioned by the genetic features related with the progression pathway in colorectal cancer. To address this issue, we have investigated MTX resistance in a series of colon cancer cell lines representative of alternative genetic pathways.

Characterization of untreated colon cancer cell lines and presence of pre-existing amplicons

Six independent colon cancer cell lines (LoVo, HCT-116, DLD-1, SW480, HT-29 and SK-CO-1) and three cell lines with a common origin (KM12C, KM12SM and KM12L4A) and established from carcinoma and liver metastases selected for varying metastatic potential in nude mice by Morikawa et al. (1988a, 1988b) were characterized cytogenetically by G-bands karyotyping and comparative genomic hybridization (CGH). Three cell lines (DLD-1, LoVo and HCT116) showed a near-diploid karyotype with a few structural alterations, in agreement with previous studies (Lengauer et al., 1998; Masramon et al., 2000; Abdel-Rahman et al., 2001; Camps et al., 2004). All three exhibited MSI. Another three cell lines (HT-29, SW480 and SK-CO-1) were aneuploid and showed multiple chromosomal reorganizations. The remaining three cell lines (KM12C, KM12SM and KM12L4A) showed a complex karyotype (KM12C showed a near-diploid DNA content and the two derivatives KM12SM and KM12L4A were near-tetraploid) and constitute the first model in which both chromosomal and MSI coexist (Camps et al., 2004). A summary of the main features of each cell line is shown in Table 1.

Table 1 - Main genetic features of colon cancer cell lines used in this study.

Full table

Specific attention was given to the presence of pre-existing DNA amplifications (as indicative of gene amplification capacity) that was detected by CGH in three aneuploid cell lines, SK-CO-1, SW480 and HT-29 (Figure 1 and Table 1). HT-29 cells exhibited two normal chromosomes 8 and a metacentric chromosome formed by a spontaneous homogeneously staining region (HSR) amplification in both arms (Figure 1). Fluorescence in situ hybridization (FISH) analysis revealed the involvement of MYC in the amplification of 8q22qter (Figure 1a). Moreover, MYC signals appeared as interspersed spots (4–5 per chromosome arm) suggesting the involvement of additional genomic regions (Figure 1a). SK-CO-1 cells showed amplifications in chromosomes 18 and 22 in CGH profiles (Figure 1b). BCL2 gene amplification (18q21) was excluded by quantitative PCR (data not shown). Additionally, a possible amplification at chromosome 14q22–23 was also visible, although masked in the CGH ideogram by the gain of the whole chromosome 14 (Figure 1b). MYC additional copies but not gene amplification were observed by FISH analysis (Figure 1b). Finally, G-banding and FISH revealed that SW480 cells present one apparently normal chromosome 8, two reorganized chromosomes 8 der(8)t(8,9)(p11,q11) (Figure 1c) and der(8)t(8,19)(q11.2;?) (Figure 1c), and a reorganized chromosome der(19)t(19;8;19;5) with a MYC HSR amplification signal (Figure 1c). SW480 HSR has approximately 8–10 copies of the MYC gene (Masramon et al., 1998).

Figure 1.

Cytogenetic characterization of cell lines. (a) HT-29 chromosome 8 analysis by G-bands karyotyping, CGH and FISH with a MYC probe (Chromosome 8q24/Alphasatellite 8 Cocktail probe (Q-Biogen)). Losses at 8p and gains at 8q are pointed with green and red arrows, respectively. (b) SK-CO-1 analysis by FISH using a MYC probe shows an increased number of normal or reorganized chromosomes 8 and their corresponding MYC gene copies. CGH analysis shows amplifications in chromosomes 22, 18 and 14 (indicated with an arrow) and chromosome 5 loss. (c) SW480 analysis by G-bands karyotyping, CGH and FISH showing the presence of normal and reorganized chromosomes 8. The amplification of 8q24 is indicated with an arrow at CGH ideogram and involves the MYC gene. Methods have been described previously (Masramon et al., 2000)

Full figure and legend (183K)

With regard to KM12C and derivative cells, no signal of gene amplification was observed, as in diploid cell lines. CGH analysis of KM12C cells detected gains at 8q24qter that correspond to an increased number of normal or reorganized chromosomes 8, as demonstrated by FISH of the MYC gene (data not shown). Within the KM12C cells, three different cell populations were characterized and distinguished by the absence or presence of reorganized chromosomes 8: two normal chromosomes 8 in one subpopulation, der(8)t(8;8)(p21;q11) in a second subpopulation and der(4)t(4;8)(q11;q11) in the last subpopulation. The KM12L4A presented four chromosomes 8 with their corresponding MYC locus. KM12SM cell line also contained the four chromosomes 8 and an additional copy of the MYC gene in a reorganized chromosome (data not shown).

Response to MTX treatment

MTX-resistant clones from each cell line were obtained by exposure to stepwise increasing concentrations of MTX from 10^-8 to 10^-6 M (Figure 2 and Table 1). SW480 and SK-CO-1 cells were unable to acquire MTX resistance. In both cases, cells died at low MTX concentrations (3 10^-8 and 10^-7, respectively). Aneuploid cells HT-29 were resistant to MTX treatment. CGH analysis of HT-29 cultured at 10^-6 M MTX showed 'de novo' amplification of the 5q12–14 region on all the surviving clones (Figure 2b). The DHFR gene was detected in the form of double minutes (DMs) by FISH (Figure 2b). A 20–40-fold amplification of the DHFR gene was demonstrated by real-time PCR (see Supplementary Information).

Figure 2.

Sub-populations resistant to MTX from each cell line were obtained by adding increasing concentrations (from 10^-8 to 10^-6 M) of this drug to five T25 flasks per cell line with DMEM culture medium (Life Technologies, Ltd, Paisley, Scotland) supplemented with 10% fetal bovine serum (Life Technologies). A limited number of resistant cells from each T25 flask and from each cell line were plated in Petri dishes under the same conditions of culture medium and MTX concentration. Two to four isolate clones from each Petri dish were collected and grown up in another T25 flask under the same conditions as described above. (a) Chromosome 5 analysis in untreated HT-29 cells. G-banding, CGH and FISH analysis using a chromosome 5q12–13 probe (Bacpac Resources, RPCI-11 human male BAC library of the Children's Hospital, Oakland, CA, USA) showed no signs of DHFR gene amplification. (b) CGH analysis of HT-29 clone 3 showed an amplification at 5q12–14. FISH analysis (right) using the DHFR probe showed new extrachromosomal DMs in addition to a single complete chromosome 5 with the corresponding DHFR copy (pointed with an arrow). (c) SW480 chromosome 5 G-banding analysis showing one complete chromosome 5 and the reorganized homolog. CGH analysis showed loss of the DHFR region (5q12–14). Chromosome 5 painting with a green specific probe showed a single normal chromosome 5 and chromosome 5 reorganizations

Full figure and legend (161K)

DNA amplification has been associated with chromosome double-strand breaks or activation of fragile sites, and these chromosomal abnormalities are caused by structural chromosomal instability in human tumors (Yunis and Soreng, 1984; Poupon et al., 1996; Coquelle et al., 1997; Singer et al., 2000; Tsushimi et al., 2001; Tanaka et al., 2002). HT-29, SW480 and SK-CO-1 cells fitted within this hypothesis because all of them displayed aneuploid complex karyotypes and exhibited pre-existing amplifications. Unexpectedly, only HT-29 cells developed resistance by DHFR amplification in the form of DM. The failure of SW480 and SK-CO-1 to develop resistance by DHFR amplification can be attributed to the presence of only one complete chromosome 5 (Figure 2c). It has been hypothesized that at least two homologous chromosomes are needed in the amplification event (Coquelle et al., 1997; Singer et al., 2000). One chromosome 5 would be reorganized while the other would remain stable. This postulate is in agreement with the observed changes in HT-29 cells, which contained three normal chromosomes 5 with their corresponding DHFR loci (as shown by FISH) (Figure 2a), while the MTX-resistant clones have lost or reorganized one or two chromosomes 5, which is paralleled by the formation of DMs containing the DHFR gene (Figure 2b). These results imply that appropriate karyotyping of tumor cells may help in the determination of responsiveness to specific drugs.

Amplification of the target gene (DHFR) has been shown to be the most important mechanism of resistance in cultured cells (Rots et al., 2000; Singer et al., 2000; Albertson et al., 2003). Nevertheless, the near-diploid cell lines DLD-1, HCT 116 and LoVo became resistant to high MTX concentrations (10^-6 M), the same as KM12C, KM12SM and KM12L4A with no amplification of the DHFR gene, as demonstrated by CGH, FISH and real-time quantitative PCR analyses (data not show). KM12C and derived cell lines exhibited a complex karyotype but did not show pre-existing amplifications. Moreover, MTX resistance was acquired without DHFR gene amplification, indicating that either amplification capability was not feasible in these cells or that other, more efficient, mechanism(s) of drug resistance were available. Interestingly, KM12 cell lines are a rare example of coexistence of chromosomal instability and MSI (Camps et al., 2004). Since all the cell lines displaying MSI developed resistance, it can be hypothesized that this type of instability is a dominant characteristic over the amplification mechanism. The mutator phenotype associated with MSI confers a high mutation rate to cells (Loeb, 2001), increasing the probability of mutational events at many of the genes involved in the MTX cytotoxic pathway, which might result in MTX resistance without DHFR amplification (Gorlick et al., 1996; Banerjee et al., 2002; Snijders et al., 2003).

Therefore, the existence of alternative mechanisms of resistance was considered in responsive cell lines without DHFR amplification. Previous reports showed a reduced affinity of the MTX:DHFR binding due to different mutations at amino-acid positions 9, 15 and 22, all of them at the DHFR exon 1 (Dicker et al., 1993; Blakley and Sorrentino, 1998). In order to examine if these mutations or new ones could be found in MTX-resistant cells, we have sequenced the exon 1 of the DHFR gene in parental and three MTX-resistant clones of HCT116, DLD-1, LoVo and KM12 cell lines. No evidence of alterations in the coding region of the exon 1 was found (data not shown). It has also been reported that RFC-base-defective transport due to loss of RFC expression is another mechanism of resistance to MTX in cultured human tumor cells (Matherly, 2001; Bosson, 2003). All samples showed mRNA expression without significant differences between parental cells and resistant clones, except for KM12C clone 4 and HCT116 clone 4, which showed a decreased level of RFC mRNA (Figure 3). It has also been suggested that a single mutation in the 3'UTR microsatellite sequence of the RFC gene might strongly affect its expression (Ruggiero et al., 2003). DLD-1, LoVo, HCT116 and KM12 cell lines and their resistant clones showed instability in the 3'UTR (T)₁₃ microsatellite sequence of RFC (data not show), but the RFC protein expression was unaffected (Figure 3), suggesting that this mechanism is unlikely to be responsible for the resistance.

Figure 3.

RFC analysis. (a) RT–PCR for RFC mRNA expression in KM12C, HT29 and HCT116 parental cell lines and their MTX resistant clones. Parental HT-29 cells were used as controls. RFC cDNA was prepared by reverse transcription from total RNA using a specific primer. PCR resulted in a 211 bp product and was analysed on a 6% polyacrylamide denaturing gel and stained with silver. Primer sequences are described in Supplementary material. All samples show normal expression levels except KM12C clone 4 and HCT116 clone 4, which have a decreased expression. 18S RNA expression was used as control. (b) Western blot RFC protein analysis in KM12C, HT29 and HCT116 parental cell lines and their MTX resistant clones showed minimal differences of RFC between untreated and MTX-treated cells. -Actin was used as normal control. Western Blot methods are described in Supplementary Information

Full figure and legend (52K)

Development of drug resistance is a major clinical problem that significantly decreases the effectiveness of chemotherapy in human cancers. Therefore, a better knowledge of the cellular mechanisms responsible for resistance to the drugs currently used in conventional chemotherapeutic regimens may be of great help in drawing alternative and more effective therapeutic approaches for unresponsive patients. Our results indicate that multiple mechanisms operate in MTX resistance in colon cancer cells, and that both the capacity to develop resistance and the mechanism(s) involved are associated with the genetic features of the tumor, namely the karyotype and the presence of a microsatellite mutator phenotype. This implies that an appropriate genetic profiling of tumors before treatment may allow the design of personalized therapeutic strategies with improved response rates.

References

1. Abdel-Rahman WM, Katsura K, Rens W, Gorman PA, Sheer D, Bicknell D, Bodmer WF, Arends MJ, Wyllie AH & Edwards PA. (2001) Proc. Natl. Acad. Sci. USA 98: 2538−2543. | Article | PubMed | ChemPort |
2. Albertson DG, Collins C, McCormick F & Gray JW. (2003) Nat. Genet. 34: 369−376. | Article | PubMed | ISI | ChemPort |
3. Banerjee D, Mayer-Kuckuk P, Capiaux G, Budak-Alpdogan T, Gorlick R & Bertino JR. (2002) Biochim. Biophys. Acta 1587: 164−173. | Article | PubMed | ISI | ChemPort |
4. Blakley RL & Sorrentino BP. (1998) Hum. Mutat. 11: 259−263. | Article | PubMed | ISI | ChemPort |
5. Bosson G. (2003) Br. J. Biomed. Sci. 60: 117−129. | PubMed | ISI | ChemPort |
6. Camps J, Morales C, Prat E, Ribas M, Capella G, Egozcue J, Peinado MA & Miro R. (2004) Int. J. Cancer 110: 869−874. | Article | PubMed | ISI | ChemPort |
7. Chen MJ, Shimada T, Moulton AD, Cline A, Humphries RK, Maizel J & Nienhuis AW. (1984) J. Biol. Chem. 259: 3933−3943. | PubMed | ISI | ChemPort |
8. Chu E, Grem JL, Johnston PG & Allegra CJ. (1996) Stem Cells 14: 41−46. | PubMed | ISI | ChemPort |
9. Cole PD, Kamen BA, Gorlick R, Banerjee D, Smith AK, Magill E & Bertino JR. (2001) Cancer Res. 61: 4599−4604. | PubMed | ISI | ChemPort |
10. Coquelle A, Pipiras E, Toledo F, Buttin G & Debatisse M. (1997) Cell 89: 215−225. | Article | PubMed | ISI | ChemPort |
11. Dicker AP, Waltham MC, Volkenandt M, Schweitzer BI, Otter GM, Schmid FA, Sirotnak FM & Bertino JR. (1993) Proc. Natl. Acad. Sci. USA 90: 11797−11801. | PubMed | ChemPort |
12. Dutrillaux B. (1995) Adv. Cancer Res. 67: 59−82. | PubMed | ISI | ChemPort |
13. Fearon ER & Vogelstein B. (1990) Cell 61: 759−767. | Article | PubMed | ISI | ChemPort |
14. Frattini M, Balestra D, Suardi S, Oggionni M, Alberici P, Radice P, Costa A, Daidone MG, Leo E, Pilotti S, Bertario L & Pierotti MA. (2004) Clin. Cancer Res. 10: 4015−4021. | PubMed | ISI | ChemPort |
15. Gayet J, Zhou XP, Duval A, Rolland S, Hoang JM, Cottu P & Hamelin R. (2001) Oncogene 20: 5025−5032. | Article | PubMed | ISI | ChemPort |
16. Goel A, Arnold CN, Niedzwiecki D, Chang DK, Ricciardiello L, Carethers JM, Dowell JM, Wasserman L, Compton C, Mayer RJ, Bertagnolli MM & Boland CR. (2003) Cancer Res. 63: 1608−1614. | PubMed | ISI | ChemPort |
17. Gorlick R, Goker E, Trippett T, Waltham M, Banerjee D & Bertino JR. (1996) N. Engl. J. Med. 335: 1041−1048. | Article | PubMed | ISI | ChemPort |
18. Grady WM & Markowitz SD. (2002) Annu. Rev. Genomics Hum. Genet. 3: 101−128. | Article | PubMed | ISI | ChemPort |
19. Heppner GH & Miller FR. (1998) Int. Rev. Cytol. 177: 1−56. | PubMed | ISI | ChemPort |
20. Leith JT & Dexter DL. (1986) Mammalian Tumor Cell Heterogeneity CRC Press Inc.: Boca Raton, FL.
21. Lengauer C, Kinzler KW & Vogelstein B. (1998) Nature 396: 643−649. | Article | PubMed | ISI | ChemPort |
22. Loeb LA. (2001) Cancer Res. 61: 3230−3239. | PubMed | ISI | ChemPort |
23. Masramon L, Arribas R, Tortola S, Perucho M & Peinado MA. (1998) Br. J. Cancer 77: 2349−2356. | PubMed | ISI | ChemPort |
24. Masramon L, Ribas M, Cifuentes P, Arribas R, Garcia F, Egozcue J, Peinado MA & Miro R. (2000) Cancer Genet. Cytogenet. 121: 17−21. | Article | PubMed | ISI | ChemPort |
25. Matherly LH. (2001) Prog. Nucleic Acid Res. Mol. Biol. 67: 131−162. | PubMed | ISI | ChemPort |
26. Meyerhardt JA & Mayer RJ. (2005) N. Engl. J. Med. 352: 476−487. | Article | PubMed | ISI | ChemPort |
27. Mitelman F, Johansson B, Mandahl N & Mertens F. (1997) Cancer Genet. Cytogenet. 95: 1−8. | Article | PubMed | ISI | ChemPort |
28. Morikawa K, Walker SM, Jessup JM & Fidler IJ. (1988a) Cancer Res. 48: 1943−1948. | PubMed | ISI | ChemPort |
29. Morikawa K, Walker SM, Nakajima M, Pathak S, Jessup JM & Fidler IJ. (1988b) Cancer Res. 48: 6863−6871. | PubMed | ISI | ChemPort |
30. Nunberg JH, Kaufman RJ, Schimke RT, Urlaub G & Chasin LA. (1978) Proc. Natl. Acad. Sci. USA 75: 5553−5556. | PubMed | ChemPort |
31. Olschwang S, Hamelin R, Laurent-Puig P, Thuille B, De Rycke Y, Li YJ, Muzeau F, Girodet J, Salmon RJ & Thomas G. (1997) Proc. Natl. Acad. Sci. USA 94: 12122−12127. | Article | PubMed | ChemPort |
32. Perucho M, Peinado MA, Ionov Y, Casares S, Malkhosyan S & Stanbridge E. (1994) Cold Spring Harb. Symp. Quant. Biol. 59: 339−348. | PubMed | ISI | ChemPort |
33. Poupon MF, Smith KA, Chernova OB, Gilbert C & Stark GR. (1996) Mol. Biol. Cell 7: 345−354. | PubMed | ISI | ChemPort |
34. Rajagopalan H & Lengauer C. (2004) Nature 432: 338−341. | Article | PubMed | ISI | ChemPort |
35. Ribas M, Masramon L, Aiza G, Capella G, Miro R & Peinado MA. (2003) FASEB J. 17: 289−291. | PubMed | ChemPort |
36. Risques RA, Moreno V, Ribas M, Marcuello E, Capella G & Peinado MA. (2003) Cancer Res. 63: 7206−7214. | PubMed | ISI | ChemPort |
37. Rots MG, Pieters R, Kaspers GJ, Veerman AJ, Peters GJ & Jansen G. (2000) Br. J. Haematol. 110: 791−800. | Article | PubMed | ISI | ChemPort |
38. Rots MG, Pieters R, Peters GJ, Noordhuis P, van Zantwijk CH, Kaspers GJ, Hahlen K, Creutzig U, Veerman AJ & Jansen G. (1999) Blood 93: 1677−1683. | PubMed | ISI | ChemPort |
39. Ruggiero T, Olivero M, Follenzi A, Naldini L, Calogero R & Di Renzo MF. (2003) Nucleic Acids Res. 31: 6561−6569. | Article | PubMed | ISI | ChemPort |
40. Singer MJ, Mesner LD, Friedman CL, Trask BJ & Hamlin JL. (2000) Proc. Natl. Acad. Sci. USA 97: 7921−7926. | Article | PubMed | ChemPort |
41. Snijders AM, Fridlyand J, Mans DA, Segraves R, Jain AN, Pinkel D & Albertson DG. (2003) Oncogene 22: 4370−4379. | Article | PubMed | ISI | ChemPort |
42. Tanaka H, Tapscott SJ, Trask BJ & Yao MC. (2002) Proc. Natl. Acad. Sci. USA 99: 8772−8777. | PubMed | ChemPort |
43. Tsushimi T, Noshima S, Oga A, Esato K & Sasaki K. (2001) Cancer Genet. Cytogenet. 126: 34−38. | Article | PubMed | ISI | ChemPort |
44. Vogelstein B & Kinzler KW. (2004) Nat. Med. 10: 789−799. | Article | PubMed | ISI | ChemPort |
45. Yunis JJ & Soreng AL. (1984) Science 226: 1199−1204. | PubMed | ISI | ChemPort |

Acknowledgements

We thank Silvia Beà and Rosa Miró for technical support. This work was supported by a grant from the Ministry of Education and Science (SAF03/5821). C Morales is an IDIBELL fellow.

Supplementary Information accompanies the paper on Oncogene website (http://www.nature.com/onc)