The TI-1 cell line was reportedly established from peripheral blood blasts from a male patient with acute myeloid leukemia (AML) French-American-British subtype M2.1 It was described as having a complex karyotype with 65 to 71 chromosomes and the XXY sex chromosomes. The chromosome aberrations included 2q+, 6p+, 9p+, 9p-, 13p+; trisomy of chromosomes 1, 4, 5, 7, 8, 11, 12, 15, 16, and 19; and 11 unidentified marker chromosomes.1
One of the authors (M.A.C.) received the TI-1 cell line from the original investigators, and because it had not been hitherto fully characterized we performed cytogenetic and molecular genetic analyses. TI-1 cells were cultured in RPMI-1640 medium with 20% fetal bovine serum (GibcoBRL, Rockville, MD). Cells were harvested and chromosomes G-banded using standard methods. Spectral karyotyping (SKY) was performed according to the manufacturer's protocol (Applied Spectral Imaging [ASI], Migdal Haemek, Israel). Fluorescence in situ hybridization (FISH) was performed using standard techniques with the LSI BCR/ABL Dual Color, Dual Fusion Translocation Probe (Vysis, Downers Grove, IL). Comparative genomic hybridization (CGH) was performed according to standard protocol.2 For analysis of copy-number changes, a combination of 3.0 standard deviations from the mean, and upper and lower thresholds of above 1.17 (gains) and below 0.83 (losses) were used.3 DNA isolation and restriction landmark genomic scanning (RLGS) were carried out according to standard protocols.4,5
Cytogenetic analysis of TI-1 revealed a complex karyotype with a modal chromosome number of 68 (range, 60-70 chromosomes) (Figure1A-B). As in the original report,1 we observed trisomy of chromosomes 1, 4, 5, 7, 8, 11, 15, 16, and 19. One copy of chromosome 12, recognized as normal by G-banding, was shown by SKY to be a cryptic der(12)t(12;21). Likewise, the chromosome recognized as a normal Y chromosome in Figure 4 of Taoka et al1 was demonstrated by SKY and FISH to be a der(22)t(9;13;22). Each of the 16 structurally altered chromosomes in Figure 4 of Taoka et al1 was present in our preparations (Figure 1A). Thus, the cells we studied had the same karyotype as TI-1 cells analyzed by Taoka et al,1 ruling out the possibility that the abnormalities arose during in vitro propagation in our laboratory.
The final karyotype description, based on G-banding, SKY, and FISH, is as follows: 63-70<3n>,XX,-X,der(2)(2pter→2q37:: hsr(9q34;22q11.2)),der(2)(2pter→2q37::2?::18?q11→18?q21:: 1p32→1pter),-3,+der(5)t(5;6)(q11;?),der(6)(6pter→6p12 ::1?::6p2?1→6qter),+der(7)t(7;7)(p1?1;q3?1),-9,del(9)(p13p24), der(9)t(9;9)(p1?3;q22),der(10)t(3;10)(p21;q23),+der(10)t(10;17)(p11;?p11)t(3;10)(p21;q23),der(12)t(12;21)(p11.2;q21),-13, der(13)ins(13;9)(p11.1;?),-14,der(17)t(9;17)(?p21;p11)x2,del(18) (q21q23),+der(18)t(?2;18)(?;q21),der(20)(20qter→20p11::1?::6?), der(21)t(1;21)(q21∼23;p13),der(22)(22pter→22q11.2::hsr(9q34; 22q11.2)::13q?::hsr(9q34;22q11.2)::13q?::hsr(9q34;22q11.2)::13q?:: hsr(9q34;22q11.2)),+der(22)(22pter→22q11.2::hsr(9q34;22q11.2) ::13q?::hsr(9q34;22q11.2)::13q?::hsr(9q34;22q11.2)::13q?::hsr (9q34;22q11.2))[cp13]/60-69,idem,der(6)t(3;6)(q2?5;p2?5)[cp10].
The CGH profile, showing 14 gains and 9 losses (Figure 1C), corresponds well with cytogenetics. The copy-number karyotype is as follows: +1p31.3-pter, +1q, +2q24.2-32.1, -2q33-35, +3p21-pter, -3p14.3-q24, +5p, +6p, +6q25.1-qter, +7q21.1-qter, -9pter-q22.3, +9q34.1-qter, +10q11.2-22.3, -10q23-qter, -12p, -13q11-21.3, +13q31-32, -14, -17p, +18pter-q12, -18q12.3-qter, +21q, +22q11.2-13.2.
The RLGS profile of TI-1 had 5 loci with DNA amplification, 2 of which had strong homology to chromosome 22q11. Amplification of RLGS locus 3C71 on 22q11 is shown in Figure 1D; amplification of 22q11 was confirmed by CGH (Figure 1C). FISH localized the amplifiedBCR/ABL fusion gene on 2 acrocentric marker chromosomes and the der(2)t(2;9;22) (Figure 1E).
Surprisingly, our data are essentially identical to those in earlier reports on standard cytogenetics,6-8 FISH withBCR/ABL probe,7-9 multicolor FISH,8,9 and CGH8-10 of the K-562 chronic myelocytic leukemia (CML) cell line, established in 1970 from a female patient with CML in blast crisis.11 Seventeen of 19 structurally aberrant chromosomes we detected were also present in K-562 cells analyzed by multicolor FISH.8,9 Although Naumann et al8 and Gribble et al9 did not detect the der(2)(2pter→2q37::2?:: 18?q11→18?q21::1p32→1pter) or der(18)t(?2;18)(?;q21), we believe these rearrangements result from a translocation between der(18)t(1;18)(p32;q21), an abnormality present in most published karyotypes of K-562,6,7,9 and a normal chromosome 2.
Three key points allowed us to conclude that TI-1 is a derivative of K-562: (1) the presence of many identical, unique and frequently very complex structural aberrations; (2) the same pattern of chromosome gains and losses, including lack of the Y chromosome, whose supposed presence was quoted by Taoka et al1 as an important feature distinguishing TI-1 cells from the K-562 cells; and (3) identical CGH profiles and chromosomal locations of amplifiedBCR/ABL in TI-1 and K-562 cells. Because our G-banded karyotype is the same as the karyotype published by Taoka et al,1 the cross-contamination most likely occurred at the original source.
Importantly, Drexler and colleagues have estimated that 18% of human tumor cell lines have intraspecies cross-contamination that occurred at the source of cell line establishment, including other cases in which K-562 was the contaminating culprit.12,13It has recently been recommended that short tandem repeat profiling be used as an international reference standard for human cell lines used in research settings.14 We wish to notify the scientific community that the TI-1 cell line is a cross-contaminant of K-562 and should no longer be used for research on AML.
Supported by grants CA82351 (B.J.W.), 5P30CA06058, CA09338 (M.A.C.), and CA88111 (C.P.) from the National Cancer Institute, Bethesda, MD