If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Institute of Cancer Sciences, Faculty of Medical and Human Sciences, University of Manchester, and Manchester Academic Health Science Centre, Manchester, United Kingdom
Institute of Cancer Sciences, Faculty of Medical and Human Sciences, University of Manchester, and Manchester Academic Health Science Centre, Manchester, United Kingdom
Institute of Cancer Sciences, Faculty of Medical and Human Sciences, University of Manchester, and Manchester Academic Health Science Centre, Manchester, United KingdomDepartment of Hematology, Manchester Royal Infirmary, Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom
Offprint requests to: Richard John Byers, Ph.D., Department of Histopathology, Institute of Cancer Sciences, Oxford Road, Manchester, M13 9WL, United Kingdom
Affiliations
Institute of Cancer Sciences, Faculty of Medical and Human Sciences, University of Manchester, and Manchester Academic Health Science Centre, Manchester, United KingdomDepartment of Hematology, Manchester Royal Infirmary, Central Manchester University Hospitals NHS Foundation Trust, Manchester, United Kingdom
In chronic myeloid leukemia (CML) cells from different stages of maturation may have differential expression of BCR-ABL at both messenger RNA (mRNA) and protein level. However, the significance of such differential expression to clinical disease behavior is unknown. Using the CML-derived, BCR-ABL expressing cell line, K562, distinct plastic-adherent (K562/Adh) and nonadherent (K562/NonAdh) subpopulations were established and then analyzed both as single cells and as bulk cell populations. BCR-ABL mRNA was upregulated in K562/Adh compared with K562/NonAdh cells in both single cell and bulk population analyses (p < 0.0001). Similarly, phosphorylation of BCR protein was upregulated in K562/Adh, compared with K562/NonAdh cells (63.42% vs. 23.1%; p = 0.007), and these two K562 subpopulations were found to express significantly different microRNA species. Furthermore, treatment with the BCR-ABL tyrosine kinase inhibitor, imatinib, reduced cell viability more rapidly in K562/NonAdh compared with K562/Adh cells (p < 0.005) both at single and bulk cell levels. This discovery of an adherent subpopulation of K562 cells with increased BCR-ABL mRNA, increased phosphorylated BCR protein expression, differential microRNA expression, and increased imatinib resistance suggests that a similar subpopulation of cells can also mediate clinical resistance to imatinib during treatment of patients with CML.
Chronic myeloid leukemia (CML) is characterized by the presence of the Philadelphia chromosomal translocation within primitive hematopoietic stem cells. This reciprocal translocation causes ABL (Abelson oncogene) on chromosome 9 to be juxtaposed with the BCR (Breakpoint Cluster Region) on chromosome 22, creating a fusion gene in which the ABL tyrosine kinase is constitutively activated. Clinically, CML is typified by an expansion of the myeloid, and often megakaryocytic, cell populations. Treatment of CML uses inhibitors of the BCR-ABL protein tyrosine kinase, most commonly imatinib mesylate (imatinib). Response is monitored clinically, morphologically, cytogenetically, and molecularly; these techniques providing increasingly sensitive levels of disease detection. The aim of clinical treatment is to achieve molecular remission, whereby significant expression of BCR-ABL as determined by reverse transcriptase polymerase chain reaction (RT-PCR) is detected in fewer than 1 in 106 cells. Rising BCR-ABL levels detected by RT-PCR analysis of bone marrow or blood heralds the return of clinical disease.
Although it is frequently assumed in CML that individual cells assayed by RT-PCR have similar levels of BCR-ABL expression, this may not be the case. Thus, RT-PCR analysis of bulk cell populations in which the majority of cells do not express BCR-ABL messenger RNA (mRNA) can return a low value, even if a small subpopulation of CML cells that have high expression of the oncogene are present. This is important because these subpopulations of cells that express high levels of BCR-ABL could be responsible for development of clinical resistance. In this regard, Copland et al. [
], using primary cultures of CML cells, found that cells resistant to imatinib treatment expressed higher levels of BCR-ABL and had characteristics of a more primitive developmental stage than did those sensitive to imatinib [
], significant upregulation of the BCR-ABL gene was found in a plastic adherent subpopulation of cells. There is, therefore, heterogeneity of BCR-ABL expression among CML cells and, importantly, there may be an adherent or developmentally primitive subpopulation of CML cells that have higher expression of BCR-ABL associated with resistance to imatinib therapy.
Single-cell methods of analysis allow more detailed investigation of heterogeneity of BCR-ABL expression and potentially imatinib-resistant cell populations in CML. The kinetics of BCR-ABL at DNA, RNA and protein levels, comparing single cells and bulk cell populations have not previously been examined. This study, therefore, has systematically investigated the heterogeneity of BCR-ABL expression in both single and bulk cell populations at DNA, mRNA, and protein levels, using the K562 cell line as a CML model cell system. Adherent and nonadherent K562 subpopulations were established and studied separately to differentially test their resistance to imatinib.
Methods
Cell culture and isolation of plastic-adherent and nonadherent cell sublines
The human CML blast crisis cell line K562 (American Type Culture Collection, Manassas, VA, USA; catalog no. CCL-243) was grown in RPMI 1640 culture medium as detailed in the Supplementary Methods (online only, available at www.exphem.org). Previous studies investigating adherence in K562 cells have used plastic adherence [
]; therefore, we used this established method to segregate cells into adherent and nonadherent subpopulations. Plastic surface-adherent (K562/Adh) and nonadherent (K562/NonAdh) cell populations from standard culture growth of the K562 cell line were generated through more than 4 months of passage as follows. First, standard suspension cultures of K562 cells were split with removal of nonadherent floating cells, after which the flask was sharply tapped to dislodge remaining adherent cells. The nonadherent and adherent cells were then each subcultured in new flasks. At the next passage, the cells in both the “nonadherent” and “adherent” flasks were split into nonadherent and adherent fractions as before. This process was repeated at each subsequent passage to enrich for nonadherent or adherent cells.
DNA analysis
Bulk cell analysis of genomic DNA by quantitative PCR
Cell pellets were prepared from K562/Adh and K562/NonAdh cells by centrifugation and DNA extracted followed by real-time quantitative PCR (qPCR) for POLR2D (a single copy reference gene) and exon 6 of the ABL fragment of BCR-ABL.
Single cell analysis by fluorescence in situ hybridization
K562/Adh and K562/NonAdh sublines were harvested by centrifugation followed by fixation using methanol–acetic acid, after which they were spread on glass slides and hybridized with a dual-fusion translocation BCR-ABL translocation probe to detect the BCR-ABL fusion gene. Hybridization signal was measured in triplicate in 20 cells each from K562/Adh and K562/NonAdh cells using a DeltaVision Deconvolution microscope system.
RNA analysis
Bulk cell analysis in bulk using real time qPCR (RT-qPCR)
K562/Adh and K562/NonAdh cells were pelleted by centrifugation and lysed with Trizol followed by chloroform–ethanol RNA extraction followed by cDNA synthesis using reverse transcriptase. All samples had an RNA integrity number (RIN) of 9 or more and an rRNA ratio [28S/18S] greater than 1.7, confirming high quality of the extracted RNA. BCR-ABL expression was measured by real-time qPCR for ABL and BCR-ABL (primers and probes in the Supplementary Methods [online only, available at www.exphem.org]) and was quantified against standard curves.
Single cell analysis using RT-qPCR
Single-cell qPCR was performed in adherence-unsorted single cells and in single K562/Adh and K562/NonAdh cells sorted directly into 384-well microtiter plates using a fluorescence activated cell sorter followed by single-cell qPCR as detailed in the Supplementary Methods (online only, available at www.exphem.org).
Protein analysis
Protein analysis of bulk cells by flow cytometry
Intracellular phospho-BCR protein was measured in K562/Adh and K562/NonAdh bulk cell populations by flow cytometry. Cells were permeabilized and incubated with a phospho-BCR (Tyr177) antibody. After incubation, cells were resuspended in secondary antibody and analyzed by flow cytometry.
Protein analysis on single cells by in situ proximity ligation assay
The Duolink proximity ligation assay (PLA; Olink Bioscience, Uppsala, Sweden) was used to detect BCR-ABL–specific phosphorylation of Phospho-BCR Y177 as a measure of the BCR phosphorylation status in single cells. BCR-ABL negative Jurkat cells were used as negative control cells. Duolink Image Tool was used to score PLA signals and single-cell data analyzed with GraphPad Prism.
Imatinib treatment
K562/Adh and K562/NonAdh were exposed to three different concentrations of imatinib (Enzo Life Sciences, NY, USA), specifically at 100 nmol/L, 10 nmol/L, and 1 nmol/L. After the exposure of samples to each concentration of imatinib for 2, 4, 24, and 48 hours, cell viability was measured using the XTT Cell Viability Assay, the absorbance of samples being measured with a spectrophotometer (ELISA Reader, Biotek Powerwave 200 96-well Monochromatic Microplate Reader; Vienna, VA, USA) at a wavelength of 460 nm. A wavelength of 690nm was used to measure nonspecific background.
MicroRNA profiling
Total RNA was extracted from K562/Adh and K562/NonAdh cells and sent for global microRNA (miRNA) profiling at Biogazelle (Zwijnaarde, Belgium) for expression of 755 miRNAs [
Serial passage and selection of plastic-adherent K562/Adh and nonadherent K562/NonAdh cells generated biologically distinct sublines
K562 cells were grown in suspension, and a small fraction was noted to be adherent to the plastic dish. These adherent cells (K562/Adh) were passaged serially and expressed as a percentage of the total number of cells at each passage (Fig. 1). Initially, adherent cells constituted 5% of the total cells, increasing with each subsequent passage and reaching a maximum at 16 weeks (mean, 9.2%; range, 8.3%–10%), after which there was no further increase in the percentage of cells adherent upon subsequent passage. In addition, at each passage, the nonadherent fraction was subpassaged. In these cultures, the proportion of adherent cells fell with each passage (data not shown). These findings indicate the presence of biologically distinct subgroups of cells distinguished by adherence.
Figure 1Percentage of adherent cells following serial passages. Percentage of adherent cells initially rose steadily over time with selection for adherent cells at each passage, and stabilized after 16 weeks at a mean of 9.2% of all cells.
Adherent K562/Adh and nonadherent K562/NonAdh cells have similar BCR-ABL fusion gene copy number in both single cells and bulk cell populations
To investigate whether the K562/Adh and K562/NonAdh cells had different copy numbers of the BCR-ABL fusion gene, qPCR was used to measure ABL gene copy number in bulk cell populations. No statistically significant difference was found for the mean relative normalized ratio (RNR) for genomic ABL DNA copy number between K562/Adh and K562/NonAdh cells. RNR was 47.73 for K562Adh and 53.40 for K562NonAdh cells (p = 0.11; Fig. 2A). Single-cell analysis of fusion gene copy number was performed using fluorescence in situ hybridization (FISH). Multiple BCR-ABL fusion gene signals (yellow) were present in interphase nuclei of K562 cells demonstrating amplification of the BCR-ABL fusion gene; one or two single BCR-ABL fusion signals were seen in interphase nuclei of HNT-34 cells (BCR-ABL positive control), and no fusion signals were detected in Jurkat cells (BCR-ABL negative control). There was no statistically significant difference in BCR-ABL fusion gene copy number between K562/Adh and K562/NonAdh interphase nuclei (mean copy numbers of 13.83 and 14.33 copies respectively; p = 0.63; two-tailed t test; Fig. 2B).
Figure 2(A) Quantitative PCR amplification of ABL Exon 6 from K562/Adh and K562/NonAdh cells. Fold change comparison in a bulk DNA copy number analysis of the exon 6 in the ABL gene. Relative normalized ratios of exon 6 of the ABL gene showed no statistically significant difference between K562/Adh and K562/NonAdh cells (47.73 and 53.40, respectively; p = 0.11). (B) FISH analyses of BCR-ABL gene amplification in K562/Adh and K562/NonAdh cells. High-resolution deconvolution microscopy was used to quantity fusion gene copy number, showing no statistically significant difference between K562/Adh and K562/NonAdh cells (mean copy numbers of 13.83 and 14.33 copies, respectively; p = 0.63).
Adherent K562/Adh cells express increased BCR-ABL mRNA transcripts compared with nonadherent K562/NonAdh cells in both single cells and bulk cell populations
To assess whether K562/Adh and K562/NonAdh cells expressed different levels of BCR-ABL RNA transcripts, bulk cell mRNA analysis was performed using RT-qPCR. The level of BCR-ABL mRNA transcripts was found to be significantly higher in the K562/Adh cells, which showed 11-fold higher expression than the K562/NonAdh cells (K562/Adh cells and K562/NonAdh cells showed 46.3- and 4.19-fold increase in BCR-ABL respectively compared with HNT-34 cells; p = 0.022; Fig. 3A). Single-cell analysis was performed using RT-qPCR and demonstrated significant upregulation of BCR-ABL mRNA transcripts in K562/Adh cells when compared with K562/NonAdh cells (mean BCR-ABL copy numbers of 53.11 and 14.06 respectively; p = 0.0013; Fig. 3B). Adherence of unsorted single K562 cells also showed a range of BCR-ABL mRNA expression (Fig. 3B), with a small fraction of cells showing a high level of expression similar to that in the K562/Adh cells. There was no significant difference in BCR-ABL mRNA expression between the adherence of unsorted K562 single cells and that of K562/Adh single cells (p = 0.496). The difference in BCR-ABL mRNA expression between adherent and nonadherent cells was similar to that seen in bulk cell measurements.
Figure 3(A) Bulk expression of BCR-ABL mRNA in K562/Adh and K562/NonAdh cells. RT-qPCR comparing the expression level of BCR-ABL mRNA in relation to a stable reference transcript (ABL) showing upregulation of BCR-ABL in K562/Adh cells compared with K562/NonAdh (p = 0.022). Results are presented as a BCR-ABL/ABL transcript ratio. (B) Single-cell expression of BCR-ABL mRNA in K562/Adh and K562/NonAdh cells. The mean BCR-ABL mRNA level was higher in K562/Adh cells compared with K562/NonAdh cells (mean BCR-ABL copy numbers of 53.11 and 14.06, respectively; p = 0.0013). BCR-ABL mRNA level was also measured in adherence unsorted single cells, which showed no significant difference in expression from that in K562/Adh cells (p = 0.496).
K562/Adh cells have higher levels of phosphorylated BCR-ABL fusion protein compared with K562 NonAdh cells in both single cells and bulk cell populations
Phosphorylation of BCR protein at tyrosine residue (Tyr177) plays a pivotal role in the transforming activity of BCR-ABL leading to CML [
]. To investigate whether there was differential BCR phosphorylation between the adherent K562/Adh cells and the non-adherent K562/NonAdh cells, expression of phospho-BCR (Tyr177) was measured by flow cytometry. Phospho-BCR was highly expressed in a distinct subpopulation of K562/Adh cells (mean of 61.9% cells were positive for phosphor-BCR [Tyr177]). In contrast high phospho-BCR expressing cells were infrequent in the non-adherent population (mean of 14.5% positive cells were positive for phosphor-BCR [Tyr177]). This difference in phospho-BCR expression between K562/Adh and K562/NonAdh cells was significantly different (p = 0.0074; Fig. 4A). Protein analysis within single cells was performed using an in situ PLA to detect phosphorylated BCR-ABL, confirming a significantly higher level of phosphorylated BCR-ABL in K562/Adh cells compared with K562 NonAdh cells (mean signal number per cell of 8.23 and 3.02, respectively; p < 0.0001; Fig. 4B and 4C).
Figure 4(A) Flow cytometric analysis of intracellular Phospho BCR in K562/Adh and K562/NonAdh cells. Flow cytometric analysis of intracellular Phospho BCR in K562/Adh cells demonstrated a distinct subpopulation of high expressing cells (mean of 61.9% cells were positive for phospho-BCR [Tyr177]), whereas high phospho-BCR–expressing cells were infrequent in the nonadherent population (mean of 14.5% positive cells were positive for phosphor-BCR [Tyr177]; p = 0.0074). (B) Proximity ligation assay for phospho-BCR and ABL in K562/Adh and K562/NonAdh cells. In situ PLA images illustrate the phosphorylation status of BCR-ABL in K562/Adh and K562/NonAdh cells. The red spots represent phosphorylated signals. Adherent cells (left panel) show higher number of phosphorylated signals than non-adherent cells (right panel). Scale bar =10 μm. (C) Quantification of the phosphorylation events in single cells. The mean phospho-BCR-ABL PLA signal number was higher in K562/Adh cells compared to K562/NonAdh cells (mean PLA signal numbers of 8.23 and 3.02, respectively; p < 0.0001).
Figure 4(A) Flow cytometric analysis of intracellular Phospho BCR in K562/Adh and K562/NonAdh cells. Flow cytometric analysis of intracellular Phospho BCR in K562/Adh cells demonstrated a distinct subpopulation of high expressing cells (mean of 61.9% cells were positive for phospho-BCR [Tyr177]), whereas high phospho-BCR–expressing cells were infrequent in the nonadherent population (mean of 14.5% positive cells were positive for phosphor-BCR [Tyr177]; p = 0.0074). (B) Proximity ligation assay for phospho-BCR and ABL in K562/Adh and K562/NonAdh cells. In situ PLA images illustrate the phosphorylation status of BCR-ABL in K562/Adh and K562/NonAdh cells. The red spots represent phosphorylated signals. Adherent cells (left panel) show higher number of phosphorylated signals than non-adherent cells (right panel). Scale bar =10 μm. (C) Quantification of the phosphorylation events in single cells. The mean phospho-BCR-ABL PLA signal number was higher in K562/Adh cells compared to K562/NonAdh cells (mean PLA signal numbers of 8.23 and 3.02, respectively; p < 0.0001).
Treatment with the BCR-ABL tyrosine kinase inhibitor imatinib reduced cell viability more rapidly in K562/NonAdh compared with K562/Adh cells
Imatinib at concentrations of 1, 10, and 100 nmol/L significantly reduced cellular viability of K562/NonAdh when compared with K562/Adh cells (p < 0.005), the adherent cells showing a fall in cell viability at 24 hours compared with 4 hours for nonadherent cells (Fig. 5). These results demonstrate that the adherent subpopulation of K562 cells with higher mRNA and protein expression of phospho-BCR-ABL had relative resistance to imatinib.
Figure 5Cell viability assay for K562/Adh and K562/NonAdh treated with imatinib. Cell viability of K562/NonAdh (A) was significantly reduced after 4 hours of imatinib treatment, compared with reduction in cell viability by imatinib after 24 hours for K562/Adh cells (B) (p < 0.005), demonstrating the relative resistance of K562/Adh to imatinib: both K562/Adh and K562/NonAdh treated with 1, 10, and 100 nmol/L imatinib as shown.
Whole genome microRNA profiling demonstrated the K562/Adh and K562/NonAdh subpopulations express significantly different miRNA species
Whole genome expression profiling of 755 miRNAs for the K562/Adh and K562/NonAdh sub-lines was assessed. This analysis demonstrated the expression of 232 miRNA (31% of miRNAs) in K562/NonAdh cells and 194 (26% of miRNAs) miRNA in K562/Adh cells. Of these, 190 miRNAs were expressed in both K562/Adh and K562/NonAdh sub-lines, 4 miRNAs were expressed in K562/Adh alone, and 42 miRNAs in K562/NonAdh cells alone. After global mean normalization of common targets, the fold changes of expression level of the 190 commonly expressed miRNAs were compared between K562/Adh and K562/NonAdh cells (Supplementary Methods, online only, available at www.exphem.org). Of these, 32 miRNA species were upregulated with a fold change greater than 2 in K562/Adh compared to K562/NonAdh (Supplementary Table E3, online only, available at www.exphem.org). Additionally, of the 190 microRNAs expressed by both K562/Adh and K563/NonAdh cells, 14 have previously been associated with BCR-ABL expression in the literature (Table 1). Of these 14 miRNAs, 9 were overexpressed in K562/NonAdh, and 5 overexpressed in K562/Adh. miR-190b was the most highly upregulated miRNA in K562/Adh cells, by 29.01-fold, whereas miR-340 was the most downregulated miRNA, being downregulated by 14-fold in K562/Adh compared with K562/NonAdh cells. The normalized relative quantities for these 14 miRNAs are shown in Figure 6.
Table 1MicroRNAs dysregulated in K562/Adh cells with previous association with CML in literature; expression level given for K562/Adh cells compared to K562/NonAdh cells
Figure 6Normalized relative quantities for published dysregulated miRNA candidates in k562/Adh and K562/NonAdh cells. The normalized relative quantities of 14 miRNA candidates are shown for the k562/Adh (black) compared with K562/NonAdh (white) cells.
The BCR-ABL–specific tyrosine kinase inhibitor imatinib has transformed the clinical treatment of CML. However, clonal evolution, involving mutations in BCR-ABL or ABL oncogene amplification, has appeared as a common cause of drug resistance [
]. Specifically, a link between increased expression of BCR-ABL and the appearance of kinase mutations has been reported, with a twofold increase in the level of BCR-ABL mRNA expression being predictive of emergence of a resistant BCR-ABL mutant clone [
Detection of BCR-ABL mutations in patients with CML treated with imatinib is virtually always accompanied by clinical resistance, and mutations in the ATP phosphate-binding loop (P-loop) are associated with a poor prognosis.
]. It has also been suggested that primitive CML cells are less responsive to tyrosine kinase inhibitors and can form a reservoir for tyrosine kinase resistant subclones [
Cell adhesion-mediated drug resistance (CAM-DR) protects the K562 chronic myelogenous leukemia cell line from apoptosis induced by BCR/ABL inhibition, cytotoxic drugs, and gamma-irradiation.
]. Detection and understanding of these resistant subclones is therefore clinically important.
To address heterogeneity of BCR-ABL expression in CML, we have studied the kinetics of BCR-ABL expression at DNA, RNA, and protein levels in bulk cell populations and single cells, using adherent and nonadherent K562 sublines as a model. Serial passage of K562 cells generated stable subfractions enriched for plastic-adherent cells and for nonadherent populations. We suggest that this differential adhesion, and associated BCR-ABL expression of the K562/Adh and K562/NonAdh cells is likely to reflect different intrinsic biology between the subclones. It is recognized that the expression of BCR-ABL is important for cell adhesion in CML [
Cell adhesion-mediated drug resistance (CAM-DR) protects the K562 chronic myelogenous leukemia cell line from apoptosis induced by BCR/ABL inhibition, cytotoxic drugs, and gamma-irradiation.
] showed a dose-dependent effect of BCR-ABL expression on adhesion in myeloid cells, which was also associated with increased malignancy. An association between BCR-ABL expression with plastic adherence has also been reported [
] studied bulk populations and compared the expression level of BCR-ABL mRNA in K562/Adh cells with that of parental K562 cells; using RT-qPCR, they found higher levels of BCR-ABL expression in the adherent cells. In the present study, we confirmed higher BCR-ABL expression using stable bulk subpopulations selected for high adherence in culture. Taken together, these results confirm the presence, in bulk cell populations, of two distinct K562 cell populations based on adherence characteristics and BCR-ABL mRNA expression, which have biological relevance and which show statistically significant differential BCR-ABL expression. Single-cell analysis in the present study demonstrated a high mean expression of BCR-ABL mRNA and protein expression in the K562/Adh cells, but also a wide variation in expression levels when compared with the K562/NonAdh cells. In particular, when assessed in terms of both mRNA and protein expression, the K562/Adh cells contained a subpopulation of single cells with very high levels of BCR-ABL expression.
In the present study, we showed that the higher expression of BCR-ABL mRNA did not arise from a change in BCR-ABL copy number, either by RT-qPCR of bulk cell populations or by FISH analysis of single cells, indicating that they arise from a common CML clone. This result has implications for understanding the acquisition of increased BCR-ABL transcription and a drug-resistant phenotype in the K562/Adh cells. Given the reported role of miRNA in the regulation of BCR-ABL expression in CML [
]. ABL is directly targeted by different miRNAs, such as miR-203 and miR-130a, and miRNAs may downregulate BCR-ABL levels, inhibiting cell proliferation [
]. The role of miRNAs in BCR-ABL expression is complex and diverse as would be expected from the wide-ranging regulatory roles of this large group of molecules. As a result, the difference in miRNA expression occurs by different mechanisms; this has been reviewed and modelled in detail in Verma et al, [
], leading to a steady state in BCR-ABL expression; this is of particular interest given the downregulation of miR-451 in the K562/Adh subline.
In the present study, 35 miRNAs were found to be upregulated by more than twofold in the K562/Adh cells. A literature review of miRNAs associated with BCR-ABL expression identified 14 of these 35 miRNAs as likely to be important in the downregulation of BCR-ABL transcripts [
]. Downregulation of mir-451 in the K562/Adh cells may therefore explain the upregulation of BCR-ABL mRNA in this subline. There is also evidence that BCR-ABL kinase activity suppresses the expression of negative regulatory miRNAs, thereby increasing BCR-ABL expression in a positive feedback loop. Because important tumor suppressor miRNAs in the BCR-ABL pathway (e.g., miR-203 [
]) were downregulated in the K562/Adh population, disruption of such a regulatory loop by upregulation of miR-203, miR-130a, or miR-451 might improve CML therapy with tyrosine kinase inhibitors.
Single-cell BCR-ABL mRNA measurement in the K562/Adh and K562/NonAdh cells in the present study has therefore demonstrated both a significantly higher mean expression by cells in the adherent subline, within which were cells with high individual levels of expression, with such cells being absent from the nonadherent subline. Interestingly, analysis of adherence-unsorted single cells also demonstrated a subfraction of cells with high levels of BCR-ABL mRNA, which we propose might have segregated into the K562/Adh subline during adherence enrichment (Fig. 3B). Taken together with the results from genomic BCR-ABL measurement, our results indicate that although the adherent and nonadherent sublines might have originated from a common CML clone—supported by equivalent BCR-ABL copy number—differential adhesive properties, differential BCR-ABL mRNA and protein levels, and imatinib resistance can arise as a result of changes to miRNA expression within a subpopulation already present in the unsorted initial bulk population, rather than being stimulated by adhesive culture. In primary CML samples, cells with higher BCR-ABL transcripts may belong to a more primitive imatinib-resistant, progenitor population [
Given heterogeneity of BCR-ABL expression, the small populations of cells that express high levels of BCR-ABL detected in our single-cell analyses might not be detected in bulk cell analyses. This finding highlights the value of measuring BCR-ABL mRNA and protein in single cells, particularly because the high-expressing population might represent a potential pool from which clinical resistance may arise. Routine measurement is not yet feasible in clinical practice, although this in future may be possible using flow cytometry or high throughput microfluidic PCR [
Flow cytometry was performed with help from Dr T. Somervaille (The Paterson Institute for Cancer Research and The University of Manchester). Proximity ligation assay was performed with help from Syed Fathollah.
Author contributions: All authors contributed to research design, or acquisition, analysis and interpretation of data. All authors contributed to drafting and editing the manuscript and all approved the final version submitted. E.G.K. designed and performed the research, collected, analyzed, and interpreted the data, and wrote the manuscript; F.M. performed the research, collected data, and wrote the manuscript; J.B. designed the research, analyzed and interpreted all data, and wrote the manuscript; A.J.M. helped with manuscript preparation and editing; R.J.B. designed the research, analyzed and interpreted the data, and wrote the manuscript; P.J.R.D. designed the research, analyzed and interpreted the data, and wrote the manuscript.
Conflict of interest disclosure
No financial interest/relationships with financial interest relating to the topic of this article have been declared.
Appendix
Supplementary methods
Cell culture and isolation of plastic-adherent and nonadherent cell sublines
The human CML blast crisis cell line K562 (American Type Culture Collection, Manassas, VA, USA) was grown in RPMI 1640 culture medium (Sigma, Poole, UK) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Invitrogen, Paisley, UK), 2 mmol/L L-glutamine (SAFC Bioscience, Lenexa, KS, USA), 100 IU/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich, Castle Hill, Australia); this was used for all cell manipulations. Cells were maintained at 37°C in 5% CO2 at a density 1 × 105 and 1 × 106 cells/mL. Surface-adherent (K562/Adh) and nonadherent (K562/NonAdh) cell populations from standard culture growth of the K562 cell line were generated through more than 4 months of passage as follows. First, standard-suspension cultures of K562 cells were split with removal of nonadherent floating cells, after which the flask was sharply tapped to dislodge remaining adherent cells. The nonadherent and adherent cells were then each subcultured in new flasks. At the next passage, the cells in both the “nonadherent” and “adherent” flasks were split into nonadherent and adherent fractions as before. This was repeated at each subsequent passage to enrich for nonadherent or adherent cells.
DNA analysis
Bulk cell analysis of genomic DNA quantitative PCR
Cell pellets (1 × 106 cells) were prepared from adherent and nonadherent sublines by centrifugation and DNA extracted using the Bioline Isolate DNA Kit (Bioline, London, UK). DNA was quantified using a NanoDrop ND-1000-Spectrophotometer, (Thermo Scientific, Waltham, MA, USA) followed by real-time quantitative PCR (qPCR) for POLR2D, used as single copy gene control, and exon 6 of the ABL fragment of BCR-ABL. qPCR was performed on a Roche LightCycler (Roche Diagnostics, Basel, Switzerland) using SYBR Green I LightCycler 480PCR master mix (Roche Diagnostics, Switzerland), forward primer and reverse primers and with thermal cycling of one cycle at 95°C for 10 minutes followed by 45 cycles of 95°C for 10 seconds and 60°C for 30 second. Genomic DNA was quantified using the following equation: Relative Normalized Ratio or 2 – ΔΔCq = 2 power of [(Cq target (Exon 6) – Cq Reference (POL R2D)) of the calibrator (Jurkat cells) – (Cq target (Exon 6) – Cq Reference (POL R2D))] of the samples (K562/Adh and K562/NonAdh)].
Single-cell analysis by fluorescence in situ hybridization
Adherent and nonadherent sublines were incubated with 0.5 μg/mL of Demecolcine (Invitrogen, Carlsbad, CA, USA) for 30 minutes and 1 × 106 cells harvested by centrifugation followed by fixation using methanol:acetic acid (3:1; Fisher Scientific, Loughborough, UK) after which they were spread on glass slides and hybridized with a dual-fusion translocation BCR-ABL translocation probe (Abbott Vysis, Chicago, IL, USA), to detect BCR-ABL fusion gene. Following post-hybridization washes, the slides were counterstained with 4,6-diaminido-2-phenylindole dihydrochloride (DAPI) plus antifade solution-mounting medium (Abbott Vysis, Chicago, IL, USA) and hybridization signal measured in triplicate in 20 cells each from K562/Adh and K562/NonAdh cells using a DeltaVision Deconvolution microscope system (Applied Precision, Issaquah, WA, USA).
RNA analysis
Bulk cell analysis in bulk using RT-qPCR.
Adherent and nonadherent cells (5–10 × 106) were pelleted by centrifugation and lysed with 1 mL of Trizol solution (Invitrogen, Carlsbad, CA, USA), followed by centrifugation with 0.2 mL of chloroform. The upper aqueous layer was aspirated, an equal volume of 70% (v/v) ethanol was added, and RNA was extracted using an RNeasy Kit (Qiagen, Hilden, Germany), followed by cDNA (cDNA) synthesis using SuperScript II Reverse Transcriptase RNAse Reverse Transcriptase (Invitrogen) and RnaseOUT (Invitrogen). BCR-ABL expression was measured by real-time qPCR for ABL, and BCR-ABL (primers and probes are detailed in Supplementary Table E1, Supplementary Table E2 [online only, available at www.exphem.org]). qPCR was performed using a Roche LightCycler 480 (Roche Diagnostics, Switzerland) using forward and reverse primers (Matabion International AG, Martinsried, Germany); Locked Nucleic Acid (LNA) probe (Universal Probe Library, Roche, Switzerland), master mix (FastStart Taq DNA polymerase, reaction buffer, deoxyribonucleoside triphosphate, and 3.2 mmol/L MgCl2) and template. Standard curves were generated using synthetic template oligonucleotides (sTO) with serial dilutions from 105 molecules down to 10 molecules from which the absolute quantities of BCR-ABL were determined and normalized against the ABL gene.
Single cell analysis using RT-qPCR
Single adherent and nonadherent cells (>105 cells/mL) were sorted directly into 384-well microtiter plates using a fluorescence activated cell sorter (BD Influx Cell Sorter, BD Biosciences, San Jose, CA, USA) for RT-qPCR. Cell viability was confirmed by positivity for 7-aminoacteomycin D. Single-cell qPCR was then performed using a Roche LightCycler 480 (LC480) and the AmpliGrid Single Cell One-Step RT-PCR Kit (Advalytix, Olympus Life Science Research Europa, Munich, Germany) with ABL and BCR-ABL primers and probes. The Start One-Step RT-qPCR reaction was heated to 42°C for 10 min before 50°C for 10 min followed by 58°C for 30 min. Samples were then heated to 95°C for 10 minutes followed by 45 cycles of 94°C for 30 seconds, 55°C for 75 seconds and 72°C for 45 seconds. Absolute quantification of transcript copy numbers in each single cell was determined from parallel reactions of template oligonucleotides (sTOs).
Protein analysis
Protein analysis of bulk cells by flow cytometry
Intracellular phospho-BCR protein was measured in adherent and nonadherent bulk cell populations by flow cytometry. Briefly, 1 × 105 of cells were washed with PBS and resuspended in fixing reagent (Reagent A of Fix & Perm Kit; Invitrogen), permeabilized with permeabilizing reagent (Reagent B of Fix & Perm Kit) and incubated with a phospho-BCR (Tyr177) antibody (Cell Signaling Technology, Beverly, MA, USA; 1:50 dilution) for 60 minutes at room temperature in the dark. After incubation, cells were washed twice in PBS/2% FBS, resuspended in secondary antibody (anti-rabbit IgG Alexa Fluor 568 conjugate [Cell Signaling Technology] 1:50 diluted in PBS) and incubated for 60 minutes at room temperature in the dark. Finally, cells were washed in PBS and analyzed by flow cytometry (CyFlow ML; Partec, Muenster, Germany).
Protein analysis on single cells by in situ proximity ligation assay
The proximity ligation assay (PLA; Duolink in situ PLA; Olink Bioscience, Uppsala, Sweden) was used to detect BCR-ABL specific phosphorylation of Phospho-BCR Y177 as a measure of the BCR phosphorylation status in single cells. Cells were deposited on glass slides and fixed with 100% (v/v) ethanol followed by incubated for 30 minutes at 37°C with Duolink Blocking solution before incubation with Phospho-BCR Y177 rabbit polyclonal antibody (1:1000; Abnova, Heidelberg, Germany) and BCR mouse monoclonal antibody (1:50, Abnova) overnight at 4°C. Sections were then washed in Wash Buffer A followed by incubation with the PLA probes (PLA PLUS Anti- mouse and PLA MINUS Anti-rabbit probes; Olink Bioscience, Uppsala, Sweden), diluted in PLA Antibody Diluent, for 1 hour at 37°C. Slides were then washed and incubated with Duolink hybridization solution for 30 minutes at 37°C, followed by another wash and incubation with Duolink Ligation solution, supplemented with Duolink Ligase, for 30 minutes at 37°C. Slides were then washed before incubation with Duolink Amplification-Polymerase solution for 100 minutes at 37°C, followed by washing in Wash Buffer B and mounting in Duolink Mounting Medium (Olink Bioscience). BCR-ABL negative Jurkat cells were used as a negative control group. The Duolink Image Tool was used to score PLA staining after signal detection with an Olympus microscope (Olympus BX51, Tokyo, Japan). Single-cell data were obtained and analyzed with GraphPad Prism software (GraphPad, La Jolla, CA, USA).
Imatinib treatment
K562/adh and K562/ NonAdh CML cells were grown in imatinib (Enzo life Science, NY, USA) at 100, 10, and 1 nmol/L, and cell viability was measured using the XTT Cell Viability Assay Kit (Roche, Switzerland). Cell viability was assayed in three technical replicates.
MicroRNA profiling
K562/Adh and K562/NonAdh cells were pelleted by centrifugation, and cell pellets were lysed with 1 mL TRIZOL Reagent (Life Technologies, Paisley, UK), followed by the addition of 0.2 mL of chloroform per 1 mL of Trizol (BDH; VWR International, Lutterworth, Leicestershire, England). Centrifugation was then used to yield three phases, after which RNA was precipitated from the upper aqueous phase with 100% isopropyl alcohol (Propan-2-01; Fisher Scientific). Samples were incubated at room temperature for 10 minutes and were centrifuged at 12,000 × g for another 10 minutes at 2–4°C, the RNA pellet washed with 1 mL of 75% (v/v) ethanol (Fisher Scientific), air-dried, and resuspended in RNase-free water (Qiagen). RNA samples underwent RT-qPCR–based global miRNA profiling at Biogazelle (Zwijnaarde, Belgium) [
Supplementary Table E3MicroRNA fold change expression in microRNAs expressed in K562/Adh and K562/NonAdh cells: fold change microRNA expression after global mean normalization of common targets in K562/Adh cells compared with K562/NonAdh cells in 190 microRNAs expressed in both K562/Adh and K562/NonAdh cells
Detection of BCR-ABL mutations in patients with CML treated with imatinib is virtually always accompanied by clinical resistance, and mutations in the ATP phosphate-binding loop (P-loop) are associated with a poor prognosis.
Cell adhesion-mediated drug resistance (CAM-DR) protects the K562 chronic myelogenous leukemia cell line from apoptosis induced by BCR/ABL inhibition, cytotoxic drugs, and gamma-irradiation.
00865-5&id=gr1.jpg){kind=link}
00865-5&id=gr2.jpg){kind=link}
00865-5&id=gr3.jpg){kind=link}
00865-5&id=gr4ab.jpg){kind=link}
00865-5&id=gr4c.jpg){kind=link}
00865-5&id=gr5.jpg){kind=link}
00865-5&id=gr6.jpg){kind=link}