Chronic myeloid leukaemia (CML) is an acquired disease resulting from the malignant transformation of a single stem cell, characterized by anaemia extreme blood granulocytosis and granulocytic immaturity (1). CML is triggered by a reciprocal translocation of the long arms of chromosomes 9 and 22, referred to as the Philadelphia chromosome (2). The breakpoint cluster region (BCR) gene on chromosome 22 and the Abelson murine leukaemia viral oncogene homolog 1 (ABL1) gene on chromosome 9 are involved in this t(9;22)(q34;q11) translocation leading to the formation of a new chimeric fusion oncogene, BCR-ABL1. Expression of this fusion oncogene produces an abnormal, elongated 8-kb ribonucleic acid (RNA) fragment, which is translated into a unique tyrosine phosphoprotein kinase, considered to be essential for the development of the leukaemic phenotype (1,3).

Imatinib mesylate (Gleevec®; Novartis, Basel, Switzerland) has been established as the standard first-line therapy for CML (4,5). Response to therapy is defined as optimal, suboptimal, or failure of treatment. Optimal response is defined as follows: Complete haematological response [normal full blood count, minimal cytogenetic response (CyR) at 3 months]; partial cytogenetic response at 6 months; complete CyR (CCyR) at 12 months and major molecular response (≥3-log reduction in BCR-ABL1 transcripts) at 18 months (6).

MicroRNAs (miRNAs) have recently been identified as a class of small functional non-coding RNAs of 18–25 nucleotides (7,8) which bind to the 3′-untranslated region of mRNAs, downregulating gene expression at the post-transcriptional level. Recent research suggests that miRNAs may play an important role in the pathology of haematological malignancies (9).

miRNAs are found in the genomes of C. elegans, Drosophila, mice, as well as humans, often with highly conserved sequences. However, in mammals, no genomic sequences that are perfectly complementary to known miRNAs have yet been identified (10). The miRNA stem-loop precursor, pre-miRNA, may be transcribed as part of a longer primary RNA termed pri-miRNA (11). miRNAs play critical roles in cellular processes, including metabolism, apoptosis, differentiation and development (8).

miRNAs are often found in polycistronic clusters containing an average of 2–3 miRNAs. These polycistronic clusters are usually located in fragile regions prone to deletions and other molecular damaging alterations (12). miRNAs are expressed in specific haematological cell types with a regulatory function in early haematopoietic differentiation, erythropoiesis, granulopoiesis, megakaryopoiesis and lymphoid development (12). An example was the demonstration of a cluster of miRNAs found to be transcribed from an intron of the C13-25 open reading frame locus at 13q31-q32. This polycistron (hsa-miR-17-92) is amplified and overexpressed in different types of B-cell lymphoma (12). Lymphogenesis was found to be promoted by hsa-miR-17-92 in a murine stem cell transplantation model (13). In addition, it was determined that this polycistron is also transcriptionally regulated by c-MYC, which may be of great significance, as oncogenic variants of ABL1 are upregulated by c-MYC expression; furthermore, c-MYC is required and acts synergistically with BCR-ABL1, resulting in transformation to CML (12).

Gordon et al found hsa-miR-10a, hsa-miR-17/92, hsa-miR-150, hsa-miR-203 and hsa-miR-328 to be the most commonly deregulated miRNAs in CML (14). A miRNA may function as a tumour suppressor or an oncogene (15). An example of this is hsa-miR-17/92, which has been shown to target BCL2, STAT5 and JAK2 (tumour suppressor), as well as the CDK inhibitor CDKN1A (p21), functioning as an oncogene. This cluster of miRNAs also exhibits increased expression in the early chronic phase of CML, but not in blast phase (BP), making it a possible therapeutic target. Downregulation of hsa-miR-328 has also been described as a crucial stage of disease progression to BP (15).

The present pilot study was performed to determine the miRNA expression profiles in a South African population of CML patients compared with the miRNA signature in CML described in the literature. This study was approved by the Research Ethics Committee of the University of Pretoria (38/2007).

Complete haematological response, CCyR and major molecular response are currently used as prognostic indicators. The RT-qPCR used to detect and measure BCR-ABL transcript levels is required for this classification (6). A growing body of evidence suggests that miRNA expression may be used as a prognostic marker. In a study performed on miRNA expression in chronic lymphocytic leukaemia (CLL), Calin et al (17) observed a significant association between a group of miRNAs and the expression of ZAP-70 and the time between diagnosis and initial treatment. The authors believe that miRNA expression may be a valuable addition to existing prognostic markers for CLL (17). Cummins and Velculescu alluded to the possibility of using miRNAs as early indicators of cancer initiation or progression (18). Wu et al also concluded that miRNAs may be used as prognostic markers in cancer (19).

The aim of the present study was to compare miRNA expression profiles in patients with newly diagnosed CML, those on therapy with imatinib mesylate, and healthy individuals. The miRNA expression data demonstrated that a significant number of miRNAs were expressed in the control group compared with the other two groups. Our results were consistent with those of Ramkissoon et al (20), in that no hsa-miR155 expression was found in healthy individuals, but was present in both groups of CML patients. One discrepant finding was the expression of hsa-miR-223 by the malignant cells isolated by Ramkissoon et al (20). We were unable to identify expression of hsa-miR223 in either of the two groups of CML patients, whereas their study only focused on expression of miR-142, miR-155, miR-181 and miR-223 (20).

Focusing on the samples from the group of patients on therapy (group 2), samples C2 and C7 were from patients who were not in remission at the time. Sample C9 was from a patient who achieved CCyR at the time the sample was collected. Of note, the patient who provided sample C7 was not responding to therapy at the time the sample was collected. The expression profiles of these three samples suggest that hsa-miR-122 and hsa-miR210 may be used as possible indicators of response to therapy; however, due to the small sample size, definitive conclusions could not be reached.

Jose-Eneriz et al investigated the expression profiles of 250 miRNAs in 8 CML patients, 3 of whom were considered to be resistant to therapy (4), and they identified 19 miRNAs differentially expressed between CML patients responding to therapy and those resistant to imatinib mesylate therapy. They found the following 18 miRNAs to be downregulated in CML patients not responding to therapy: hsa-miR-7, hsa-miR-23a, hsamiR-26a, hsa-miR-29a, hsa-miR-29c, hsa-miR-30b, hsamiR-30c, hsa-miR-100, hsa-miR-126#, hsa-miR-134, hsa-miR-141, hsa-miR-183, hsa-miR-196b, hsa-miR-199a, hsa-miR-224, hsa-miR-326, hsa-miR-422b and hsa-miR-520a. Only hsa-miR-191 was upregulated in CML patients resistant to therapy (4). In contrast, our results showed 49 miRNAs to be upregulated compared with the 18 miRNAs that were downregulated (4). However, the results of both studies were based on small sample sizes. In addition, although both studies had the same objectives, the methodology differed, making comparison difficult. However, the results of both studies suggest that there is a significant difference in the miRNA expression profile patterns between the two groups of CML patients.

As the sample size of this pilot study was small, more in-depth research on a large scale is required to validate these results. Further research in this field is required to elucidate the possible role of miRNA expression as a prognostic marker. Novel miRNAs may be identified that may be of value as prognostic markers in CML, which may lead to individualized therapy and improved patient outcome.