Copy number alterations and copy-neutral loss of heterozygosity in Ukrainian patients with primary myelofibrosis

Poluben L.*1, Bryke Ch.R.2, Hsu Y.2, Shumeiko O.3, Neumerzhitska L.1, Klimuk B.1, Rybchenko L.1, Klymenko S.1, Balk S.P.2, Fraenkel P.G.2

Summary. Aim: To examine frequencies and spectrum of genomic alterations in Ukrainian patients diagnosed with primary myelofibrosis (PMF). Materials and Methods: We enrolled 30 Ukrainian patients diagnosed with PMF who were previously tested for usual mutations in mye­loproliferative neoplasms driver genes (JAK2, MPL and CALR). Genomic DNA samples were obtained from peripheral blood leukocytes of these patients. Copy number alterations and copy-neutral loss of heterozygosity (cnLOH) were assessed using a high-density CytoScan HD microarray platform. Statistical significance was evaluated by the Fisher exact test. Results: We identified frequent genomic alterations, but no significant difference in the rates of copy-number loss, copy-number gain, cnLOH, or multiple genomic alterations were found in the groups of PMF patients that were positive for one of the usual mutations in driver genes or negative for such mutations (33.3% and 55.6%, p = 0.4181, 19.0% and 11.1%, p = 1.0000, 61.9% and 44.4%, p = 0.4434, 33.3% and 55.6%, p = 0.4181, respectively). The most frequent alterations were cnLOH at 1p36-1p22, 9p24.3-9p13.3 and 11q12.3-11q25; copy number loss at 7q21-7q36.3 and 13q12.3-13q14.3. Copy number alterations and cnLOH commonly affected the EZH2, LAMB4, CBL, CUX1, ATM, RB1 and TP53 genes, in addition to JAK2, MPL and CALR. Conclusion: We demonstrated the spectrum of genomic alterations in the groups of the Ukrainian PMF patients with or without the usual mutations in the specific driver genes. We identified several potential genes, which may be involved in the myeloproliferative neoplasms development and their phenotype modification (EZH2, LAMB4, CBL, CUX1, ATM, RB1 and TP53).

DOI: 10.32471/exp-oncology.2312-8852.vol-41-no-1.12540

Submitted: November 1, 2018.
*Correspondence: E-mail: larysa.poluben@gmail.com
Abbreviations used: cnLOH — copy-neutral loss of heterozygo­sity; ET – essential thrombocythemia; MPN – myeloproliferative neoplasm; PMF – primary myelofibrosis; PV – polycythemia vera; SNP – single-nucleotide polymorphism.

Primary myelofibrosis (PMF) is a clonal disorder of early hematopoietic stem cells manifesting as bone marrow fibrosis and pancytopenia and classified as BCR-ABL-negative myeloproliferative neoplasm (MPN) [1]. PMF patients are characterized by a severe disease course and worse prognosis, compared with patients of other MPN subtypes: polycythemia vera (PV) and essential thrombocythemia (ET). MPNs are driven by acquired somatic mutations of JAK2, CALR and MPL genes in more than 85% of cases [2].

Studies of chromosomal alterations in MPN patients revealed that they occurred more frequently in PMF patients (50%) compared with PV (15%) and ET (5%) patients. Some alterations were detected recurrently: deletions of 20q, 18q, 13q and 12p; trisomy of chromosomes 8 and 9; copy number gain at 1p, 9p, 17q; and various translocations [3, 4]. Chromosomal alterations in PMF patients, such as trisomy of chromosome 8, deletions of 7, 7q, 5, 5q, 12p, inversion of chromosome 3 are considered as an unfavorable karyotype and used for the risk stratification for these patients [5].

Improved methods for genomic alterations studies, from standard karyotyping with limited sensiti­vity to the high-resolution single-nucleotide polymorphism (SNP) arrays, allowed detection of short regions of copy number alterations and copy-neutral loss of heterozygosity (cnLOH). Thus, oncogenic microalterations can be studied, in particular, in MPN patients. This increases the search power for identification of potentially involved genes in MPN development (as 10–15% of MPN patients are negative for known mutations of the specific driver genes) [2, 5–7].

The aim of the study was to examine frequencies and spectrum of genomic alterations (copy number alterations and cnLOH) in Ukrainian PMF patients.

MATERIALS AND METHODS

30 Ukrainian PMF patients, previously tested for mutations in MPN driver genes (JAK2, MPL and CALR), were enrolled in the study. There were 13 JAK2 V617F-positive, three MPL W515-positive and five positive for CALR gene mutations PMF patients. Nine PMF patients were negative for these mutations. Each patient signed an informed consent in accordance with the Declaration of Helsinki. The study was approved by the local Ethical Committee at the National Research Center for Radiation Medicine (Kyiv, Ukraine). DNA samples were obtained from peripheral blood leukocytes of PMF patients, using a Quiamp DNA extraction kit (Qiagen, Hilgen, Germany). Copy number alterations and cnLOH were assessed, using a High-density Affymetrix CytoScan HD oligo-SNP microarray platform (Affymetrix, Santa Clara, CA, US). Digested and labeled genomic DNA was hybridized to 2.67 million probes according to manufacturer’s recommendations. The Chromosome Analysis Suite (ChAS) software version 3.1 (Affymetrix) was used to analyze the data. The genome assembly version GRCh37/hg19 was used as a reference. All genomic alterations were visually inspected and confirmed, and regions with poor quality were excluded. The regions with at least 50 markers (over 200 kb (50 markers over 100 kb for leukemia regions) were considered for gains, 30 markers over 50 kb (15 markers over 20 kb for leukemia regions) — for losses, and a minimum length of 5 Mb (3 Mb for leukemia regions) — for cnLOH. An online catalog of human genes and genetic disorders Online Mendelian Inheritance in Man (OMIM) was used to identify leukemia-related genes within the altered regions. Statistical significance was evaluated by the Fishers exact test.

RESULTS AND DISCUSSION

Copy-number alterations and cnLOH were frequently found in Ukrainian PMF patients (Table, Fig. 1). There were 71.4% (15/21) of cases with genomic alterations among PMF patients positive for one of known driver gene (JAK2, MPL or CALR) mutations and 55.6% (5/9) of cases — among PMF patients negative for these mutations (p = 0.4311). However, here was no significant difference in the rates of copy-number loss, copy-number gain, cnLOH, or multiple genomic alterations, respectively, in these groups of PMF patients (33.3% and 55.6%, p = 0.4181; 19.0% and 11.1%,= 1.0000; 61.9% and 44.4%, = 0.4434; 33.3% and 55.6%, = 0.4181). The most frequently altered regions were cnLOH at 1p36-1p22 (N = 6), 9p24.3-9p13.3 (N = 3) and 11q12.3-11q25 (N = 3); copy number loss at 7q21-7q36.3 (N = 5) and 13q12.3-13q14.3 (N = 2). Our findings are consistent with published data [3, 4, 7]. CnLOH at 9р, 1р and 19р duplicated the usual driver gene (JAK2, MPL or CALR, respectively) mutations in 33.3% (7/21) of cases among the subset of PMF patients that were positive for one of these mutations. The length of these altered regions varied from 13.7 to 93.6 Mb (Table, Fig. 2). A shorter genomic fragment (1 Mb) of cnLOH with duplicated MPL gene mutation was visually detected in a PMF patient (ID 842) and considered as leukemogenic.

Table. Copy-number alterations and cnLOH in PMF patients
Chr Patients’s ID Driver gene mutated Region altered Type
of alteration
Size, Mb Genes within the regions Altered DNA burden, %*
1 740 TN 1p36.32p36.22, 1q32.3q41 cnLOH 12.8 RBP7 100
818 TN 1p36.33p34.3 cnLOH 34.8 RAP1GAP, RBP7 26
615 JAK2 1q21.1q44 GAIN 103.8 RBBP5, RBM34, TP53BP2 85
842 MPL 1p34 cnLOH 1 MPL 100
848 MPL 1p36.33p33 cnLOH 48.2 CSF3R, MPL, RAP1GAP, RBP7 100
983 JAK2 1p33p32.3 cnLOH 3   100
702 TN 1p36.36p22.1 cnLOH 93.6 CSF3R, JAK1, MPL, RAP1GAP, RBMXL1, RBP7, RPL5 100
2 638 JAK2 2p22.2, 2p23.3, 2p25.3, 2q35 LOSS 3.3 DNMT3A, TP53I3 55–59
702 TN 2q22.3q23.3 LOSS 3.9   54
3 852 JAK2 3p14.2 LOSS 0.3   100
4 740 TN 4q31.23q31.3 cnLOH 4.9 FBXW7 100
1131 JAK2 4p16.3 LOSS 1.8   61
926 TN 4q31.3 LOSS 0.364   100
724 JAK2 4q22.3q23 cnLOH 3.5 RAP1GDS1 100
5 724 JAK2 5p13.2q11.2 cnLOH 9.7   100
638 JAK2 5p, 5q multiple alterations LOSS 134.4 IRF1, RBM22, NPM1, DDX41 24–88
852 JAK2 5p15.2p14.3 cnLOH 6.1   100
6 740 TN 6q21q22.31 cnLOH 15.2   100
7 740 TN 7q21.3 cnLOH 4.3   100
818 TN 7q22.3q36.2 LOSS 47.9 BRAF, EZH2, LAMB4, POT1 21
846 TN 7q35q36.2 LOSS 6.8 EZH2 92
638 JAK2 7q11.23, 7q11.23q21.11, 7q21.2q21.3 GAIN 12.9 RBM48 41–65
638 JAK2 7q21.11q21.2, 7q21.3q36.3 LOSS 71.7 BRAF, CUX1, EZH2, LAMB4, POT1, RBM33 56, 57
702 TN 7q21.3q31.31 LOSS 25.1 CUX1, EZH2, LAMB4 55
842 MPL 7q36.1q36.2 cnLOH 3.3   100
8 846 TN 8p23.3q24.3 GAIN 146.4 CSMD1, RAD21-AS1, RBM12B, RUNX1T1, TP53INP1 76
9 615 JAK2 9p24.3p13.3 cnLOH 35.9 FANCG, JAK2 100
818 TN 9q32q33.1 LOSS 5.5   20
724 JAK2 9p24.3p13.1, 9q34.2q34.3 cnLOH 42.4 FANCG, JAK2 30, 100
539 JAK2 9p24.3p23 cnLOH 13.7 JAK2 41
702 TN 9q21.11q21.13 cnLOH 6.5   100
11 740 TN 11p15.5p15.4 LOSS 1.7   71
740 TN 11q13.2q25 cnLOH 67.7 ATM, CBL, 100
1131 JAK2 11q12.3q13.2, 11q13.3q25 cnLOH 72.1 ATM, CBL, RBM7, TP53AIP1, RBM14, RBM4B 26, 100
1008 JAK2 11q23.3q24.1 cnLOH 6.4 CBL 100
12 615 JAK2 12p13.33p11.1 LOSS 34.7 AEBP2, GPRC5A, KRAS 93
638 JAK2 12q multiple alterations LOSS 12.5 SH2B3, NCOR2 47–61
926 TN 12q21.2q21.31 cnLOH 4.1   100
13 702 TN 13q12.3q14.3 LOSS 19.8 BRCA2, RB1 80
638 JAK2 13q14.13q14.3 LOSS 4.8 RB1 35
15 740 TN 15q23q24.2 cnLOH 7.7   100
743 CALR 15q13.3 GAIN 0.433   91
1014 JAK2 15q24.1q24.2 LOSS 1.4   46
16 743 CALR 16q23.1 LOSS 0.174   81
17 818 TN 17p13.3q21.2 LOSS 39.6 PRPF8, RAP1GAP2, SUZ12, TP53 19
638 JAK2 17p13.1p11.2 cnLOH 9.1   100
638 JAK2 17p13.3p13.1, 17q21.31q21.32, 17q21.33 LOSS 13.3 PRPF8, RAP1GAP2, TP53 50–55
638 JAK2 17q23.2q25.3 GAIN 22.5 RBFOX3, SRSF2 54
18 702 TN 18q12.2q21.1 cnLOH 12.4 SETBP1 100
19 538 CALR 19p13.3p12 cnLOH 22.9 CALR, CALR3, ELANE, JAK3, ZSWIM4 100
740 TN 19q12q13.12 cnLOH 3.5 CEBPA 100
20 904 CALR 20p13 GAIN 0.226   81
842 MPL 20p13p12.3 cnLOH 7 RAD21L1, RBCK1 100
702 TN 20q11.21q13.13 LOSS 18.8 ASXL1, RBL1, RBM12, RBM39, RBPJL, TP53INP2, TP53RK 80
743 CALR 20q13.13q13.33 cnLOH 15.7 CBLN4, CTCFL, DIDO1, GNAS, RTEL1, TP53RK 100
21 724 JAK2 21q11.2 LOSS 0.335   53
22 703 JAK2 22q12.1q12.3 cnLOH 4.4   100
Note: Chr – chromosome; GAIN – copy-number gain; LOSS — copy-number loss; Mb – megabase.
*Percentage of DNA with identified cnLOH, GAIN or LOSS among studied DNA sample.
113 Copy number alterations and copy neutral loss of heterozygosity in Ukrainian patients with primary myelofibrosis
Fig. 1. Genomic alterations in Ukrainian PMF patients positive (a) and negative (b) for usual mutations in MPN driver genes (JAK2, MPL and CALR). LOSS — copy number loss; GAIN — copy number gain
221 Copy number alterations and copy neutral loss of heterozygosity in Ukrainian patients with primary myelofibrosis
Fig. 2. The length of altered regions: a — cnLOH at chromosomes carrying usual mutations of known MPN driver genes; b — the region with cnLOH at 9p24.3p13.1 (35.9 Mb) in JAK2 V617F-positive PMF patient (patient’s ID 615)

In PMF patients, the most frequently affected genes due to copy number alterations and cnLOH in addition to JAK2, MPL and CALR were EZH2, LAMB4, CBL, CUX1, ATM, RB1 and TP53 genes (Fig. 3). Copy number losses of EZH2 at 7q36.1 were detected in three PMF patients negative for usual driver gene mutations and in one JAK2 V617F-positive PMF patient. Epigenetic regulator EZH2 is a member of Polycomb Repressive Complex 2 which is involved in H3K27 trimethylation. Loss of function mutations and cytogenetic alterations of EZH2 are frequently observed in MPN patients. Recent studies showed that EZH2 alterations may be early events in leukemogenesis [8, 9].

 Copy number alterations and copy neutral loss of heterozygosity in Ukrainian patients with primary myelofibrosis
Fig. 3. Frequencies of affected genes with cnLOH and LOSS in Ukrainian PMF patients positive and negative for usual mutations in MPN driver genes (JAK2, MPL and CALR). LOSS — copy number loss; GAIN — copy number gain

Other studies demonstrated that EZH2 loss can dramatically modify the myeloproliferative phenotype reducing survival in the presence of JAK2 V617F mutation. During disease initiation stage, the cooperation between EZH2 alterations and JAK2 V617F mutations increases the ability of JAK2 V617F-positive stem cells to self-renewal [10]. Interestingly, the mentioned JAK2 V617-positive PMF patient (ID 638) with coexisting EZH2 loss at 7q36.1 had multiple chromosome alterations indicating genomic instability potentially caused by this harmful initial combination (Fig. 4). This patient had additional copy number losses at 2p (involving epigenetic regulator DNMT3A and TP53I3 gene which cooperates with p53 in cell death control); 5p and 5q; 12q (involving SH2B3 gene which assists in JAK2-signaling regulation); 13q and 17q (involving well-studied across different malignancies RB1 and TP53 genes, respectively). Altered DNA burden for these regions with copy number losses ranged from 24 to 88%, but most of them were closed to 40–50%, suggesting their relation to the same leukemogenic cell clone. Proto-­oncogene CBL encodes E3 ubiquitin ligase which negatively regulated JAK2-signaling due to JAK2 molecules ubiquitination and degradation. Even though, there is no evidence confirming ability of CBL gene mutations to drive disease, it was shown that they increase cell proliferation due to hypersensitivity to cytokines [11]. In the study we observed two JAK2 V617-positive PMF cases with CBL homozygous loss, indicating that impaired ubiquitination of signaling molecules might give advantages to myeloproliferation. The LAMB4 gene variants were reported in studies on myeloid neoplasms previously, but their biological function remains unknown in MPN [12]. cnLOH of DNA-damage response ATM gene and copy number loss of TP53 suggest their contribution to the disease evolution due to loss of DNA repair function.

 Copy number alterations and copy neutral loss of heterozygosity in Ukrainian patients with primary myelofibrosis
Fig. 4. Multiple genomic alterations (patient’s ID 638)

Another interesting gene was POT1 which was deleted in two cases of PMF patients who harbored multiple genomic alterations (Table 1). This gene encodes a nuclear protein involved in the telomere maintenance, regulating its lengths, protecting chromosome ends from illegitimate recombination and abnormal chromosome segregation. Significantly shortened telomeres, activation of telomerase, and altered expression of telomere-associated proteins are common features of various hematologic malignancies [13].

Recent study reported the use of an oligonucleotide that targets the RNA template of a human telomerase reverse transcriptase and inhibits its activity in some PMF patients [14]. Thus, the role of impaired POT1 protein is most likely implemented in cooperation with other damaging genomic alterations.

CONCLUSIONS

The study demonstrates the spectrum of genomic alterations in Ukrainian PMF patients that are positive and negative for usual mutations in MPN driver genes (JAK2, MPL or CALR). We have identified several genes potentially involved in the disease development and phenotype modification (EZH2, LAMB4, CBL, CUX1, ATM, RB1 and TP53).

REFERENCES

No Comments » Add comments
Leave a comment

ERROR: si-captcha.php plugin says GD image support not detected in PHP!

Contact your web host and ask them why GD image support is not enabled for PHP.

ERROR: si-captcha.php plugin says imagepng function not detected in PHP!

Contact your web host and ask them why imagepng function is not enabled for PHP.