記住我
ORIGINAL ARTICLE
Year : 2020 | Volume
: 45
| Issue : 3 | Page : 121-128
GRAF gene expression in patients with acute and chronic myeloid leukaemia: impact on the prognosis
Fetnat M Tolbaa1, Safia M Diab1, Doaa M El-Ghanam2, Ghada M Mahmoud1, Deena A El-Shabrawy1
1 Department of Clinical Pathology, Benha University, Benha, Egypt
2 Department of Clinical Pathology, Mansoura University, Mansoura, Egypt
Correspondence Address:
Deena A El-Shabrawy
Department of Clinical Pathology, Benha University, Benha
Egypt
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/ejh.ejh_31_17
Objective We aimed to assess the expression level of GRAF gene in patients with acute myeloid leukemia (AML) and chronic myeloid leukemia (CML) and to clarify its prognosticsignificance on the clinical outcome.
Background GTPase regulator associated with focal adhesion kinase (GRAF) is a newly identified protein specifically binding to focal adhesion kinase and negatively regulates the small GTP-binding protein RhoA, which is well known for its growth-promoting effect in RAS-mediated malignant transformation. GRAF is recognized as a tumor suppressor gene that binds to focal adhesion kinase.
Methods In this study, we investigated the expression of the GRAF transcript by realtime quantitative PCR in 37 AML patients, 35 CML patients and 15 healthy and sex-matched controls, and to clarify its prognostic significance on the clinical outcome.
Results We found that the GRAF expression is low in patients with AML and CML. A high GRAF expression is a favorable prognostic marker in patients with AML and CML. In AML, high GRAF transcript.
Conclusion High GRAF expression is a favorable prognostic marker in AML patients and a protective factor against CML progression.
Keywords: acute myeloid leukemia, chronic myeloid leukemia, GTPase regulator associated with the focal adhesion kinase, real-time quantitative PCR, tumor suppressor proteins
Acute myeloid leukemia (AML) is a cytogenetically and molecularly heterogeneous disease characterized by clonal proliferation of myeloid precursors and maturation arrest of myeloid cells in the bone marrow (BM) and impaired production of normal blood cells [1]. Chronic myeloid leukemia (CML) is a progressive and often fatal myeloproliferative neoplasm. The hallmark of CML is an acquired chromosomal translocation known as the Philadelphia chromosome, which results in the synthesis of the breakpoint cluster Region-Abelson murine leukaemia (BCR-ABL) fusion oncoprotein, a constitutively active tyrosine kinase [2]. The introduction of imatinib, a tyrosine kinase inhibitor that is specific for BCR-ABL, was a breakthrough in CML therapy
GTPase regulator associated with the focal adhesion kinase (GRAF), a putative tumor suppressor gene, which is located at chromosome 5q31 and its protein is ubiquitously expressed in various tissues [3], and is found inactivated in hematopoietic malignancies by either genetic or epigenetic abnormalities [4].
GRAF is a protein specifically binding to the proline-rich region in the COOH terminus of the focal adhesion molecule (FAK) and negatively regulates the small GTP-binding protein RhoA, which is well known for its growth-promoting effect in RAS-mediated malignant transformation [5],[6].
The GRAF gene is the human homologue of an isolated avian cDNA. The avian GRAF protein binds to the C-terminal domain of p125 FAK, one of the tyrosine kinases predicted to be a critical component of the integrin signalling transduction pathway, in a Src homology 3 (SH3) domain-dependent manner and stimulate the GTPase activity of the GTP-binding protein RhoA. Thus, GRAF acts as a negative regulator of RhoA, which is well known for its growth-promoting effect in RAS-mediated malignant transformation [5]. Mutations and deletions of GRAF gene were found in some cases with AML or myelodysplastic syndrome with a deletion 5q [3].
Furthermore, Bojesen and colleagues found that GRAF gene promoter was methylated in AML and myelodysplastic syndrome. The suppressed GRAF expression could be restored in leukemic cell lines by treatment with a demethylating agent and an inhibitor of histone deacetylases. GRAF mRNA is decreased in myeloid malignancies. However, the GRAF expression level could improve the stratification of prognostication in patients with myeloid diseases [7].
The aim of this study to assess the expression level of GRAF gene in patients with AML and CML and to clarify its prognostic significance on the clinical outcome.
Patients and methodsThis study was conducted on 37 newly diagnosed de novo adult AML patients representing various French–American–British (FAB) subtypes (19 men and 18 women) aged from 19 to 60 years; 35 CML cases representing various phases (17 men and 18 women) aged from 25 to 72 years and 15 healthy-matched and sex-matched controls. These were patients admitted to the Oncology Center Mansoura University Hospital during 2012 to 2014. An informed consent was obtained from the patients prior to their enrollment in this study. Patients were diagnosed based on standard diagnostic methods; morphological based on FAB classification, as well as cytochemical, immunological and cytogenetic. Distribution of AML patients based on FAB classification was: six had M1 (16.22%), nine had M2 (24.32%), 12 had M4 (32.43%), six had M5 (16.22%) and four had M6 (10.81%). All procedures performed in studies involving human participants were in accordance with the ethical standards of the Benha University Faculty of Medicine Research Committee and with the 1964 Helsinki Declaration and its later amendments.
AML patients received the standard ‘3+7’ induction chemotherapy protocol: doxorubicin (45 mg/m2/day) for 3 days and cytarabine (100 mg/m2/day as a continuous 24 h intravenous infusion) for 7 days. BM aspiration was done between 21 and 28 days after initiation of chemotherapy to demonstrate morphological remission. Consolidation is composed of three to four courses of high-dose cytosine arabinoside (3 g/m2 every 12 h on days 1, 3 and 5; total, 18 g/m2). BM samples were collected from the patients initially at diagnosis and at day 28 of induction therapy. Patients were followed up once every 3 months with clinical examination and complete blood counts. BM aspirate was done if there was any doubt of a relapse on clinical examination or peripheral smear.
Among the CML patients 24 (68.6%) were in chronic phase (CP); six (17.1%) were in accelerated phase and five (14.3%) cases were in blastic phase. CML patients were treated regularly with (400–600 mg) imatinib. Patients were regularly monitored on an outpatient basis; biweekly physical examinations, blood counts and biochemistry were obtained during the first month of imatinib therapy and then monthly until a cytogenetic response was achieved, and then every 3 months thereafter until a complete cytogenetic response was confirmed; BM evaluation was performed every 3 months.
The expression levels of the GRAF transcript were determined using real-time quantitative PCR: RNA extraction from 0.5 ml EDTA-anticoagulated peripheral blood samples, cDNA formation and real-time quantitative PCR execution.
RNA extraction: Extraction done using QIAamp RNA Blood Minikit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). Highly purified RNA was extracted and stored at −20°C in RNase-free water until used in cDNA reverse transcription. The concentration, quality and purity of RNA were measured with a UV spectrophotometer at 260/280 nm. The integrity and distribution of RNA were checked by electrophoresis on a 1.5% agarose gel.cDNA reverse transcription: To synthesize single-stranded cDNA from total RNA we used High Capacity cDNA Reverse Transcription Kit (Applied Biosystem, Foster City, California, USA). cDNA synthesis reaction was performed with 10 μl of RNA sample+10 μl Master Mix (2 μl buffer, 0.8 μl dNTPs, 2 μl random primers, 1.0 μl reverse transcriptase, 1.0 μl RNase inhibitor, 3.2 μl nuclease-free water). Then it was incubated at 25°C for 10 min, 42°C for 1 h, the reaction was inactivated by incubated at 95°C for 5 min and then stored at −20. The reverse transcription reactions product (cDNA) was used for quantitative PCR.Quantitative real-time PCR (QRT-PCR): For quantification of GRAF gene and a reference gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH), real-time quantitative pCR (RT-PCR) was performed using TaqMan Gene Expression Master Mix Kit (Applied Biosystem) and ABI PRISM 7000 Sequence Detection System (Thermo Fisher Scientific, 168 Third Avenue, Waltham, MA, USA). RT-PCR was performed in a MicroAmp optical 96-well plate with 5 μl cDNA, 1 μl of forward primer, 1 μl of reverse primer, 1 μl probe, 1.5 μl endogenous control (GADPH), 12.5 μl Master Mix, 3 μl distilled water. The thermal cycle conditions were set as follows: denaturation (95°C for 15 s), annealing and extension (60°C for 1 min), enzyme activation (95°C for 10 min, then incubation for optimal enzyme activity (50°C for 2 min) for 40 cycles.The primer sequences were as follows: GRAF forward 5′-ATTCCAGCAGCAGCTTACA-3′, GRAF reverse 5′-GATGAGGTGGGCA TAGGG-3′,
GADPH forward 5′-GAAGGTGAAGGTCGGAGTC-3′, GADPH reverse 5′-GAAGATGGTGATGGGATTTC-3′.
Data analysis
Relative quantification of (GRAF) gene normalized to a reference (GAPDH) gene
We chose the normal peripheral blood cells as the calibrator and the cancerous peripheral blood cells of patient’s cells as the test sample. To determine the relative expression of a target gene in the test sample and calibrator sample using reference gene, GAPDH, as the normalizer, the expression levels of both the target and the reference genes have been determined using RT-QPCR, we determined the expression level of the target gene (GRAF) in the test sample relative to the calibrator sample using the comparative cycle time (CT) method [8]. The CT of the target (GRAF) gene was normalized to that of the reference (GAPDH) gene, for both the test sample and the calibrator (normal) sample:
ΔCTtest=CTtarget, test−CTref, test,
ΔCTcalibrator=CTtarget, calibrator−CTref, calibrator,
the ΔCT of the test sample was normalized to the DCT of the calibrator:
ΔΔCT=DCTtest−DCTcalibrator,
the expression ratio was then calculated:
normalized expression ratio=2–ΔΔCT.
The result obtained is the fold increase (or decrease) of the target gene in the test sample relative to the calibrator sample. Normalizing the expression of the target gene to that of the reference gene compensates for any difference in the amount of sample tissue.
ResultsAcute myeloid leukemia patients
Clinical outcome of acute myeloid leukemia patients
Twenty-four (64.9%) patients of AML cases achieved complete remission (CR), seven (18.9%) were refractory, six (16.2%) died during induction therapy, 10 (27.0%) relapsed after CR and a total of 16 (43.2%) cases died during the duration of the study. The overall survival (OS) and disease-free survival (DFS) was 21.12 and 23.01 months, respectively. Cumulative proportion of AML cases surviving at 24 months: the OS was 62.6%, while 59.3% of them were DFS.
GRAF gene expression
We found no significant difference in GRAF expression in different age and sex groups. No relation was found between GRAF expression and clinical manifestations, and laboratory findings including total leukocyte count (TLC), hemoglobin (Hb), platelets, percentage of blasts cells, FAB classification and immunophenotyping of the AML cases (P>0.05 for each). Data are not shown.
ROC analysis was conducted to identify the optimal GRAF expression level for the prediction of achieving CR in AML patients. From this curve, the best cut-off values were established for the prediction of achieving CR in AML patients was 0.007, with a sensitivity of 95.8% and a specificity of 92.3%. The area under the curve was 0.939 [95% confidence interval (CI)=0.854–1.024, P<0.001] ([Table 1],[Table 2],[Table 3] and [Figure 1]).
Table 1 ΔCT values of GRAF gene and GRAF gene expression in studied groups
Table 2 GRAF expression and the clinical outcome in acute myeloid leukemia patients
Table 3 Area under the curve of receiver operating characteristic curve of acute myeloid leukemia complete remission
Figure 1 A receiver operating curve (ROC) analysis of GTPase regulator associated with the focal adhesion kinase expression levels for the prediction of achieving complete remission in acute myeloid leukemia patients. Blue line is ROC, green line is reference line.Regarding GRAF expression and OS and DFS in the AML group, we found that the patients with GRAF expression above median survived longer than those with GRAF expression less or equal to median. The OS was significantly longer in those with GRAF expression above median (69.9%, mean 27.4 months) versus those with GRAF expression below median (42.2%, 15.7 months) (P=0.048). DFS was longer in those with GRAF expression above median, although non-statistically significant (P=0.680) ([Table 4] and [Figure 2] and [Figure 3]).
Figure 2 Overall survival and GTPase regulator associated with the focal adhesion kinase expression in acute myeloid leukemia cases.
Figure 3 GTPase regulator associated with the focal adhesion kinase expression and disease-free survival in acute myeloid leukemia cases.The prediction of survival was done using multivariate analysis applying age, TLC and GRAF expression (below vs. above median) as covariates. Multivariate analysis did not show any significant differences in DFS and OS regarding any covariate (P>0.05). Notably, GRAF expression had low hazard ratio (HR) in DFS and OS (HR=0.691, 0.440, respectively). This may indicate a protective prognostic effect of GRAF expression in AML cases for DFS and OS, although this was not statistically significant ([Table 5]).
Table 5 Disease-free survival and overall survival as dependent parameters studied with other covariates (multivariate analysis)Chronic myeloid leukemia patients
Clinical outcome of chronic myeloid leukemia patients
Response to imatinib was achieved in 25 (71.4%) of the cases, while 10 (28.6%) of the cases were resistant to chemotherapy.
GRAF gene expression in chronic myeloid leukemia patients
No significant differences were found between GRAF expression and degrees of splenomegaly and hepatomegaly, TLC and Hb (P>0.05). However, platelets, basophilia and higher peripheral blasts showed significant negative correlation with GRAF expression (P=0.031, <0.001 and 0.002, respectively).
Regarding GRAF expression and phases of CML cases, there were significant differences between GRAF expression in control versus blastic, accelerated and combined (blastic)+accelerated groups (P=0.004, 0.005 and <0.0001 respectively). Moreover, there were significant differences between GRAF expression in CP versus blastic, accelerated and combined (blastic+accelerated) groups (P<0.0001 for each). No significant increase was found between control versus CP and blastic versus accelerated phase (P=0.335 and 0.991, respectively) ([Table 6]). Regarding the response to treatment, there was significant increase in GRAF expression between those who were chemotherapy responders and those who were resistant (P=0.004) ([Table 7]).
Table 6 GRAF expression in different phases of chronic myeloid leukemia cases
Table 7 GRAF expression in chronic myeloid leukemia patients regarding treatment responseROC analysis was conducted to identify the optimal GRAF expression level for the prediction of resistance in CML patients. From this curve, the best cut-off values were established for the prediction of resistance in CML patients was 3.406, with a sensitivity of 80% and a specificity of 67%. The area under the curve was 0.824 (95% CI=0.668–0.980, P=0.003) ([Table 8] and [Figure 4]).
Table 8 Area under the curve of receiver operating characteristic curve of chronic myeloid leukemia resistance
Figure 4 A receiver operating curve (ROC) analysis of GTPase regulator associated with the focal adhesion kinase expression levels for the prediction of resistance in chronic myeloid leukemia patients. Blue line is ROC, green line is reference line.Univariate and multivariate analyses were performed to identify potential prognostic factors associated with the development of blastic and accelerated phases in CML phase. Univariate analyses showed that increased TLC, basophilia and decreased GRAF expression (equal to or below median) are associated with progression of CML to accelerated or blastic phases. Prediction of risk was done using logistic regression analysis. Multivariate analysis shows that GRAF expression equal to or below median is the only independent prognostic factor to CML progression (P=0.006), and GRAF expression above median is a protective factor against CML progression (OR=0.010, 95% CI=0.0001–0.265). Notably, increased TLC and basophilia were associated with the probability of development of accelerated and blastic phases by univariate analysis, but not by multivariate analysis ([Table 9]).
Table 9 Chronic myeloid leukemia risk to develop accelerated and blastic phases as dependent parameters studied with other covariates Discussion Myeloid malignancies are clonal diseases of hematopoietic stem or progenitor cells. They result from genetic and epigenetic alterations that perturb key processes such as self-renewal, proliferation and differentiation [9].
A GRAF protein contains a centrally located GTPase activating protein domain, followed by a serine/proline-rich domain and a carboxy-terminal SH3 domain [10]. GRAF is a newly identified protein specifically binding to the proline-rich region in the COOH terminus of FAK and negatively regulates the small GTP-binding protein RhoA, which is well known for its growth-promoting effect in RAS-mediated malignant transformation [5]. Rho family GTPases play a role in the growth control besides regulating the organization of the actin cytoskeleton [11].
Physiologically, RhoA has been shown to be involved in the regulation of apoptosis, migration, proliferation and differentiation of cells [12],[13],[14].
Clinical studies have shown the correlation of increased expression of RhoA and invasion, metastasis and progression of several solid tumours including the liver, bladder, oesophageal, head and neck, ovary, gastric, testicular, lung and breast carcinomas [11]. FAK is a cytoplasmic tyrosine kinase identified as a key mediator of signalling by integrins, a major family of cell surface receptors for extracellular matrix, as well as other receptors in both normal and cancer cells [15].
The SH3 domain of the GRAF protein can bind to the proline-rich region in the COOH terminus of FAK [14],[16]. Aberrant expression of FAK is frequent in AML cells and it enhances the migration of leukemic cells from marrow to the circulating compartment, confers drug resistance and negatively influences the clinical outcome [17].
In this study, the level of GRAF expression was significantly decreased in patients with AML compared with the control group (P<0.0001). This was in accordance with previous studies that have been reported that the GRAF transcript was significantly decreased in AML patients [10],[11],[14]. On the contrary, another study reported that the GRAF transcript has not been detected in the primary leukemic cells of AML [18].
In this study, no significant differences were found between GRAF expression level and age, sex, haematologic parameters (including TLC, Hb, platelets and blasts), FAB subtypes immunophenotyping and clinical features (P>0.05 for each). This was in agreement with some previous studies [1],[10],[11].
In this study we found that AML patients with high GRAF transcript levels (mean 1.428) had significantly higher CR rate (P<0.005). This was in accordance with another study, which found that AML patients with higher GRAF expression (high GRAF, 73.5%) showed a significantly higher CR rate (P=0.02) and a lower incidence of primary resistant disease (P=0.01) [14].
In the present study, there was a highly significant difference between GRAF expression and clinical outcome in AML patients who achieved continuous CR, refractory, died in induction therapy and those who relapsed (P<0.0001). We found high GRAF expression in AML patients who achieved continuous CR (mean 1.428), very low GRAF expression in refractory patients (mean 0.003), low GRAF expression in patients who died in induction period (mean 0.017) and relapsed patients (mean 0.228).
As regards the performance characteristics for the prediction of achieving CR in AML in the present study, the ROC curve analysis showed that the best cut-off value for GRAF was 0.007; at this point the sensitivity was 95.8% and the specificity was 92.3% with an area under the curve being 0.939 and (P<0.001).
Interestingly, AML patients with a high expression of GRAF had longer OS (27.4 months) than those with a low expression (15.7 months) (P=0.048). These findings were in concordance with a previous study, which followed up a group of CML patients for 30 months and reported that patients with high GRAF expression had a significantly longer OS than those with a low GRAF expression (P=0.03) and the DFS was longer in those with higher GRAF expression (23.8 months), although nonsignificant (P=0.680) [14]. This might indicate that a high GRAF transcript level could be protective for patients with AML.
Notably, GRAF expression had a low HR in DFS, OS (HR=0.691 and 0.440, respectively). This may indicate a protective prognostic effect of GRAF expression in AML cases for DFS and OS. These results might indicate that the determination of the GRAF expression level may contribute to a more detailed risk stratification of AML patients.
Regarding the CML cases, no significant differences were found between GRAF expression in CML cases and control group (P>0.05). This compares favourably with other studies dealing with GRAF expression in CML patients, as they found no difference in the GRAF transcript amount between CML patients and controls (P>0.05) [11].
No significant correlations were found in CML groups between GRAF expression versus splenomegaly and hepatomegaly (P>0.05). As regards the laboratory findings, no significant correlations were found between GRAF expression in CML patients and TLC, Hb (P>0.05). However, platelets, basophilia and higher peripheral blasts showed significant negative correlation with GRAF expression (P=0.031, <0.001 and 0.002, respectively).In the present study, there was no significant difference in GRAF expression level between CML patients in CP and controls (P>0.05). This is compatible with a previous study that reported no significant difference in GRAF transcript amount between CML patients at CP and controls (P>0.05) [11].
However, there was significant difference in GRAF expression level between control versus blastic phase (P=0.004) with high GRAF expression (mean 12.226) in control, and low GRAF expression (mean 0.281) in blastic phase. Moreover, there was significant difference in GRAF expression level between control versus accelerated phase (P=0.005), with high GRAF expression (mean 12.226) in control, and low GRAF expression (mean 0.254) in accelerated phase. There was significant difference in GRAF expression level between control versus blastic plus accelerated phase (P<0.0001), with high GRAF expression (mean 12.226) in control, and low GRAF expression in accelerated and blastic phase. The mechanisms responsible for the disease progression of CML remained poorly understood. Some studies have suggested that several alterations promote this progress, including differentiation arrest caused by the suppression of translation of the transcription factor, CCAAT/enhancer-binding protein (CEBPa) induced by the BCR-ABL oncoprotein in CML cell, increasing genomic instability in CML cell resulting from the reduced capability of genome surveillance system, telomere shortening and loss of tumor suppressor gene such as P53, retinoblastoma 1, CDKN2A, DAPK1 and others [19].
Interestingly, Qian and colleagues found that GRAF transcript was further downregulated during CML progression. Moreover, p210 Bcr-Abl, containing a centrally located Rho-specific guanine nucleotide exchange factors (RhoGEF) domain, affects the actin cytoskeleton assembly and thereby the cellular adhesion and migration by RhoA signalling pathway.
In this study, we demonstrated a significant difference in GRAF expression between those who were chemotherapy responders and those who were resistant (P=0.004), with high GRAF expression in chemotherapy responders (mean 7.161), and low Graf expression in chemotherapy resistance (mean 1.633).
As regards the performance characteristics to identify the optimal GRAF expression level for the prediction of resistance in CML patient, ROC curve analysis showed that the best cut-off value for GRAF was 3.406, at this point the sensitivity was 80% and a specificity was 67%. The area under the curve was 0.824 and P<0.003. Univariate and multivariable analyses were performed to identify potential prognostic factors associated with the development of blastic and accelerated phases in CML phase. Univariate analysis showed that increased TLC, basophilia and decreased GRAF expression are associated with progression of CML to accelerated or blastic phases. Multivariate analysis showed that GRAF expression is the only independent prognostic factor to CML progression (P=0.006), and high GRAF expression is a protective factor against CML progression. Notably, increased TLC and basophilia were associated with the probability of development of accelerated and blastic phases by univariate analysis, but not by multivariate analysis.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References
留言 (0)