Relationship of oxidative stress in the resistance to imatinib in Tunisian patients with chronic myeloid leukemia: A retrospective study

Abstract Background This work aimed to evaluate oxidative stress in chronic myeloid leukemia (CML) patients treated with tunisian (IM) vs controls and in CML patients with resistance to IM vs patients without resistance to IM. Methods The study included 40 CML patients and 34 controls. Of 40 patients with CML, 26 patients were developed in resistance to IM. The oxidant/antioxidant markers were evaluated by spectrophotometric methods for all used samples. Results For CML patients, increased malondialdehyde (MDA) and advanced oxidation protein products (AOPP) levels were found compared to controls (P < .001; P = .01). Higher catalase (CAT) activity (P = .048) and lower superoxide dismutase (SOD) and glutathione peroxidase (GPx) activities, reduced Glutathione (GSH) and vitamin C levels were found in CML patients (P < .001). The comparison between the resistant vs no‐resistant CML patients revealed higher MDA level (P = .02) and CAT and SOD activities in IM‐resistant patients (P = .04, P = .03). GPx activity was reduced (P = .04). Furthermore, increased mean ratio of MDA/GSH, MDA/GPx, and SOD/(GPx + CAT) was found in IM‐resistant patients as compared with no‐resistant (P = .01, P = .01, P = .035). The mean ratio of GPx/GSH in the IM‐resistant CML patients was lower than in IM no‐resistant one (P = .039). For IM‐resistant patients, we found negative correlation between MDA level and the ratio SOD/(CAT + GPx) (r = −0.46, P = .002); and positive correlation between SOD and (CAT + GPx) activities (r = 0.38, P = .06) and between GSH level and GPx activity (r = 0.53, P = .01). Conclusions Our results have shown a highly disturbed oxidative profile in IM‐resistant CML patients as compared to no‐resistant. The H2O2 has a key role in the resistance to IM treatment.


| INTRODUC TI ON
Chronic myeloid leukemia (CML) is a malignant myeloproliferative disorder caused by the BCR-ABL dysfunctional protein. 1,2 The development of inhibitors specifically of the BCR-ABL tyrosine kinase activity as a therapeutic agent has revolutionized CML treatment. [3][4][5] Imatinib mesylate (IM) is the first generation of tyrosine kinase inhibitors (TKIs) that inhibit the activity of the BCR-ABL tyrosine kinase by blocking the ATP-binding site and consequently inducts apoptosis in the CML cells. [6][7][8] Notwithstanding the excellent results obtained with IM for the therapy of CML, a minority of CML patients acquired resistance to IM, which has become an emerging problem resolved only in part by second and third generations of TKIs. 9,10 Various mechanisms of IM resistance were identified, including overexpression of the BCR-ABL gene, mutations in the BCR-ABL kinase domain and genetic variations and/or altered expression of IM gene transporters. [11][12][13][14] Among them, the reactivation of BCR-ABL protein by mutations in the kinase domain, such as T315I mutation, was highly documented as one of the most prevalent mechanisms leading to IM resistance. 14 H Zhang et al demonstrated that the oxidation of redox-sensitive cysteine residues of BCR-ABL protein by reactive oxygen species (ROS) can lead to a change in the three-dimensional structure of the ABL kinase domain of this oncoprotein that is the domain of IM. 3 The precise mechanisms underlying the resistance to IM therapy remain largely elusive and controversial. This study opens new perspectives for future works to solve the relapse issue of CML patients treated with IM.
The impact of oxidative stress on CML improvement is not clear.

Previous investigations have shown that the transformation of BCR-
ABL oncogene can promote the generation of ROS and redox imbalance. [15][16][17] However, others have incriminated the overproduction of ROS in the CML cells and exogenous ROS in the development of CML as well as in the resistance to IM. 8,9,[18][19][20][21][22][23][24][25] Therefore, a systemic analysis of the markers of oxidative stress in plasma of Tunisian patients with CML has been undertaken. The present study was conducted to evaluate the markers of oxidative stress: (a) in CML patients treated with TKI and healthy subjects in order to understand the association of ROS in CML, and (b) in CML patients IM-resistants and IM no-resistants to explain the mechanism of IM resistance.

| Ethics statement
The experimental protocol was established in accordance with the guidelines of the Declaration of Helsinki, and informed consent was obtained from patients.

| Patients
Forty CML patients ( Patient's demographic and biological data (complete blood count and spleen volume), blood disorders, and therapy failure were obtained from hospital records.
The BCR-ABL gene mutation testing was performed exclusively on patients who showed IM resistance.
Thirty-four healthy subjects (25 men and 9 women) were collected from the Regional Blood Transfusion Center, and those with any systemic or chronic disease were excluded from this study.

| Plasma samples
Blood was collected by venepuncture from CML patients and controls in tubes comprising ethylenediaminetetraacetic acid (EDTA).
The plasma layer was separated by centrifugation at 4042 g for 10 minutes. Samples were immediately frozen and stored at −80°C in a small aliquot until analysis.

| Protein concentration
The protein content was assessed by the Bradford assay 27 using bovine serum albumin (BSA) as a standard. The protocol consisted in adding 5 µL of plasma (1:100) with 295 µL of Bradford reagent. After gentle vortex, the samples were measured at 595 nm. Protein concentration was expressed as mg/mL.

| Malondialdehyde level
The lipid peroxidation in the plasma was evaluated by malondialdehyde (MDA) level, using thiobarbituric acid reactive species (TBARS) assay. 28 The plasma was heated with the reagent (containing 0.8% thiobarbituric acid, 15% trichloroacetic acid, and 25% hydrochloric acid) during 15 minutes at 95°C. After cooling, the sample was centrifuged and measured spectrophotometrically at 532 nm. Plasmatic MDA level was expressed as n moles MDA/mg of total protein.

| Advanced oxidation protein products level
The advanced oxidation protein products (AOPP) level was measured

| Ascorbic acid level
The ascorbic acid (vitamin C) level was determined based on the method of Jacota and Dani. 30 To 100 µL of plasma, 400 µL of 10% trichloroacetic acid was added. After cooling, the sample was centrifuged at 4000 rpm for 5 minutes. The extract was diluted to 1:2 and then added to 200 µL of diluted Folin reagent. After 10 minutes, the absorption maximum of the blue colored complex developed by the interaction of ascorbic acid with Folin reagent was read at a wavelength of 769 nm. The results were expressed as ng of ascorbic acid/ mg of total protein.

| Reduced glutathione level
The glutathione (GSH) level was determined by the method of After deproteinizing plasma by 4% sulfosalicylic acid (v/v), the extract was added to 160 µL of PBS (0.1 mol/L) and 64 µL of Ellman's reagent. The sample was rapidly covered with aluminum foil and then incubated for 15 minutes. This latter was measurable spectrophotometrically at a wavelength of 412 nm. The plasmatic GSH level was expressed as n moles GSH/mg of total protein.

| Glutathione peroxidase activity
The activity of glutathione peroxidase (GPx) was performed by the The reaction was initiated by adding 500 µL of hydrogen peroxide (H 2 O 2 ) (5 mmol/L). One min later, this reaction was stopped by 1 mL of 5% trichloroacetic acid. After centrifugation, the extract was incubated with 1 mL of PBS (0.1 mol/L) and 500 µL of DTNB (10 mmol/L) in the dark for 5 minutes. The optical density was determined at 412 nm. GPx activity was expressed as n moles of disappeared GSH/ min/mg of total proteins.

| Catalase activity
The activity of Catalase (CAT) was quantified by the procedure of Aebi based on the decomposition rate of H 2 O 2 . 33 This reaction was expressed as U/mg of protein by using wavelength 240 nm with The protocol consisted in disposing 285 µL of PBS (0.1 mol/L) in two spectrophotometer cuvettes and adding 15 µL of H 2 O 2 except to an only one cuvette. After measurement at 240 nm, 5 µL of the plasma was added in the first and second cuvette. The optical density was also read immediately at 240 nm.

| Superoxide dismutase activity
The activity of Superoxide dismutase (SOD) was determined by measuring its capacity to inhibit the photochemical reduction of nitroblue tetrazolium (NBT) according to Beauchamp and Fridovich. 34 The experimental protocol consisted in incubating a mixture con- •− to blue colored formazan was followed at 580 nm. One unit of SOD activity is described as the amount of enzyme that inhibits the rate of NBT photoreduction rate by 50% under the specified conditions. The SOD activity was expressed as U/mg of protein.

| Statistical analysis
Statistical differences between groups were performed using the unpaired t test. The correlation study was assessed by the Pearson correlation test. The significance was obtained for a P-value < .05.
All results are expressed as mean ± standard error mean (SEM). The statistical analyses and the figures were performed using GraphPad Prism 6. were also included as a control group in the study. No significant differences were found for these characteristics between controls and CML patient groups (P > .05).

| Oxidation markers
Two markers of oxidative stress were evaluated in plasma of CML patients and controls. The results showed significant increased MDA as well as AOPP levels in plasma of CML patients compared to controls (P < .001 and P = .01 respectively) Figure 1A, B respectively).

| Antioxidant defense
The CAT activity was higher in CML patients' plasma compared to the control group (P = .048; Figure 2B). In contrast, SOD and GPx activities were found to be lower in plasma of CML patients as compared to controls (P < .001; Figure 2A, C respectively). Moreover, decreased levels of GSH and vitamin C were found in patients in comparison to controls (P < .001; Figure 2D, E respectively).

| Oxidation markers
The comparison between the IM no-resistant and IM-resistant patients showed a higher rate of MAD in IM-resistant (P = .02; Figure 1A).
However, no difference was found for the AOPP level ( Figure 1B).

| Antioxidant defense
The comparison by the unpaired t test revealed that CAT and SOD activities were increased in IM-resistant patients with regard to the IM no-resistant patients (P = .04, P = .03 respectively; Figure 2A, B respectively). In contrast, GPx activity was reduced in IM-resistant patients when compared to IM no-resistant one (P = .04; Figure 2C).

No significant differences were observed between IM-resistant and
IM no-resistant patients concerning the GSH and the vitamin C levels (P = .98, P = .32, respectively; Figure 2D, E, respectively).
Our study showed significant increase in the mean ratio of MDA/ GSH, MDA/GPx, SOD/GPx, and SOD/ (GPx + CAT) in IM-resistant patients when compared with no-resistant (P = .01, P = .01, P = .005, P = .035, respectively; Table 2). The mean ratio of GPx/GSH in IMresistant CML patients was lower than that of IM no-resistant one (P = .039; Table 2).

| Correlation studies
The correlation study was conducted between oxidative stress  with Gleevec displayed reduced SOD activity compared to a control group. 44 Several investigators have documented the low SOD activity in many other types of leukemia such as acute myeloid leukemia (AML) and chronic lymphocytic leukemia (CLL) and opted for an inhibition of the enzyme by the high rate of ROS. 45,46 Indeed, it was

| D ISCUSS I ON
shown that significant amounts of (O 2 •− ) were generated by leukocytes isolated from AML patients with reduced SOD activity. 46 The   54 In the CML IM resistance case, the mutation T315I was frequently observed and even the second-generation BCR-ABL inhibitors remain ineffective. 48 Among the proposed approaches is the use of oxida- Further large-scale investigations are mandatory to elucidate the fact that targeting ROS level could be a novel approach to overcome IM resistance.

| CON CLUS IONS
So far, the research we carried out has been an attempt to explore the relationship of oxidative stress in the resistance to IM a key role in the resistance to IM treatment, which could contribute to the development and/or the progression to more severe conditions.

ACK N OWLED G M ENTS
The authors are grateful to Professor Moufida Bouyahia for assistance in writing this article.