Patients diagnosed with CML and treated at the Institute of Hematology and Blood Transfusion, Prague, between 2010 and 2015 were included in the present study for prospective follow-up. In total, 36 patients newly diagnosed with CML in CP were enrolled. All but 2 patients were treated with 400 mg IM/day for at least 12 months. In one patient, this dose was reduced to 300 mg for two months due to side-effects and in one patient, IM was applied in a 600 mg dose from the seventh month of therapy. Thirty-three patients were followed up for at least 24 months and the remaining 3 patients for only 9, 12 and 13 months due to their death. Altogether, 5 patients died during the study, 2 of them due to progression of CML. Peripheral blood samples were taken at diagnosis and regular trimonthly monitoring for BCR-ABL1 was conducted during IM therapy. Blood samples of age- and gender-matched healthy controls were also collected. The characteristics of the study cohort are summarized in Table I. The response to IM treatment (optimal, warning or failure) was evaluated according to the European LeukemiaNet (ELN) recommendations based on the BCR-ABL1 transcript levels (20). The present study was approved by the Institutional Ethics Committee of the Institute of Hematology and Blood Transfusion, Prague and conducted according to the Declaration of Helsinki. All subjects signed the informed consent to participate in the present study.

Total RNA was prepared from peripheral blood leukocytes using TRIzol reagent, and cDNA was synthesized by M-MLV reverse transcriptase (Promega, Madison, WI, USA) using random hexamer primers (21). The level of mRNA expression of BCR-ABL1 and genes encoding centrosomal proteins (called centrosomal genes in the present study) were analyzed trimonthly since the start of IM treatment by reverse-transcriptase quantitative real-time polymerase chain reaction (RT-qPCR). The β-glucuronidase gene (GUSB) that is used for standardized BCR-ABL1 monitoring in international scale (IS) units served as a reference gene also for expression analysis of centrosomal genes. BCR-ABL1 quantification was standardized within the European Treatment and Outcome Study (EUTOS) for the CML project of ELN and data were reported in the IS units. Primers and probes for BCR-ABL1 and GUSB were applied according to the Europe Against Cancer recommendations, and commercial plasmid standards were used to obtain calibration curves (Ipsogen, Marseille, France). The transcripts of centrosomal genes were amplified in the following TaqMan Gene Expression Assays (Applied Biosystems, Foster City, CA, USA): Hs01582073_m1 for AURKA, Hs00234864_m1 for HMMR, Hs00901772_m1 for ESPL1 and Hs00153444_m1 for PLK1. The qPCR was performed on Rotor-Gene 3000A (Qiagen, Hilden, Germany). Briefly, 10 µl of the PCR mixture contained 5 µl of TaqMan Universal Master Mix II, without UNG (Applied Biosystems), 0.5 µl of TaqMan Gene Expression Assay, and 2 µl of 3-fold diluted cDNA. After initial incubation at 50°C for 2 min and polymerase activation at 95°C for 10 min, each of the 40 cycles consisted of denaturation at 94°C for 15 sec, and primer annealing and elongation at 60°C for 60 sec. After amplification, analyses were performed using the Rotor-Gene software. The relative quantification of mRNA expression was evaluated by the ΔΔCq method (22). The mean gene expression in the healthy controls was used as a calibrator.

Antigens for the detection of antibodies in an enzyme-linked immunosorbent assay (ELISA) were produced using a glutathione S-transferase (GST) gene fusion system in E. coli. Due to the size limitation of this system, partial sequences of the human genes AURKA (accession no. BC001280; nucleotides 600–1212 of the coding sequence), HMMR (BC108904; 1551–2178), PLK1 (BC014846; 4–670), and ESPL1 (NM_012291; 4–2106) were amplified from plasmids obtained from the Harvard Medical School (Boston, MA, USA) by Phusion High-Fidelity DNA polymerase (Finnzymes, Espoo, Finland) using the following primers containing XhoI restriction sites (underlined sequences): AURKA, 5′-TTAGCTCGAGGTCCATGATGCTACCAGAGTCTAC-3′ (forward) and 5′-TTAGCTCGAGAGACTGTTTGCTAGCTGATTCTTTG-3′ (reverse); PLK1, 5′-TGTACTCGAGAGTGCTGCAGTGACTGCAGGG-3′ (forward) and 5′-TTAGCTCGAGGCTCAGCACCTCGGGAGCTATG-3′ (reverse); ESPL1, 5′-TGTACTCGAGAGGAGCTT CAAAAGAGTCAACTTTGG-3′ (forward) and 5′-TTAGCTCGAGTCTCCGATCCCGCTCGATAC-3′ (reverse); HMMR, 5′-TGTGCTCGAGACCAAGTCAGCACTAAAGG (forward) and 5′-TTAGCTCGAGCTTCCATGATTCTTGACACTCCATAG-3′ (reverse).

After digestion with XhoI (New England Biolabs, Ipswich, MA, USA), the PCR products were cloned into the pGEX-5X-1 vector carrying the GST gene (GE Healthcare, Little Chalfont, UK) and the cloned sequences were verified using the BigDye Terminator v1.1 Cycle Sequencing kit (Applied Biosystems). For the production of recombinant GST-fusion proteins, E. coli BL21-CodonPlus (DE3)-RIPL competent cells (Agilent Technologies, Santa Clara, CA, USA) were transformed with the constructed plasmids and incubated at 37°C in Luria-Bertani medium (Duchefa, Haarlem, The Netherlands) with 100 µg/ml ampicillin and 2% glucose (Sigma-Aldrich, St. Louis, MO, USA). At an OD₆₀₀ of 1.2–1.6, the production of recombinant proteins was induced by adding 0.3 mM isopropyl-β-D-thio-galactoside (IPTG; Duchefa) and the cells were incubated at 30°C for 6 h. The bacterial pellets were resuspended in lysis buffer: 20 mM Tris-HCl (pH 8.0), 200 mM NaCl, 2 mM DTT, 200 mM protease inhibitor phenylmethylsulfonyl fluoride (PMSF; Serva, Heidelberg, Germany), and 0.5% N-lauroylsarcosine (Sigma-Aldrich) in the case of soluble GST-fusion proteins, HMMR, PLK1 and ESPL1. Since the GST-fusion protein AURKA was insoluble in this lysis buffer it was dissolved in a buffer with 2% N-lauroylsarcosine [10 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1 mM EDTA, 7 mM DTT, 200 mM PMSF and 2% N-lauroylsarcosine] (23). The resuspended cells were sonicated three times for 10–15 sec on ice and then lysed three times in the EmulsiFlex-C5 high pressure homogenizer (Avestin, Ottawa, Canada). The lysates were centrifugated (4°C, 30 min, 16,000 × g) and the supernatants were mixed 1:1 with glycerol (Sigma-Aldrich) and stored at −20°C. The production of recombinant proteins was verified by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting.

Polysorb 96-well plastic plates (Nunc, Roskilde, Denmark) were coated overnight at 4°C with glutathione casein (200 ng/well) in 50 mM carbonate buffer (pH 9.6) as previously described (24). Human sera were diluted 1:50 in blocking buffer [0.2% casein (Sigma-Aldrich) and 0.05% Tween-20 (Serva) in phosphate-buffered saline (PBS)] and incubated with the lysate containing the fusion antigen (Ag) or only the GST tag at room temperature for 3 h. Control samples were incubated without lysates (without Ag). The wells of the coated plates were incubated with 200 µl of blocking buffer at 37°C for 1 h and then with 50 µl of the bacterial lysates containing the GST fusion proteins diluted in blocking buffer (1 µg/µl total proteins) at 37°C for 1 h. The plates were washed five times with washing buffer (0.05% Tween-20 in PBS) and the pre-incubated human sera were added and incubated at 37°C for 1 h. After washing, the plates were loaded with donkey anti-human IgG polyclonal antibodies conjugated to horseradish peroxidase (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) diluted 1:7,500 in blocking buffer, and incubated at 37°C for 1 h. The plates were washed and stained with tetramethylbenzidine substrate (Sigma-Aldrich). The absorbance was measured at 450 and 630 nm. The specificity of the antibody reactions was determined by the inhibition of the reactions with the homologous antigens in comparison with the effect of the control GST protein using the formula: (A_{without Ag} - A_{with Ag})/(A_{without Ag} - A_{with GST}).

Comparison of groups was performed using the Mann-Whitney test. The Spearman correlation coefficient was calculated to evaluate the relationship between independent variances. Survival outcomes were compared using the log-rank test. Failure was defined as death, progression to AP or BC, failure to achieve BCR-ABL1 transcript levels <10% at 6 months or <1% at 12 months, or major molecular response (MMR) loss. Results were considered significantly different at P<0.05. Calculations were performed using GraphPad Prism 5.02 software (GraphPad Software, San Diego, CA, USA).

Significantly enhanced expression of all four evaluated centrosomal genes in CML patients was found in total peripheral blood leukocytes at diagnosis in comparison with the level in the healthy controls (Fig. 1A). When patients were stratified based on the response to one-year treatment with IM, patients with optimal response had lower expression of the centrosomal genes at diagnosis than patients with warning or failure, but this difference was significant only for AURKA and HMMR in the patients with treatment failure (Fig. 1A).

As the expression of all studied centrosomal genes has been suggested to be upregulated by BCR-ABL1, the correlation of mRNA expression at diagnosis between centrosomal genes and BCR-ABL1 was also analyzed. However, a statistically significant correlation is shown only for AURKA (Fig. 1B).

During IM treatment, peripheral blood samples were regularly taken from patients and the mRNA expression of BCR-ABL1 and centrosomal genes was quantified. We found four patterns of gene expression kinetics (Fig. 2): in pattern 1 (15 patients), the expression of centrosomal genes reached the basal level (defined by a decrease in relative expression below 2 for all genes; i.e. the mRNA levels were not twice as high as those in the controls) after three months of IM treatment and MMR was achieved within 12 months. In pattern 2 (15 patients), the expression of centrosomal genes was also quickly reduced, but MMR was not recorded within 12 months of treatment. Pattern 3, characterized by the delayed decrease in centrosomal gene expression (i.e. the basal level was not achieved within three months after IM start) and by MMR within 12 months, was found in one patient who showed MMR three months after an increase in IM dose to 600 mg/day. In 5 patients with pattern 4, the basal level of transcripts of the centrosomal genes was detected with a delay (5–9 months after IM start) and MMR was not achieved within 12 months. Both patients who died of CML showed this expression pattern.

To analyze the association of the expression of centrosomal genes with the response to IM treatment, patients were stratified into a low-risk group with a decrease in centrosomal gene expression at the first follow-up after IM start (patterns 1 and 2, 30 patients) and a high-risk group with a delayed decrease (patterns 3 and 4, 6 patients). For comparison, early molecular response (EMR) at three months was evaluated and patients were divided into a low-risk group (BCR-ABL1 ≤10%, 19 patients) and a high-risk group (BCR-ABL1 >10%, 17 patients). While EMR predicted better cumulative incidence of MMR and failure-free survival (FFS), the delayed reduction in centrosomal gene expression in the course of IM treatment was significantly associated with worse overall survival (OS) (Fig. 3).

In enrollment sera of 29 patients, the antibodies against AURKA, HMMR, PLK1 and ESPL1 antigens were determined by ELISA using GST-fusion recombinant proteins produced in E. coli. Due to nonspecific reactions, we did not quantify antibodies by sera titration, but the sera were incubated with antigen-carrying fusion proteins or control GST and the relative inhibition of specific reactions was calculated. For ESPL1 and PLK1, a statistically significant increase was found in comparison with sera from blood donors (Fig. 4A). After stratification of sera into low and high antibody groups based on median production, no significant differences were found for individual antigens (data not shown), but when the overall antibody production against all four antigens was evaluated (means were calculated from the values of all four antigens), significantly higher cumulative incidence of MMR and FFS were shown for patients with high production of antibodies (Fig. 4B).