Epigenetic regulation of cathepsin L expression in chronic myeloid leukaemia

Abstract

The expression and significance of cathepsin L (CTSL) has been extensively studied in solid tumours. However no such information in chronic myeloid leukaemia (CML) was available. We investigated the activity and expression of this protease in peripheral blood mononuclear cells (PBMCs) of 47 adult CML patients. Thirty adults suffering from systemic diseases and 50 healthy volunteers served as controls. The mRNA levels of CTSL, its specific endogenous inhibitor cystatin C and transcriptional up-regulator vascular endothelial growth factor (VEGF) were quantitated by real-time qPCR. CTSL protease activity and its mRNA expression were significantly higher in CML chronic phase (CP) patients compared to CML accelerated phase/blast crisis (AP/BC) patients and controls (P≤ 0.001). VEGF whose expression was most pronounced in CP and declined (P≤ 0.001) in the advanced phases of the malignancy exhibited a strong positive correlation with CTSL expression (r= 0.97; P≤ 0.001). Cystatin C expression was significantly lower (P≤ 0.001) in CML and displayed inverse correlation with CTSL (r=-0.713; P≤ 0.001) activity. CTSL promoter was significantly hypomethylated in CML CP compared to CML AP/BC patients as well as controls. K562, a BC CML cell line displayed CTSL activity, expression and methylation status of CTSL promoter that was comparable to CML AP/BC patients. Treatment of these cells or PBMCs isolated from CML AP/BC patients with 5'-aza-cytidine resulted in a dramatic increase in CSTL activity and/or expression thereby demonstrating the role of promoter methylation in the stage specific expression of CTSL in CML. Differential expression of CTSL in CML at various stages of malignancy may prove useful in identification of the high-risk patients thereby facilitating better management of disease.

Fig 1

Activity and expression of CTSL expression in CML patients. (A) Box plot representation of CTSL activity in CML patients. (B) Relative abundance of CTSL mRNA in CML patients and controls. (a) Significantly higher compared to CML AP/BC as well as K562 cells; (b) significantly higher compared to PCs; (c) significantly higher compared to normal healthy controls (NC) (P # 0.001, Mann-Whitney U-test). (C) Correlation between activity and mRNA expression of CTSL in CML patients (Pearson correlation analysis) (D) CTSL expression in various phases of CML. Immuno-blotting of CTSL protein in PBMCs isolated from representative CML CP, CML AP/BC and NC samples was carried out as described in Materials and methods. Simultaneously, a-tubulin was also detected in the same samples to serve as loading control.

Fig 2

Expression of VEGF and cystatin C in CML. (A) Relative abundance of VEGF mRNA in CML patients and controls. (a) Signifi-cantly higher compared to CML AP/BC; (b) significantly higher compared to PCs; (c) significantly higher compared to NC (P≤ 0.001, Mann-Whitney U-test). (B) Correlation of VEGF with CTSL mRNA in the CML cohort. (Pearson correlation analysis) (C) Increase in CTSL activity by exogenous VEGF. PBMCs isolated from CML CP and NCs were seeded (80% confluent) in six-well plates followed by serum starvation for 12 hrs. Then 40 ng/ml of recombinant VEGF 165 was added to these cells in fresh serum free RPMI 1640 containing 0.1% bovine serum albumin. After 48 hrs of VEGF treatment, cells were lysed and CTSL activity in the lysates was measured as described in ‘Materials and methods’. PBMCs treated with PBS were processed identically and served as controls. Values presented are mean ± S.E. from four independent estimations. Results were analysed by Student’s t-test and values significantly different from respective controls have been marked by ‘a’. (D) Relative abundance of cystatin C mRNA in CML patients and controls; (a) significantly lower compared to PCs; (b) significantly lower compared to NC (P≤ 0.001, Mann-Whitney U-test). (E) Correlation of cystatin C mRNA with CTSL activity in the CML cohort (Pearson correlation analysis).

Fig 2

Expression of VEGF and cystatin C in CML. (A) Relative abundance of VEGF mRNA in CML patients and controls. (a) Signifi-cantly higher compared to CML AP/BC; (b) significantly higher compared to PCs; (c) significantly higher compared to NC (P≤ 0.001, Mann-Whitney U-test). (B) Correlation of VEGF with CTSL mRNA in the CML cohort. (Pearson correlation analysis) (C) Increase in CTSL activity by exogenous VEGF. PBMCs isolated from CML CP and NCs were seeded (80% confluent) in six-well plates followed by serum starvation for 12 hrs. Then 40 ng/ml of recombinant VEGF 165 was added to these cells in fresh serum free RPMI 1640 containing 0.1% bovine serum albumin. After 48 hrs of VEGF treatment, cells were lysed and CTSL activity in the lysates was measured as described in ‘Materials and methods’. PBMCs treated with PBS were processed identically and served as controls. Values presented are mean ± S.E. from four independent estimations. Results were analysed by Student’s t-test and values significantly different from respective controls have been marked by ‘a’. (D) Relative abundance of cystatin C mRNA in CML patients and controls; (a) significantly lower compared to PCs; (b) significantly lower compared to NC (P≤ 0.001, Mann-Whitney U-test). (E) Correlation of cystatin C mRNA with CTSL activity in the CML cohort (Pearson correlation analysis).

Fig 3

Bcr-Abl expression and its correlation with VEGF as well as CTSL. Correlation between Bcr-Abl and (A) VEGF or (B) CTSL mRNA expression in CML CP patients. Pearson correlation coefficient (r) indicated a strong positive association of Bcr-Abl with VEGF (A) as well as CTSL mRNA (B) expression in CML CP patients. Correlation between Bcr-Abl and (C) VEGF (D) CTSL mRNA expression in CML AP/BC patients. Pearson correlation coefficient (r) indicated no association of Bcr-Abl with VEGF (C) or CTSL (D) in CML AP/BC patients.

Fig 4

CTSL promoter methylation in leukaemia. Methylation status of the 650 bp CTSL promoter proximal to the transcription initiation site was performed in representative samples (n= 10) from CML CP, CML AP/BC, PCs and NC groups. Similarly, the methylation status of this promoter in K562, a cell line derived from BC CML patient, was assessed. Each CpG dinucleotide is represented by a circle. Open circle represents an unmethylated site whereas methylated sites have been represented by filled circles.

Fig 5

Quantitative analysis of CTSL in 5′-aza-cytidine treated K562 cells. (A) Effect of demethylating agent 5′-aza-cytidine on mRNA levels of CTSL and cystatin C. Total cellular RNA isolated from K562 cells grown in the presence of 5′-aza-cytidine for different time periods was reverse transcribed and subjected to PCR using primers specific for CTSL, cystatin C or 18S RNA. The amplified products were resolved on 1.2% agarose gel and visualized under UV after staining with ethidium bromide. (B) K562 cells were grown in the presence or absence of 5′-aza-cytidine for different time periods. Cell lysates containing equal amount of protein were immuno-blotted for CTSL protein using monoclonal antibodies against it. Immuno-blotting for α-tubulin served as loading control. Lane 1: Vehicle treated K562 cells at 72 hrs; lane 2: vehicle treated K562 cells at 0 hrs; lane 3: 5′-aza-cytidine treated K562 cells at 72 hrs; lane 4: 5′-aza-cytidine treated K562 cells at 48 hrs; lane 5: 5′-aza-cytidine treated K562 cells at 24 hrs; lane 6: 5′-aza-cytidine treated K562 cells at 12 hrs; lane 7: untreated K562 cells. (C) Total RNA isolated from untreated K562 cells or after treatment with 5 μM 5′-aza-cytidine for 72 hrs was reverse transcribed and subjected to real-time PCR using amplimers specific for CTSL mRNA. a – significantly higher compared to untreated K562 cells (P≤ 0.05, Student’s test). (D) Cell lysates were prepared from untreated or 5 μM 5′-aza-cytidine treated K562 cells. CTSL-specific assay was performed with equal amount of total protein. a – significantly higher compared to untreated K562 cells (P≤ 0.05, Student’s test). (E) 5′-aza-cytidine treatment increases CTSL expression in CML AP/BC. A total of 3 × 10⁷ PBMCs isolated from CML AP/BC patients were plated in each well of a six-well dish and treated with 5 μM 5′-aza-cytidine or vehicle control. After 72 hrs the PBMCs were lysed and CTSL in the cell lysate was assessed by Western blotting and compared with its expression in untreated PBMCs isolated from CML CP patients. Immuno-blotting for α-tubulin served as loading control.

Fig 5

Quantitative analysis of CTSL in 5′-aza-cytidine treated K562 cells. (A) Effect of demethylating agent 5′-aza-cytidine on mRNA levels of CTSL and cystatin C. Total cellular RNA isolated from K562 cells grown in the presence of 5′-aza-cytidine for different time periods was reverse transcribed and subjected to PCR using primers specific for CTSL, cystatin C or 18S RNA. The amplified products were resolved on 1.2% agarose gel and visualized under UV after staining with ethidium bromide. (B) K562 cells were grown in the presence or absence of 5′-aza-cytidine for different time periods. Cell lysates containing equal amount of protein were immuno-blotted for CTSL protein using monoclonal antibodies against it. Immuno-blotting for α-tubulin served as loading control. Lane 1: Vehicle treated K562 cells at 72 hrs; lane 2: vehicle treated K562 cells at 0 hrs; lane 3: 5′-aza-cytidine treated K562 cells at 72 hrs; lane 4: 5′-aza-cytidine treated K562 cells at 48 hrs; lane 5: 5′-aza-cytidine treated K562 cells at 24 hrs; lane 6: 5′-aza-cytidine treated K562 cells at 12 hrs; lane 7: untreated K562 cells. (C) Total RNA isolated from untreated K562 cells or after treatment with 5 μM 5′-aza-cytidine for 72 hrs was reverse transcribed and subjected to real-time PCR using amplimers specific for CTSL mRNA. a – significantly higher compared to untreated K562 cells (P≤ 0.05, Student’s test). (D) Cell lysates were prepared from untreated or 5 μM 5′-aza-cytidine treated K562 cells. CTSL-specific assay was performed with equal amount of total protein. a – significantly higher compared to untreated K562 cells (P≤ 0.05, Student’s test). (E) 5′-aza-cytidine treatment increases CTSL expression in CML AP/BC. A total of 3 × 10⁷ PBMCs isolated from CML AP/BC patients were plated in each well of a six-well dish and treated with 5 μM 5′-aza-cytidine or vehicle control. After 72 hrs the PBMCs were lysed and CTSL in the cell lysate was assessed by Western blotting and compared with its expression in untreated PBMCs isolated from CML CP patients. Immuno-blotting for α-tubulin served as loading control.