Leukemia Research 39 (2015) 236–241
Contents lists available at ScienceDirect
Leukemia Research journal homepage: www.elsevier.com/locate/leukres
PDGFB hypomethylation is a favourable prognostic biomarker in primary myelofibrosis Claudia Augello a,1 , Umberto Gianelli a,b,1 , Rossella Falcone c , Silvia Tabano a , Federica Savi d , Eleonora Bonaparte c , Michele Ciboddo c , Leda Paganini a,c , Antonina Parafioriti e , Dario Ricca c , Silvia Lonati f,g , Daniele Cattaneo f,g , Nicola Stefano Fracchiolla f,g , Alessandra Iurlo h , Agostino Cortelezzi f,g , Silvano Bosari a,c , Monica Miozzo a,c,∗ , Silvia Maria Sirchia i a
Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milano, Italy Hematopathology Section, Division of Pathology, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico, Milano, Italy Division of Pathology, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico Milano, Italy d Division of Pathology, Ospedale San Paolo, Milano, Italy e Pathology Department, Istituto Ortopedico G. Pini, Milano, Italy f Hematology and Transplantation Unit, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico, Policlinico, Milano, Italy g Department of Clinical Sciences and Community Health, Università degli Studi di Milano, Milano, Italy h Oncohematology Unit of the Elderly, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico Milano, Italy i Department of Health Sciences, Università degli Studi di Milano, Milano, Italy b c
a r t i c l e
i n f o
Article history: Received 11 August 2014 Received in revised form 13 November 2014 Accepted 21 November 2014 Available online 2 December 2014 Keywords: Primary myelofibrosis DNA methylation PDGFB FGF2 LINE-1
a b s t r a c t Primary myelofibrosis (PMF) is a myeloproliferative neoplasm characterised by the clonal proliferation of the haematopoietic precursors together with the progressive development of bone marrow fibrosis. This stromal alteration is an important clinical issue and specific prognostic markers are not currently available. In bone marrow biopsies from 58 PMF patients, we explored the methylation pattern of genes encoding cytokines involved in the stromal reaction, namely platelet-derived growth factor-beta (PDGFB), transforming growth factor-beta (TGFB) and basic fibroblast growth factor (FGF2). We also evaluated the methylation profile of the Long Interspersed Nucleotide Element 1 (LINE-1). PDGFB, FGF2 and LINE-1, but not TGFB, were significantly differently methylated in PMF compared to controls. Significantly, PDGFB hypomethylation (75th percentile), FGF2 (0–21%: 60%: >75th percentile). The PFS curves were calculated according to the Kaplan–Meier method and compared by the log-rank test. p < 0.05 was considered to be statistically significant. All analyses were performed using the statistical software GraphPad Prism (version 5.0).
3. Results 3.1. PDGFB methylation levels are heterogeneous and dynamic in PMF The methylation status of the PDGFB, FGF2 and TGFB promoters was investigated in BMBs from 58 PMF patients. The percentage methylation of PDGFB and FGF2 in the PMF BMB samples ranged from unmethylated to hypermethylated (PDGFB range: 3–95%, mean: 39%; FGF2 range: 1–96%, mean: 41%), whereas in the controls the methylation values for both genes were clustered in a more restricted interval (PDGFB range: 28–46%, mean: 35%, cases vs. controls: p value is not significant; FGF2 range: 16–43%, mean: 28%; cases vs. controls: p = 0.02) (Fig. 1A, Supplementary Table 1). The TGFB promoter was unmethylated (1–5%) in both cases and controls and, therefore, was not considered for further analyses. Analysis of PDGFB and FGF2 methylation levels according to PMF grade showed that the methylation levels of both PDGFB and FGF2 were significantly higher in PMF MF0 cases compared to the controls (p = 0.005 and p = 0.007 for PDGFB and FGF2, respectively; Fig. 1B and C). MF1 and MF2 cases showed heterogeneous methylation of both genes, including a subgroup of each showing hypomethylation. For PDGFB, the difference in methylation between the MF0 cases and the MF1 and MF2 cases were all significant (p = 0.002 and p = 0.003, respectively; Fig. 1B, Supplementary Table 1). The dynamicity of the methylation patterns of PDGFB and FGF2 during disease progression was also observed by the analysis of
seven PMF cases in which sequential biopsies have been taken at different stages of the disease in the course of follow-up (Fig. 1D, Supplementary Table 1). 3.2. LINE-1 methylation profile differs in PMF patients and controls The methylation levels of LINE-1 in the PMF BMBs ranged from 37% to 99% (mean: 70%; Fig. 1A, Supplementary Table 1) and showed a pronounced heterogeneous distribution of values in all PMF grades, as shown in Fig. 1A and Supplementary Fig. S1. The degree of LINE-1 methylation in the PMF cases was significantly different from that of the controls (range: 52–64%, mean: 57%; p = 0.007). Supplementary Fig. S1 related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.leukres.2014.11.012. Interestingly, the methylation levels of LINE-1 was not significantly different among increasing grades of marrow fibrosis and there was not a correlation between the methylation level of PDGFB and FGF2 and that of LINE-1 (data not shown). 3.3. JAK2 and CALR mutational status do not correlate with PDGFB and FGF2 methylation levels The JAK2 and CALR genotyping results are reported in Supplementary Table 1. IDH1 and IDH2 mutations were not detected in any of the patients. There was no association between JAK2 and CALR mutational status and methylation results of PDGFB and FGF2 (data no shown). 3.4. PDGFB and FGF2 immunohistochemical analysis do not correlate with their methylation status First, we examined the distribution of PDGFB protein expression in normal and in neoplastic BMB according to cell lineages. In normal BMB, we found cytoplasmic PDGFB immunoreactivity only in megakaryocytes; PDGFB protein was not expressed in other BM cells. When we evaluated PDGFB expression in neoplastic BM, PDGFB immunoreactivity was detected as cytoplasmic granular positivity in all megakaryocytes and in a variable percentage (ranging from 3% to 40%) of other bone marrow cells (myeloid precursors, histiocytes, macrophages and endothelial cells) (Supplementary Fig. S2). In particular, 70% of analysed PMF BMBs showed PDGFB protein expression in a small percentage (less than 10%) of bone marrow cells, while the remaining cases expressed PDGFB protein in a higher percentage (ranging from 10% to 40%) of cells. Supplementary Fig. S2 related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.leukres.2014.11.012. In BMB control, FGF2 protein expression was not detectable in megakaryocytes, it was observed in cytoplasm of different cell types (stromal elements, macrophages, erythroblasts and
C. Augello et al. / Leukemia Research 39 (2015) 236–241
239
Fig. 1. PDGFB, FGF2 and LINE-1 methylation levels in PMF. (A) Distribution of PDGFB, FGF2 and LINE-1 methylation values in PMF cases and controls. The horizontal segments indicate the percentiles at 75◦ , 50◦ and 25◦ (from the top to the bottom), the vertical segments indicate the interval values. (B–C) Dot-plot of the percentage of PDGFB (B) and FGF2 (C) promoters methylation in BMBs from PMF cases (stages MF0-3) and controls. Bars represent mean ± SEM. The dashed lines represent the range of control’s methylation values. (D) PDGFB and FGF2 promoter methylation percentages in BMB samples from seven PMF patients during disease progression. ***p < 0.0005, **p < 0.005, * p < 0.05 (unpaired t test).
endothelial cells). We also evaluated the FGF2 immunoreactivity in normal samples and we observed that the median value was 24.2 ± 12.6 (SD) = 36.8. Based on FGF2 median value of controls, in PMF we distinguished low FGF2 expressors [≤36.8] and high FGF2 expressors [>36.8]. In particular, 36% showed high FGF2 expression. Moreover, 31% of PMF cases expressed the FGF2 protein in megakaryocytes nucleus. The immunoreactivity of PDGFB and FGF2 is not associated with the fibrosis grade. No evidence of a correlation was found between the immunoreactivity of PDGFB and FGF2 and their methylation status in PMF cases (data no shown). Finally, there is no association between marrow cellularity and megakaryocyte count, and PDGFB and FGF2 methylation. 3.5. PDGFB hypomethylated correlate with PMF prognosis We explored whether there was an association between the clinical parameters of disease progression (fibrosis, IPSS and DIPSS) and PDGFB and FGF2 methylation level percentiles. The results showed that PDGFB hypomethylation (
Contents lists available at ScienceDirect
Leukemia Research journal homepage: www.elsevier.com/locate/leukres
PDGFB hypomethylation is a favourable prognostic biomarker in primary myelofibrosis Claudia Augello a,1 , Umberto Gianelli a,b,1 , Rossella Falcone c , Silvia Tabano a , Federica Savi d , Eleonora Bonaparte c , Michele Ciboddo c , Leda Paganini a,c , Antonina Parafioriti e , Dario Ricca c , Silvia Lonati f,g , Daniele Cattaneo f,g , Nicola Stefano Fracchiolla f,g , Alessandra Iurlo h , Agostino Cortelezzi f,g , Silvano Bosari a,c , Monica Miozzo a,c,∗ , Silvia Maria Sirchia i a
Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milano, Italy Hematopathology Section, Division of Pathology, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico, Milano, Italy Division of Pathology, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico Milano, Italy d Division of Pathology, Ospedale San Paolo, Milano, Italy e Pathology Department, Istituto Ortopedico G. Pini, Milano, Italy f Hematology and Transplantation Unit, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico, Policlinico, Milano, Italy g Department of Clinical Sciences and Community Health, Università degli Studi di Milano, Milano, Italy h Oncohematology Unit of the Elderly, IRCCS Fondazione Ca’ Granda Ospedale Maggiore Policlinico Milano, Italy i Department of Health Sciences, Università degli Studi di Milano, Milano, Italy b c
a r t i c l e
i n f o
Article history: Received 11 August 2014 Received in revised form 13 November 2014 Accepted 21 November 2014 Available online 2 December 2014 Keywords: Primary myelofibrosis DNA methylation PDGFB FGF2 LINE-1
a b s t r a c t Primary myelofibrosis (PMF) is a myeloproliferative neoplasm characterised by the clonal proliferation of the haematopoietic precursors together with the progressive development of bone marrow fibrosis. This stromal alteration is an important clinical issue and specific prognostic markers are not currently available. In bone marrow biopsies from 58 PMF patients, we explored the methylation pattern of genes encoding cytokines involved in the stromal reaction, namely platelet-derived growth factor-beta (PDGFB), transforming growth factor-beta (TGFB) and basic fibroblast growth factor (FGF2). We also evaluated the methylation profile of the Long Interspersed Nucleotide Element 1 (LINE-1). PDGFB, FGF2 and LINE-1, but not TGFB, were significantly differently methylated in PMF compared to controls. Significantly, PDGFB hypomethylation (75th percentile), FGF2 (0–21%: 60%: >75th percentile). The PFS curves were calculated according to the Kaplan–Meier method and compared by the log-rank test. p < 0.05 was considered to be statistically significant. All analyses were performed using the statistical software GraphPad Prism (version 5.0).
3. Results 3.1. PDGFB methylation levels are heterogeneous and dynamic in PMF The methylation status of the PDGFB, FGF2 and TGFB promoters was investigated in BMBs from 58 PMF patients. The percentage methylation of PDGFB and FGF2 in the PMF BMB samples ranged from unmethylated to hypermethylated (PDGFB range: 3–95%, mean: 39%; FGF2 range: 1–96%, mean: 41%), whereas in the controls the methylation values for both genes were clustered in a more restricted interval (PDGFB range: 28–46%, mean: 35%, cases vs. controls: p value is not significant; FGF2 range: 16–43%, mean: 28%; cases vs. controls: p = 0.02) (Fig. 1A, Supplementary Table 1). The TGFB promoter was unmethylated (1–5%) in both cases and controls and, therefore, was not considered for further analyses. Analysis of PDGFB and FGF2 methylation levels according to PMF grade showed that the methylation levels of both PDGFB and FGF2 were significantly higher in PMF MF0 cases compared to the controls (p = 0.005 and p = 0.007 for PDGFB and FGF2, respectively; Fig. 1B and C). MF1 and MF2 cases showed heterogeneous methylation of both genes, including a subgroup of each showing hypomethylation. For PDGFB, the difference in methylation between the MF0 cases and the MF1 and MF2 cases were all significant (p = 0.002 and p = 0.003, respectively; Fig. 1B, Supplementary Table 1). The dynamicity of the methylation patterns of PDGFB and FGF2 during disease progression was also observed by the analysis of
seven PMF cases in which sequential biopsies have been taken at different stages of the disease in the course of follow-up (Fig. 1D, Supplementary Table 1). 3.2. LINE-1 methylation profile differs in PMF patients and controls The methylation levels of LINE-1 in the PMF BMBs ranged from 37% to 99% (mean: 70%; Fig. 1A, Supplementary Table 1) and showed a pronounced heterogeneous distribution of values in all PMF grades, as shown in Fig. 1A and Supplementary Fig. S1. The degree of LINE-1 methylation in the PMF cases was significantly different from that of the controls (range: 52–64%, mean: 57%; p = 0.007). Supplementary Fig. S1 related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.leukres.2014.11.012. Interestingly, the methylation levels of LINE-1 was not significantly different among increasing grades of marrow fibrosis and there was not a correlation between the methylation level of PDGFB and FGF2 and that of LINE-1 (data not shown). 3.3. JAK2 and CALR mutational status do not correlate with PDGFB and FGF2 methylation levels The JAK2 and CALR genotyping results are reported in Supplementary Table 1. IDH1 and IDH2 mutations were not detected in any of the patients. There was no association between JAK2 and CALR mutational status and methylation results of PDGFB and FGF2 (data no shown). 3.4. PDGFB and FGF2 immunohistochemical analysis do not correlate with their methylation status First, we examined the distribution of PDGFB protein expression in normal and in neoplastic BMB according to cell lineages. In normal BMB, we found cytoplasmic PDGFB immunoreactivity only in megakaryocytes; PDGFB protein was not expressed in other BM cells. When we evaluated PDGFB expression in neoplastic BM, PDGFB immunoreactivity was detected as cytoplasmic granular positivity in all megakaryocytes and in a variable percentage (ranging from 3% to 40%) of other bone marrow cells (myeloid precursors, histiocytes, macrophages and endothelial cells) (Supplementary Fig. S2). In particular, 70% of analysed PMF BMBs showed PDGFB protein expression in a small percentage (less than 10%) of bone marrow cells, while the remaining cases expressed PDGFB protein in a higher percentage (ranging from 10% to 40%) of cells. Supplementary Fig. S2 related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.leukres.2014.11.012. In BMB control, FGF2 protein expression was not detectable in megakaryocytes, it was observed in cytoplasm of different cell types (stromal elements, macrophages, erythroblasts and
C. Augello et al. / Leukemia Research 39 (2015) 236–241
239
Fig. 1. PDGFB, FGF2 and LINE-1 methylation levels in PMF. (A) Distribution of PDGFB, FGF2 and LINE-1 methylation values in PMF cases and controls. The horizontal segments indicate the percentiles at 75◦ , 50◦ and 25◦ (from the top to the bottom), the vertical segments indicate the interval values. (B–C) Dot-plot of the percentage of PDGFB (B) and FGF2 (C) promoters methylation in BMBs from PMF cases (stages MF0-3) and controls. Bars represent mean ± SEM. The dashed lines represent the range of control’s methylation values. (D) PDGFB and FGF2 promoter methylation percentages in BMB samples from seven PMF patients during disease progression. ***p < 0.0005, **p < 0.005, * p < 0.05 (unpaired t test).
endothelial cells). We also evaluated the FGF2 immunoreactivity in normal samples and we observed that the median value was 24.2 ± 12.6 (SD) = 36.8. Based on FGF2 median value of controls, in PMF we distinguished low FGF2 expressors [≤36.8] and high FGF2 expressors [>36.8]. In particular, 36% showed high FGF2 expression. Moreover, 31% of PMF cases expressed the FGF2 protein in megakaryocytes nucleus. The immunoreactivity of PDGFB and FGF2 is not associated with the fibrosis grade. No evidence of a correlation was found between the immunoreactivity of PDGFB and FGF2 and their methylation status in PMF cases (data no shown). Finally, there is no association between marrow cellularity and megakaryocyte count, and PDGFB and FGF2 methylation. 3.5. PDGFB hypomethylated correlate with PMF prognosis We explored whether there was an association between the clinical parameters of disease progression (fibrosis, IPSS and DIPSS) and PDGFB and FGF2 methylation level percentiles. The results showed that PDGFB hypomethylation (
