Molecular Profiling of CD34 Cells in Idiopathic Myelofibrosis Identifies a Set - 《stemcells》干细胞学杂志干细胞之家



免疫细胞治疗专区	欢迎关注干细胞微信公众号

12 3 4 5 6 7 8 9 10 ... 24 下一页

返回列表

查看: 364617\|回复: 238	go Molecular Profiling of CD34 Cells in Idiopathic Myelofibrosis Identifies a Set [复制链接]

江边孤钓

版主

Rank: 7 Rank: 7 Rank: 7

积分: 0
威望: 0
包包: 483

楼主

发表于 2009-3-5 01:00 |只看该作者 |倒序浏览 |打印

作者：Paola Guglielmellia, Roberta Zinib, Costanza Bogania, Simona Salatib, Alessandro Pancrazzia, Elisa Bianchib, Francesco Mannellia, Sergio Ferrarib, Marie-Caroline Le Bousse-Kerdilsc, Alberto Bosia, Giovanni Barosid, Anna Rita Migliaccioe, Rossella Manfredinib, Alessandro M. Vannucchia作者单位：aDepartment of Hematology, Azienda Ospedaliera-Universitaria Careggi, University of Florence, Florence, Italy;bDepartment of Biomedical Sciences, Biological Chemistry Section, University of Modena and Reggio Emilia, Modena, Italy;cINSERM U, University Paris , Institut Andr Lwoff, Villejuif Cedex, Fr & W: V2 W: T9 e3 y- Z

   / H. D0 ^) {* {# I/ m


         9 l0 F7 |. r4 M6 J5 q
         $ ]# x& K! l; N3 ?. @/ V
         " q5 D- s" p/ |& v! `* `





      【摘要】0 l" d3 B/ u+ s4 e( o1 N
   This study was aimed at the characterization of a gene expression signature of the pluripotent hematopoietic CD34  stem cell in idiopathic myelofibrosis (IM), which would eventually provide novel pathogenetic insights and/or diagnostic/prognostic information. Aberrantly regulated genes were revealed by transcriptome comparative microarray analysis of normal and IM CD34  cells; selected genes were also assayed in granulocytes. One-hundred seventy four differentially expressed genes were identified and in part validated by quantitative polymerase chain reaction. Altered gene expression was corroborated by the detection of abnormally high CD9 or CD164, and low CXCR4, membrane protein expression in IM CD34  cells. According to class prediction analysis, a set of eight genes (CD9, GAS2, DLK1, CDH1, WT1, NFE2, HMGA2, and CXCR4) properly recognized IM from normal CD34  cells. These genes were aberrantly regulated also in IM granulocytes that could be reliably differentiated from control polycythemia vera and essential thrombocythemia granulocytes in 100% and 81% of cases, respectively. Abnormal expression of HMGA2 and CXCR4 in IM granulocytes was dependent on the presence and the mutational status of JAK2V617F mutation. The expression levels of both CD9 and DLK1 were associated with the platelet count, whereas higher WT1 expression levels identified IM patients with more active disease, as revealed by elevated CD34  cell count and higher severity score. In conclusion, molecular profiling of IM CD34  cells uncovered a limited number of genes with altered expression that, beyond their putative role in disease pathogenesis, are associated with patients' clinical characteristics and may have potential prognostic application.
      【关键词】 Idiopathic myelofibrosis CD  cells Gene expression profiling WT JAKVF mutation" m, b3 n2 t" Y3 C- l
   INTRODUCTION

Myelofibrosis with myeloid metaplasia, also known as chronic idiopathic myelofibrosis (IM) .
& p8 h/ A$ ~: m, V8 o% E. I4 P" l
With the aid of microarray technology  allowed us to study a cell population representative of the disease's molecular aberrations, obviating difficulties in collecting BM aspirates because of dry tap due to fibrosis; these cells were compared with those obtained from the BM of healthy subjects. Characterization of the transcriptome of IM CD34  cells allowed the identification of 174 genes that were aberrantly regulated and might putatively represent a molecular signature of the disease. Furthermore, a set of eight differentially expressed genes was characterized that, according to class prediction analysis, properly distinguished 100% of IM versus normal CD34  cells. To evaluate whether the abnormal expression of these eight genes was also maintained in the mature progeny of CD34  cells, we measured gene expression levels in the granulocytes and found a 100% discrimination power of IM from normal granulocytes; additionally, an 81% correct prediction was obtained when IM granulocytes were compared with those isolated from patients with PV or ET. Among the eight genes comprising the prediction set, the abnormal expression profile of HMGA2 and CXCR4 in IM granulocytes was associated with the presence of JAK2V617F mutation, whereas that of NFE2, CD9, CDH1, GAS2, WT1, and DLK1 was not. Finally, we found that the expression levels of DLK1 and CD9 were related to the platelet count and that high expression levels of WT1 identified patients with more active disease, as indicated by elevated number of circulating CD34  cells and higher severity score.& C/ u9 n) V9 C- }" f

MATERIALS AND METHODS

Subjects

The diagnosis of IM was made according to the necessary criteria identified by the Italian Consensus Conference criteria (diffuse marrow fibrosis and absence of BCR-ABL rearrangement) and an algorithm based on variable combination of accessory criteria represented by splenomegaly, tear-drop erythrocytes, circulating immature myeloid cells and erythroblasts, and clusters of abnormal Mks in the BM ; the overall grading of "severity" score was 0¨C6. The study had received the approval from the local Ethical Committee, and informed consent was obtained from the subjects involved at the time of sample collection.
: z; Y9 j& H& C3 P
CD34  Cell Enumeration and Purification) _( Z# E$ v0 J$ E6 @

The number of CD34  cells in the PB of IM subjects was determined using 50 µl of EDTA-anticoagulated blood; cells were stained with CD45-fluorescein isothiocyanate (FITC)-/CD34-phycoerythrin (PE)-conjugated monoclonal antibodies (both from BD Biosciences, San Jose, CA, http://www.bdbiosciences.com) and propidium iodide for excluding dead cells. At least 200,000 events were acquired on a FACScan flow cytometer (BD Biosciences) and analyzed by CellQuest software; the percentage of positive cells was calculated according to the guidelines from the International Society of Hematotherapy and Graft Engineering . The absolute number of circulating CD34  cells per liter was calculated by normalizing to the total leukocyte count.
$ Z. P1 P& T6 `1 O- ]7 Z
CD34  cells were purified from 30¨C50 ml of PB collected from IM patients or from 5 ml of BM aspirates obtained in preservative-free heparin from healthy donors. Mononuclear cells were separated over a Ficoll-Paque gradient (Lymphoprep; Nycomed Pharma, Asker, Norway, http://www.nycomed.com) and processed through two sequential steps of immunomagnetic CD34  selection (Miltenyi Biotec, Bergisch Gladbach, Germany, http://www.miltenyibiotec.com) . Purity of the isolated CD34  cell population was evaluated by flow cytometry after labeling with PE-HPCA2 anti-CD34 monoclonal antibody (BD Biosciences). Aliquots of CD34  cells were immediately resuspended in lysis buffer for RNA purification.

RNA Extraction and Microarray Data Analysis
* ]/ ?( D" T) V# }
Total RNA was extracted using Trizol (Invitrogen Ltd, Paisley, U.K., http://www.invitrogen.com). Disposable RNA chips (Agilent RNA 6000 Nano LabChip kit; Agilent Technologies, Waldbrunn, Germany, http://www.home.agilent.com) were used to determine the concentration and purity/integrity of RNA with the Agilent 2100 Bioanalyzer. To obtain enough RNA to perform the Affymetrix analysis, 0.6 µg of RNA from CD34  cells of five different normal donors or IM patients were pooled, obtaining three pools of 3 µg of RNA each. The biotin-labeled target synthesis reactions, as well as the Affymetrix HG-U133A GeneChip arrays hybridization, staining, and scanning, were performed using Affymetrix standard protocols (Affymetrix, Santa Clara, CA, http://www.affymetrix.com). Briefly, biotin-labeled cRNA was purified using RNeasy spin columns (Qiagen Inc., Valencia, CA, http://www1.qiagen.com), and 20 µg was fragmented following the Affymetrix GeneChip protocol. The assessment of cRNA concentration/quality and fragmentation was performed with Agilent RNA chips. Fragmented cRNA was then hybridized to an identical lot of Affymetrix HG-U133A GeneChip arrays for 16 hours. GeneChips were washed and stained using the instrument's standard eukaryotic GE WS2v4 protocol, using antibody-mediated signal amplification. GeneChips were finally scanned using the GCS3000 Affymetrix GeneChip scanner .

The GeneChip Operating Software (GCOS) absolute analysis algorithm was used to determine the amount of transcript mRNA (signal) using the GCOS global scaling option. GCOS-generated data were uploaded onto GeneSpring software version 7.2 (Agilent Technologies). Two additional normalization procedures were performed. First, each signal was divided for the 50th percentile of all normalized above-10 signals in that sample and was then divided by the median value in all samples. A "low-level" filtering procedure was performed to reduce noise: genes showing an "absent" call in all conditions, as well as genes showing a normalized intensity between 0.5 and 2, were removed. The unsupervised analysis was performed on this "low-level filtered" gene list using the "condition tree" option and applying the Pearson correlation equation. For supervised analysis, the "low-level filtered" gene list was further processed to select only those genes that showed a fold change more than 2 or less than 0.5 between normal and IM samples. Furthermore, a Welch analysis of variance (ANOVA) test (parametric test, with variances not assumed equal, p value cutoff = .05), using the Benjamini and Hochberg method to control the family-wise error rate, was performed on the list of genes deriving from the twofold change filter analysis.; Q' G' D) P$ S! s
' m0 w7 ^8 N. z
Class prediction analyses were performed using the Support Vector Machine algorithm as implemented in GeneSpring (polynomial dot product  kernel function; diagonal scaling factor: 0).4 s; j1 H. R; ]: ?0 X2 t
' R4 ]3 m6 @2 t4 f' W
Real-Time Quantitative Polymerase Chain Reaction
+ A7 r+ ~1 D# }
cDNA was reverse-transcribed from total RNA (100 ng per sample) obtained from CD34  cells using a High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA, https://www2.appliedbiosystems.com). TaqMan polymerase chain reaction (PCR) was carried out with the TaqMan Universal PCR master mix, using either custom TaqMan low-density arrays or TaqMan gene expression assays (all reagents from Applied Biosystems), by means of an ABI Prism 7900 HT sequence detection system (Applied Biosystems). Assays were performed in quadruplicate. Gene expression profiling was achieved using the comparative cycle threshold (CT) method of relative quantitation using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the housekeeping gene. To normalize data, CT was calculated for each sample using the mean of its CT values subtracted from the mean CT value measured in the entire population of healthy subjects, considered as a calibrator; relative quantitation (RQ) value was expressed as 2¨CCT. Normalized CT values were uploaded onto GeneSpring using the real-time data transformation.; O& n: K8 o7 D3 U% u

For the analysis of gene expression levels in granulocytes, PB cells were separated by differential centrifugation over a Ficoll-Paque gradient, and after removal of contaminating red cells by hypotonic lysis, the cell pellet was resuspended in Trizol for RNA extraction; analysis of cytosmears showed that more than 98% of cells were granulocytes. cDNA was reverse-transcribed with random hexamers and MuLV (murine leukemia virus) reverse transcriptase (Applied Biosystems) and processed for real-time PCR as above, except that RNase P was used as the reference gene.7 c- x1 U5 [- C% [1 |  k! X
, W+ P* s( N, G0 t% U
Flow Cytometry Analysis of CD9, CXCR4 (CD184), and CD164 Expression on Circulating CD34  Cells, u! y8 S# x5 F: D% s; @4 K6 ^

The cell surface expression of CD9, CXCR4 (CD184), and CD164 was analyzed by flow cytometry (fluorescence-activated cell sorting ) on fresh (within 3 hours from drawing) PB samples from normal subjects or patients with IM, PV, or ET. Samples were incubated for 30 minutes at 4¡ãC with FITC-conjugated anti-CD34, PerCyp-conjugated anti-CD45, and PE-conjugated anti-CD9, -CXCR4, or -CD164 antibodies (all from BD Biosciences), followed by red-cell lysis and washing. Appropriate isotope controls were used for each sample; cellular debris was excluded in a side scatter/forward scatter plot. A minimum of 200,000 events were acquired. Results were expressed both as the percentage of CD45 /CD34  cells coexpressing CD9, CXCR4, or CD164 and as the ratio of geometric mean fluorescence intensity (MFI) by dividing the value of specific antibodies with the corresponding isotope control antibody.
6 n3 V/ v( J8 M% k  o7 z
Analysis of JAK2V617F Mutation

The analysis of JAK2 mutation was performed by an allele-specific PCR, starting from 75 ng DNA purified from granulocytes using the QIAmp DNA blood Kit (Qiagen) .4 o/ a. _5 j# n
1 [  r: n6 d) g
Statistical Analysis# J0 P' K2 X! \' b8 ], V( A- U6 V

Comparison between groups was performed by the Mann-Whitney U or Fisher test; associations between clinical characteristics and experimental data (logarithmically transformed) were assessed by Spearman's or Wilcoxon-Mann-Whitney test, as appropriate, using the SPSS (StatSoft, Inc., Tulsa, OK, http://www.statsoft.com) or GraphPad InStat software (GraphPad Software, Inc., San Diego, http://www.graphpad.com) for computation. Logistic regression analysis was performed according to the SPSS software. The chosen level of significance from two-sided tests was p

RESULTS

Patient Characteristics

The hematologic and clinical characteristics of the 88 IM patients included in the study are presented in Table 1; for comparison, data concerning the 35 patients with PV or ET are shown. The median number of circulating CD34  cells in IM patients was 60.0 x 106/l, significantly higher than in PV or ET (p * f) j3 o2 S1 B) A! t

Table 1. Clinical characteristics, at the time of sample collection, of the patients with IM and the other chronic myeloproliferative disorders included in the study6 Z  B5 G& N( R8 P/ E7 s

Gene Expression Profile of CD34  Cells from IM Patients

We screened for abnormally regulated genes in pooled CD34  cells purified from 15 IM subjects and 15 healthy controls (three distinct pools each). Purity of CD34  cells was always more than 98%. All of the microarray analysis data have been deposited in the Gene Expression Omnibus MIAME-compliant public database at http://www.ncbi.nlm.nih.gov/geo (supplemental online Table S1).

Microarray data analysis showed that mRNA complexity was comparable in normal and IM samples, given that the number of sequences called "present" by Affymetrix GCOS absolute analysis algorithm was 9,821 and 10,946 in normal and IM CD34  cells, respectively. An unsupervised clustering analysis, performed on samples that had been managed using a low-level filtered gene approach, paired the transcript profiles of CD34  cells from IM patients or normal subjects, respectively (supplemental online Fig. S1). Using the filtering procedure and the ANOVA analysis described in Materials and Methods, we finally identified 218 probesets, corresponding to 174 well-characterized genes, that showed significant differences in expression levels between the two kinds of cell samples. Functional analysis revealed that several of these 174 genes might be putatively involved in megakaryocytic commitment, BM fibrosis, oncogenesis, or hematopoietic stem cell adhesion and migration; a list of the 77 best-characterized genes under this respect is presented in supplemental online Fig. S2 (see also Discussion).

To validate array data, we designed a TaqMan low-density array containing 47 GAPDH TaqMan gene expression assays (supplemental online Table S2) that had been selected based on either the statistically significant difference in gene expression levels or their putative pathogenetic role in IM. TaqMan assays were carried out in an independent cohort of CD34  cells from eight IM patients and five healthy subjects and allowed to validate 36/47 genes (77%) (Fig. 1).

Figure 1. Clustering analysis of the 36 genes representing the "validation set". Thirty-six genes, from a total of 47 chosen from the 174 abnormally regulated genes in IM CD34  cells, were validated using a quantitative reverse transcription-polymerase chain reaction technology on a new set of eight IM patients and five healthy controls. GeneSymbol is listed on the right side of dendrogram. Clustering has been performed applying the "gene tree" and "condition tree" clustering algorithms provided by GeneSpring and the standard correlation equation. For each gene, the columns marked in dark blue refer to the results originally obtained from microarray analysis of the three pools of IM patients and controls, whereas those marked in turquoise refer to TaqMan data. Gene coloring was based on normalized signals, as shown at the bottom. Abbreviations: CTR, control; IM, idiopathic myelofibrosis.

Differential Expression of CD9, CXCR4 (CD184), and CD164 by FACS Analysis on CD34  Cells from Patients with IM" r  C7 @2 z1 L& Y8 Q% v1 _% f! H

Among the products of the 36 validated genes described above, we selected CD9, CXCR4, and CD164 for cytofluorimetric analysis because these proteins might play some role in the pathogenesis of IM. All of these membrane receptors are possibly involved in hematopoietic stem/progenitor cell adhesion and migration; in particular, CD9 is a tetraspanin , the receptor for SDF-1, intervenes in the mechanisms of CD34  cell homing and migration. According to gene expression analysis, the levels of CD9 and CD164 mRNA were increased in IM CD34  cells, whereas those of CXCR4 were decreased, as compared with controls. The results of the FACS analysis on circulating CD34  cells from IM patients (n = 20), normal healthy subjects (n = 15), and patients with PV or ET (n = 20 each) are shown in Figure 2.. f+ b, O) Y0 F! G7 n% F% O

Figure 2. Fluorescence-activated cell sorting analysis of the expression of CD9, CXCR4 (CD184), and CD164 on CD34  cells from patients with IM, PV, or ET and healthy controls. Panels on the left depict the percentage of electronically gated CD45 /CD34  cells that coexpress CD9, CD184, or CD164, whereas the MFI level is represented in the panels on the right. All of these analyses were performed on PB CD34  cells; data are representative of IM, PV, or ET patients (20 each) and 15 healthy controls. PE-labeled antibodies. Boxes represent the interquartile range that contains 50% of the subjects, the horizontal line in the box marks the median, and bars show the range of values. Data obtained in IM patients were compared with both controls and PV or ET patients for calculation of significance level. *, p 7 u( o, H7 m8 p2 [# X
& _0 v- P5 }- J0 a1 ?3 v- s
The fraction of gated CD45 /CD34  cells coexpressing CD9 in IM patients was significantly higher than in healthy subjects and PV or ET patients (p

On the other hand, the percentage of PB CD45 /CD34  cells coexpressing CXCR4 was significantly lower in IM patients than in controls (p 9 ?! ]' H0 k( p: Q, Z1 |+ V

IM-Derived CD34  Cells Can Be Reliably Distinguished from Normal Ones Based on TaqMan Low-Density Array Using a Set of Eight Genes! V5 I, {/ b4 \. p3 ]  l+ B

With the aim of identifying a smaller set of genes that might successfully distinguish IM from normal CD34  cells, we selected 8 genes among the 36 TaqMan validated genes on the basis of both their abnormal expression levels and their putative pathogenetic role in IM (listed in supplemental online Table S3). The performance of this set of genes was then assessed by employing class prediction analysis with the Support Vector machine using 15 samples (5 controls and 10 IM) as the training set and 27 samples (4 controls and 23 IM) as the test set. The results of the cross-validation of the training set as well as the outcome of test prediction are detailed in supplemental online Table S4; we found no incorrect prediction in the cross-validation of the training set or in the test set (27/27 = 100% validation).5 B3 k7 q# |, c& e5 D

The Eight Genes Comprising the Prediction Set in CD34  Cells Are also Abnormally Expressed in IM Granulocytes and Allow Differentiation of IM from Normal, PV, or ET Granulocytes
$ s* ]0 I7 o% w
We then asked whether abnormal expression of the eight gene markers considered above could also be demonstrated in granulocytes, which would represent a more convenient source for analysis than CD34  cells. Using real-time RT-PCR, we found that CD9, NFE2, GAS2, DLK1, WT1, HMGA2, and CDH1 were all significantly increased in IM granulocytes compared with healthy controls, whereas CXCR4 was reduced (Fig. 3). Abnormally high expression levels of CD9, GAS2, CDH1, and NFE2 were also measured in the granulocytes from patients with PV or ET, whereas the levels of CXCR4, DLK1, WT1, and HMGA2 in these patients did not differ significantly from controls.
! g  w7 f) j2 T8 `. v  Q2 w' @9 s
Figure 3. Expression levels of the eight genes comprising the prediction set in the granulocytes from patients with chronic myeloproliferative disorder and healthy controls. Gene expression levels were measured by real-time reverse transcription-polymerase chain reaction starting from granulocyte RNA and were expressed as RQ log10. Boxes represent the interquartile range that contains 50% of the subjects, the horizontal line in the box marks the median, and bars show the range of values. Data are representative of 25 IM, 30 PV, and 30 ET patients (except for WT1, CD9, DLK1, and HMGA2, in which the number of IM patients examined comprised between 45 and 60). The significance levels of the differences among patient categories are detailed in the table inside. Abbreviations: Ctr, control; ET, essential thrombocythemia; IM, idiopathic myelofibrosis; n.s., not significant; PV, polycythemia vera; RQ, relative quantitation.
" n9 e$ l* U/ Y& t$ d' W4 s* h
This same set of genes was used to perform class prediction analysis on IM and control granulocytes, using 17 samples as a training set and 23 samples as a test set. As detailed in supplemental online Table S5, no incorrect prediction for either IM patients or controls was made in the training or test prediction set. This analysis was also carried out on the granulocytes obtained from patients with the different CMPDs; 30 patients (10 each with IM, PV, or ET) were used as a training set and 55 patients (16 IM, 19 PV, and 20 ET) as a test set. The results of class prediction analysis (shown in supplemental online Table S6) demonstrated 100% correct prediction in the cross-validation of the training set, whereas three incorrect predictions were made in the test set prediction (52/55 = 95% validation of all cases), and concerned three IM patients out of 16 (19%) that were misclassified as PV/ET.

Effect of the Presence of JAK2V617F Mutation on Gene Marker Expression Levels4 Z9 q  O$ I& ^) D$ V

To evaluate whether abnormal expression of any of the eight gene markers correlated with the presence of the JAK2V617F mutation, IM patients were grouped according to their mutational status and the respective levels of gene expression were considered (Fig. 4). We found that abnormal expression of HMGA2 and CXCR4 was associated with the presence of JAK2V617F; in the case of CXCR4, the expression levels were significantly lower in homozygotes than in heterozygotes, suggesting dependence of gene expression on the number of mutated alleles. On the other hand, the expression levels of CD9, DLK1, GAS2, CDH1, NFE2, and WT1 were not significantly different in JAK2V617F-mutated patients as compared with wild-type.

Figure 4. Effects of JAK2V617F mutation on the expression of the genes comprising the prediction set in idiopathic myelofibrosis granulocytes. Gene expression levels were measured by real-time reverse transcription-polymerase chain reaction starting from granulocyte RNA and were expressed as RQ log10. Boxes represent the interquartile range that contains 50% of the subjects, the horizontal line in the box marks the median, and bars show the range of values. Patients were categorized according to the absence (WT) or the presence of JAK2V617F mutation (V617F) and whether the mutation was found in the heterozygous or homozygous form. *, p

Association of CD9 and DLK1 Expression Levels with Platelet Count in IM Patients7 ~9 V/ Q* G$ n8 ~- a5 g: W+ f0 ^# |
4 O9 G2 [6 D) s4 ~0 r$ C$ x
To establish possible correlations of the aberrantly regulated genes comprised within the eight-gene set with hematologic phenotype of IM patients, expression levels in granulocytes were correlated with hemoglobin, white blood cell, and platelet counts. We found that CD9 and DLK1 expression levels were directly (r = .56, p 8 I: E  C$ \& |  o) m# y

Figure 5. Clinical correlates of CD9, DLK1, and WT1 gene expression levels in patients with idiopathic myelofibrosis (IM). Gene expression levels were measured by real-time reverse transcription-polymerase chain reaction starting from granulocyte RNA and were expressed as RQ log10. (A, B): Gene expression levels were plotted against platelet count at the time of blood sampling in 62 and 54 IM patients, respectively, for CD9 (A) and DLK1 (B). In plots (C) and (D), WT1 expression levels were plotted against CD34  cell count (n = 46) and clinical severity score (n = 60), respectively, calculated at the time of blood sampling. Boxes in (D) represent the interquartile range that contains 50% of the subjects, the horizontal line in the box marks the median, and bars show the range of values. Twenty-seven and 33 patients were in the low and high score groups, respectively. Abbreviation: RQ, relative quantitation.! g5 h/ w0 a2 i  C: o

Association of WT1 Expression Levels with Disease Severity

To address whether abnormal gene expression was associated with clinical characteristics of IM patients, we used the Dupriez score (at diagnosis) as a prognostic index and the CD34  cell count and the severity score (both calculated at the time of blood sampling for this study) as indexes of disease severity. In univariate analysis, we found a strong association of increased expression of WT1 with both the CD34  cell count (r = .61, p
4 K2 V  N$ @! b6 u5 O- e& m: g
DISCUSSION
0 `- M$ ^9 B- i) y% Q; B
To investigate pathogenetic mechanisms of IM, and possibly to identify molecular markers associated with the disease and/or with clinical aspects, we have characterized the transcriptome profile of CD34  cells purified from the PB of IM patients. We obtained CD34  cells from the BM of healthy donors as controls, because their very low number in steady-state PB precluded cell enrichment in amounts sufficient for microarray analysis. Although some differences in the global gene profile between these two sources of cells may be expected .1 ~! x$ Z. ~3 ^4 Y, _* p7 b

Gene profiling analysis identified 218 transcripts that correspond to 174 well-characterized genes and show aberrant expression in IM versus normal CD34  cells; these transcripts, based on their known function(s), are possibly involved in some pathogenetic steps of this disorder (supplemental online Fig. S2) and would deserve further investigation. Among those that are involved in the control of the development of Mks, which represent the most obviously involved cell lineage in IM, are CD9, which influences Mk differentiation . Of interest for IM pathogenesis, the expression of some oncogenes and neoplastic marker genes, such as ETS2 (already reported as upregulated in PV), DLK1, MYC, LEPR, WT1, PDZK3, PIM, TEM6, GRB10, TNKS, PLAG1, and TPBG, was upregulated, whereas tumor-suppressor genes AIM-1, DP-1, and BRCA1 were downregulated. Finally, it is worth mentioning the activation of genes that are not normally expressed in the hematopoietic system, such as KLRG1, GLS, BTNA-3 and -2, HFL1, KRT18, and CKB (for a list of pertinent references, please refer to supplemental online Table S7).0 p1 K9 i0 y+ q% s* @

Gene profiling analysis in IM CD34  cells has been reported by Jones et al. , for the discrimination of IM versus control CD34  cells according to class prediction analysis. This novel gene set includes CD9, CDH1, CXCR4, GAS2, NFE2, DLK1, HMGA2, and WT1.
6 G5 g1 x" X* f! |! T# A* j
The possibility that abnormal gene expression found in CD34  cells was also maintained in their mature progeny was addressed by studying PB granulocytes, which would also represent a more accessible source of cells for further analysis. Indeed, the eight-gene set that properly distinguished IM from normal CD34  cells also proved to be aberrantly regulated in the granulocytes of patients with IM and, with some differences (Fig. 3), in PV or ET granulocytes. According to the results of class prediction analysis using this eight-gene set, IM granulocytes could be efficiently distinguished from either control or PV and ET granulocytes in 100% and 81% of cases, respectively. A notable exception was the CDH1 gene, which showed opposite changes from CD34  cells (where it was downregulated) to granulocytes (where expression levels were significantly higher than controls). Reduced expression of E-cadherin, the product of CDH1, influences cell-cell adhesion interactions and may facilitate cancer metastasis .

The three RNA pools prepared from IM patients and used for microarray analysis contained an almost 50% mixture of JAK2V617F-mutated (n = 7) and -nonmutated (n = 8) patients, mirroring the overall incidence of the mutation in this disease; therefore, the list of differentially expressed genes actually reflects the abnormal transcriptome of IM CD34  cells including both JAK2V617F-dependent and -independent genes. Further analyses, using cells prepared from either mutated or wild-type IM patients, might help in identifying differences of gene expression attributable to the mutation per se; meanwhile, based on the results of gene expression in granulocytes, HMGA2 and CXCR4 should be added to the growing list of genes whose expression is affected by the presence of JAK2V617F mutation .9 c" ?' Y3 j- D* }+ y( ]3 h
* D" q" J: N: S' w, u9 j+ \
WT1, the Wilms' tumor 1 gene, deserves particular interest because high WT1 expression levels in IM cells were found to be associated with signs of disease activity, such as the number of CD34  cells in the PB or the severity score (Fig. 5C, 5D). Although WT1 is considered a tumor-suppressor gene, the WT1 wild-type form is overexpressed in a variety of human tumors, including myelodysplasia (MDS) and leukemias .4 i  s/ a% W1 k: n7 o" g; O0 ?

Figure 6. A working model linking the abnormal expression of WT1, DLK1, CXCR4, and CDH1 in the regulation of CD34  cells and fibrosis development in idiopathic myelofibrosis. Abbreviation: HSC, hematopoietic stem cell.2 }  U+ w2 y- ]( j4 Z

Collectively, the results of this study led to the characterization of a limited set of genes specifically associated with both the CD34  cells and mature granulocytes of IM patients. These genes, beyond being potentially implicated in some pathogenetic aspects of the disease and amenable to future studies, have intriguingly been found to associate with some clinical characteristics; in particular, it is anticipated that determination of WT1 expression levels in IM granulocytes might suitably employ WT1 as the first molecular marker of disease activity and, prospectively, prove useful to evaluate response to therapy .( J! x+ x4 |3 V1 C& D

DISCLOSURES, W' ^! V. [2 E9 N# a8 k1 f2 u* R
$ L! h8 g3 u$ d' R
The authors indicate no potential conflicts of interest.' M; J3 N/ p8 a. U

ACKNOWLEDGMENTS
; _2 v. W$ x+ }5 e. P% m9 f
We thank all colleagues who referred patients, and we also thank patients for their willingness to contribute to the study. Dr. G. Longo helped with statistical analysis. This study was supported by the Italian Ministry of Health (Progetti di Ricerca di Interesse Nazionale) and The National Stem Cell Project; Associazione Italiana per la Ricerca sul Cancro, Milano; Ente Cassa di Risparmio di Firenze; Grants RBNE0189JJ 006/RBNE01SP72 003 from the Italian Ministry of Industrial and University Research; institutional funds from the University of Illinois; the Association pour la Recherche contre le Cancer (ARC) number 9806, the Groupement d'Interet Scientifique (GIS)-Institut des Maladies Rares number 03/GIS/PB/SJ/n¡ã35. A.P. was the recipient of a fellowship from Associazione Italiana per le Leucemie Firenze. R. M. and A.M.V. contributed equally to the study./ l6 @9 F/ e3 b  U) H0 T
      【参考文献】$ K% c& ^( |; q2 Z# @

Vardiman JW, Harris NL, Brunning RD. The World Health Organization (WHO) classification of the myeloid neoplasms. Blood 2002;100:2292¨C2302.
) t  Z) W5 e5 A
Barosi G. Myelofibrosis with myeloid metaplasia: Diagnostic definition and prognostic classification for clinical studies and treatment guidelines. J Clin Oncol 1999;17:2954¨C2970.

Tefferi A. Myelofibrosis with myeloid metaplasia. N Engl J Med 2000;342:1255¨C1265.0 u/ O3 U2 W& y7 {: h: j
+ m) q. a  ]. w, q6 j' E* ^6 c1 h
Xu M, Bruno E, Chao J et al. The constitutive mobilization of bone marrow-repopulating cells into the peripheral blood in idiopathic myelofibrosis. Blood 2005;105:1699¨C1705.7 H, `# X5 X% x" O
( E, `* B' N% _
Xu M, Bruno E, Chao J et al. The constitutive mobilization of CD34  cells into the peripheral blood in idiopathic myelofibrosis may be due to the action of a number of proteases. Blood 2005;105:4508¨C4515.& T. X$ Q. d$ V: Z! y6 {

Barosi G, Viarengo G, Pecci A et al. Diagnostic and clinical relevance of the number of circulating CD34( ) cells in myelofibrosis with myeloid metaplasia. Blood 2001;98:3249¨C3255.2 ~# z2 L. d' Y: X

Arora B, Sirhan S, Hoyer JD et al. Peripheral blood CD34 count in myelofibrosis with myeloid metaplasia: A prospective evaluation of prognostic value in 94 patients. Br J Haematol 2005;128:42¨C48.6 F' Y( T8 `& Q: Z% \  u
2 ?+ T( w1 h0 G
Mesa RA, Li CY, Ketterling RP et al. Leukemic transformation in myelofibrosis with myeloid metaplasia: A single-institution experience with 91 cases. Blood 2005;105:973¨C977.
# o" f' k( |5 M( Z6 Y0 M
Castro-Malaspina H, Jhanwar SC. Properties of myelofibrosis-derived fibroblasts. Prog Clin Biol Res 1984;154:307¨C322./ r9 Y. e+ z; R( p  i& d
2 h( V7 j/ K8 f5 i9 c2 g; J8 ~" V
Le Bousse-Kerdiles MC, Martyre MC. Involvement of the fibrogenic cytokines, TGF-beta and bFGF, in the pathogenesis of idiopathic myelofibrosis. Pathol Biol (Paris) 2001;49:153¨C157., j) q% ~; M8 X9 |1 e! }
/ P5 H8 v. p. E6 n) \/ ]+ `
Villeval JL, Cohen-Solal K, Tulliez M et al. High thrombopoietin production by hematopoietic cells induces a fatal myeloproliferative syndrome in mice. Blood 1997;90:4369¨C4383.
6 T4 ]3 W  |5 _; x
Yan XQ, Lacey D, Hill D et al. A model of myelofibrosis and osteosclerosis in mice induced by overexpressing thrombopoietin (mpl ligand): Reversal of disease by bone marrow transplantation. Blood 1996;88:402¨C409.
% s$ Q9 j' A- ^1 H
Vannucchi AM, Bianchi L, Cellai C et al. Development of myelofibrosis in mice genetically impaired for GATA-1 expression (GATA-1(low) mice). Blood 2002;100:1123¨C1132.

Pikman Y, Lee BH, Mercher T et al. MPLW515L Is a novel somatic activating mutation in myelofibrosis with myeloid metaplasia. PLoS Med 2006;3:e270.
5 |5 K. Y! \/ G, c1 |
Vannucchi AM, Pancrazzi A, Guglielmelli P et al. Abnormalities of GATA-1 in megakaryocytes from patients with idiopathic myelofibrosis. Am J Pathol 2005;167:849¨C858.

Levine RL, Wadleigh M, Cools J et al. Activating mutation in the tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Cancer Cell 2005;7:387¨C397.
# l0 C% U6 {8 g0 i& W8 @
Baxter EJ, Scott LM, Campbell PJ et al. Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet 2005;365:1054¨C1061.

James C, Ugo V, Le Couedic JP et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature 2005;434:1144¨C1148.( a* _# A) ]( Q# y# L) J
+ p* J/ z; `5 @' A4 A/ m; k; L
Kralovics R, Passamonti F, Buser AS et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders. N Engl J Med 2005;352:1779¨C1790.& W  d; B, N& U+ J* e/ @

Lippert E, Boissinot M, Kralovics R et al. The JAK2¨CV617F mutation is frequently present at diagnosis in patients with essential thrombocythemia and polycythemia vera. Blood 2006;108:1865¨C1867.. [2 \+ ?9 V. [/ Q. A( J0 z

Antonioli E, Guglielmelli P, Pancrazzi A et al. Clinical implications of the JAK2 V617F mutation in essential thrombocythemia. Leukemia 2005;19:1847¨C1849.
, T3 _1 }# r  z. i; T
Lacout C, Pisani DF, Tulliez M et al. JAK2V617F expression in murine hematopoietic cells leads to MPD mimicking human PV with secondary myelofibrosis. Blood 2006;108:1652¨C1660./ `, n& {4 V5 m- }* @1 c

Wernig G, Mercher T, Okabe R et al. Expression of Jak2V617F causes a polycythemia vera-like disease with associated myelofibrosis in a murine bone marrow transplant model. Blood 2006;107:4274¨C4281.

Cazzola M, Skoda R. Gain of function, loss of control - a molecular basis for chronic myeloproliferative disorders. Haematologica 2005;90:871¨C874.* x( |) K! P& u7 [

Kaushansky K. On the molecular origins of the chronic myeloproliferative disorders: It all makes sense. Blood 2005;105:4187¨C4190.

Ebert BL, Golub TR. Genomic approaches to hematologic malignancies. Blood 2004;104:923¨C932.; L: Y4 h% m$ R

Bilhou-Nabera C, Brigaudeau C, Clay D et al. Does cytogenetic mosaicism in CD34 CD38low cells reflect the persistence of normal primitive hematopoietic progenitors in myeloid metaplasia with myelofibrosis? Blood 2003;102:1551¨C1552.2 N$ X, }, v3 C' C2 v: k; ]* s

Barosi G, Ambrosetti A, Finelli C et al. The Italian Consensus Conference on Diagnostic Criteria for Myelofibrosis with Myeloid MetaplasiaBr J Haematol 1999;104: 730¨C737.& W* T. V; d) U: f  O1 \

Dupriez B, Morel P, Demory JL et al. Prognostic factors in agnogenic myeloid metaplasia: A report on 195 cases with a new scoring system. Blood 1996;88:1013¨C1018.3 m( V: U& \. o3 ]# i
3 @" |! L) @$ ~. ~, H) U( o- h  I
Marchetti M, Barosi G, Balestri F et al. Low-dose thalidomide ameliorates cytopenias and splenomegaly in myelofibrosis with myeloid metaplasia: A phase II trial. J Clin Oncol 2004;22:424¨C431.
" O* e$ I9 v$ w7 k' Z
Keeney M, Chin-Yee I, Weir K et al. Single platform flow cytometric absolute CD34  cell counts based on the ISHAGE guidelines. International Society of Hematotherapy and Graft Engineering. Cytometry 1998;34:61¨C70.

Manfredini R, Zini R, Salati S et al. The kinetic status of hematopoietic stem cell subpopulations underlies a differential expression of genes involved in self-renewal, commitment, and engraftment. STEM CELLS 2005;23:496¨C506.$ s, d9 ^, y. I8 d7 I
7 K5 O; p2 L7 t. i
Boucheix C, Rubinstein E. Tetraspanins. Cell Mol Life Sci 2001;58:1189¨C1205.$ h* ]) e, Q0 j! ^7 p1 i, W8 V
4 E$ T$ W: ~5 o  }2 l
Zannettino AC, Buhring HJ, Niutta S et al. The sialomucin CD164 (MGC-24v) is an adhesive glycoprotein expressed by human hematopoietic progenitors and bone marrow stromal cells that serves as a potent negative regulator of hematopoiesis. Blood 1998;92:2613¨C2628.
9 t. _! c/ h2 O! t
Burger JA, Kipps TJ. CXCR4: A key receptor in the crosstalk between tumor cells and their microenvironment. Blood 2006;107:1761¨C1767.

Steidl U, Kronenwett R, Rohr UP et al. Gene expression profiling identifies significant differences between the molecular phenotypes of bone marrow-derived and circulating human CD34  hematopoietic stem cells. Blood 2002;99:2037¨C2044.+ S; v5 @( W8 G

Clay D, Rubinstein E, Mishal Z et al. CD9 and megakaryocyte differentiation. Blood 2001;97:1982¨C1989.

Jones LC, Tefferi A, Vuong PT et al. Detection of aberrant gene expression in CD34  hematopoietic stem cells from patients with agnogenic myeloid metaplasia using oligonucleotide microarrays. STEM CELLS 2005;23:631¨C637." k# [$ h0 B2 Z. {! s4 l) ^

Nelson WJ, Nusse R. Convergence of Wnt, beta-catenin, and cadherin pathways. Science 2004;303:1483¨C1487.: y% r: k& T" u7 \1 g, S; L( Z
9 p2 j7 s! m5 f5 e
Falanga A, Marchetti M, Barbui T et al. Pathogenesis of thrombosis in essential thrombocythemia and polycythemia vera: The role of neutrophils. Semin Hematol 2005;42:239¨C247.& b/ J0 B! R- R1 O7 H7 X% `, S
  t4 o; j$ r% ]$ g
Gene Expression Omnibus profiles. Available at http://www.ncbi.nlm.nih.gov/geo, GDS781 October 4, 2005.; ~7 E$ [+ O: p( l" W

Tefferi A, Sirhan S, Lasho TL et al. Concomitant neutrophil JAK2 mutation screening and PRV-1 expression analysis in myeloproliferative disorders and secondary polycythaemia. Br J Haematol 2005;131:166¨C171.
( N. `$ ]; v' n, Q% i7 f
Goerttler PS, Steimle C, Marz E et al. The Jak2V617F mutation, PRV-1 overexpression and EEC formation define a similar cohort of MPD patients. Blood 2005;106:2862¨C2864.

Kralovics R, Teo SS, Buser AS et al. Altered gene expression in myeloproliferative disorders correlates with activation of signaling by the V617F mutation of Jak2. Blood 2005;106:3374¨C3376.! V: A0 n- N, Y- J: U

Keilholz U, Menssen HD, Gaiger A et al. Wilms' tumour gene 1 (WT1) in human neoplasia. Leukemia 2005;19:1318¨C1323./ o: v7 J) U; v! x5 b- d

Cilloni D, Gottardi E, Messa F et al. Significant correlation between the degree of WT1 expression and the International Prognostic Scoring System Score in patients with myelodysplastic syndromes. J Clin Oncol 2003;21:1988¨C1995./ F5 B3 o4 ?1 C' Q# X4 _

Inoue K, Sugiyama H, Ogawa H et al. WT1 as a new prognostic factor and a new marker for the detection of minimal residual disease in acute leukemia. Blood 1994;84:3071¨C3079.. p- h! d* s5 P9 C) `8 b

Inoue K, Tamaki H, Ogawa H et al. Wilms' tumor gene (WT1) competes with differentiation-inducing signal in hematopoietic progenitor cells. Blood 1998;91:2969¨C2976.+ }& b5 f/ s2 m0 E- I, F: e

Tsuboi A, Oka Y, Ogawa H et al. Constitutive expression of the Wilms' tumor gene WT1 inhibits the differentiation of myeloid progenitor cells but promotes their proliferation in response to granulocyte-colony stimulating factor (G-CSF). Leuk Res 1999;23:499¨C505.
% s# F/ k! W  |9 p2 h
Jomgeow T, Oji Y, Tsuji N et al. Wilms' tumor gene WT1 17AA(-)/KTS(-) isoform induces morphological changes and promotes cell migration and invasion in vitro. Cancer Sci 2006;97:259¨C270.
: |8 Q# @- n8 D  g; h  f7 a
Sakajiri S, O'Kelly J, Yin D et al. Dlk1 in normal and abnormal hematopoiesis. Leukemia 2005;19:1404¨C1410.

Miyazato A, Ueno S, Ohmine K et al. Identification of myelodysplastic syndrome-specific genes by DNA microarray analysis with purified hematopoietic stem cell fraction. Blood 2001;98:422¨C427.

Huang CC, Chuang JH, Huang LL et al. The human Delta-like 1 homologue is implicated in the progression of liver fibrosis in biliary atresia. J Pathol 2004;202:172¨C179.+ ~9 ~% |. o: s5 `6 G  J
0 ~" O; A! s9 M- |. Y4 E
Oka Y, Tsuboi A, Taguchi T et al. Induction of WT1 (Wilms' tumor gene)-specific cytotoxic T lymphocytes by WT1 peptide vaccine and the resultant cancer regression. Proc Natl Acad Sci U S A 2004;101:13885¨C13890.  [- U$ O. o# l3 O: T

Barosi G, Bordessoule D, Briere J et al. Response criteria for myelofibrosis with myeloid metaplasia: results of an initiative of the European Myelofibrosis Network (EUMNET). Blood 2005;106:2849¨C2853.

收藏0 分享0 顶0 踩0