Original Article | Published:


Gene expression profiling of B lymphocytes and plasma cells from Waldenström's macroglobulinemia: comparison with expression patterns of the same cell counterparts from chronic lymphocytic leukemia, multiple myeloma and normal individuals

Leukemia volume 21, pages 541549 (2007) | Download Citation



The tumoral clone of Waldenström's macroglobulinemia (WM) shows a wide morphological heterogeneity, which ranges from B lymphocytes (BL) to plasma cells (PC). By means of genome-wide expression profiling we have been able to identify genes exclusively deregulated in BL and PC from WM, but with a similar expression pattern in their corresponding cell counterparts from chronic lymphocytic leukemia (CLL) and multiple myeloma (MM), as well as normal individuals. The differentially expressed genes have important functions in B-cell differentiation and oncogenesis. Thus, two of the genes downregulated in WM-BL were IL4R, which plays a relevant role in CLL B-cell survival, and BACH2, which participates in the development of class-switched PC. Interestingly, one of the upregulated genes in WM-BL was IL6. A set of four genes was able to discriminate clonal BL from WM and CLL: LEF1 (WNT/β-catenin pathway), MARCKS, ATXN1 and FMOD. We also found deregulation of genes involved in plasma cell differentiation such as PAX5, which was overexpressed in WM-PC, and IRF4 and BLIMP1, which were underexpressed. In addition, three of the target genes activated by PAX5CD79, BLNK and SYK – were upregulated in WM-PC. In summary, these results indicate that both PC and BL from WM are genetically different from the MM and CLL cell counterpart.


Waldenstrom's macroglobulinemia (WM) is a B-lymphoproliferative disorder (BLPD) characterized by the proliferation in the bone marrow of clonal B lymphocytes (BL) with lymphoplasmacytic differentiation. The tumoral clone of WM shows a wide morphological heterogeneity, which ranges from BL to plasma cells (PC), including a lymphoplasmacytic population that defines the disease.1, 2 Our group has demonstrated that the PCs are part of the malignant clone as assessed by cytoplasmic light-chain restriction,3 along with molecular biology and fluorescent in situ hybridization (FISH) studies (data not shown), which have shown that this PC compartment carries the same pattern of immunoglobulin gene rearrangement and the same cytogenetic abnormalities as those present in clonal BL.

The complexity of malignant clone indicates that WM is a singular entity that displays similarities and differences with other BLPD. Thus, as compared to multiple myeloma (MM), both diseases share a clonal PC component responsible for monoclonal paraprotein production, IgM subtype in WM and non-IgM in MM, except for the rare IgM MM variant. On the other hand, IgM paraprotein can be observed in many subtypes of NHL, particularly low/intermediate-grade NHL, as well as in chronic lymphocytic leukemia (CLL). Moreover, about 20% of IgM monoclonal gammapathy of unknown significance eventually develop WM, lymphoma, CLL or amyloidosis.4 The information provided by conventional cytogenetic studies in WM is rather limited because of the predominance of normal metaphases. In spite of this, the studies carried out so far reveal a particular cytogenetic profile that differs from other BLPD. Unlike MM and many other BLPD, translocations involving the 14q32 region are extremely infrequent in WM.5 Therefore, WM is part of those BLPD, like CLL, in which Ig heavy-chain rearrangements are not involved in their pathogenesis. The only recurrent chromosomal abnormality identified by conventional cytogenetics and FISH studies is deletion of 6q.6

The differences between clonal cell compartments of WM and their cell counterpart in other lymphoproliferative disorders have not yet been sufficiently explored7 (Zhan et al., Blood 2002; abstract No. 1227 and Gutiérrez et al., Blood 2005; abstract No. 503) mainly because of the difficulty in selecting pure clonal cell populations in sufficient amounts to apply genetic analysis. The development of multiparameter flow cytometry sorting, which allows the separation of cell populations with high purity, and the reproducibility of gene expression microarray analysis in mRNA amplified samples prompted us to investigate the gene expression profile of both clonal B cells and PC from WM, and to compare it with BL from CLL and PC from MM, as well as with both normal B lymphocytes (NBL) and PC from healthy donors.

Materials and methods


Bone marrow (BM) samples from 10 patients with WM, 12 with MM and 11 with CLL were included in the study. All samples corresponded to newly diagnosed untreated patients. In addition, eight NBL samples from peripheral blood and five normal plasma cells (NPC) from BM of healthy donors were also selected in order to relate the deregulation of gene expression profiling of clonal populations to normal condition. The study was approved by the local research ethics committee and written informed consent was obtained from all patients and healthy donors.

Multiparameter flow cytometry sorting

The isolation of clonal BL and PC populations was carried out by multiparameter flow cytometry sorting with the appropriate monoclonal antibodies (MoA). The combination of MoA for isolation of clonal BL and PC from WM patients included Kappa or Lambda-fluorescein isothiocyanate (DAKO, Glostrup, Denmark), CD10-PE (Immunotech, Marseille, France), CD38-PerCP-Cy5.5 (Pharmingen, San Diego, CA, USA), CD19-PE-Cy7, CD34-APC and CD45-APC-Cy7 (BDB, San José, CA, USA). For PC from MM and normal BM, we used CD38-APC (BDB, San José, CA, USA), for BL from CLL patients, we used CD19-PE and CD5-APC (BDB, San José, CA, USA); for NBL, only CD19-PE was used. The purity obtained for clonal cells was more than 95%.

Genome-wide expression profiling

RNA isolation, labeling and microarray hybridization have been previously reported.8 Total RNA was extracted from purified cell populations using the RNeasy Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. The RNA integrity was assessed using Agilent 2100 Bioanalyzer (Agilent, Palo Alto, CA, USA). Labeling and hybridizations were performed according to protocols from Affymetrix. Briefly, 100 ng of total RNA was amplified and labeled using the GeneChip two-cycle cDNA synthesis kit and GeneChip IVT labeling kit (Affymetrix Inc., Santa Clara, CA, USA) and then hybridized to Human Genome U133A microarray (Affymetrix), after quality checking on GeneChips Test3 Arrays. Washing and scanning were performed using Fluidics Station 400 and GeneChip Scanner (Affymetrix Inc.).

Genome-wide expression data analysis

Data pre-processing and differential expression analyses

Expression value for each probe set was calculated using RMAExpress program that uses RMA (Robust Multi-array Average) algorithm.9 For a first supervised analysis (see below), gene-filtering methods were applied following the detection calls calculated using MAS 5.0 algorithm (Affymetrix). Probe sets with absent calls as well as those showing minimal variation across all samples (maximum–minimum log2 variation<2.5) were filtered out. In a second, more exhaustive analysis, the complete set of microarrays without gene filtering was introduced in a multivariable matrix where each of the six cell states (NBL, WM-BL, CLL-BL, NPC, WM-PC and MM-PC) were introduced to perform the differential expression analyses (see below).

Unsupervised cluster analysis

To classify the samples, hierarchical clustering (Cluster and TreeView software), based on the average-linkage method with the centered correlation metric, was used.10 To double classify the samples and genes in an unsupervised method, hierarchical clustering based on the complete linkage with centered correlation metric was used within R (version 2.3.1, 2006-06-01, http://www.r-project.org/). Multidimensional scaling method (BRB Array Tools version 3.0) was performed using Euclidean distance.11

Supervised analysis

Two types of differential expression analyses were carried out on the data. First, Significant Analysis of Microarrays (SAM) algorithm was used to identify genes with statistically significant changes in expression between different classes. All data were permutated over 100 cycles by using the two-class (unpaired) and multiclass response format, not considering equal variances. Significant genes were selected based on the lowest false discovery ratio (FDR) and controlling the q-value for the gene list.12 FDR level was <2% in all the class comparisons.

A second more complete differential expression analysis was performed using the multivariable states matrix quoted above and applying a empirical Bayesian method for differential expression calculation,13 correcting for multiple testing with the method of Reiner et al.14 and selecting genes always with a P-value of differential expression below 0.01. All these methods were implemented using in R the Linear Models for Microarray Data (LIMMA) package and other packages from BioConductor (http://www.bioconductor.org/) for genomic and microarrays data handling.

Prediction analysis of microarrays (PAM, R), by using nearest-shrunken centroids and crossvalidation assay, was carried out to identify the smallest subsets of genes that were able to accurately predict classes.15

Gene function analysis

The probe sets were functionally annotated and grouped according to their biological function using GeneOntology biological process descriptions. The functional analysis to identify the most relevant biological mechanisms, pathways and functional categories in the data sets of genes selected by statistical analysis, was generated through the use of Ingenuity Pathways Analysis (Ingenuity Systems, Mountain View, CA, USA).

Results and discussion

Initially, we investigated whether the selected B populations, BL from WM and from CLL and PC from WM and MM displayed specific expression profiles that were clearly distinguishable from each other. The hierarchical clustering analysis using over six thousand genes displayed a dendrogram with two major branches: the first one included two clusters that grouped all the clonal BL samples from WM (WM-BL) and a clearly separate cluster containing CLL samples; the second branch included all clonal PC samples divided into two clusters, one containing the clonal PC from WM (WM-PC) and the other, the PC from MM (MM-PC) (Figure 1). Multidimensional scaling analysis identified similar clusters according to the different cell populations (Supplementary Figure 1). When we introduced the normal cell populations (NBL and NPC) in a second clustering analysis, the heat map showed that CLL and MM samples were maintained in completely separate clusters. However, some of the WM-BL and WM-PC samples clustered next to their respective normal cell counterpart, suggesting a higher similarity level between WM clonal cells and normal cells than between CLL, or MM cells and their normal cell counterparts (Supplementary Figure 2). This finding was confirmed using supervised differential expression analysis, which identified a number of deregulated genes significantly higher in CLL vs NBL and MM vs NPC comparisons (data not shown) than in WM-BL vs NBL and WM-PC vs NPC.

Figure 1
Figure 1

Dendrogram of hierarchical cluster analysis, based on the expression of 6057 genes passing a variation filter, generated by using Cluster and TreeView programs. WM-BL (blue), WM-PC (green), CLL (red) and MM (orange) samples were included.

After confirming in the unsupervised analysis that clonal cell populations from WM were distributed in clusters clearly separated from CLL and MM cells, we investigated the transcriptomic signature of WM related to normal status. The expression profile of WM using the entire malignant population (CD19+ and CD138+ cells) has recently been published.7 In our study, the separate isolation of CD19+ B cells and CD138+ PC gave us the opportunity to search for the expression profile that defined clonal BL of WM (WM-BL) with relation to NBL and the molecular fingerprint of clonal PC of WM (WM-PC) with respect to NPC. The comparative analyses of the gene expression profile of WM-BL vs NBL identified a set of 171 genes differentially expressed between the two groups (109 genes were downregulated in WM-BL and 62 were upregulated). The gene function analysis revealed that most of these genes were included in cellular growth and proliferation, cell death and immune system categories. The largest network identified by ingenuity pathways analysis program contained 29 differentially expressed genes (Supplementary Figure 3). Interestingly, three members of the activator protein 1 (AP-l) group, JUN, FOSB and BATF, were overexpressed in WM-BL. Both FOSB and BATF form stable heterodimers with JUN proteins. This complex system is involved in cell proliferation, transformation and death, which turn it into an attractive pathway for further investigation.16 Moreover, genes involved in B-cell development, such as BTK, CD69, CD83, IRF8 and ITPR1, were deregulated in WM-BL: the overexpression of CD69 and CD83 in WM could be useful for diagnostic applications based on immunophenotypic analysis.17 Of special interest as a potential therapeutic target is the upregulation in WM-BL of CEACAM1 gene, which exhibits potent angiogenic properties and is a major effector of vascular endothelial growth factor (VEGF).18 When comparing the expression profile of WM-PC with that of NPC, a total of 498 genes were differentially expressed, all of them upregulated in the WM-PC group. Further analysis of these genes derived from Ingenuity networks showed that the three most significant functional categories were RNA post-transcriptional modification, DNA replication and cellular assembly and organization (Supplementary Figure 4).

The definition of WM molecular transcriptomic signature prompted us to explore a possible relationship between WM-BL and WM-PC. In fact, immunophenotypic analysis of WM carried out by our group has shown that phenotypic pattern exhibited by WM-PC is different from that observed in both normal and myelomatous PC, particularly because of WM-PC retaining intermediate features between clonal BL and PC. In order to explore genes that reflect this derailment in the differentiation pathway of the clonal BL into PC, we performed a multivariable analysis that included selection of significant genes with differential expression fulfilling a double conditional of being differential between NBL and NPC but not between WM-BL and WM-PC. This double conditional was not empty but quite interestingly a significant amount of genes were found. This analysis showed a total of 37 genes whose expression level in WM-PC was intermediate between WM-BL and, NPC and MM-PC. This set of deregulated genes included genes involved in plasma cell differentiation. PAX5, whose transcription must be repressed to allow PC differentiation,19 had significantly higher expression levels in WM-PC than in MM-PC. In other words, in WM-PC PAX5 expression had an intermediate level between WM-BL and PC from MM and NPC. An increased expression level of PAX5 has also been described in PC from CD20-positive myelomas.20 By contrast, BLIMP1 and IRF4 genes, which have an essential role in PC differentiation,21 had significantly lower expression levels in WM-PC than they did in MM-PC and NPC. In addition, we have observed that three of the target genes activated by PAX5CD79, BLNK and SYK – were upregulated in WM-PC, with an expression level more similar to WM-BL. It is possible that the lack of PAX5 repression in PC from WM is responsible for the upregulation of CD79, BLNK and SYK, which are characteristic markers of B lymphoid cell identity but which are lost in the maturation process into PC22 (Supplementary Figure 5). Of particular interest is the high expression of topoisomerase II beta (Top2B) in WM-BL and to a lesser extent in WM-PC in comparison with MM-PC and NPC. Accordingly, topoisomerase II inhibitors could be a more useful therapeutic approach in WM treatment than in MM patients.

The second approach was to look for genes exclusively deregulated in each lymphoid malignancy in order to identify gene categories and molecular pathways involved in WM development, but not in CLL or MM, as well as to discard pathways used in CLL and MM but not in WM. For this purpose, both NBL and NPC were also introduced into the analysis. Thus, gene deregulation always referred to normal condition. To avoid false-positive genes as much as possible and to make sure that the selected genes could be considered as specific markers, we used a conservative approach in the analysis with stringent FDR (0.2%). Deregulated genes were grouped in four sections as described below.

Genes exclusively deregulated in BL from CLL but with a similar expression pattern in BL from WM and NBL

A total of 68 functionally annotated genes (Table 1) were identified as genes exclusively deregulated in BL from CLL (CLL-BL), although they had the same expression profile in WM (WM-BL) and NBL. Most of these genes were upregulated in CLL and were assigned to two functional categories, cell signaling and regulation of transcription categories. Among the genes involved in cell signaling, of particular interest is CCR7 gene, which showed high expression levels in CLL-BL compared with WM-BL and NBL. This chemokine receptor is known to regulate trafficking of normal B, T and dendritic cells into the lymph node, and an increased expression of CCR7 has been invoked as a mechanism through which the CLL B cells can move from PB to BM and lymph nodes.23, 24, 25

Table 1: Genes exclusively deregulated in BL from CLL but with a similar expression pattern in WM-BL and NBL

The seven genes (ID3, SFMBT1, PBX3, HIVEP2, ZFP64, RXRA, LEF1) involved in the regulation of transcription category were upregulated in CLL. LEF1 was the most significant overexpressed gene in CLL-BL samples compared with WM-BL and NBL. Lymphoid enhancer factor-1 (LEF-1) is a member of the lymphoid enhancer factor/T-cell factor (LEF/TCF) family of high-mobility group (HGM) transcription factors.26 LEF/TCF proteins have been shown to interact with β-catenin as downstream mediators of WNT signals-transduction pathway.27 Although LEF-1 was originally identified in pre-B and T cells, its function in BL development remains unknown.28, 29

Genes exclusively deregulated in BL from WM but with a similar expression pattern in BL from CLL and NBL

In contrast to CLL, the number of genes exclusively deregulated in WM-BL was much lower, only 19 significant genes (Table 2), and with a more balanced distribution between over and underexpressed genes. The two most significantly deregulated functional categories in WM-BL were the molecular transport and cellular proliferation categories. These included genes such as one calcium binding protein (S100A8), the oncogene FOSB and the hematopoietic cell kinase (HCK), all upregulated in WM. So, they may be useful markers for WM as well as a starting point for investigating therapeutic interventions. Interestingly, one of the upregulated genes in WM-BL was IL6. This finding is consistent with the recent report by Chng et al.7 and previous data showing high serum IL6 concentrations in patients with WM.30

Table 2: Genes exclusively deregulated in BL from WM but with a similar expression pattern in CLL and NBL

Two of the genes (IL4R and BACH2) downregulated in WM-BL compared with CLL-BL and normal BL have important functions in B-cell development. IL4/IL4R pathway plays an important role in CLL B-cell survival.31 The significant underexpression of IL4R in WM-BL suggests that this pathway is not directly involved in WM development. Interestingly, BACH2, which is abundantly expressed in the early stages of B-cell differentiation and turned off in terminally differentiated cells,32 was significantly downregulated in WM-BL compared with CLL-BL and NBL. This finding supports the recent study that showed that BACH2 is dispensable for the development of IgM PC but is crucial for the development of class-switched PC.33

In order to ascertain which genes were able to predict whether clonal BL belong to WM or CLL, a class prediction analysis using nearest-shrunken centroids method was performed (Figure 2a). This analysis revealed that a set of four genes (LEF1, MARCKS, ATXN1 and FMOD) was able to discriminate clonal BL from both diseases, with perfect accuracy (Figure 2b). LEF1 involved in WNT/βcatenin pathway, ATXN1, a gene not involved in cancer so far, and FMOD a protein from the extracellular matrix, were all upregulated in CLL compared with WM. Previous reports suggest that expression of FMOD in hematological malignancies seems to be exclusive for CLL and mantle cell lymphoma.34, 35 In contrast, MARCKS, a gene implicated in suppression of proliferation, was upregulated in WM.36

Figure 2
Figure 2

(a) ‘Nearest-shrunken centroids’ analysis. Shrunken differences for the 20 genes having at least one non-zero difference are depicted. Left-sided bars indicate lower expression in subgroups relative to overall centroid; right-sided bars indicate higher expression in subgroups relative to overall centroid. The genes with non-zero components in each class are almost mutually exclusive. The length of horizontal bars represents the difference between the overall centroid and each of the two subgroup centroids. Genes are ordered according to the greatest difference at the top and the smallest difference at the bottom. (b) Graphics of the four genes that discriminate clonal BL from WM and CLL with an accuracy better than 99%. The y axis shows the log2 intensity value. Red and blue circled symbols represent CLL and WM samples, respectively.

Genes exclusively deregulated in PC from MM but with a similar expression profile in PC from WM and NPC

Again, using a stringent FDR, a total of 42 genes were identified as genes exclusively deregulated in PC from MM (MM-PC), although they had the same expression profile in PC from WM (WM-PC) and normal PC (Table 3). Most of these genes were upregulated in MM. Among the top 15 MM-PC upregulated genes, there was a novel septin family member (SEPTIN10), a protein implicated in response to oxidative stress (SEPP1), proteins involved in protein processing and folding (CANX, PSENEN) and the integrin-associated protein, CD47, which modulates cell activation and adhesion.37, 38 Four of the only five genes downregulated in MM-PC compared with WM-PC and NPC encoded Ig molecules, whereas the other was a member of a tumor necrosis factor receptor superfamily, TNFRSF7/CD27, which is involved in PC differentiation and may activate death signals.39, 40 It should be noted that when we increased the FDR up to 2%, which is more useful for detecting possible gene interactions although the risk of selecting false positive-genes increases, we found that several genes implicated in WNT signaling had differential expression between PC from WM and from MM. Two of them, DKK1 and FRZB, which have been previously described as fundamental genes in bone disease in MM,41 had an absent expression in WM-PC, whereas they exhibited a high expression level in MM-PC. This could be an explanation for the lack of bone lesions in WM despite the presence of clonal PC.

Table 3: Genes exclusively deregulated in PC from MM but with a similar expression profile in WM-PC and NPC

Genes exclusively deregulated in PC from WM but with a similar expression profile in PC from MM and NPC

Only 13 significant deregulated genes were identified in this group (Table 4), all of them overexpressed in WM-PC (particularly the tyrosine kinase SYK). Most of these 13 genes were present with a higher or similar intensity level in WM-BL, supporting a close relationship between clonal WM-PC and WM-BL. This finding, along with the aforementioned singular deregulation found in some of the crucial genes involved in PC differentiation, strongly supports the notion that PC from WM result from an incomplete maturation process of clonal BL. By contrast, both a member of the CD20 family MS4A342 and the oncogene MYB showed a higher expression in WM-PC than in WM-BL.

Table 4: Genes exclusively deregulated in PC from WM but with a similar expression profile in MM and NPC

Finally, we also investigated the relationship between the three BLPD included in the present study. Using a supervised approach, we looked for genes differentially expressed between malignant and normal cells. For this purpose, we have investigated both: (1) genes similarly deregulated in clonal BL from WM and CLL in comparison with NBL, and (2) genes similarly deregulated in clonal PC from WM and MM in comparison with NPC. A set of 28 genes were deregulated in the same direction in WM-BL and CLL-BL compared with NBL (Supplementary Figure 6). One of the most significant overexpressed genes, both in WM and CLL, was the IL10 receptor (IL10RA). The IL10 signaling pathway has been involved in the expansion of malignant B cells from other BLPD such as mantle cell lymphoma.36, 43 To our knowledge, this is the first time that a deregulation of this pathway has been associated with WM, and this could represent a possible target for therapeutic intervention. By contrast, the IL13 receptor (IL13RA1) was downregulated in both diseases compared with NBL. IL13 and IL4 receptors are closely related. It has been proposed that the IL4 receptor in B cells is composed of three chains (IL4RA, IL13RA1 and the gamma common chain), which make it capable of transducing an IL13- or IL4-dependent proliferative signal.44 Our data demonstrate a downregulation of both IL4R and IL13RA in WM, suggesting that these growth factor pathways are not responsible for B-cell survival in WM. A different scenario could be hypothesized for CLL because gene expression profiling shows a high expression for IL4R and a very low expression for IL13RA, which adds more controversy to the still insufficiently characterized IL4 and IL13 receptors.31

Concerning the relationship between WM and MM, 68 genes were upregulated in PC from both diseases (Supplementary Figure 7). Most of these genes encoded for proteins involved in molecular transport (SCP2, PRDX1, APEX1, UGP2), RNA trafficking (RAN, XPO1), RNA post-transcriptional modification (heterogeneous nuclear ribonucleoproteins) and ribosomal proteins. This suggests that WM-PC display the same secretory machinery as MM-PC.

Overall, our results indicate that both PC and BL from WM are genetically different from the MM and CLL cell counterpart. The differentially expressed genes have important functions in B-cell differentiation and oncogenesis. At the same time, a multivariable combined analysis revealed a set of genes characteristic of WM in both BL and PC cell types, providing a distinct transcription signature of this disease.


  1. 1.

    , , , , . Waldenstrom's macroglobulinemia: clinical features, complications, and management. J Clin Oncol 2000; 18: 214–226.

  2. 2.

    , , , , , et al. Clinicopathological definition of Waldenstrom's macroglobulinemia: consensus panel recommendations from the Second International Workshop on Waldenstrom's Macroglobulinemia. Semin Oncol 2003; 30: 110–115.

  3. 3.

    , , , , , et al. Immunophenotypic analysis of Waldenstrom's macroglobulinemia. Semin Oncol 2003; 30: 187–195.

  4. 4.

    , , . Waldenstrom macroglobulinaemia. Lancet Oncol 2003; 4: 679–685.

  5. 5.

    , , , , . 14q32 Translocations discriminate IgM multiple myeloma from Waldenstrom's macroglobulinemia. Semin Oncol 2003; 30: 153–155.

  6. 6.

    , , , , , et al. Waldenstrom macroglobulinemia neoplastic cells lack immunoglobulin heavy chain locus translocations but have frequent 6q deletions. Blood 2002; 100: 2996–3001.

  7. 7.

    , , , , , et al. Gene expression profiling of Waldenstrom's macroglobulinemia reveals a phenotype more similar to chronic lymphocytic leukemia than multiple myeloma. Blood 2006; 108: 2755–2763.

  8. 8.

    , , , , , et al. Gene expression profile reveals deregulation of genes with relevant functions in the different subclasses of acute myeloid leukemia. Leukemia 2005; 19: 402–409.

  9. 9.

    , , , , , . Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 2003; 31: e15.

  10. 10.

    , , , . Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 1998; 95: 14863–14868.

  11. 11.

    , . Modern Multidimensional Scaling: Theory and Applications. Springer Verlag: New York, 1997.

  12. 12.

    , , . Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 2001; 98: 5116–5121.

  13. 13.

    . Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 2004; 3: Article 3.

  14. 14.

    , , . Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 2003; 19: 368–375.

  15. 15.

    , , , . Diagnosis of multiple cancer types by shrunken centroids of gene expression. Proc Natl Acad Sci USA 2002; 99: 6567–6572.

  16. 16.

    , . AP-1 in cell proliferation and survival. Oncogene 2001; 20: 2390–2400.

  17. 17.

    , , , , , . The soluble form of CD83 is present at elevated levels in a number of hematological malignancies. Leuk Res 2004; 28: 237–241.

  18. 18.

    , , , , , et al. CEA-related cell adhesion molecule 1: a potent angiogenic factor and a major effector of vascular endothelial growth factor. Mol Cell 2000; 5: 311–320.

  19. 19.

    , . Plasma cell differentiation and multiple myeloma. Curr Opin Immunol 2004; 16: 226–234.

  20. 20.

    , , , , , . Expression of PAX5 in CD20-positive multiple myeloma assessed by immunohistochemistry and oligonucleotide microarray. Mod Pathol 2004; 17: 1217–1222.

  21. 21.

    , . Regulation of plasma-cell development. Nat Rev Immunol 2005; 5: 230–242.

  22. 22.

    , , . Regulatory mechanisms that determine the development and function of plasma cells. Annu Rev Immunol 2003; 21: 205–230.

  23. 23.

    , , , , , et al. ZAP-70 expression is associated with enhanced ability to respond to migratory and survival signals in B-cell chronic lymphocytic leukemia (B-CLL). Blood 2006; 107: 3584–3592.

  24. 24.

    , , , . The chemokine receptor CCR7 and alpha4 integrin are important for migration of chronic lymphocytic leukemia cells into lymph nodes. Blood 2002; 99: 2977–2984.

  25. 25.

    , . Chemokine receptor expression in B-cell lymphoproliferative disorders. Leuk Lymphoma 2004; 45: 2491–2496.

  26. 26.

    , . Wnt signalling in stem cells and cancer. Nature 2005; 434: 843–850.

  27. 27.

    , , , . Wnt signaling in B-cell neoplasia. Oncogene 2003; 22: 1536–1545.

  28. 28.

    , . Variation of LEF-1 mRNA expression in low-grade B-cell non-Hodgkin's lymphoma. Leuk Res 2006; 30: 29–32.

  29. 29.

    , , , , , et al. Activation of the Wnt signaling pathway in chronic lymphocytic leukemia. Proc Natl Acad Sci USA 2004; 101: 3118–3123.

  30. 30.

    , , , , , . Serum levels of IL-6 and its soluble receptor (sIL-6R) in Waldenstrom's macroglobulinemia. Eur J Haematol 2001; 66: 1–6.

  31. 31.

    , . IL-4 biology: impact on normal and leukemic CLL B cells. Leuk Lymphoma 2003; 44: 897–903.

  32. 32.

    , , , , , et al. Prognostic significance of BACH2 expression in diffuse large B-cell lymphoma: a study of the Osaka Lymphoma Study Group. J Clin Oncol 2005; 23: 8012–8017.

  33. 33.

    , , , , , et al. The transcriptional programme of antibody class switching involves the repressor Bach2. Nature 2004; 429: 566–571.

  34. 34.

    , , , , , et al. Fibromodulin as a novel tumor-associated antigen (TAA) in chronic lymphocytic leukemia (CLL), which allows expansion of specific CD8+ autologous T lymphocytes. Blood 2005; 105: 1566–1573.

  35. 35.

    , , , , , et al. Fibromodulin, an extracellular matrix protein: characterization of its unique gene and protein expression in B-cell chronic lymphocytic leukemia and mantle cell lymphoma. Blood 2005; 105: 4828–4835.

  36. 36.

    , , , , . Mantle cell lymphomas express a distinct genetic signature affecting lymphocyte trafficking and growth regulation as compared with subpopulations of normal human B cells. Cancer Res 2002; 62: 4398–4405.

  37. 37.

    , , , , . Transendothelial migration of neutrophils involves integrin-associated protein (CD47). Proc Natl Acad Sci USA 1995; 92: 3978–3982.

  38. 38.

    , , , , , et al. Mechanisms of CD47-induced caspase-independent cell death in normal and leukemic cells: link between phosphatidylserine exposure and cytoskeleton organization. Blood 2002; 100: 2882–2890.

  39. 39.

    , , , , , et al. A possible role for the loss of CD27–CD70 interaction in myelomagenesis. Br J Haematol 2003; 120: 223–234.

  40. 40.

    , , , , , et al. CD27, a member of the tumor necrosis factor receptor family, induces apoptosis and binds to Siva, a proapoptotic protein. Proc Natl Acad Sci USA 1997; 94: 6346–6351.

  41. 41.

    , , , , , et al. The role of the Wnt-signaling antagonist DKK1 in the development of osteolytic lesions in multiple myeloma. N Engl J Med 2003; 349: 2483–2494.

  42. 42.

    , . Identification of a CD20-, FcepsilonRIbeta-, and HTm4-related gene family: sixteen new MS4A family members expressed in human and mouse. Genomics 2001; 72: 119–127.

  43. 43.

    , , , , , et al. The molecular signature of mantle cell lymphoma reveals multiple signals favoring cell survival. Cancer Res 2003; 63: 8226–8232.

  44. 44.

    . IL-13 effector functions. Annu Rev Immunol 2003; 21: 425–456.

Download references


We to thank Mark Anderson from the University Technology Transfer Office, and Isabel Isidro and Teresa Prieto for technical assistance. This study was partially supported by Spanish Myeloma Network Program (G03/136), ‘Ministerio de Ciencia y Tecnología’ grant (SAF 04/06587) and ‘Junta de Castilla y León’ grant (106/A/06).

Author information


  1. Servicio de Hematología, Hospital Universitario de Salamanca and Centro de Investigación del Cáncer (CIC), Universidad de Salamanca-CSIC, Salamanca, Spain

    • N C Gutiérrez
    • , E M Ocio
    • , P Maiso
    • , M Delgado
    • , M J Arcos
    • , J M Hernández
    •  & J F San Miguel
  2. Grupo de Investigación Bioinformática, Centro de Investigation del Cáncer (CIC), Universidad de Salamanca-CSIC, Salamanca, Spain

    • J de las Rivas
  3. Unidad de Genómica, Centro de Investigation del Cáncer (CIC), Universidad de Salamanca-CSIC, Salamanca, Spain

    • E Fermiñán
  4. Servicio General de Citometría, Centro de Investigacion del Cáncer (CIC), Universidad de Salamanca-CSIC, Salamanca, Spain

    • M L Sánchez


  1. Search for N C Gutiérrez in:

  2. Search for E M Ocio in:

  3. Search for J de las Rivas in:

  4. Search for P Maiso in:

  5. Search for M Delgado in:

  6. Search for E Fermiñán in:

  7. Search for M J Arcos in:

  8. Search for M L Sánchez in:

  9. Search for J M Hernández in:

  10. Search for J F San Miguel in:

Corresponding author

Correspondence to J F San Miguel.

Supplementary information

About this article

Publication history







Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)

Further reading

  • MARCKS regulates tonic and chronic active B cell receptor signaling

    • Chenguang Xu
    • , Yan Fang
    • , Zhiyong Yang
    • , Yukai Jing
    • , Yonghui Zhang
    • , Chaohong Liu
    •  & Wanli Liu

    Leukemia (2018)

  • Unraveling the heterogeneity of IgM monoclonal gammopathies: a gene mutational and gene expression study

    • Cristina Jiménez
    • , María Isabel Prieto-Conde
    • , María García-Álvarez
    • , Miguel Alcoceba
    • , Fernando Escalante
    • , María del Carmen Chillón
    • , Alfonso García de Coca
    • , Ana Balanzategui
    • , Alberto Cantalapiedra
    • , Carlos Aguilar
    • , Rocío Corral
    • , Tomás González-López
    • , Luis A. Marín
    • , Abelardo Bárez
    • , Noemí Puig
    • , Aránzazu García-Mateo
    • , Norma C. Gutiérrez
    • , María Eugenia Sarasquete
    • , Marcos González
    •  & Ramón García-Sanz

    Annals of Hematology (2018)

  • Two high-risk susceptibility loci at 6p25.3 and 14q32.13 for Waldenström macroglobulinemia

    • Mary L. McMaster
    • , Sonja I. Berndt
    • , Jianqing Zhang
    • , Susan L. Slager
    • , Shengchao Alfred Li
    • , Claire M. Vajdic
    • , Karin E. Smedby
    • , Huihuang Yan
    • , Brenda M. Birmann
    • , Elizabeth E. Brown
    • , Alex Smith
    • , Geffen Kleinstern
    • , Mervin M. Fansler
    • , Christine Mayr
    • , Bin Zhu
    • , Charles C. Chung
    • , Ju-Hyun Park
    • , Laurie Burdette
    • , Belynda D. Hicks
    • , Amy Hutchinson
    • , Lauren R. Teras
    • , Hans-Olov Adami
    • , Paige M. Bracci
    • , James McKay
    • , Alain Monnereau
    • , Brian K. Link
    • , Roel C. H. Vermeulen
    • , Stephen M. Ansell
    • , Ann Maria
    • , W. Ryan Diver
    • , Mads Melbye
    • , Akinyemi I. Ojesina
    • , Peter Kraft
    • , Paolo Boffetta
    • , Jacqueline Clavel
    • , Edward Giovannucci
    • , Caroline M. Besson
    • , Federico Canzian
    • , Ruth C. Travis
    • , Paolo Vineis
    • , Elisabete Weiderpass
    • , Rebecca Montalvan
    • , Zhaoming Wang
    • , Meredith Yeager
    • , Nikolaus Becker
    • , Yolanda Benavente
    • , Paul Brennan
    • , Lenka Foretova
    • , Marc Maynadie
    • , Alexandra Nieters
    • , Silvia de Sanjose
    • , Anthony Staines
    • , Lucia Conde
    • , Jacques Riby
    • , Bengt Glimelius
    • , Henrik Hjalgrim
    • , Nisha Pradhan
    • , Andrew L. Feldman
    • , Anne J. Novak
    • , Charles Lawrence
    • , Bryan A. Bassig
    • , Qing Lan
    • , Tongzhang Zheng
    • , Kari E. North
    • , Lesley F. Tinker
    • , Wendy Cozen
    • , Richard K. Severson
    • , Jonathan N. Hofmann
    • , Yawei Zhang
    • , Rebecca D. Jackson
    • , Lindsay M. Morton
    • , Mark P. Purdue
    • , Nilanjan Chatterjee
    • , Kenneth Offit
    • , James R. Cerhan
    • , Stephen J. Chanock
    • , Nathaniel Rothman
    • , Joseph Vijai
    • , Lynn R. Goldin
    • , Christine F. Skibola
    •  & Neil E. Caporaso

    Nature Communications (2018)

  • Exploring the nature of prediagnostic blood transcriptome markers of chronic lymphocytic leukemia by assessing their overlap with the transcriptome at the clinical stage

    • Jelle Vlaanderen
    • , Max Leenders
    • , Marc Chadeau-Hyam
    • , Lützen Portengen
    • , Soterios A. Kyrtopoulos
    • , Ingvar A. Bergdahl
    • , Ann-Sofie Johansson
    • , Dennie D.G.A.J. Hebels
    • , Theo M.C.M. de Kok
    • , Paolo Vineis
    •  & Roel C.H. Vermeulen

    BMC Genomics (2017)

  • DEPTOR maintains plasma cell differentiation and favorably affects prognosis in multiple myeloma

    • Dalia Quwaider
    • , Luis A. Corchete
    • , Irena Misiewicz-Krzeminska
    • , María E. Sarasquete
    • , José J. Pérez
    • , Patryk Krzeminski
    • , Noemí Puig
    • , María Victoria Mateos
    • , Ramón García-Sanz
    • , Ana B. Herrero
    •  & Norma C. Gutiérrez

    Journal of Hematology & Oncology (2017)