ALBERT

Description: Ig mutation status discriminates aggressive from more indolent chronic lymphocytic leukemia (CLL). Several gene expression studies distinguished CLL from other leukemia but failed to discriminate these two types of patients. Nevertheless, differentially expressed genes including zap70 were identified and further validated as an important prognostic factor. Zap70 is a normal component from the T cell receptor pathway, and may have a critical role in the B cell receptor (BCR) signaling pathway in CLL. We questioned whether zap70 might alter response to BCR signaling in CLL cells and explain the difference in biologic characteristics of this disease. To examine this we crosslinked the BCR in CLL cells from 12 patients and B cells from 6 healthy donors. We examined the response to BCR stimulation on cell proliferation, apoptosis and gene expression profile. Here we analyzed the temporal gene expression after BCR crosslinking at different time points, in addition to the basal level of gene expression. Based on the results of a pilot study examining multiple different time points, four time points were analyzed in all cases ranging from 60 min to 390 min and gene expression of stimulated (S) and unstimulated (US) CLL and control cells analyzed for a total of 170 HU133 plus2.0 DNA-chips. We defined the expression data from the logarithmic ratio log(S/US) for each time point. In keeping with previous findings, the basal gene expression distinguished CLL samples from healthy B-cells, but did not distinguish mutated from unmutated patients. We defined the linear combination first for the initial time point and then for all four time points to score for expression over time. This temporal gene profile of BCR engagement was based on filtering of 2850 genes that now distinguish healthy B-cells from CLL-cells, but is now capable also of distinguishing a specific BCR signature of a group of patients. Of note, this distinguished patients based less upon Ig gene mutation status but on zap70 protein level, with a mean protein expression difference of 9 fold between these two groups. This more aggressive pattern reveals changes in gene expression in important genetic pathways associated with BCR crosslinking, including JUN and NFkB. Stimulating a crucial pathway for CLL biology identified temporal gene expression pathways linked with the aggressiveness of this leukemia. This suggests that a fuller understanding of the biologic differences in leukemia cells may be identified by a more dynamic approach to gene profiling by examination of response to specific stimuli, rather than examination at the basal level. Based upon these results, ongoing studies examine the biologic impact of gene silencing of zap70 using RNAi on BCR stimulation in CLL cells.

Description: Inferring a temporal gene network from a crucial signaling pathway in leukemic cells is a leading problem in oncology. We built a temporal transcriptional network of B-cell receptor (BCR) crosslinking in CLL and healthy B-cells, a critical step to understand the dynamics of BCR gene expression and to understand gene regulation at a system level. CLL cells have defects in apoptosis and the BCR pathway appears crucial in this process, leading to differential signaling and cell response according to the Ig gene mutational status and zap70 expression. We built this network by analysis of the gene expression profile after BCR crosslinking in six mutated (M) and six unmutated (UM) CLL cells and six healthy B-cells. After a pilot study examining multiple time points, total RNA was purified at four time points (60 to 390 min) from stimulated (S) and unstimulated (US) cells, for a total of 170 HU133plus2.0 DNA-chips analyzed. The logarithmic ratio data log(S/US) for each time point and each patient and the linear combination for four time points were analyzed to score expression over time. This temporal clustering discriminates healthy B-cells from CLL-cells, but now also distinguishes two groups of patients, one mainly UM with higher zap70 protein levels. BCR engagement induces different gene expression for this group of aggressive CLL. We built a temporal model of gene expression for these three groups using two iterative steps. The first step is a modified K-means clustering approach using log(S/US) of temporal gene expression. This results in groups of genes with common temporal structure whose expression exhibits significant differences after BCR stimulation. The number of considered genes were then reduced, keeping a small number with the largest increase in expression within each group. Most of these genes are important in BCR transcription including JUN, DUSP1 and NFkB and most first wave genes are transcription factors. The second step is to construct predictive models of gene expression, considering only causal linear predictive models. Specifically the expression of an output gene at each time is predicted using a weighted linear combination of the expression of another gene at past time points. The method groups pairs of genes by common predictive model. While paired genes may reside in different initial clusters, upon convergence they are clustered by which predictive models they use. The procedure first assigns random pairings of genes and then we iterate between two steps, computing the best predictive model using a regularized least squares algorithm emphasizing sparse models and optimal gene pairings using a modified Hungarian bipartite graph matching approach. In practice the method converges in a small number of iterations. To refine and test this model we use RNAi to silence genes in the first wave of transcription after BCR stimulation and study the impact on the model. From the global gene regulatory network, we aim to predict the minimal number of gene to silence to influence the global structure of the BCR regulatory network. Influencing the transcriptional structure of aggressive CLL toward that of indolent CLL and healthy B-cells is a first step to reprogram the transcriptional response of leukemic cells.

Description: Gene expression in cells is a dynamic process but is usually examined at a single time point. We used gene expression profiling over time to build temporal models of gene transcription after B-cell receptor (BCR) signaling in healthy and malignant B cells and chose this as a model since BCR cross-linking induces both cell proliferation and apoptosis, with increased apoptosis in chronic lymphocytic leukemia (CLL) compared to healthy B cells. To determine the basis for this, we examined the global temporal gene expression profile for BCR stimulation and developed a linear combination method to summarize the effect of BCR simulation over all the time points for all patients. Functional learning identified common early events in healthy B cells and CLL cells. Although healthy and malignant B cells share a common genetic pattern early after BCR signaling, a specific genetic program is engaged by the malignant cells at later time points after BCR stimulation. These findings identify the molecular basis for the different functional consequences of BCR cross-linking in healthy and malignant B cells. Analysis of gene expression profiling over time may be used to identify genes that might be rational targets to perturb these pathways.

Description: B-cell receptor (BCR) engagement is widely acknowledged to sustain aberrant cell behavior and uncontrolled monoclonal proliferation in chronic lymphocytic leukemia (CLL), as well as in most leukemias and lymphomas arising from mature B lymphocytes. The precise mechanisms by which BCR signaling controls neoplastic B-cell proliferation are ill characterized. In this work, primary leukemic cells of untreated patients at initial stage of CLL (Binet stage A / Rai 0) presenting biological characteristics of aggressive form of the disease (unmutated IGHV genes and ZAP70 protein expression) were studied. Proliferation of CLL cells was induced ex vivo in six CLL samples using anti-IgM, together with mandatory co-stimulating factors (CD40L, IL-4 and IL-21) (Schleiss, Sci Rep, 2019). Using this model, we generated a unique set of 108 transcriptional and proteomic profiles during four days after activation (9 points from T0 to 96 hours after stimulation). A total of 23,348 transcripts and 50,503 unique peptides, the latter corresponding to 4,664 unique proteins, were identified and quantified. Statistical analysis of genes and proteins expression patterns identified a structured proliferative signature. In unsupervised analysis, principal component analysis (PCA) representation of the whole transcriptional profile showed the temporality of the response. Moreover, unsupervised temporal gene expression analysis using iterative optimization revealed clusters of temporal patterns exhibiting structured expression modulations during the time course after cell stimulation. Although the overall proteomic response appeared less structured than the transcriptional one at early time points after stimulation, samples showed a tendency for segregation with respect of their proliferative response at the two later time points. Hierarchical clustering, aimed to search for correlations between the transcriptional profiles, confirmed the temporal organization of the samples, while proliferative and non-proliferative cells were still distinguishable. Also, we identified a CLL cell activation signature corresponding to 3,097 differentially expressed genes (DE) and 1,209 differentially-abundant (DA) proteins. This unique dataset provided a unique opportunity to model the proliferative program of CLL cells after BCR engagement. We used the reverse engineering approach based on regression and system of equations previous published (Vallat et al., PNAS 2013). Thus, a temporal label was assigned to each gene and protein, restricting its potential link to actors with later temporal labels. The temporal model inferred with the genes and proteins datasets of proliferating CLL cells resulted in a regulatory network composed of 2,167 genes and 1,074 proteins representing 2,848 unique symbols, among which 395 gene-protein pairs, connected by 53,131 oriented links. Different temporal layers of actors were identified. At the earliest time-point after cell activation a first group was identified, involving transcriptional repression, negative regulation of BCR signaling, apoptotic process actors and a second group involving G1/S transition, DNA-replication genes. Later, expression of G0/G1 switch genes and regulators of cell proliferation and differentiation were identified. In conclusion, using a large dataset of temporal transcriptional and proteomic measurements coupled with mathematical modelling, we were able to unravel the molecular program downstream the signaling cascade activated by the engagement of the BCR and triggering primary CLL cell proliferation. This program was proven to organize around a limited number of hub genes and proteins whose sequential commitment drives the cellular response leading to proliferation days after cell activation. These hubs represent potential candidates for the development of novel therapeutic strategies for the treatment of aggressive CLL. Disclosures No relevant conflicts of interest to declare.

Description: B cell chronic lymphocytic leukemia (CLL), the most prevalent B cell malignancy in adults, is characterized by expansion of monoclonal mature B lymphocytes. Despite advances in treatment, the disease remains incurable warranting further efforts to identify novel molecular targets in CLL. B cell receptor (BCR) signaling contributes to apoptosis resistance in CLL limiting the efficacy of therapeutic approaches. In this study we investigated the expression of spleen tyrosine kinase (SYK), a key component of the BCR signaling pathway, in CLL and its role in apoptosis. Gene expression profiling identified enhanced expression of SYK and downstream pathways in CLL compared to healthy B cells. Immunoblotting showed increased expression and phosphorylation of SYK, PLCγ2, STAT3, and ERK1/2 in CLL compared to healthy B cells suggesting enhanced activation of these mediators in CLL. Separate analyses according to prognostic parameters revealed 1.8-fold higher SYK protein level in unmutated compared to mutated CLL cells determined by densitometric analysis (n=22, p=0.0031). These findings may well explain the higher BCR signaling capacity in the unmutated CLL subset. Various SYK inhibitors (piceatannol, curcumin, SYK Inhibitor II, and IV (Calbiochem)) reduced phosphorylation of the SYK downstream targets PLCγ2, STAT3, and ERK1/2 in a time- and dose-dependent manner and induced apoptosis in the CLL cell lines Mec-1 and EHEB and primary CLL cells. SYK Inhibitor II revealed highest cytotoxic effects on primary CLL cells, but did not significantly reduce the viability of healthy B cells. Thus, apoptotic effects of this inhibitor were analyzed in a larger cohort of patient samples along with the well-established SYK inhibitor R406 (Rigel Inc.). After 48 hour treatment, relative viability of CLL cells was reduced to 76% and 44% for SYK Inhibitor II (4 mM and 10 mM) and to 66% for R406 (4 mM), respectively (n=38, p

Description: Chronic lymphocytic leukemia (CLL) is characterized by a highly variable clinical course with 2 extreme subsets: indolent, ZAP70− and mutated immunoglobulin heavy chain gene (M-CLL); and aggressive, ZAP70+ and unmutated immunoglobulin heavy chain (UM-CLL). Given the long-term suspicion of antigenic stimulation as a primum movens in the disease, the role of the B-cell receptor has been extensively studied in various experimental settings; albeit scarcely in a comparative dynamic proteomic approach. Here we use a quantitative 2-dimensional fluorescence difference gel electrophoresis technology to compare 48 proteomic profiles of the 2 CLL subsets before and after anti-IgM ligation. Differentially expressed proteins were subsequently identified by mass spectrometry. We show that unstimulated M- and UM-CLL cells display distinct proteomic profiles. Furthermore, anti-IgM stimulation induces a specific proteomic response, more pronounced in the more aggressive CLL. Statistical analyses demonstrate several significant protein variations according to stimulation conditions. Finally, we identify an intermediate form of M-CLL cells, with an indolent profile (ZAP70−) but sharing aggressive proteomic profiles alike UM-CLL cells. Collectively, this first quantitative and dynamic proteome analysis of CLL further dissects the complex molecular pathway after B-cell receptor stimulation and depicts distinct proteomic profiles, which could lead to novel molecular stratification of the disease.

Description: Motivation With the growth of big data, variable selection has become one of the critical challenges in statistics. Although many methods have been proposed in the literature their performance in terms of recall (sensitivity) and precision (PPV) is limited in a context where the number of variables by far exceeds the number of observations or in a highly correlated setting. Results In this article, we propose a general algorithm which improves the precision of any existing variable selection method. This algorithm is based on highly intensive simulations and takes into account the correlation structure of the data. Our algorithm can either produce a confidence index for variable selection or be used in an experimental design planning perspective. We demonstrate the performance of our algorithm on both simulated and real data. We then apply it in two different ways to improve biological network reverse-engineering. Availability Code is available as the SelectBoost package on the CRAN, https://cran.r-project.org/package=SelectBoost. Some network reverse-engineering functionalities are available in the Patterns CRAN package, https://cran.r-project.org/package=Patterns. Supplementary information Supplementary data are available at Bioinformatics online.

Description: Abnormal expression of the tyrosine kinase ZAP-70 by tumoral B cells in chronic lymphocytic leukemia (CLL) is associated with bad prognosis, related to B cell receptor (BCR) hypersignalling, clonal expansion and autoimmune cytopenia (AIC) occurrence, these latter being mostly induced by polyclonal IgG from the residual non tumoral B cells. We previously shown that ZAP-70 is expressed by these non tumoral B cells in CLL, positively associated with its expression in CLL B cells and with AIC occurrence (Ghergus et al. Poster ASH 2017). Here, we show for the first time a potential role of ZAP-70 expression in tolerance breakdown in CLL and in an original knock in mouse model overexpressing ZAP-70 conditionally in B cells. First, to assess a potential molecular link between ZAP-70+ CLL and non tumoral B cells, an analysis of their BCR repertoire has performed on FACS-sorting CD19+CD5-IgM-IgD- (non tumoral) and CD19+CD5+IgM-IgD- (tumoral) single B cells from blood samples of CLL patients with AIC. ZAP-70 positivity was screened by RT-PCR, and variable regions of heavy (IGVH) and light (IGVK/VL) immunoglobulin genes amplified by RT-PCR on ZAP-70+ cells. To date, analysis of 24 BCR sequences from 7 patients showed that non tumoral ZAP-70+ B cells were polyclonal, without stereotypy, using different V(D)J and CDR3 in comparison with those of the corresponding CLL B cells. IGVH of non tumoral ZAP-70+ B were mostly mutated, of replacement type, suggesting antigenic contact, contrary to CLL B cells. To determine potential autoreactivity of the non-tumoral ZAP-70+ B cells, IGVH and corresponding IGVK/VL were amplified for production of recombinant antibodies (rAb). To date, among 17 rAB from 7 different patients, 2/13 (15.4%) have an antinuclear autoreactivity on HEp-2 cells and 4/17 (23.5%) were polyreactive on ELISA (DNA, lipopolysaccharide, insulin), compared respectively to 6% and 4,3% of control B cells (Wardemann et al., Science 2003). Production of 7 additional rAb and tests for anti-erythrocytes and anti-platelets reactivity are in process. To study functional consequences of early ZAP-70 expression in B cells in vivo, we generated a knock in Zap-70+/Mb1-Cre+mouse model (KI ZAP), to induce conditional expression of ZAP-70 in the B cell compartment from the proB stage, with KI Zap-70+/Mb1-Cre-mice as controls (CTRL). The ZAP-70 mRNAs levels in B cells from KI ZAP mice were on average 20 times higher than that in CTRL B cells. Up to 20 months-old, KI ZAP mice did not develop signs of lymphoproliferation. KI ZAP mice had hypo-IgG since 16 weeks-old (p

Description: Unraveling the molecular basis for the different functional consequences of BCR cross-linking is essential to understand the leukemogenesis process of CLL. There is increasing evidence that an antigen driven process is crucial in CLL proliferation, based on restriction of the IgVH repertory as well as shared antigen-binding motifs used by mutated or unmutated CLL B-cells. BCR activation leads to a signaling cascade reinforced in the more aggressive CLL form by the ZAP70 protein tyrosine kinase and constitutive phosphorylation of HS1 protein. We have previously shown, using gene expression profiling over time, that crosslinking of the BCR induces a specific temporal gene expression program in the leukemic cells, leading to a complex balance disorder between proliferation and cell death. These results demonstrated a core BCR gene expression shared among all B cell (and across species). Furthermore, each cell category (healthy B-cells, indolent and aggressive CLL cells) also showed a specific temporal gene expression after BCR cross-linking. This analysis has revealed complex expression disorder of multiple genes coding proteins involved in proliferation and death regulation (Vallat et al. Blood 07). To further investigate the functional consequences of BCR cross-linking we also examined modulation of the BCR proteome, since a dynamic description of the BCR functional consequences in CLL cells would not be complete without parallel temporal proteomic analysis. We now have cross-linked the BCR in freshly isolated B-cells from CLL patients. We have isolated total proteins at 2 early, one intermediate and a later time point 30 min later than the time points selected in the previous mRNA experiments. At each analysis time point both stimulated and control unstimulated cells were examined. A 2D electrophoresis analysis (pH 3–10, 150–20 kDa) revealed between 1100 and 1800 (mean 1400) polypetidic spots at each time point after BCR engagement. A total of 600 different polypeptides show a specific pattern of expression over time after stimulation, up- (450 polypeptides) or down-regulated (150 polypeptides). We are currently identifying these different proteins of interest by mass spectrometry (Maldi-TOF and LC-MS/MS) and learning these results in parallel of the previous genomic results. Preliminary results show the complementarity of the two analyses. Many of the proteins identified correspond to changes in mRNA belonging to the temporal gene program previously described after BCR engagement, thereby validating the gene expression data at the protein level, but some proteins show a specific temporal pattern of expression whereas their corresponding coding mRNA were not present in the temporal gene expression. Such complementarities in this dynamical approach allow us to refine the molecular basis for the functional BCR cross-linking consequences in CLL and to identify putative therapeutic targets for intervention in this disease. Ongoing experiments are examining the consequences of intervention to alter expression of critical components within the functional consequences of BCR cross-linking in CLL, and potentially the consequences of antigen driven leukemogenesis.