Unit: Immuno-Hematology and Immunopathology

Director: Guillaume DIGHIERO

The work from our laboratory was devoted to: 1) the study of the role of Immunoglobulin (Ig) genes mutational status associated to the classical prognostic parameters in refining prognostic prediction in CLL; 2) the study of the mechanisms involved in the low expression of the B cell receptor (BCR) and of the functional impairment of BCR signaling in this disease; 3) the study of AID (activation induced cytidine deaminase) expression in this disease and 4) the study of gene expression profiles in CLL.


Chronic lymphocytic leukemia (CLL), the most frequent form of leukemia in Western Countries, is characterized by a progressive accumulation of functionally incompetent, long-lived small mature monoclonal B lymphocytes, with a characteristic phenotype:CD5+,CD23+, low BCR expression resulting from low surface immunoglogulin and CD79b expression. It is far from uniform in presentation and clinical course. About one-third of patients never require treatment and have a long survival; in another third an initial indolent phase is followed by progression of the disease; the remaining third of patients have aggressive disease at the outset and need immediate treatment.

1) Study of the role of Ig gene mutational status associated to the classical prognostic parameters in refining prognostic prediction in CLL (Yuri Vasconcelos and G. Dighiero)

The development of the Rai and Binet staging systems has allowed the division of patients with chronic lymphocytic leukemia into three prognostic groups: good, intermediate and poor prognosis. However, neither of these staging systems is able to accurately predict evolution of these patients. An important long term study from our group (Dighiero et al, 1998) demonstrated that about half of patients included in the good prognosis group never evolve, do not require treatment and die of causes unrelated to CLL. The other half however, will evolve, require treatment and die of CLL related causes. Given that the absence of mutations in V genes is strongly linked to poor prognosis, we have undertaken a study on 146 patients aiming at defining whether the mutational profile of Ig was able to better predict prognosis in CLL patients when associated to the Binet's clinical staging. Our results demonstrated that both the Binet's clinical staging and the mutational pattern of Ig genes are strong, complementary and independent prognostic factors. The addition of both parameters results in a high improvement of prognosis assessment in this disease (Vasconcelos et al, 2003).

1) What are the mechanisms underlying underexpression of the BCR and impaired functional response in CLL B cells? (Responsibles: B. Payelle-Brogard and F. Vuillier in collaboration with C. Magnac G. Dumas).

The BCR consists of two heterodimers CD79a/CD79b bound non-covalently with the surface membrane Ig (SIg). In CLL, both SIg and CD79b have been reported to be consistently under-expressed, though the mechanisms accounting for this phenomenon have not been elucidated. In addition, CLL B cells display very poor signaling upon stimulation through the BCR. Since anergic B cells have been demonstrated to express low levels of the BCR and to poorly signal upon BCR stimulation, CLL B cells may correspond to malignant transformation of an anergic B cell.

In a previous work, we have studied the complete sequence of the B29 gene, encoding for CD79b, in 20 different familial CLL patients and demonstrated the absence of genetic abnormalities underlying the low surface expression of the CD79b. Since genetic alterations could not account for under-expression of the BCR in CLL, we have undertaken transcriptional and post-transcriptional analysis of this gene. Our results demonstrate that: 1) there is no major defect in transcriptional expression of the BCR (Payelle-Brogard et al, 1999). 2) Analysis of intracellular expression of the BCR components showed adequate synthesis, despite consistent low membrane expression of the receptor. 3) Neither a genetic defect in the transmembrane domain of SIg which it associates with CD79a/CD79b nor a genetic abnormality in the chaperone protein calnexin involved in folding and assembly of the BCR were found. 4) In contrast a defect in the processing of nascent IgM, demonstrated by its failure to become resistant to endoglycosidase H treatment was found (Payelle-Brogard et al, 2002 and 2003). This defect indicates that they are retained in the endoplasmic reticulum. 5) In addition, a constant defect in assembly of Ig M and CD79b chains of the BCR could be demonstrated in CLL B cells, by using confocal microscopy techniques, which allowed to demonstrate its retention in the Reticulum Endoplasmic compartment and its inability to progress to the cell surface. These findings demonstrate that a post-transcriptional defect located at the BCR intracellular assembly and/or trafficking levels could be involved in the low expression of the BCR in B-CLL and are reminiscent to those previously reported in a transgenic murine model of anergic B cells. This project led by B. Payelle-Brogard has received the scientific and financial support of the Fondation contre la Leucémie.

Figure illustrate this defect: whereas normal B cells succeed to merge the µ (red) and the CD79b (green) molecules at the membrane (yellow color), this is very poorly achieved by CLL B cells.

Recent advances in membrane biology have led to the identification of glycosphingolipid and cholesterol-rich plasma membrane microdomains, or lipid rafts, that have been proposed to function as platforms for both signal transduction and membrane trafficking. To better define the functional abnormality of CLL BCR, we have decided to study the traffic of the BCR molecules in CLL B cells, upon stimulation by the BCR pathway. Following the proteomic characterization of lipid rafts in CLL previous to any stimulation, we will proceed in a second step, to the study of the traffic of the different components of BCR following stimulation with anti-µ antibodies. This project led by F. Vuillier, has received the scientific and financial support of the "Fondation contre le Leucémie".

2) AID expression in CLL (Responsibles: G. Dighiero with P. Oppezzo and G. Dumas).

After rearrangement, Ig variable genes are diversified by somatic hypermutation (SHM), while the effector function of the constant domain is modified by class switch recombination (CSR). These processes depend on activation induced cytidine deaminase (AID), a putative RNA-editing enzyme, expressed in B-cells from secondary lymphoid organs upon CD40 ligand (CD40L) stimulation. Since the absence of AID expression in one form of the hyper-IgM syndrome in humans and in AID targeted mice, abolishes both CSR and SHM, this protein is supposed to play a major role in both processes. Honjo and colleagues postulated that this enzyme might be involved in them, through the production of similar or identical molecular substrates for RNA editing. On the other hand, recent reports suggest that SHM and CSR could all be initiated by the direct action of AID on DNA through a DNA-deamination mechanism of antibody diversification. However, at present its precise mode of action remains unknown.

Since some CLL B-cells can undergo CSR without somatic mutations, they may constitute a useful model to dissect AID function. In this work, we have studied AID expression, the presence of mutations in pre-switch µ DNA region, CSR and SHM in 65 CLL patients. Our results demonstrate that: 1) CLL B-cells can constitutively express AID, which expression is associated with mutations in the pre-switch region and clonally related isotype-switched transcripts. Interestingly, this phenomenon is almost exclusively restricted to CLL B cells exhibiting Ig V genes in unmutated form and 2) in CLL without constitutive AID expression, its induction upon stimulation results in pre-switch mutations and CSR process. However, despite that either constitutive or induced AID expression were able to induce somatic mutations in the pre-switch µ region and to induce active CSR, somatic hypermutations in the variable domain could never be obtained. Overall, our results show a dissociation between SHM and CSR in CLL and suggest that, at least in this disease, AID would require additional help to carry out SHM process.

The important question addressed is whether "ectopic" expression of AID in these CLL patients, is recruiting the same AID partners as in normal germinal center B cells . To assess these questions we propose to: a) Identify putative protein/s factors partners of AID protein by using the "Two Hybrid system and b) Characterize structurally and functionally AID and its putative partners through a collaboration with the Structural Biochemistry Laboratory of the Pasteur Institute. For this we are attempting to produce AID in its native form and to solve the tri-dimensional structure of AID alone and in complex with its putative partners.

3) Micro-array studies of gene profile expression in different CLL prognostic groups (Responsibles: G. Dighiero with Yuri Vasconcelos and Frédéric Davi (Hôpital Pitié-Salpêtrière), Florence Cymbalista (Hôtel-Dieu de Paris) and Xavier Troussard (CHU de Caen) on behalf of the French Cooperative Group on CLL.

Chronic lymphocytic leukemia (CLL) is a disease with a highly variable clinical course. During recent years the development of biologic markers has allowed a better definition of prognosis in CLL. Different serological, phenotypic and cytogenetic markers have been proposed to help better prediction of survival and progression of this disease. So far, the best biological parameter capable of predicting its evolution is the mutational status of the Ig genes, with Ig mutated cases having a far better prognostic than those with unmutated genes. This may suggest that there are two types of CLL: one arises from relatively less differentiated (immunologically naive) B cells with unmutated heavy chain genes, and displays poor prognosis; the other evolves from more differentiated B cells (memory B cells) with somatically mutated heavy chain genes, and has a good prognosis. However, recent data derived from gene expression profiling analysis failed to clearly distinguish unmutated and mutated cases and favor the view that all cases of CLL have a common cell origin and/or a common mechanism of malignant transformation. We have used the microarray technology to determine whether gene expression profile could distinguish these two groups of CLL.

Eighteen cases of CLL, including 9 indolent forms of the disease ( Binet stage A; i.e. at least 5 years without treatment and any evolution) and mutated Ig genes and 9 cases with stage B or C aggressive disease and unmutated Ig genes, were studied. For these latter cases, only cells collected before therapy were analyzed. Leukemic cells were purified by negative selection, using anti-CD3, -CD14, -CD16, -CD56 antibodies, yielding > 98% pure CLL cell populations. RNAs were extracted and gene expression analyzed using the Affymetrix human U133 Genechip with 22283 probe sets.

In agreement with previous reports indicating that Ig mutated and unmutated cases have globally the same gene expression pattern, a supervised statistical analysis showed that only 85 genes were differentially expressed by a factor >2 between the two groups of CLL. Among these 85 genes, our results showed homogeneous overexpression of ZAP-70, lipoprotein lipase (LPL), BCL-7A, VLA-4, dystrophin (DMD) and gravin (AKAP12) in the aggressive unmutated cases, while stable mutated cases overexpressed T cell function-related genes HLA-DQB1, CTLA-4, TCF7, ADAMDEC1, as well as Wnt-3 and JUN oncogene. However, in contrast to previous reports, a non supervised hierarchical clustering analysis could clearly separate the stable mutated group from the aggressive unmutated one, except for one case. Microarray results could subsequently be confirmed by using Real Time PCR.

These results show that gene expression profiling can distinguish CLL cases with a stable evolution and mutated Ig genes from those with unmutated Ig genes and progressive disease. Thus, monitoring the expression of a very limited number of genes might suffice to identify patients displaying an indolent disease from patients exhibiting an aggressive one.

The difficulty to extend the study of Ig genes to routine practice has led to a relentless hunt for surrogate markers. Thus, our results derived from microarrays studies indicating a higher expression level of LPL and ZAP-70 genes among unmutated CLL cases, and of ADAM29 among mutated CLLS, led us to investigate which of these genes (individually or in combination) could better predict the IgVH mutational status in 130 CLL patients. LPL and ADAM29 expressions were assessed by RQ-PCR, while ZAP-70 was analyzed by flow cytometry. Overall, LPL/ADAM29 ratio (L/A) provided the best predictive values for IgVH mutational status in the whole series. When compared to ZAP-70, L/A exhibited similar predictive ability. Although results with these markers were found discordant in about 20% of cases, double positive (ZAP+ L/A+) and negative (ZAP- L/A-) cases corresponded to respectively UM and MT patients with almost 100% of accuracy. Thus, the simultaneous study of L/A and ZAP-70 is capable to dispense IgVH sequencing in 80% of CLL patients.

Keywords: Autoimmunity, Chronic Lymphocytic Leukemia, Immunopathology

Activity Reports 2003 - Institut Pasteur

Page Top research Institut Pasteur homepage

If you have problems with this Web page, please write to rescom@pasteur.fr