Saturday, August 06, 2011

CD38 + CD49d: a complex situation in CLL

My former colleagues at Kings College Hospital have looked at CD38 and CD49d in CLL and their results are published in last month's BJHaem.

This study has shown that the level of CD38 and CD49d expression are strongly correlated in CLL, both within and between patients. They provide convincing evidence that the CLL cell surface membrane contains a preformed complex in which CD38 is physically associated with CD49d, CD44 and MMP9.

CD38 is a membrane protein whose level of expression varies during B cell differentiation. Its expression on individual CLL cells changes during the cell's circulation and sometimes during the disease. Ligation, by agonistic antibody or fibroblasts transfected with its ligand CD31 leads to a calcium flux, increased proliferation and improved viability. Thus in-vivo interactions between CD38-positive CLL cells and endothelial CD31 generate signals that promote tumour viability and proliferation, which may partly explain why high expression of CD38 by tumour cells predicts for adverse outcome in CLL.

In addition to its role in promoting survival and proliferation, CD38 is involved in the migration of CLL cells into the tissues. Expression of CD38 and ZAP-70 identifies CLL cells with enhanced migration to the chemokine CXCL12 and a functional link between the two has thus been proposed.

The migration of lymphocytes into tissues is begun by chemokine-induced activation of integrins, which increases their affinity for endothelial ligands such as ICAM1 and VCAM1, resulting in firm adhesion and migration though the endothelial barrier. In some cells integrins are arranged in supramolecular complexes termed podosomes that are linked to the cytoskeleton and contain other proteins involved in tissue invasion such as matrix metalloproteinases. Recent evidence suggests that CLL cells, but not normal B cells, possess podosomes containing 4β1 integrin, VLA4 – of which CD49d is the subunit - , a splice variant of CD44 and MMP-9. Interestingly, the level of expression of CD44v, CD49d and MMP-9 have all been linked to the presence of bulky lymphadenopathy and/or adverse outcome in CLL.

In addition to their role in tissue invasion, interactions between VLA4 and its ligand VCAM1 have also been shown to promote the viability of CLL cells and high levels of CD49d are associated with increased numbers of VCAM1 expressing microvessels within lymph nodes. These observations are paralleled by work from Kings showing that CD38-positive cases of CLL have increased numbers of microvessels in lymph nodes and a more recent study in which interactions with vascular endothelial cells were shown to inhibit CLL cell apoptosis through NFκB-mediated regulation of Bcl-2 family proteins. Incubation with endothelial cells also resulted in NFκB-dependent up-regulation of CD38 and CD49d at both the protein and transcription level.

It is clear that CD38 and CD49d share numerous features in CLL. They are co-ordinately regulated by NFκB activation, mediate pro-survival signalling and migration into the tissues, are associated with increased lymph node vascularity, predict adverse outcome and their expression levels are strongly correlated both between and within individual patients with CLL.

The findings of this study provide a structural basis for these striking observations. Given the known properties of the individual molecules and their functional overlap it seems very likely that the CD38/CD49d/MMP9/CD44 complex plays a role in migration into the tissues and in pro-survival signalling. Accumulation of leukaemic cells in the tissues is responsible for many of the clinical features of CLL and defines more advanced and progressive disease with an adverse outcome. Tissue invasion also exposes the tumour cells to other components of the leukaemic microenvironment such as T cells, nurse-like cells and other stromal elements that promote expansion of the leukaemic clone. The identification of the CD38/CD49d complex suggests a biological basis for the correlation between expression of these molecules and prognosis. As the structure is not present in normal B cells, it also presents a novel leukaemia-specific target for the therapy of this common disorder.

1 comment:

Taiji Scholar said...

The complex formation of CD38, CD49d, CD44 and MMP-9 shown by Buggins and colleagues is very interesting. And given the known properties of the individual molecules and their functional overlap it seems very likely that the CD38/CD49d/ MMP9/CD44 complex plays a role in migration into the tissues and in pro-survival signaling (Buggins et al., 2011). Recently, we showed that circulating B-cell chronic lymphocytic leukemia cells display impaired migration to lymph nodes and bone marrow (Hartmann et al., 2009). Nevertheless, only residual VLA-4 positive CLL subsets could reenter BM, whereas both normal and CLL cells homed to the mice spleen in an LFA-1-independent and VLA-4-independent manner. Because VLA-4 and CD38 expression have been described to be highly associated in CLL (Gattei et al., 2008; Zucchetto et al., 2009), we sought to determine their functional consequences in the context of supportive bone marrow (BM) niches. For this we employed in vivo adoptive transfers of human leukemic cells into immunodeficient mice, in vitro flow cytometrical phenotyping of peripheral blood (PB) - and BM-derived CLL cells, and stromal cell co-culture experiments. Moreover the extent of BM infiltration in CLL patients was correlated to the VLA-4 and CD38 status. Our data suggest that VLA-4 but not CD38 expression plays a prominent role in the homing of high-risk CLL cells to supportive BM niches and in human BM infiltration but only a limited role in supportive CLL cell-stromal cell interactions in vitro.

Among other things we found out that:
• VLA-4 and CD38 expression are associated in CLL cells. Percent VLA-4+ correlated to percent CD38+ CLL cells of 144 CLL patients (Spearman’s Rho = .6293, P,.0001).
• VLA-4, and not CD38, is essential for BM homing.
• VLA-4 and CD38 are higher expressed in BM than in PB.
• Ki-67 expression in BM aspirates is higher in CLL samples from high-risk patients, and VLA-4 negative and CD38 positive subclones express higher levels of Ki-67 in BM.
• VLA-4 and CD38 high-risk CLL samples show higher spontaneous apoptosis rates.
• CLL cells from patients with different risk-groups are equally protected by stromal cells but cells from high-risk patients are able to adhere in higher numbers.

In summary, Brachtl et al., 2011 (Plos One) suggest that VLA-4, rather than CD38, is mainly responsible for the recirculation of high-risk CLL cells into BM and for high BM infiltration observed in CLL patients. VLA-4 seems to be necessary to position those cells that are highly dependent on accessory survival signals at the appropriate supportive niche.