mRNA Profiling of CLL Cells Derived from the Blood, Bone Marrow and Lymph Node

Background

Chemo-immunotherapy remains the backbone of therapy for patients with chronic lymphocytic leukaemia (CLL), with evidence pointing towards long term remission and even cure in patients with mutated IGHV status receiving this therapy frontline (1). However, relapse is common especially in those harbouring abnormalities in TP53 or ATM, unmutated IGHV status or a complex karyotype (2).

It has become increasingly apparent that the bone marrow (BM) and lymph node (LN) play important roles in promoting the survival and proliferation of CLL cells. Signalling pathways triggered by interactions within these niches, such as the B cell receptor (BCR) pathway, and intracellular proteins such as Bcl-2 are vitally important in the biology of CLL. Novel therapeutic agents, such as ibrutinib, which target components of the BCR pathway, and the Bcl-2 inhibitor venetoclax, have demonstrated the potential of targeted therapies in CLL (3, 4). Novel therapeutic approaches must target the proliferative, drug-resistant compartments of disease within these microenvironments.

The NanoString® nCounter platform enables mRNA profiling of archival samples, including formalin-fixed, paraffin-embedded tissue (FFPE). We have previously demonstrated the utility of this technology by comparing the mRNA expression profile of CLL cells derived from the peripheral blood (PB), archival BM and LN tissue as well as PB-derived CLL cells following in vitro co-culture with a human stromal cell line under either normoxic or hypoxic conditions. Here we present an update on our previous work with increased sample numbers in each of the tissues or culture conditions.

Methods

RNA was extracted from FFPE BM trephines (n = 5) and LN sections (n = 5), using the QIAGEN RNeasy FFPE Kit. All biopsies analysed were comprised of > 80 % lymphocytes, as determined by microscopic review.

RNA from PBMC fractions (n = 5) was isolated either immediately or following co-culture with HS5 stromal cells for 24 h under normoxic (n = 5) or hypoxic (n = 5) conditions using the QIAGEN RNeasy Mini Kit.

RNA from all preparations was quantified using a NanoDrop™ spectrophotometer. A total of 200ng of FFPE-derived RNA and 100ng of PBMC-derived RNA was analysed per sample on the NanoString® platform using a 260 gene panel.

Three-fold changes in mRNA expression were considered significant.

Results

Of the 260 genes profiled, 89 were upregulated in the BM samples and 52 in the LN samples compared to expression in PB-derived CLL cells. Changes were seen in genes encoding for proteins involved in chemotaxis (CXCL9), the regulation of apoptosis (BCL2L1), surface receptors (FLT3) and genes associated with intracellular signalling, metabolism and cell division. 35 genes were downregulated in the LN samples and 31 in the BM samples. These changes were seen in genes coding for surface receptors (ROR1 and CXCR4), genes coding for intracellular signalling proteins (RAF1) and genes coding for transcription factors (JUN and FOS).

Co-culture of PBMCs with HS5 cells induced similar changes to those observed in our comparison of the PBMCs and BM samples; genes coding for 61.5% and 50.0% of the mRNA expression changes observed in the LN were observed in PBMCs cultured under normoxic and hypoxic conditions respectively. A similar comparison of the BM samples identified concordant changes in expression of 46.5% and 39.2% of genes under normoxic and hypoxic conditions respectively. Importantly, changes observed in genes coding for the anti-apoptotic protein MCL1, the surface receptors CXCR4 and ROR1 and the transcription factors ATF, FOS and JUN were consistent across samples from LN, BM and the in vitro model.

In summary, we have utilised the NanoString® nCounter platform to profile PB, BM and LN-derived CLL cells and have identified panels of genes that are either up or down-regulated in cells derived from these microenvironments. Furthermore, the high concordance between RNA changes in the in vitro model and the primary tissue suggest the HS5 co-culture system mimics aspects of the tumour microenvironment. These data provide a better understanding of how CLL cells populate and proliferate in the tumour microenvironment and may lead to novel therapeutic strategies.