SEARCH

SEARCH BY CITATION

Keywords:

  • chronic lymphocytic leukaemia;
  • microenvironment;
  • proliferation;
  • survival;
  • co-culture systems

Summary

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

Interactions in the tumour microenvironment can promote chronic lymphocytic leukaemia (CLL) cell survival, proliferation and drug resistance. A detailed comparison of three co-culture systems designed to mimic the CLL lymph node and vascular microenvironments were performed; two were mouse fibroblast cell lines transfected with human CD40LG or CD31 and the third was a human microvascular endothelial cell line, HMEC-1. All three co-culture systems markedly enhanced CLL cell survival and induced a consistent change in CLL cell phenotype, characterized by increased expression of CD38, CD69, CD44 and ITGA4 (CD49d); this phenotype was absent following co-culture on untransfected mouse fibroblasts. In contrast to HMEC-1 cells, the CD40LG and CD31-expressing fibroblasts also induced ZAP70 expression and marked CLL cell proliferation as evidenced by carboxyfluorescein succinimidyl ester labelling and increased Ki-67 expression. Taken together, our data show that co-culture on different stroma induced a remarkably similar activation phenotype in CLL cells but only the CD40LG and CD31-expressing fibroblasts increased ZAP70 expression and CLL cell proliferation, indicating that ZAP70 may play a critical role in this process. This comparative study reveals a number of striking similarities between the co-culture systems tested but also highlights important differences that should be considered when selecting which system to use for in-vitro investigations.

Chronic lymphocytic leukaemia (CLL) is a highly heterogenous disease with a very variable clinical outcome. It is now appreciated that it is a highly proliferative disorder with significant tumour cell turnover every day (Messmer et al, 2005) yet despite this, primary CLL cells are notoriously difficult to culture in-vitro and drug testing models are hindered by the poor survival of these cells. This raises the question as to what signals are provided in-vivo that enables these cells to survive and proliferate. Recent work has highlighted the role of accessory cells within the tumour microenvironment in the survival and induction of proliferation in these malignant cells (Patten et al, 2008; Buggins et al, 2010; Ferretti et al, 2011; Herishanu et al, 2011). As a result of this, a variety of co-culture systems have been developed to mimic the tumour microenvironment. However, to date no study has systematically characterized the effects of these co-culture systems or provided a direct head to head comparison in terms of CLL cell survival, proliferation and phenotype (Patten et al, 2008; Plander et al, 2009; Buggins et al, 2010; Coscia et al, 2011; Ferretti et al, 2011; Pepper et al, 2011).

Tumour proliferation is believed to mainly occur in pseudofollicles, which develop in the lymph nodes, bone marrow and spleen (Schmid & Isaacson, 1994; Patten et al, 2008). Interactions with T-lymphocytes, the microvasculature, soluble factors and other stromal elements are all thought to play a major role in the survival and expansion of the tumour cells. In keeping with this concept, lymph node biopsies from CLL patients with aggressive disease contain activated T-lymphocytes. We have previously demonstrated that proliferating CLL cells co-localize with activated CD4+ T-cells in the lymph node (Patten et al, 2008) and ligation of CD40 on CLL cells by its ligand CD40LG (expressed by activated T-lymphocytes) has recently been shown to induce differential responses in terms of up-regulation of surface markers and induction of chemokines (Scielzo et al, 2011). Lymph nodes of CLL patients with aggressive disease also contain large numbers of CD31+ vessels (Patten et al, 2008) and CD31+ nurse-like cells (Deaglio et al, 2005). In a recent study, we demonstrated that interactions with endothelial cells can promote the survival of CLL cells (Buggins et al, 2010) and induce the expression of CD38 and ITGA4 (CD49d) on the tumour cells; both of these molecules are associated with aggressive disease and inferior clinical outcome (Damle et al, 1999; Shanafelt et al, 2008; Majid et al, 2011).

These studies indicate that it is interactions with accessory cells, such as activated T-lymphocytes and endothelial cells, that play a role in sustaining CLL cells in-vivo. Therefore there is a need to model these interactions in-vitro in order to define the critical molecular interactions that promote survival and proliferation in this disease. This study compared three different co-culture systems designed to mimic the lymph node and vascular microenvironments, and compared and contrasted their effects on CLL cells. Two of the model systems utilized were mouse embryonic fibroblasts transfected with human CD40LG or human CD31 (CD40L-TF and CD31-TF) and the third was a microvascular human endothelial cell line, HMEC-1. The aim of these experiments was to compare the effects of these different co-culture systems in order to identify the most appropriate in-vitro model system for mimicking the tumour microenvironment.

Materials and methods

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

Patient samples

Peripheral blood mononuclear cells (PBMCs) from 42 patients with confirmed CLL (Table 1) were isolated by density gradient separation (Histopaque-1077; Sigma, Poole, UK). Cells were either cultured fresh or cryopreserved in RPMI 1640 medium, 40% fetal bovine serum (FBS) and 10% dimethylsulfoxide (Sigma) as previously described (Patten et al, 2008). Ethical approval was obtained from the local institutional review board of King's College Hospital and South East Wales ethics committee, respectively. In every case, informed written consent was obtained according to the Declaration of Helsinki.

Table 1. Characteristics of CLL patients in the study
IDBinet stageCD38 (%)ZAP70 (%)IGHV mutation statusIGHV gene usageFISH
1B57·0NDUnmutatedIGHV1-69Normal (TP53 mutation)
2 85·0NDUnmutatedIGHV5-51Trisomy 12
3C22·0NDUnmutatedIGHV4-4ND
4B98·0NDUnmutatedIGHV1-69Trisomy 12
5B3·0NDMutatedIGHV3-23del 13q14·3
6A1·0NDNDNDND
7A48·0NDMutatedIGHV4-34Normal
8ANDNDNDNDND
9C52·0NDUnmutatedIGHV1-2del 13q14·3
10A49·0NDNDNDND
11A1·0NDNDNDdel 13q14·3
12A2·0NDUnmutatedIGHV2-70del 13q14·3
13B44·0NDNDNDNormal
14B98·0NDMutatedIGHV1-8del 17p
15A49·0NDUnmutatedIGHV4-34del 13q14·3/del 17p
16B43·0NDMutatedIGHV1-3ND
17A43·0NDUnmutatedIGHV4-39ND
18A1·5NDNDNDND
19A35·0NDNDNDNormal
20B96·0NDMutatedIGHV3-21del 13q14·3
21A3·922·2MutatedIGHV2-5Normal
22A9·810·4MutatedIGHV1-2del 13q14·3
23C5·73·4MutatedIGHV4-59Normal
24A84·813·2MutatedIGHV4-34Normal
25A84·21·0UnmutatedIGHV3-9del 11q
26A16·41·0MutatedIGHV3-64del 13q14·3
27A46·031·0MutatedIGHV3-74Normal
28A6·11·4MutatedIGHV3-74Normal
29A100·087·0MutatedIGHV3-7del 13q14·3
30A10·392·0MutatedIGHV3-33ND
31A2·32·0MutatedIGHV3-7del 13q14·3
32A99·780·4MutatedIGHV1-3ND
33A13·011·0MutatedIGHV3-9del 13q14·3
34A8·94·6MutatedIGHV3-21del 13q14·3
35C28·758·6UnmutatedIGHV3-30del 13q14·3/del 17p
36B70·433·1UnmutatedIGHV3-74ND
37A12·020·0MutatedIGHV3-21Trisomy 12
38B99·580·0MutatedIGHV3-21del 11q
39A5·034·0MutatedIGHV4-34ND
40A0·910·3MutatedIGHV3-23Normal
41A36·034·0UnmutatedIGHV3-53del 13q14·3
42C47·025·0UnmutatedIGHV1-69del 11q

Liquid culture conditions

Chronic lymphocytic leukaemia PBMCs were cultured at 2 × 106/ml in either HMEC-1 recommended medium [M199, 10% (v/v) FBS, L-glutamine (2 mmol/l final), penicillin (2000 units per ml), streptomycin (2 mg/ml), endothelial cell growth supplement (ECGS) (10 μg/ml), hydrocortisone (1 μg/ml), 2-mercaptoethanol (5 μmol/l), human epidermal growth factor (hEGF) (10 ng/ml), ascorbic acid in M199, (1 μg/ml), vascular endothelial growth factor (VEGF) (0·5 ng/ml), insulin-like growth factor 1 (IGF1) (10 ng/ml) supplemented with 1% bovine serum albumin (Sigma)] or Dulbecco's Modified Eagle's Media (DMEM) complete media (DMEM containing 10% fetal calf serum, 2% penicillin plus streptomycin, 1% L-glutamine and 5 ng/ml interleukin 4 [Miltenyi Biotech, Bisley, UK]). Cells were incubated at 37°C in a fully humidified atmosphere of 5% CO2.

Co-culture conditions

Human microvascular endothelial cells (HMEC-1, Centres for Disease Control and Prevention, Atlanta, GA) were seeded at 105/ml in 24-well plates in the recommended medium and incubated overnight to allow cells to adhere. CLL PBMCs were cultured alone and on HMEC-1 cells at 2 × 106/ml as previously described (Buggins et al, 2010) and harvested at the time points indicated. Untransfected mouse fibroblasts (NTL) and genetically modified fibroblasts expressing CD31 (CD31-TF a kind gift from Dr Silvia Deaglio, University of Torino & Human Genetics Foundation, Turin, Italy) or CD40LG (CD40L-TF from Dr Aneela Majid, Medical Research Council Toxicology Unit, Leicester University, Leicester, UK) were seeded at 2 × 106/ml in DMEM complete media. The cells were left overnight to adhere to the plates. 2 × 106 CLL cells were added and the co-cultures were left at 37°C, 5% CO2 for up to 21 d. The supplemented DMEM media was changed every 3–4 d and cells harvested at the time points indicated. All co-cultures were seeded at confluence to ensure optimal contact with primary CLL cells.

Flow cytometry

The phenotype of CLL PBMCs at time 0 and at 24 h following co-culture with HMEC-1 cells, CD40L-TF, CD31-TF, NTL fibroblasts or control medium was analysed by 5-colour flow cytometry. CD38, CD44, ITGA4, ZAP70, CD69, CD11c, CD103 and CD138 expression were determined using CD19-Pacific Blue (PB), CD5-phycoerythrin-cyanin 7 (PECy7), CD38-pycoerythrin (PE), CD44-PE, CD69-allophycocyanin (APC), ZAP70-fluorescein isothiocyanate (FITC) (all eBioscience, Hatfield, UK), ITGA4-FITC (Serotec, Kidlington, UK) and CD11c-FITC, CD103-PE and CD138-APC (Dako, Ely, UK). Apoptosis of cells was assessed after 7 d co-culture using flow cytometry following labelling with CD19-PB, CD5-PeCy7, Annexin V-FITC (Becton Dickinson, Oxford, UK), and 7-amino-actinomycin D (7-AAD; Becton Dickinson) according to the manufacturer's instructions. Proliferation was analysed at various time points by measuring Ki-67 expression using flow cytometry. Cells were labelled with CD19-PB, CD5-PeCy7 before treatment with Fix and Perm reagent (eBioscience) supplemented with 5% Nonidet P-40 and labelling with Ki-67-FITC (Becton Dickinson) or matched isotype control. All antibodies and clones used are shown in Table 2.

Table 2. Antibodies used for flow cytometry
AntibodyClone or identifierSupplier
CD38-PEHB7eBioscience
ITGA4-FITCBu49Serotec
CD19-Pacific blueHIB19eBioscience
CD5-PE Cy7UCHT2eBioscience
CD44-PEIM7eBioscience
CD69-APCFN50eBioscience
ZAP70-FITCIE7·2eBioscience
Ki-67-FITC51-36524Becton Dickinson
CD11c -FITCKB90Dako
CD103-PEBer-ACT8Dako
CD138-APCMI 15Dako

CFSE labelling of primary CLL cells

Primary CLL cells were labelled with 10 mmol/l carboxyfluorescein succinimidyl ester (CFSE) (Invitrogen, Carlsbad, CA, USA) in 1 ml of DMEM supplemented with 1% FBS. Cells were incubated at 37°C for 10 min before being washed twice in DMEM supplemented with 10% FBS. CLL cells were then seeded into the various co-culture systems at 2 × 106 cells/ml. Cell proliferation was assessed after 7 and 14 d in culture by the decrease in CFSE labelling (compared to day 0 cells) and was quantified using FlowJo 9.3.3 software (TreeStar Inc., Ashland, OR, USA).

Statistical analysis

All statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA, USA). All of the paired data were tested for normality and considered Gaussian, so data sets were compared using the paired t-test. P-values < 0·05 were considered to be statistically significant.

Results

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

We have previously reported that HMEC-1 cells promote the survival of CLL cells in-vitro compared with liquid culture (Buggins et al, 2010). We therefore initially compared the survival index of this co-culture system with NTL mouse fibroblasts and that of CD31-TF or CD40L-TF. Primary CLL cells from 42 different CLL patients were analysed for viability following co-culture alone or with NTL mouse fibroblasts, CD40L-TF or CD31-TF or HMEC-1 cells for 7 d. All four co-culture systems promoted survival of the CLL cells when compared with liquid culture (NTL = 0·01, CD40L-TF = 0·0018, CD31-TF = 0·0009 and HMEC-1 < 0·0001) with HMEC-1 cells providing the greatest cytoprotection (Fig 1).

image

Figure 1. Human microvascular endothelial cell line-1 cells and both CD31-TF and CD40L-TF prevent apoptosis of CLL cells. Primary CLL cells from 21 different CLL patients were cultured alone or co-cultured with untransfected mouse fibroblasts (NTL), CD40L-TF, CD31-TF or HMEC-1 cells for 7 d. CD19+CD5+CLL cells were gated on and apoptosis measured by Annexin V/7-amino-actinomycin D positivity. All the co-culture systems prevented CLL cell apoptosis with HMEC-1 cells giving the best survival.

Download figure to PowerPoint

We then determined whether the different co-culture systems induced similar phenotypic changes in CLL cells by analysing them at an earlier time point of 24 h, chosen because time course studies had shown that the maximal changes in the phenotypic markers included in this study were observed this time point (data not shown). The antibody panels were chosen to quantify the effect of co-culture on the expression of markers associated with adhesion (CD38, CD44 and ITGA4), activation (CD38 and CD69), migration (ITGA4) and hairy cell leukaemia/plasmacytoid differentiation (CD103, CD11c and CD138). We also analysed the expression of ZAP70 due to its association with progressive disease (Rassenti et al, 2004). We have already shown that CD38 and ITGA4 are up-regulated following co-culture with HMEC-1 cells (Buggins et al, 2010). Here, we showed that both the transfected fibroblast cell lines did the same, with the CD31-TF cells being the most potent stimulator (CD38: CD40L-TF < 0·0001, CD31-TF < 0·0001 and HMEC-1 = 0·0464. ITGA4: CD40L-TF < 0·0001, CD31-TF < 0·0001 and HMEC-1 = 0·031; Figs 2A and 3A,B). This is perhaps not surprising as CD31 is the only known ligand for CD38 and ITGA4 is known to co-localize with CD38 (Buggins et al, 2011). In contrast, the NTL mouse fibroblasts failed to significantly induce these phenotypic changes in the absence of the human ligands but maintained the expression of the antigens at similar levels to those measured at time 0.

image

Figure 2. Representation flow cytometry plots demonstrating the phenotypic changes induced in CLL cells following co-culture. Primary CLL cells from 35 different CLL patients were co-cultured alone, on NTL mouse fibroblasts or with CD40L-TF, CD31-TF or HMEC-1 cells for 24 h. At time 0 and time 24 h, cells were stained for flow cytometry and CD19+CD5+CLL cells were gated on and levels of CD38, CD69 and ZAP70 measured. (A) and (B) are representative flow cytometry plots showing that co-culture with CD31-TF caused increased levels of CLL CD38 expression and co-culture with HMEC-1 cells increased levels of the activation marker CD69. (C) illustrates that co-culture with CD31-TF induced an increase in CLL ZAP70 expression but co-culture on HMEC-1 cells had no effect on ZAP70 expression. MFI, mean fluorescence intensity.

Download figure to PowerPoint

In addition, CD44 and CD69 were also up-regulated by all three co-culture systems to a similar level (CD44: CD40L-TF < 0·0001, CD31-TF < 0·0001 and HMEC-1 = 0·0004 CD69: CD40L-TF < 0·0001, CD31-TF < 0·0001 and HMEC-1 < 0·0001 Figs 2B and 3C,D). Again, the NTL mouse fibroblasts failed to significantly induce these phenotypic changes above the expression of the antigens as measured at time 0. Interestingly, ZAP70, which has been associated with a similar expression pattern as CD38 (Damle et al, 2007), was only up-regulated by the transfected fibroblasts and unaffected by co-culture with HMEC-1 cells (CD40L-TF < 0·0001, CD31-TF < 0·0001 and HMEC-1 = 0·437; Figs 2C and 3E). The two markers of differentiation towards a hairy cell phenotype (CD103 and CD11c) were not significantly induced by co-culture with NTL, HMEC-1 and CD31-TF cells. In contrast, the CD40L-TF cells induced a small but significant increase in in CD11c but not CD103 (= 0·008 and = 0·092; Figure S1A and S1B respectively). Although the change in CD11c was statistically significant, the mean fluorescence intensity values for this antigen remained very low compared with other antigens measured in this study, even after co-culture with CD40L-TF cells. As CLL cells cultured on the CD40L-TF and CD31-TF co-culture systems had a notable increase in the forward scatter, suggesting they could potentially be in transformation to a plasmacytoid phenotype, we looked at CLL expression of the plasmacytoid marker CD138 in these systems. Neither system induced significant changes in the expression of the CD138 (Figure S1C).

image

Figure 3. Human microvascular endothelial cell line-1 cells and both CD31-TF and CD40L-TF increase expression of CD38, ITGA4, CD44 and CD69 by CLL cells. Primary CLL cells from 35 different CLL patients were cultured alone, or co-cultured on NTL, CD40L-TF, CD31-TF or HMEC-1 cells for 24 h. At time 0 and time 24 h, cells were stained for flow cytometry and CD19+CD5+CLL cells were gated on and levels of CD38, ITGA4, CD44, CD69 and ZAP70 measured. (A–E) show composite bar charts demonstrating the results from all the primary CLL cells assayed. All four co-culture cell types induced increased levels of CLL expression of (A) CD38, (B) ITGA4, (C) CD44 and (D) CD69 with the most notable effect on CD38 and CD49 induced by CD31-TF. (E) Only CD31-TF and CD40L-TF were able to induce an increase in CLL expression of ZAP70. The NTL (control) mouse fibroblasts did not induce the same phenotypic changes. MFI, mean fluorescence intensity.

Download figure to PowerPoint

To analyse the proliferation of the CLL cells in the different co-culture systems, longer-term co-culture assays of primary CLL cells from six different patients on all the co-culture systems were set up. In all of the systems the CLL cells clustered around the confluent ‘feeder’ cells, but on the HMEC-1 cells and the NTL mouse fibroblasts they remained the same size, whereas on the transfected fibroblasts they showed signs of blasting and proliferation (Fig 4A). As it has already been shown that the proliferation marker Ki-67 is upregulated in the CLL cells from the lymph nodes compared to the peripheral blood in these patients (Herishanu et al, 2011), we compared Ki-67 levels of CLL cells in our different systems. Analysis of Ki-67 on day 7 clearly showed that HMEC-1 cells did not induce Ki-67 expression by CLL cells in all six patients tested, whereas both sets of transfected fibroblasts induced Ki-67 expression at day 7, with CD40L-TF cells inducing a further increase at day 21 (Fig 4B,C). It is interesting to note that CD31-TF cells induced proliferation whereas HMEC-1 cells, which express very low levels of CD31, could not. CD31-TF cells have a much higher density of CD31 expression than HMEC-1 cells (data not shown), which suggests that ligand density may be an important factor in driving CLL proliferation, at least in the context of CD31 signalling.

image

Figure 4. Both CD31-TF and CD40L-TF induce proliferation in CLL cells but HMEC-1 cells do not. Primary CLL cells from six different patients were co-cultured for up to 21 d alone or with CD40L-TF, CD31-TF or HMEC-1 cells. CLL cell samples were removed at day 0, 7 and 21 and CD5/CD19+ cells analysed for expression of Ki-67. (A) Photographs of primary CLL cells clustered on CD40L-TF, CD31-TF and HMEC-1 cells taken on a Zeiss Axio microscope using a ×100 objective lens. (B) A representative figure demonstrating the increase in Ki-67 expression in CD19+ CLL cells following co-culture with CD31-TF compared to liquid culture and HMEC-1 co-culture. (C) At day 7 both CD40L-TF and CD31-TF induced an increase in expression of Ki-67 in CLL cells from all six patients tested compared to liquid culture. At day 21 CD40L-TF induced a further increase in Ki-67 expression; this was less marked in those cells co-cultured with the CD31-TF. HMEC-1 cells did not induce Ki-67 expression in any CLL cells at the same time-points.

Download figure to PowerPoint

These results suggest that, although CD40L-TF cells afforded the least cytoprotection to CLL cells when compared with HMEC-1 cells or CD31-TF cells, they were the most potent inducers of CLL proliferation. Indeed, these two phenomena may be linked; cells with a higher rate of turnover may have an increased propensity to die. To further assess the effect on cell division we compared the number of divisions seen in CFSE-labelled CLL cells co-cultured on NTL mouse fibroblasts, CD40L-TF or CD31-TF. As expected, following 21 d in culture, the CLL cells co-cultured on NTL fibroblasts showed little evidence of CLL cell proliferation (Fig 5A). In contrast CD40L-TF and CD31-TF co-cultures induced significant proliferation with CD40L-TF inducing more cell divisions than those co-cultured on CD31-TF (Fig 5B,C, respectively). It is noteworthy that there appeared to be no difference in the ability of CD40L-TF to induce proliferation in samples derived from patients in different Binet stages (Fig 5D).

image

Figure 5. CD40L-TF induced more CLL cell proliferation than CD31-TF. CLL cells from 18 different patients were labelled with carboxyfluorescein succinimidyl ester and co-cultured on (A) NTL mouse fibroblasts, (B) CD40L-TF or (C) CD31-TF. The representative figures demonstrate that after 2 weeks of co-culture CLL cells cultured on NTL fibroblasts showed very little cell division. Numbers of cell divisions was calculated using FlowJo software and are shown on each figure. In contrast, the sample CLL cells cultured with CD40L-TF and CD31-TF showed significant cell division with those on CD40L-TF showing the most proliferation. (D) Shows that the extent of proliferation induced by CD40L-TF was not dependent on Binet stage.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

It is now widely accepted that interactions in the tumour microenvironment can promote CLL cell survival, proliferation and drug resistance (Patten et al, 2005, 2008; Plander et al, 2009; Buggins et al, 2010; Pepper et al, 2011). Although the key molecular events that drive these processes are not fully defined, we have previously shown that lymph nodes from patients with aggressive disease contain increased numbers of CD31+ vessels and activated T-lymphocytes and that interactions with vascular endothelial cells enhance CLL cell survival. The present study describes a detailed comparison of three co-culture systems designed to mimic the CLL lymph node and vascular microenvironments; two utilized mouse embryonic fibroblasts transfected with human CD40LG (expressed by activated T-lymphocytes) or human CD31 (CD38 ligand) and the third was a microvascular endothelial cell line, HMEC-1.

All three systems were highly cytoprotective to CLL cells when compared with liquid culture. Interestingly the human microvascular cell line, HMEC-1, induced the best survival signals. This is in keeping with previous work, which suggests that interaction with the microvasculature plays a key role in the maintenance of CLL cell survival in-vivo (Zucchetto et al, 2009; Buggins et al, 2010). The fact that the CD31-TF cells were not as effective stimulators of survival (despite having higher CD31 antigen density) suggests that CD31 is not the sole anti-apoptotic signal in the microvasculature and/or that CD31 ligand density needs to be lower for optimal cytoprotection. The CD40L-TF cells provided the least cytoprotection of all the co-culture systems under evaluation. This indicates that the CD40-CD40LG interaction is not a major player with regard to the prevention of apoptosis in the microenvironment and/or the expression of CD40LG in this model is too high for optimal anti-apoptotic signalling.

Surprisingly, all three co-culture systems induced a remarkably similar CLL cell phenotype and were able to up-regulate markers of adhesion, activation and migration. This suggests that co-culture mediated phenotypic changes are not ligand-specific but rather represent a default setting for CLL cells following interaction with many of the cells and receptors associated with the tumour microenvironment. Importantly, the NTL mouse fibroblasts used as controls in this study failed to induce these phenotypic changes suggesting that interaction of CLL cells with human CD31 and CD40LG are critical to the effects seen in CD40L-TF and CD31-TF. The notable exception in terms of phenotypic change was the ability of CD31-TF cells and CD40L-TF cells to induce ZAP70 expression. In contrast, the HMEC-1 cells were unable to induce the expression of this tyrosine kinase suggesting that endothelial cell interactions may not be the principle determinant of the maintenance ZAP70 expression in CLL cells. Given that endothelial cells are unable to induce both ZAP70 expression and CLL cell proliferation, this data suggests that there is a possible role for this tyrosine kinase in CLL cell division.

In contrast to its relatively modest effect on in-vitro survival, CD40LG appeared to provide the strongest and most sustained proliferation signals to CLL cells. On the contrary, HMEC-1 cells failed to induce any proliferation under the conditions tested. This is perhaps not surprising as CD40LG is expressed by activated T cells and co-culture with these is known to induce CLL cell division (Patten et al, 2005, 2008). Indeed, one hypothesis for the inferior cytoprotection afforded by CD40L-TF cells is that the strong proliferative response they induce causes CLL cells to become more susceptible to cell death. Ligation of CD38 is also known to induce proliferation so it was of considerable interest that the CD31-TF also induced CLL cell division. However, HMEC-1 cells, which express low levels of CD31, were unable to cause even a marginal increase in CLL cell Ki-67 expression. This raises the question as to the role of ligand density in these interactions and highlights the possibility that many ligand/receptor interactions may induce CLL cell proliferation providing the density is sufficient. Therefore, one interpretation of our data is that HMEC-1 cells express insufficient CD31 to induce CLL cell proliferation but the low level stimulation they provide is ideal for the prevention of apoptosis. This hypothesis is supported by our previous work demonstrating that the anti-apoptotic effect of endothelial cells on CLL cells is via the induction of NF-KB (nuclear factor κB) regulated genes such as BCL2, BCL2L1, MCL1, CD38 and ITGA4 (Buggins et al, 2010). CLL cell co-culture with endothelial cells promoted their survival over a 7-d period that was associated with a small but statistically significant increased CD38 expression. In contrast, our previous work has shown that when CLL cells are co-cultured with activated autologous T cells they induce a dramatic increase in CD38 expression (Patten et al, 2008). However, over a 7-d period these CLL cells died by apoptosis, suggesting that this level of activation is incompatible with long-term CLL cell survival (unpublished data).

These results highlight the importance of the choice of co-culture model and, for the laboratory constructed systems, the importance of designing systems with ligand densities at a physiological level. As culture of CLL cells in-vitro is problematic due to the dependence of these tumour cells on interactions with the surrounding tissues, it is vital that a model that accurately mimics the microenvironment is established. Our findings indicate that the best model system for mimicking the lymph node microenvironment is the CD40L-TF cells; this co-culture system induces an activation phenotype and a strong proliferative signature consistent with recently published gene expression profiles from CLL cells derived from lymph nodes (Herishanu et al, 2011). In contrast, when modelling the CLL cell interactions in the microvasculature our data suggest that the HMEC-1 cells represent the most realistic model system, characterized by enhanced survival in the absence of proliferation. It seems likely that further refinement of these in-vitro models will lead to improved in-vitro drug testing platforms.

Acknowledgements

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

This work was supported by grants from Leukaemia & Lymphoma Research, the Medical Research Council and Leukaemia Research Appeal for Wales. CP is also supported by the National Institute for Social Care and Health Research (NISCHR) through the Cancer Genetics Biomedical Research Unit.

Authorship contribution

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

EH carried out the experimental work, analysed data and edited the manuscript; LP carried out the experimental work, analysed data and edited the manuscript; LM carried out the experimental work and analysed data; SR carried out the experimental work and analysed data; VW carried out the experimental work and analysed data; PB analysed data and edited the manuscript; NSBT analysed data and edited the manuscript; DY provided clinical samples, analysed data and edited the manuscript; SD provided clinical samples, analysed data and edited the manuscript; CF provided clinical samples, analysed data and edited the manuscript; AGSB and CP jointly conceived and supervized the study, analysed the data and wrote the manuscript.

Conflict of interest

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information

EH, LP, LM, SR, VW, PB, NSBT, DY, SD, CF, AGSB and CP declare no conflict of interests.

References

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information
  • Buggins, A.G., Pepper, C., Patten, P.E., Hewamana, S., Gohil, S., Moorhead, J., Folarin, N., Yallop, D., Thomas, N.S., Mufti, G.J., Fegan, C. & Devereux, S. (2010) Interaction with vascular endothelium enhances survival in primary chronic lymphocytic leukemia cells via NF-kappaB activation and de novo gene transcription. Cancer Research, 70, 75237533.
  • Buggins, A.G., Levi, A., Gohil, S., Fishlock, K., Patten, P.E., Calle, Y., Yallop, D. & Devereux, S. (2011) Evidence for a macromolecular complex in poor prognosis CLL that contains CD38, CD49d, CD44 and MMP-9. British Journal of Haematology, 154, 216222.
  • Coscia, M., Pantaleoni, F., Riganti, C., Vitale, C., Rigoni, M., Peola, S., Castella, B., Foglietta, M., Griggio, V., Drandi, D., Ladetto, M., Bosia, A., Boccadoro, M. & Massaia, M. (2011) IGHV unmutated CLL B cells are more prone to spontaneous apoptosis and subject to environmental prosurvival signals than mutated CLL B cells. Leukemia, 25, 828837.
  • Damle, R.N., Wasil, T., Fais, F., Ghiotto, F., Valetto, A., Allen, S.L., Buchbinder, A., Budman, D., Dittmar, K., Kolitz, J., Lichtman, S.M., Schulman, P., Vinciguerra, V.P., Rai, K.R., Ferrarini, M. & Chiorazzi, N. (1999) Ig V gene mutation status and CD38 expression as novel prognostic indicators in chronic lymphocytic leukemia. Blood, 94, 18401847.
  • Damle, R.N., Temburni, S., Calissano, C., Yancopoulos, S., Banapour, T., Sison, C., Allen, S.L., Rai, K.R. & Chiorazzi, N. (2007) CD38 expression labels an activated subset within chronic lymphocytic leukemia clones enriched in proliferating B cells. Blood, 110, 33523359.
  • Deaglio, S., Vaisitti, T., Bergui, L., Bonello, L., Horenstein, A.L., Tamagnone, L., Boumsell, L. & Malavasi, F. (2005) CD38 and CD100 lead a network of surface receptors relaying positive signals for B-CLL growth and survival. Blood, 105, 30423050.
  • Ferretti, E., Bertolotto, M., Deaglio, S., Tripodo, C., Ribatti, D., Audrito, V., Blengio, F., Matis, S., Zupo, S., Rossi, D., Ottonello, L., Gaidano, G., Malavasi, F., Pistoia, V. & Corcione, A. (2011) A novel role of the CX(3)CR1/CX(3)CL1 system in the cross-talk between chronic lymphocytic leukemia cells and tumor microenvironment. Leukemia, 25, 12681277.
  • Herishanu, Y., Perez-Galan, P., Liu, D., Biancotto, A., Pittaluga, S., Vire, B., Gibellini, F., Njuguna, N., Lee, E., Stennett, L., Raghavachari, N., Liu, P., McCoy, J.P., Raffeld, M., Stetler-Stevenson, M., Yuan, C., Sherry, R., Arthur, D.C., Maric, I., White, T., Marti, G.E., Munson, P., Wilson, W.H. & Wiestner, A. (2011) The lymph node microenvironment promotes B-cell receptor signaling, NF-kappaB activation, and tumor proliferation in chronic lymphocytic leukemia. Blood, 117, 563574.
  • Majid, A., Lin, T.T., Best, G., Fishlock, K., Hewamana, S., Pratt, G., Yallop, D., Buggins, A.G., Wagner, S., Kennedy, B.J., Miall, F., Hills, R., Devereux, S., Oscier, D.G., Dyer, M.J., Fegan, C. & Pepper, C. (2011) CD49d is an independent prognostic marker that is associated with CXCR4 expression in CLL. Leukemia Research, 35, 750756.
  • Messmer, B.T., Messmer, D., Allen, S.L., Kolitz, J.E., Kudalkar, P., Cesar, D., Murphy, E.J., Koduru, P., Ferrarini, M., Zupo, S., Cutrona, G., Damle, R.N., Wasil, T., Rai, K.R., Hellerstein, M.K. & Chiorazzi, N. (2005) In vivo measurements document the dynamic cellular kinetics of chronic lymphocytic leukemia B cells. Journal of Clinical Investigation, 115, 755764.
  • Patten, P., Devereux, S., Buggins, A., Bonyhadi, M., Frohlich, M. & Berenson, R.J. (2005) Effect of CD3/CD28 bead-activated and expanded T cells on leukemic B cells in chronic lymphocytic leukemia. Journal of Immunology, 174, 65626563 author reply 6563.
  • Patten, P.E., Buggins, A.G., Richards, J., Wotherspoon, A., Salisbury, J., Mufti, G.J., Hamblin, T.J. & Devereux, S. (2008) CD38 expression in chronic lymphocytic leukemia is regulated by the tumor microenvironment. Blood, 111, 51735181.
  • Pepper, C., Mahdi, J.G., Buggins, A.G., Hewamana, S., Walsby, E., Mahdi, E., Al-Haza'a, A., Mahdi, A.J., Lin, T.T., Pearce, L., Morgan, L., Bowen, I.D., Brennan, P. & Fegan, C. (2011) Two novel aspirin analogues show selective cytotoxicity in primary chronic lymphocytic leukaemia cells that is associated with dual inhibition of Rel A and COX-2. Cell Proliferation, 44, 380390.
  • Plander, M., Seegers, S., Ugocsai, P., Diermeier-Daucher, S., Ivanyi, J., Schmitz, G., Hofstadter, F., Schwarz, S., Orso, E., Knuchel, R. & Brockhoff, G. (2009) Different proliferative and survival capacity of CLL-cells in a newly established in vitro model for pseudofollicles. Leukemia, 23, 21182128.
  • Rassenti, L.Z., Huynh, L., Toy, T.L., Chen, L., Keating, M.J., Gribben, J.G., Neuberg, D.S., Flinn, I.W., Rai, K.R., Byrd, J.C., Kay, N.E., Greaves, A., Weiss, A. & Kipps, T.J. (2004) ZAP-70 compared with immunoglobulin heavy-chain gene mutation status as a predictor of disease progression in chronic lymphocytic leukemia. New England Journal of Medicine, 351, 893901.
  • Schmid, C. & Isaacson, P.G. (1994) Proliferation centres in B-cell malignant lymphoma, lymphocytic (B-CLL): an immunophenotypic study. Histopathology, 24, 445451.
  • Scielzo, C., Apollonio, B., Scarfo, L., Janus, A., Muzio, M., Ten Hacken, E., Ghia, P. & Caligaris-Cappio, F. (2011) The functional in vitro response to CD40 ligation reflects a different clinical outcome in patients with chronic lymphocytic leukemia. Leukemia, 25, 17601767.
  • Shanafelt, T.D., Geyer, S.M., Bone, N.D., Tschumper, R.C., Witzig, T.E., Nowakowski, G.S., Zent, C.S., Call, T.G., Laplant, B., Dewald, G.W., Jelinek, D.F. & Kay, N.E. (2008) CD49d expression is an independent predictor of overall survival in patients with chronic lymphocytic leukaemia: a prognostic parameter with therapeutic potential. British Journal of Haematology, 140, 537546.
  • Zucchetto, A., Benedetti, D., Tripodo, C., Bomben, R., Dal Bo, M., Marconi, D., Bossi, F., Lorenzon, D., Degan, M., Rossi, F.M., Rossi, D., Bulian, P., Franco, V., Del Poeta, G., Deaglio, S., Gaidano, G., Tedesco, F., Malavasi, F. & Gattei, V. (2009) CD38/CD31, the CCL3 and CCL4 chemokines, and CD49d/vascular cell adhesion molecule-1 are interchained by sequential events sustaining chronic lymphocytic leukemia cell survival. Cancer Research, 69, 40014009.

Supporting Information

  1. Top of page
  2. Summary
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. Authorship contribution
  8. Conflict of interest
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
bjh9191-sup-0001-FigS1.docWord document491KFig S1. The three cell lines induced little evidence of increase expression of the differentiation markers CD11c, CD103 and CD138 on CLL cells.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.