The VEGF receptor, neuropilin-1, represents a promising novel target for chronic lymphocytic leukemia patients


Correspondence to: Krzysztof Giannopoulos, Department of Experimental Hematooncology, Medical University of Lublin, Chodzki 4a, 20–950 Lublin, Poland, Tel.: +48817564812, Fax: +48817564813, E-mail:


Angiogenesis has been shown to substantially contribute to the progression of chronic lymphocytic leukemia (CLL). Neuropilin-1 (NRP1) represents a receptor for vascular endothelial growth factor (VEGF), which has been reported to be overexpressed in several malignancies. In our study, we characterized mRNA levels of VEGF receptors including NRP1 in a large cohort of CLL patients (n = 114), additionally we performed a detailed characterization of NRP1 expression on B cells, plasmacytoid dendritic cells (PDCs) and regulatory T cells (Tregs). The expression of NRP1 was significantly higher on leukemic lymphocytes compared to control B lymphocytes on mRNA and protein levels (22.72% vs. 0.2%, p = 0.0003, respectively), Tregs (42.6% vs. 16.05%, p = 0.0003) and PDCs (100% vs. 98% p < 0.0001). In functional studies, we found higher NRP1 expression on CLL cells after stimulation with VEGF. The correlation between expression of VEGF receptors: FLT1, NRP1 and FOXP3 expression (r2 = 0.53, p < 0.0001 and r2 = 0.49, p < 0.0001, respectively) was observed. Earlier we described the specific Treg reduction during the therapy with thalidomide in vivo. Now we observe the reduction of the NRP1 expression on Tregs in vitro, thereby suggesting a possible target of thalidomide action. In conclusion, NRP1 might represent an interesting link between angiogenesis and tolerance mechanisms and represents interesting target for therapy.


Chronic lymphocytic leukemia (CLL) is the most common type of leukemia found in older patients in the Western hemisphere. The clinical outcome is highly heterogeneous, encompassing cases progressing dramatically as well as patients who will never require therapy and who will achieve a life-span similar to age-matched healthy subjects.[1]

Angiogenesis is a fundamental process by which new blood vessels are formed.[2] Vascular endothelial growth factor (VEGF) is a crucial regulator of normal and abnormal angiogenesis and is able to induce proliferation and migration of endothelial cells of vessels. VEGF is secreted by macrophages, fibroblasts, myocytes as well as some cancer cells.[3, 4] VEGF binds to three types of tyrosine kinase receptors VEGFR1 (FLT1), VEGFR2 (kinase domain region [KDR], also known as FLK1), and VEGFR3 (FLT4) and two nonenzymatic receptors Neuropilin-1 (NRP1) and −2 (NRP2).[5] During tumorigenesis, VEGF levels increase along with disease progression and higher VEGF concentrations have been shown to be correlated with progressive disease as well as unfavorable prognosis. This dependence was observed in several cancer types including breast cancer,[6] nephroblastoma,[7] endometrial cancer[8] and non-small cell lung cancer.[9]

In context of hematology, overexpression of VEGF has been demonstrated in many hematological malignancies such as acute myeloid leukemia (AML), acute lymphocytic leukemia,[10] multiple myeloma[11] and CLL.[12] The increased levels of VEGF in plasma, which is released by the tumor, might be connected with its paracrine effects on vascular endothelial angiogenesis.[13] Angiogenesis also substantially contributes to the progression of CLL and the production of VEGF were mainly attributed to CLL cells of the proliferating compartment located in the lymph nodes and the bone marrow. It was also suggested that VEGF bound to receptors drives CLL cells to proliferate.[12, 14, 15]

Neuropilins (NRPs) are a group of multifunctional, non-tyrosine kinase receptors for two different ligand families of extracellular ligands—class 3 semaphorins and several members of the VEGF family. NRPs play an important role in various biological processes, including angiogenesis, tumorigenesis and immunological response. NRP1 may indirectly mediate effects on tumor progression by affecting angiogenesis or directly through effects on tumor cells.[16, 17] Recently, Nowakowski et al. reported NRP1 to be also expressed in CLL in a pilot study on a small cohort of patients.[18] Interestingly, NRP1 could be also expressed on regulatory T cells (Tregs) as well as plasmacytoid dendritic cells (PDCs). Both, Tregs and PDCs represent cells involved in tolerance mechanisms which could be involved in the tumor escape mechanism.[19]

As NRP1 might represent an interesting link between angiogenesis and immune tolerance and thereby promising target for future therapies, we characterized NRP1 and other VEGF receptors in CLL. Furthermore, we hypothesized that NRP1 could be a target for antiangiogenetic therapy and responsible for observed specific reduction of Tregs after thalidomide therapy in vivo.[20]

Material and Methods

Patient characteristics

Peripheral blood was collected from 152 CLL patients treated in the Department of Hematooncology, Medical University of Lublin. For 114 patients, assessment of mRNA expression was performed with quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) (group A). A group of 38 patients was analyzed with flow cytometry (group B). Functional studies were performed for eight patients. The effect of thalidomide to the level of VEGF was evaluated in 22 CLL patients treated in a clinical trial using thalidomide and fludarabine. The control samples of peripheral blood were obtained from seven healthy volunteers (HVs). The clinical characteristics of the patients are summarized in Table 1. This study was approved by the Local Ethics Committee, and the patients were informed about the use of their blood for scientific purposes.

Table 1. Clinical characteristic of CLL patients of groups A and B
  1. A—group of 114 patients analyzed by qRT-PCR for NRP1, FLT1, KDR and FOXP3; B—group of 38 patients analyzed by flow cytometry method.

Number of patients11438
Binet stage
ZAP-70 (cut-off 20%)
Not available81
CD38 (cut-off 30%)
Not available82
no del11q− or del17p−11236
del11− or del17p−22
IGVH mutation status
Not available2823

Study protocol

The effect of thalidomide with respect to the level of circulating VEGF was evaluated in 22 CLL patients treated in a clinical trial utilizing thalidomide and fludarabine. Patients received thalidomide at a daily dose of 100 mg p.o. starting from day 0 (D0). Fludarabine was added for 5 consecutive days every 28 days, starting from day 7 (D7) i.v. at a dose of 25 mg/m2 for up to six cycles. Patient received 100 mg acetylsalicylic acid for thrombosis prophylaxis. Samples were collected on D0 and D7 to evaluate the effect of thalidomide monotherapy. The patients' characteristics and clinical results of thalidomide and fludarabine therapy were published earlier.[20]

Cell isolation

Peripheral blood mononuclear cells (PBMCs) from CLL patients and HVs were isolated by Ficoll (Biochrom AG, Berlin, Germany) density gradient centrifugation and cryopreserved at −80°C to the time of analysis. The viability of obtained PBMCs was always >95%, as determined by trypan blue staining. Viable cells were quantified in a Neubauer chamber. Plasma was obtained from the blood samples of 22 patients using EDTA as anticoagulant, aliquoted and stored at −80°C for enzyme-linked immunosorbent assay (ELISA) test.

mRNA preparation and reverse transcription

For total RNA isolation from PBMCs QIAamp RNA Blood Mini Kit (Qiagen, Venlo, Netherlands) was used according to the manufacturer's instructions. From each sample, 1 µg of total RNA was reverse transcribed to 20 µl of cDNA using QuantiTect Reverse Transcription Kit (Qiagen). For quantitative RT-PCR reactions, 1 µl of cDNA of each sample was used.

Quantitative reverse transcriptase-polymerase chain reaction

For 114 cases (group A) quantitative RT-PCR (qRT-PCR) measurements of the mRNA expression of VEGF receptors including NRP1, FLT1, KDR1 and FOXP3 (forkhead box P3). qRT-PCR was performed using TaqMan Gene Expression Assays methodology according to the manufacturer protocol (Applied Biosystems, Foster City). As a constitutively expressed housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used. Thermocycling program was set for 40 cycles of 15 sec at 95°C, 1 min at 60°C on the ABI Prism 7300 Sequence Detector (Applied Biosystems). Expression levels were calculated as an inverse ratio of the difference in cycle threshold (ΔCt), where ΔCt is the Ct value of the target receptors minus Ct value of GAPDH.

Flow cytometry analysis of the expression of NRP1

For the flow cytometry analysis of the surface expression of the NRP1, 38 CLL patients were assessed (group B). The expression of NRP1 on leukemic cells, PDCs, Tregs was examined by multicolor flow cytometry using the following antibodies: phycoerythrin (PE)-conjugated anti-CD304 monoclonal antibody (moAb) (BDCA-4/NRP1), fluorescein isothiocyanate (FITC)-conjugated anti-CD5 moAb, peridinin-chlorophyll proteins (PerCP)-conjugated anti-CD19 moAb, anti-FOXP3-PE, anti-CD25-FITC, anti-CD4-PerCP, anti-BDCA2-FITC and anti-CD123-PE-Cy-5. After incubation, cells were washed twice and analyzed by flow cytometry for the expression of NRP1. A minimum of 100,000 cells were collected and analyzed. Flow cytometry analysis was performed on a FACSCalibur (BD Biosciences) and analyzed using the Summit Software (Dako Cytomation, Glostrup, Denmark).

Determination of plasma VEGF levels

VEGF was detected with the Quantikine ELISA (R&D Systems) in 22 CLL patients before and after thalidomide treatment. The test uses monoclonal antibody specific for VEGF precoated onto a microplate and an enzyme-linked polyclonal antibody raised against recombinant human VEGF. For each analysis, 100 µl of plasma was used. Standards and samples were pipetted in triplicates. A standard curve was created by plotting the mean the VEGF concentration. Concentrations were reported as pg/ml.

Leukemic B cell stimulation with different VEGF concentrations

To assess functional connection between VEGF and its receptor NRP1, CLL cells derived from eight patients were stimulated with different concentrations of VEGF. CD19-positive cells from PBMCs were separated in magnetic field according to the manufacturers protocol (MACS Milltenyi Biotec). Purified CD19+ CLL cells were cultured at 1 × 106 cells/ml for 24 hr with different VEGF concentrations (0.1, 0.5 and 1.0 ng/ml). Cells were incubated in a standard medium consisting RPMI-1640 (Biochrom, Berlin, Germany) supplemented with 10% heat-inactivated fetal bovine serum, 50 units/ml penicillin, 50 µg/ml streptomycin and 100 µg/ml neomycin at 37°C in a humidified atmosphere of 95% air and 5% CO2. After cell culture, the NRP1 surface expression on CD5+CD19+ CLL cells was measured using flow cytometry.

CD4+CD25high T cells isolation and incubation with thalidomide

For functional analyses in CLL patients, we isolated CD4+CD25high T cells from PBMCs. CD4 microbeads were used for negative isolation of CD4+ T cells according to the manufacturers protocol (Milltenyi Biotec). Next, CD4+CD25high T cells were positively selected using CD25 microbeads (Milltenyi Biotec). Purified CD4+CD25high T cells were incubated at a concentration of 1 × 106 cells/ml for 24 hr with thalidomide at a concentration of 10 µg/ml or without thalidomide in control samples. Cells were cultured at 37°C in humidified atmosphere containing 5% CO2. After cell culture, the NRP1 surface expression on CD4+CD25high T cells was measured using flow cytometry.

Statistical analysis

All results are presented as median values with range. The U Mann–Whitney test was used to evaluate the differences between subgroups of patients. Correlations of variables were computed with the Spearman rank correlation coefficient.


The mRNA expression of VEGF receptors and FOXP3 expression in CLL patients

The mRNA levels of the VEGF receptors FLT1, KDR, NRP1 and FOXP3 were measured in 114 CLL patients (group A). We observed a correlation between NRP1 expression and FOXP3 expression (r2 = 0.49, p < 0.0001) (Fig. 1a). In the same group of patients, correlation of FLT1 expression and FOXP3 expression was observed (r2 = 0.53, p < 0.0001) (Fig. 1b).

Figure 1.

Characterization of neuropilin-1 (NRP1) mRNA expression in chronic lymphocytic leukemia (CLL) patients. Figure displays characterization of the mRNA expression of VEGF receptors and forkhead box P3 (FOXP3) in a cohort of 114 CLL. The expression of FOXP3 correlates with expression of NRP1 (r2 = 0.49, p < 0.0001, a) and flt/fms-like tyrosine kinase (FLT1) (r2 = 0.53, p < 0.0001, b). CLL patients without KDR have higher level of FOXP3 compared to patents with expression of KDR (0.13 vs. 0.11, p = 0.008, c). The NRP1 level was higher in CLL patients with mutated status of the IGVH gene compared to patients with unmutated IGVH gene status (ΔCt 0.081 vs. 0.074, p = 0.02, d). [Color figure can be viewed in the online issue, which is available at]

In patients without KDR expression, we found higher level of FOXP3 compared to patients with expression of KDR (1/ΔCt of 0.13 vs. 0.11, respectively). This difference was statistical significant (p = 0.008) (Fig. 1c).

In the group with unmutated immunoglobulin heavy variable (IGVH) status, the median expression of NRP1 (1/ΔCt) was 0.074, in the group with mutated IGVH status we observed higher expression of NRP1 (1/ΔCt of 0.081). This difference reached statistically significant (p = 0.02) (Fig. 1d). There was no correlation between NRP1 expression and expression of ZAP-70 and CD38 (data not shown).

Protein expression of NRP1 on B cells of CLL patients and HVs

Using five parameters flow cytometric assessment, we found increased expression of NRP1 on leukemic lymphocytes of all CLL cases. The median expression of NRP1 on leukemic lymphocytes was significantly higher (median = 22.72%, range 3.2–63.1%) compared to NRP1 expression on lymphocytes derived from HVs (median = 0.2%, range 0.1–0.6%; p = 0.0003) (Fig. 2a). There was no significant difference between patients of different Binet stages, between CD38 positive versus negative patients as well as between ZAP-70 positive versus negative patients (data not shown).

Figure 2.

Surface expression of neuropilin-1 (NRP1) in chronic lymphocytic leukemia (CLL). Panels display a flow cytometric analysis of NRP1 expression on B cells, regulatory T cells (Tregs) and plasmacytoid dendritic cells (PDCs) from CLL patients and HVs. Median expression of NRP1 was higher on leukemic lymphocytes than on control B lymphocytes derived from HVs (22.72% vs. 0.2%, p = 0.0003) (a). Median expression of NRP1 on Tregs was higher in CLL patients (median = 42.64%, range 10–100%) as compared to HVs (median = 16.05%, range 0.0–29.22%; p = 0.0003) (b). Median expression of NRP1 was higher on PDCs in CLL patients as compared to HVs (100% vs. 98%; p < 0.0001) (c).

Expression of NRP1 on Tregs and PDCs in CLL patients and HVs

We found expression of NRP1 on Tregs as well as PDCs derived from the CLL patients. The median expression of NRP1 on Tregs was significantly higher in CLL patients (median = 42.64%, range 10–100%) as compared to HVs (median = 16.05%, range 0.0–29.22%; p = 0.0003) (Fig. 2b). NRP1 was expressed on almost all PDCs in CLL patients with median expression 100% (range: 98.2–100%, p < 0.0001) (Fig. 2c).

VEGF induce expression of NRP1 in CLL patients

In functional studies, we found that NRP1 expression is regulated by the VEGF levels in CLL. The increased expression of NRP1 was found after incubation of CLL cells with different VEGF concentrations. Although VEGF levels of 0.1–0.5 ng/ml, which are in the range of VEGF concentration observed in CLL patient samples, effectively induced expression of NRP1, higher (1, 10 and 80 ng/ml) VEGF concentrations inhibited NRP1 expression. The highest expression of NRP1 on CLL cells was observed after stimulation with the VEGF at the concentration of 0.5 ng/ml (mean fluorescence intensity [MFI] before and after stimulation 1.5 vs. 4.2, p = 0.008) (Fig. 3).

Figure 3.

Neuropilin-1 (NRP1) expression is regulated by vascular endothelial growth factor (VEGF). In functional studies, magnetically separated leukemic cells from eight CLL patients were cultured for 24 hr with different VEGF concentrations (0.1, 0.5 and 1.0 ng/ml) and analyzed with flow cytometry afterward. The VEGF level of 0.5 ng/ml, which is in the range of VEGF concentrations observed in CLL patients, effectively induced expression of NRP1. MFI of NRP-1 was higher on CLL cells after stimulation with the VEGF than before stimulation (4.2 vs. 1.5, p = 0.008).

Plasma concentrations of VEGF during thalidomide treatment in vivo

Levels of VEGF were evaluated on day 0 (D0—before initiation of therapy) and day +7 (D7—thalidomide monotherapy). We observed reduced levels of VEGF in 68% (15/22) of the CLL cases after treatment with thalidomide (median = 128.2 pg/ml, range 26.42–857.0 pg/ml vs. median = 84.43 pg/ml, range 13.74–789.1 pg/ml, p = 0.0597; Fig. 4).There were no differences in VEGF levels reduction in groups of patients who reduced CLL cells after 7 days thalidomide monotherapy (thalidomide responders) versus thalidomide nonresponders nor between groups of patients who respond to the thalidomide + fludarabine regimen (FT responders) versus FT non-responders (data not shown).

Figure 4.

Vascular endothelial growth factor (VEGF) levels before and after with thalidomide monotherapy in vivo. Figure displays results from ELISA test for VEGF before and after thalidomide monotherapy in patients with CLL in vivo. Median VEGF levels were lower in patients after 7 days of treatment than in patients before therapy (84.43 pg/ml vs. 128.2 pg/ml).

The effect of thalidomide on the expression of NRP1 on CD4+CD25high cells

As the NRP1 could be responsible for the observed specific reduction of Tregs after thalidomide in vivo, we analyzed expression of NRP1 on CD4+ T cells incubated for 24 hr with the thalidomide at concentration of 10 µg/ml in vitro. Thalidomide decreased percentage of NRP1 in CD4+ T cells in vitro experiment (Patient 1 39.3% vs. 20%, Patient 2 60.6% vs. 45.4%, Patient 3 80.7% vs. 40.1%; Fig. 5a). In detailed analysis, we characterized expression of NRP1 in particular CD4+ T cells subpopulations: CD4+CD25+ activated T helpers (Thact) and CD4+CD25high T cells (Tregs). Interestingly, the highest expression with subsequent highest reduction of NRP1 was found for Tregs (Figs. 5b and 5c).

Figure 5.

Reduction of neuropilin-1 (NRP1) expression on whole population of CD4+ T cells of chronic lymphocytic leukemia (CLL) patient before (A2) and after (A3) incubation with thalidomide in vitro. Detailed analysis of NRP1 expression in particular CD4+ T cells: CD4+CD25+ activated T helpers (Thact) and CD4+CD25high T cells (Tregs) subpopulations before (b) and after (c) incubation with thalidomide. In functional studies, magnetically separated CD4+CD25high T cells from three CLL patients were cultured with thalidomide at a concentration of 10 µg/ml or without thalidomide in control samples and analyzed with flow cytometry afterward. The thalidomide effectively downregulated Tregs and also reduced the NRP1 expression on Tregs in in vitro experiment. As compared to isotype control (A1), the expression of NRP1 was lower on CD4+ T cells after incubation with of thalidomide (A3) than on CD4+ T cells before stimulation (A2) (20% vs. 39.3%). Panels (b) and (c) display detailed analysis of NRP1 expression in particular CD4+ T cells subpopulations: CD4+CD25+ activated T helpers (Thact) and CD4+CD25high T cells (Tregs) before incubation (b) and incubation with thalidomide for 24 hr (c). [Color figure can be viewed in the online issue, which is available at]


VEGF, one of the best described angiogenetic factor, plays a significant role in both physiological and pathological angiogenesis.[4] Overexpression of VEGF is associated with the tumor progression and poor prognosis in many tumors. The blocking of VEGF transduction reduces tumor neoangiogenesis. However, it was noted that anti-VEGF therapy showed low efficacy when applied to certain types of cancers alone. A significantly higher level of efficacy was demonstrated in case of the anti-NRP1 antibody treatment combined with anti-VEGF antibodies, suggesting that both inhibition of VEGF and its functional receptor effectively inhibit tumor growth.[21]

NRP1 represents a receptor for VEGF, which has been reported to be overexpressed in several malignancies including prostate cancer,[22] lung cancer,[23] colon[24] and breast cancer.[9] The expression of NRP1 correlated with the tumor growth and its invasiveness. The increased expression of NRP1 was also demonstrated among hematological malignancies and increased NRP1 expression was associated with higher mortality rate in acute myeloid leukemia.[11] Lu et al.[25] found that expression of NRP1 was directly correlated with blast percentage in the bone marrow of patients with AML, suggesting NRP1 correlates with disease severity. Regarding CLL, there has been only one report on the expression of NRP1 so far. Authors found NRP1 expression in 7/10 early-stage disease patients by Western-blot analysis, and in 2/5 by flow cytometric analysis, while normal B lymphocytes did not express NRP1.[18] In accordance, our study provides results of a analysis of NRP1 expression in a larger cohort of CLL patients. On the mRNA level, we found strong expression of NRP1 in 114 CLL patients including patients in different Binet stages of disease. Furthermore, in flow cytometry analysis, we were able to demonstrate NRP1 overexpression not only on CLL cells but also on Tregs and PDCs.

Recently, NRP1 was reported to play an important role in the healthy immune system. Tordjman et al.[26] showed that NRP1 mediates interactions between activated PDCs and resting T cells that are essential for initiation of the primary immune response. NRP1 was also found to be expressed on the surface of Tregs, which are a subset of CD4+ T lymphocytes playing a key role in the regulation of immune responses and the progression of cancer. Tregs are characterized by the high levels of surface CD25 antigen and the transcription factor FOXP3.[27, 28] Sarris et al.[19] suggest that the increased expression of NRP1 is associated with a prolonged time of interaction between the PDCs and Tregs in response to antigen, which may lead to tolerance. Notably, the levels of NRP1 on Tregs and PDCs were higher in CLL patients than HV suggesting that NRP1 might be involved in the induction of tolerance mechanisms in CLL. On one hand deregulated tolerance in CLL could reflect the ineffective response to the autoreactive clone of B lymphocytes that proliferate, as recent data strongly suggest that CLL might originate from autoreactive B lymphocytes. Therefore, NRP1 might represent another hallmark of immunosuppression in CLL. On the other hand, NRP1 might represent an interesting link between angiogenesis and tolerance mechanism. This potential functional link was further supported by findings of current study, that is, the correlation between expression levels of certain VEGF receptors and FOXP3 in a cohort of 114 CLL patients as well as negative regulation of NRP1 on Tregs after thalidomide. Thus, our report provides important evidence that NRP1 might be a novel target for antiangiogenic therapy of CLL.

It was shown that an increase in the circulation of Treg cells has been associated with cancer development and the progression of many types of cancer, including CLL. In our previous study, we reported that higher frequencies of Treg cells correlated with decreased T-cell response to tumor antigens expressed on CLL cells.[29] In accordance, increased frequencies of CD4+CD25high cells in CLL were also described by Beyer et al.[30] In our clinical trial, in which we treated CLL patients with thalidomide and fludarabine, we found that thalidomide monotherapy in first 7 days tended to decrease CLL cells and subsequently specifically reduced Tregs.[20, 31] Similarly, lenalidomide reduced Tregs effectively in a more recent study.[32] However, the mechanism of the specific reduction of Tregs during the therapy with immunomodulatory drugs is still unknown. Immunomodulatory drugs are known to inhibit angiogenesis, with that regard, NRP1 which is VEGF receptor might be a possible target responsible for this effect. As we proved, thalidomide tends to reduce VEGF levels in patients treated with thalidomide in vivo and thus downregulation of NRP1 might be either a direct or indirect effect of antiangiogenic therapy with immunomodulatory drugs. Indeed, in vitro, in functional analyses, during which we added thalidomide to Tregs, we found the highest reduction in NRP1 expression on CD4+CD25high cells. As NRP1 is expressed on CLL and immune cells responsible for the tolerance mechanisms, it might represent an interesting target for future therapeutic interventions in CLL.

In conclusion, we characterized NRP1 expression in CLL cells, and for the first time demonstrated that Tregs as well as PDCs in CLL patients also express high levels of NRP1. Furthermore, our data provide evidence that NRP1 expression might be regulated by VEGF levels in CLL. Thus, NRP1 might be involved not only in angiogenesis but also in tolerance mechanisms. In accordance, NRP1 might represent an interesting target for therapeutic approaches in CLL, as one would target both leukemic cells as well as the immune cells such as Tregs and PDCs that play a pivotal role in tumor escape mechanisms.