Nanoparticles Engineered with Rituximab and Loaded with Nutlin-3 Show Promising Therapeutic Activity in B-Leukemic Xenografts

The nongenotoxic activator of the p53 pathway Nutlin-3 induces the rapid accumulation of p53 protein and the consequent induction of cell-cycle arrest, senescence, and apoptosis in a cell-type–specific manner (1). Previous studies have shown that Nutlin-3 has little toxicity in animal models, suggesting that the selective nongenotoxic p53 activation by Nutlin-3 might represent an alternative to the current cytotoxic chemotherapy for a variety of p53^wild-type cancers, including hematologic malignancies (2, 3). The attractiveness of Nutlin-3 with respect to conventional chemotherapy is represented by the lower risk of Nutlin-3 to induce drug resistance and the absence of genotoxic damage. In this respect, it is noteworthy that Nutlin-3 is currently under phase I clinical trial as oral formulation, RO5045337 (protocol ID: NCT00623870), in several Centers located in Canada, United States, United Kingdom, and Italy. This study will determine the maximum tolerated dose of RO5045337 and the optimal associated 4-weekly dosing schedule of RO5045337, administered as monotherapy in patients with hematologic neoplasms.

Nutlin-3 is a cis-imidazoline compound soluble in organic solvents such as ethanol, dimethyl sulfoxide (DMSO), and dimethyl formamide, whereas it is sparingly soluble in aqueous buffers. So far, administration of Nutlin-3 in mouse models of disease has been conducted mostly per os, with a huge amount of drugs required (200–400 mg/kg twice a day) and some difficulties in giving the correct amount to each mouse (1, 4–7). To improve the delivery and efficacy of this drug for the treatment of solid tumors, recent studies have used Nutlin-3 encapsulated into poly(lactide-co-glycolide) (PLGA) nanoparticles (NP; refs. 8, 9), PLGA being the most used polymer for the preparation of drug delivery systems (10, 11).

Standard treatments for B-cell malignancies, such as B-chronic lymphocytic leukemia (B-CLL), the most common lymphoid malignancy in Western countries, include mono- or polychemotherapies, usually combined with monoclonal antibodies (Ab), such as anti-CD20 rituximab (12). In this respect, a recent meta-analysis showed that patients affected by B-CLL receiving chemotherapy plus rituximab benefit in terms of overall survival (OS) as well as progression-free survival (PFS) compared with those with chemotherapy alone, especially fludarabine or fludarabine and cyclophosphamide (R-FC regimen; ref. 13).

On these bases, the aim of our study was to assess in vitro and in vivo the use of Nutlin-3 encapsulated in PLGA NP (NP-Nut) and in rituximab-engineered NP (NP-Rt-Nut), with the hope of proposing an innovative strategy able to combine the therapeutic efficacy of Nutlin-3 with the specific targeting toward CD20⁺ malignant cells of rituximab. For this purpose, we have used the SCID-JVM-2 xenograft as a model of p53^wild-type B-CLL (14, 15), and more in general of B-cell malignancies, taking advantage of the fact that the CD20⁺ JVM-2 cell line has been well-characterized for the response to Nutlin-3 and other antileukemic molecules (16, 17).

Nutlin-3 was purchased from Cayman Chemical. Nutlin-3–loaded PLGA NPs (NP-Nut) were formulated by oil-in-water single emulsion–solvent evaporation technique with slight modifications (18, 19). In brief, a solution of 100 mg PLGA RG 503H (Boehring-Ingelheim) and 10 mg Nutlin-3 (10% w/w dry weight of polymer) in 3 mL of chloroform was emulsified in 12 mL of 1% w/v aqueous solution of polyvinyl alcohol (PVA; degree of hydrolysation 86–89 Mol%, viscosity of the 4% w/w water solution at 20°C 3 mPas; Fluka) to form an oil-in-water emulsion (19). Rituximab (anti-human CD20 Ab-Mabthera; Roche)-engineered NPs were prepared starting from unloaded (NP-Rt) or from Nutlin-3 loaded (NP-Rt-Nut) PLGA NPs applying the methodology for Ab surface engineering of NP previously described (19–21). Further details on NP-Rt preparation as well as on NP and NP-Rt physicochemical characterization are reported in Supplementary Materials and Methods.

JVM-2 human leukemic cells (American Type Culture Collection) were cultured in RPMI-1640 (Lonza) containing 15% FBS (Gibco BRL), 4.5 g/L glucose, 1 mmol/L sodium pyruvate, 1.5 g/L sodium bicarbonate, and 10 mmol/L HEPES (all from Gibco). For the different functional assays described in the subsequent paragraphs, cells were seeded at the concentration of 10⁶ cells/mL before treatment with NPs, NP-Nut, NP-Rt, and NP-Rt-Nut, all used at a concentration in the range of 0.06 to 0.07 mg of NP/mL and normalized for the drug content. As positive controls, and for comparison, cells were also treated with either free Nutlin-3 (Cayman Chemical) or free rituximab (anti-human CD20 Ab-Mabthera; Roche), based on the functional assay.

Analysis of p53 activation was conducted by Western blot analysis. For this purpose, JVM-2 cells were treated with different NP preparations or with free Nutlin-3, used as positive control, for 24 hours and lysed in ice-cold radioimmunoprecipitation assay (RIPA) buffer (50 mmol/L Tris, pH 7.5, 150 mmol/L NaCl, 0.1% SDS, 1% Nonidet P-40, 0.25% sodium deoxycholate) supplemented with protease inhibitors (Roche) on ice for 1 hour, as previously described (22). Protein determination was conducted by using the BCA Protein Assay (Thermo Scientific). Equal amounts of protein for each sample were migrated in acrylamide gels and blotted onto nitrocellulose filters (23). Hybridizations were conducted by using anti-p53 (DO-1), anti-MDM2 (SMP14), anti-p21 (C-19), and antitubulin monoclonal Ab (all from Santa Cruz Biotechnology). After incubation with peroxidase-conjugated anti-mouse or anti-rabbit IgG (Sigma-Aldrich), specific reactions were revealed with the ECL Lightning Detection Kit (Perkin Elmer).

At 24 and 48 hours posttreatment with the different NP preparations or with free Nutlin-3 used as positive control, cells were analyzed for total cell viability evaluated by Trypan blue dye exclusion, as previously described (24). In parallel, the degree of apoptosis was determined by Annexin V/7-aminoactinomycin D (7-AAD) double staining (BD Biosciences Pharmingen) and flow cytometric analysis (FACScan, Becton Dickinson), as previously described (25, 26). For analysis of cell-cycle profile, cells were incubated with 50 μmol/L 5-bromodeoxyuridine (BrdUrd; Sigma Aldrich) at 37°C for 1 hour, then anti-BrdUrd Ab (BD Biosciences Pharmingen) was bound to BrdUrd, and the complex was detected by fluorescein isothiocyanate (FITC)–conjugated secondary Ab (Beckman-Coulter). Cells were stained with 50 μg/mL propidium iodide (PI; Sigma-Aldrich) and analyzed by flow cytometry, as previously described (27).

For complement-dependent cytotoxicity (CDC) assay, experiments were carried out as previously described (28, 29) with some modifications. After cell exposure to the different NP preparations and after the addition of pooled normal AB HS (25%), cultures were incubated at 37°C for 24 and 48 hours before determination of the percentage of induced dead cells.

Comprehensive description of materials and methods used for analysis of complement activation, B-CLL mouse xenograft models, histopathologic and immunophenotypical analysis, and statistical analysis are reported in Supplementary Materials and Methods.

The composition and the main physicochemical characteristics of the different NP preparations are reported in Supplementary Table S1. Of note, the Nutlin-3 content in 100 mg of NP formulation ranged from 5.0 ± 0.7 mg (encapsulation efficiency, EE of 59% ± 7%) for NP-Nut, to 4.0 ± 0.8 mg (EE of 44% ± 8%) for rituximab-engineered NPs (NP-Rt-Nut). This difference, although not significant, could be the consequence of the diffusion of Nutlin-3, probably from the outer portion of the NPs, during the conjugation of the antibody and the following purification procedures. To support the presence of rituximab on NP surface, beside the conventional method of analysis of the diameter size based on photon correlation spectroscopy, NP preparations were analyzed by atomic force microscopy (AFM; Fig. 1A). With this method, surface alterations after conjugation with rituximab arise from the comparison of images taken on antibody-free and antibody-engineered NPs. Indeed, the topographical AFM images showed the antibody-free NP (NP) formed by separated elements and rituximab-engineered NP (NP-Rt) characterized by larger aggregates of irregular shape (Fig. 1A). Moreover, the surface of NP-Rt appeared rough and heterogeneous with presence of gaps and more complex areas (as indicated by the arrows in Fig. 1A). Analysis of the height of NPs in the 3-dimensional (3D) images showed that, accordingly with the size measured by PCS analysis, the diameters of antibody-free NPs were in the range of 150 to 200 nm, corresponding to elements with spherical shape. On the other hand, NP-Rt were characterized by an irregular contour and a diameter distributed into a wider range (100–300 nm).

The coupling of rituximab on the NP surface was indirectly suggested also by electron spectroscopy for chemical analysis (ESCA), conducted to measure the elemental composition of the NPs (Fig. 1B). Analyzing the atom spectra of antibody-engineered NP (NP-Rt) in comparison with antibody-free NP, it emerged that the spectra obtained after the conjugation with rituximab showed N signal (for nitrogen atoms), suggesting the presence of the antibody on NP surface, with coverage of about 7.4% ± 0.6% (percentage of derivatization). ESCA analysis was conducted also on NP obtained stopping the derivatization after the functionalization with NHS and EDC (that contain atoms of N), just before the addition of rituximab. Only nonquantifiable traces of N were observed in these samples, and the spectrum of the derivatized NP was superimposable with that of antibody-free NPs (data not shown), thus showing that the contribution of N in NP-Rt spectrum was really due to the presence of rituximab on NP surface.

The presence of functional rituximab on the surface of the NP-Rt was definitively documented by assessing the binding to CD20 antigen on the cellular surface. As shown in Fig. 1C, only NP-Rt, but not antibody-free NPs, were able to bind specifically the cellular surface of CD20⁺ (JMV-2) leukemic cells, whereas no significant binding was detected on CD20⁻ (OCI) leukemic cells. Moreover, analysis by SDS-PAGE and silver staining (Fig. 1D) indicated a content of rituximab on the NP-Rt estimated at approximately 80 ± 10 ng of antibody/100 μg of NPs (corresponding to approximately 3.089 femtomoles of antibody/NP). In parallel, the cellular internalization of PLGA NPs by endocytosis (30, 31) was confirmed by microscopic examination upon exposure of cells to NPs for 24 hours at 37°C, followed by intracellular analyses (Supplementary Fig. S1A and S1B).

To ascertain the functionality of the different NP preparations, NP-Nut, NP-Rt, and NP-Rt-Nut were analyzed for their ability to activate the p53 pathway in vitro using the p53^wild-type JVM-2 leukemic cell line as a model system of B-cell neoplasms. As shown in Fig. 2A, NP-Nut and NP-Rt-Nut showed a comparable ability to promote the accumulation of p53 protein and to induce the expression of p53-specific targets, MDM2 and p21, thus confirming that NP preparations contained functional Nutlin-3. Moreover, NP-Nut and free Nutlin-3 exhibited a comparable ability in terms of p53 induction, whereas no effect on p53 was observed upon exposure to empty control NP (Supplementary Fig. S2).

In parallel, the same NP preparations were tested for their capacity to trigger the complement cascade in the CD20⁺ JVM-2 cells. In the presence of human serum, while the antibody-free NPs (and NP-Nut) were unable to promote C3 fragment deposition (Fig. 2B), NP-Rt clearly induced the deposition on the cell surface of the C3 fragment (Fig. 2B). NP-Rt-Nut showed a profile of complement deposition indistinguishable from that of NP-Rt alone (data not shown). These data were remarkable because they indicated that, despite being engineered on the NP surface, rituximab was still able to trigger complement activation. To distinguish between classical or alternative complement activation pathways, we next analyzed whether NP-Rt were able to trigger C4 fragment deposition. The validation of the classical pathway of complement activation by NP-Rt (and NP-Rt-Nut), but not by NPs, came from the analysis of C4 fragment deposition, selectively induced by exposure to NPs engineered with rituximab (Fig. 2B).

The biologic effects mediated by Nutlin-3 encapsulated into NPs were next evaluated on JVM-2 cell cultures, by analysis of cell viability (Fig. 3A), apoptosis (Fig. 3B), and cell cycle (Fig. 3C). The preparations loaded with Nutlin-3 (NP-Nut and NP-Rt-Nut) were able to induce cytotoxic events resulting in reduction of cell viability >60% (Fig. 3A), arising from the combination of apoptosis induction and cell-cycle block (Fig. 3B and C). The results were quantitatively and qualitatively comparable for both NP-Nut and NP-Rt-Nut preparations, whereas NP alone and NP-Rt showed no significant cytotoxic effects with respect to untreated cultures (Fig. 3A–C). No interference due to the antibody presence was observed. Because it is known in literature that the anti-CD20 Ab rituximab does not exhibit significant cytotoxic effect on CD20⁺ B cells when used in vitro under standard culture conditions (32), cytotoxicity of rituximab-engineered NPs (NP-Rt and NP-Rt-Nut) was next evaluated in complement-dependent cytotoxicity (CDC) assays, conducted using normal human serum (25%) as source of complement (Fig. 3D). These experiments clearly indicated that rituximab exposed on NP surface (NP-Rt), by activating the complement cascade, was able to trigger more than 40% cell killing (Fig. 3D). Similar cell killing activity was observed upon exposure to NP-Rt-Nut (Fig. 3D).

For the in vivo evaluation of the therapeutic potential of NP-Nut, NP-Rt, and NP-Rt-Nut, we adopted a xenograft model generated in SCID mice upon subcutaneous injection with a predetermined optimal number (10⁷) of CD20⁺ JVM-2 leukemic cells (Fig. 4A). JVM-2 xenografts were characterized by subcutaneous tumors, which started to become palpable and measurable by external observation approximately 2 weeks after cell injections and steadily progressed until mice death with a median survival of 38 days after cell injection. Histopathologic examination of the subcutaneous masses showed that tumors had a solid pattern of growth of CD20⁺ cells (Fig. 4B), with rare infiltration of CD20⁺ cells in the liver, whereas other organs, such as kidneys and spleen, were unaffected (data not shown). We were first interested in determining the anti-leukemic effects of NP-Nut with respect to free Nutlin-3. Therefore, when tumors reached 50 mm³, JVM-2 xenograft mice were treated with empty NPs, NP-Nut, free Nutlin-3, or control vehicle. No differences in mean survival were observed between NP and vehicle mice (as documented in Supplementary Fig. S3), which are therefore reported in Fig. 4C as a single survival line (Controls) for better clarity. Of interest, treatment with NP-Nut promoted a significant (P < 0.05) increase in survival as compared not only to control xenografts but also with respect to mice treated with free Nutlin-3 (Fig. 4C and Table 1). On the other hand, free Nutlin-3 was unable to significantly improve survival with respect to the controls (Fig. 4C and Table 1).

Having established that NP-Nut exhibits antileukemic activity in vivo, we next investigated whether NPs functionalized with rituximab might increase the therapeutic efficacy of the NP loaded with Nutlin-3. In this second round of experiments, we observed that the antileukemic efficiency of NP-Rt in our xenograft model was comparable to that reported for NP-Nut in terms of mean survival (Table 1). Strikingly, treatment of JVM-2 xenografts with NP-Rt-Nut slowed down tumor growth kinetics (Fig. 4D) and further enhanced overall survival, with a significant (P < 0.05) increase in comparison to the treatments with either NP-Rt (Fig. 4E and Table 1) or NP-Nut (Table 1).

In the last years, the studies conducted by our group have been aimed to investigate the potential therapeutic effects of the innovative nongenotoxic activator of the p53 pathway Nutlin-3 for the treatment of hematologic malignancies including p53^wild-type B-CLL (2, 3). The studies conducted so far by our and other groups of investigators have strongly suggested that the preclinical evaluation of Nutlin-3 appears highly warranted (2, 3). Interestingly, to overcome the potential problems related to the poor solubility of Nutlin-3, it has been proposed to use PLGA NPs, which showed improved delivery and efficacy of Nutlin-3 in vitro in solid tumor cell models (8, 9). However, to the best of our knowledge, NPs loaded with Nutlin-3 have never been examined in animal models before the present study.

The use of polymeric NPs for drug delivery offers several obvious advantages, such as the possibility to deliver in the target site a great number of active molecules for each unit due to the drug loading and to the capacity of protecting the active molecules against chemical or enzymatic degradation (33). Another important advantage is the lack of systemic toxicity associated with the use of PLGA for drug delivery (10, 31). Indeed, PLGA is one of the most successfully used biodegradable polymers; its hydrolysis leads to metabolite monomers, lactic acid and glycolic acid, which are endogenous and easily metabolized by the body via the Krebs cycle. PLGA is approved by the U.S. Food and Drug Administration (FDA) and European Medicine Agency (EMA) in various drug delivery systems in humans (31). Active targeting of NPs to tumor cells has been successfully obtained by conjugating ligands to the surface of NPs. Particularly, antibody-engineered NPs have been shown to significantly enhance the efficacy of multiple anticancer drugs in several in vivo models of solid tumors (e.g., breast cancer, colon cancer and malignant gliomas; refs. 34, 35). Therefore, in consideration of the wide experience in using rituximab in the therapeutic approaches for the treatment of B-cell malignancies and in particular for B-CLL (12, 13), and taking into account the well-known properties of PLGA for drug delivery systems (10), we have planned to engineer PLGA NP with rituximab (NP-Rt). Subsequently, NPs loaded with Nutlin-3 were engineered with rituximab (NP-Rt-Nut). Each preparative NP-Nut, NP-Rt, and NP-Rt-Nut was chemically and functionally characterized evaluating in vitro the ability to activate the p53 pathway, to induce cellular apoptosis and inhibition of cell-cycle progression, as well as to trigger complement cascade activation. These in vitro functional studies were conducted by using the same cell type (CD20⁺ JVM-2) used for the xenograft transplants in which the in vivo analyses of the NP preparations were carried out.

The major findings of our study can be recapitulated as follows. First, the chemical and structural characteristics of all NPs tested in this study made them promising for subsequent in vivo applications. In particular, the hydrodynamic diameter around 200 nm is reported as optimal for efficient cellular uptake (36), whereas z-potential close to neutrality can prevent macrophage phagocytosis when carriers are administered in vivo (37). Second, by combining the morphologic and functional analyses, we documented the conjugation of rituximab on NP surface supporting the possibility of obtaining targeted carriers able to deliver antineoplastic drugs to B tumor cells. Third, NP loaded with Nutlin-3 exhibited good drug content (at about 5 mg/100 mg of formulation) and encapsulation efficiency (close to 50%). This percentage is lower with respect to that reported by previous authors (8, 9) and is likely a consequence of the repeated centrifugation required to purify the samples especially both from PVA and from the reagents used during the coupling with antibody. Finally, NP-Nut and NP-Rt-Nut were able to functionally activate the p53 pathway in vitro, as also shown by the induction of cell-cycle arrest and of apoptosis in JVM-2 cells. On the other hand, NP-Rt and NP-Rt-Nut, but not antibody-free NP or NP-Nut, were able to activate the classical complement cascade promoting CDC in vitro. Of great interest, in vivo administration of NP-Rt and/or NP-Nut in JVM-2 xenografts exhibited a comparable therapeutic efficacy (in terms of survival rate) that was significantly higher with respect to control mice. Moreover, the antileukemic efficacy of Nutlin-3 encapsulated in NPs was further enhanced (in a significant manner) by using the rituximab-engineered NPs loaded with Nutlin-3 (NP-Rt-Nut), in keeping with their in vitro ability to activate both the p53 pathway and the complement cascade, suggesting the occurrence of mechanisms of cell death summarized in Fig. 5.

For many years, nanotechnology has been recognized as an important potential tool for cancer therapy. An appropriate carrier should be able to protect drugs, such as Nutlin-3, from metabolic inactivation, ameliorating the delivery by using parenteral administration. In addition, the carrier should be particularly suitable for tumor targeting. In this context, we could show for the first time that an approach based on the combination of rituximab and Nutlin-3 on the same NPs was not only feasible but also gave significant advantages in terms of survival of the treated animals, probably due to the better targeting of the B neoplastic cells and to the combination of the proapoptotic activity of Nutlin-3 coupled to the ability of rituximab to activate the complement-mediated cell death also when engineered on the surface of PLGA NPs (Fig. 5).

No potential conflicts of interest were disclosed.

Conception and design: R. Voltan, P. Secchiero, M.A. Vandelli, G. Zauli

Development of methodology: B. Ruozi, C. Agostinis

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): R. Voltan, P. Secchiero, B. Ruozi, C. Agostinis, L. Caruso

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): R. Voltan, P. Secchiero, C. Agostinis, L. Caruso, G. Zauli

Writing, review, and/or revision of the manuscript: R. Voltan, P. Secchiero, B. Ruozi, F. Forni, G. Zauli

Study supervision: P. Secchiero, M.A. Vandelli, G. Zauli

This study was supported by grants from Italian Association for Cancer Research (AIRC IG 11465 to G. Zauli) and from MIUR-FIRB (RBAP11Z4Z9_002 to G. Zauli; RBAP10447J_002 to P. Secchiero).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.