Theranostics 2020, Vol. 10, Issue 12 5169 Ivyspring International Publisher Theranostics 2020; 10(12): 5169-5180. doi: 10.7150/thno.43231 Research Paper Selective delivery of T22-PE24-H6 to CXCR4+ diffuse large B-cell lymphoma cells leads to wide therapeutic index in a disseminated mouse model Aïda Falgàs1,2,3, Victor Pallarès1,2,3, Naroa Serna4,5, Laura Sánchez-García4,5, Jorge Sierra2,6, Alberto Gallardo1, Lorena Alba-Castellón1, Patricia Álamo1,3, Ugutz Unzueta1,3, Antonio Villaverde3,4,5, Esther Vázquez3,4,5, Ramon Mangues1,2,3 and Isolda Casanova1,2,3 1. Biomedical Research Institute Sant Pau (IIB-Sant Pau), Hospital de la Santa Creu i Sant Pau, Barcelona Spain. 2. Josep Carreras Leukaemia Research Institute (IJC), Barcelona, Spain. 3. CIBER de Bioingeniería Biomateriales y Nanomedicina (CIBER-BBN), Barcelona, Spain. 4. Institute of Biotechnology and Biomedicine (IBB), Universitat Autònoma de Barcelona, Barcelona, Spain. 5. Department of Genetics and Microbiology, Universitat Autònoma de Barcelona, Barcelona, Spain. 6. Department of Hematology, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain.  Corresponding authors: R. Mangues: Biomedical Research Institute Sant Pau (IIB-Sant Pau), Josep Carreras Leukaemia Research Institute (IJC) and CIBER-BBN, Barcelona, Spain. Tel.: +34 935537918. Email address: rmangues@santpau.cat or E. Vazquez: Institute of Biotechnology and Biomedicine (IBB) and Department of Genetics and Microbiology, Universitat Autònoma de Barcelona and CIBER-BBN, Barcelona, Spain. Tel.: +34 5812148. Email address: esther.vazquez@uab.es © The author(s). This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/). See http://ivyspring.com/terms for full terms and conditions. Received: 2019.12.19; Accepted: 2020.03.17; Published: 2020.04.06 Abstract Background: Novel therapeutic strategies are urgently needed to reduce relapse rates and enhance survival in Diffuse Large B-Cell Lymphoma (DLBCL) patients. CXCR4-overexpressing cancer cells are good targets for therapy because of their association with dissemination and relapse in R-CHOP treated DLBCL patients. Immunotoxins that incorporate bacterial toxins are potentially effective in treating haematological neoplasias, but show a narrow therapeutic index due to the induction of severe side effects. Therefore, when considering the delivery of these toxins as cancer therapeutics, there is a need not only to increase their uptake in the target cancer cells, and their stability in blood, but also to reduce their systemic toxicity. We have developed a therapeutic nanostructured protein T22-PE24-H6 that incorporates exotoxin A from Pseudomonas aeruginosa, which selectively targets lymphoma cells because of its specific interaction with a highly overexpressed CXCR4 receptor (CXCR4+) in DLBCL. Methods: T22-PE24-H6 cytotoxicity and its dependence on the CXCR4 receptor were evaluated in DLBCL cell lines using cell viability assays. Different in vitro experiments (mitochondrial membrane potential, Western Blot, Annexin V and DAPI staining) were conducted to determine T22-PE24-H6 cell death mechanisms. In vivo imaging and therapeutic effect studies were performed in a disseminated DLBCL mouse model that mimics organ infiltration in DLBCL patients. Finally, immunohistochemistry and histopathology analyses were used to evaluate the antineoplastic effect and systemic toxicity. Results: In vitro, T22-PE24-H6 induced selective cell death of CXCR4+ DLBCL cells by activating the apoptotic pathway. In addition, repeated T22-PE24-H6 intravenous administration in a CXCR4+ DLBCL-disseminated mouse model showed a significant reduction of lymphoma burden in organs clinically affected by DLBCL cells (lymph nodes and bone marrow). Finally, we did not observe systemic toxicity associated to the nanoparticle treatment in non-DLBCL-infiltrated organs. Conclusion: We have demonstrated here a potent T22-PE24-H6 antineoplastic effect, especially in blocking dissemination in a CXCR4+ DLBCL model without associated toxicity. Thereby, T22-PE24-H6 promises to become an effective alternative to treat CXCR4+ disseminated refractory or relapsed DLBCL patients. Key words: targeted nanoparticle, CXCR4 receptor, PE24 exotoxin, DLBCL http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5170 Introduction increase the therapeutic window of immunotoxins. Our approach is highlighted in Figure 1. Firstly, we Over the last years, different immunotoxins that evaluate the in vitro cytotoxicity of T22-PE24-H6 in incorporate highly cytotoxic bacterial proteins have different CXCR4+ DLBCL cell lines and its entered clinical trials [1,2]. Only two of them, dependence on CXCR4 receptor expression. In moxetumomab pasudotox-tdfk and tagraxofusp-erzs, addition, we analyze the cell death type induced by obtained approval from the FDA for the treatment of T22-PE24-H6 and, most importantly, we evaluate the hairy cell leukemia (HCL) and blastic plasmacytoid in vivo T22-PE24-H6 antineoplastic effect in DLBCL- dendritic cell neoplasm (BPDCN), respectively, infiltrating organs, lymph nodes (LNs) and bone despite of the induction of severe side effects [3–5]. In marrow (BM), and its systemic toxicity in a fact, another immunotoxin (denileukin diftitox), disseminated mouse model. This novel approach aims approved to treat patients with cutaneous T-cell to increase the cure rates and reduce the toxicity in lymphoma (CTCL), was withdrawn from the market CXCR4+ DLBCL patients. in 2014 because of life-threatening toxicity [6,7]. In this context, clinical translation of antibody drug Methods conjugates (ADCs) and especially immunotoxins is being limited by severe off-target toxicity that leads to Production and purification of therapeutic a narrow therapeutic window, most likely because T22-PE24-H6 polypeptidic nanoparticle only 0.001-0.01% of the injected dose (ID) reaches the All the details of T22-PE24-H6 nanoarchitecture tumor [8–10]. and production were described in our previous work Aiming to enhance tumor cell uptake while [11]. reducing toxicity on normal tissues, we have developed a multimeric and self-assembled CXCR4- Cell culture and cytotoxicity assays targeted protein-based nanoparticle for the selective Human DLBCL cell lines (Toledo, SUDHL-6, delivery of toxins [11]. This approach increases toxin U-2932 and SUDHL-2) were cultured at 37 ºC in a 5% loading capacity and the number of targeting ligands CO2 humidified atmosphere. Toledo, SUDHL-6 and per nanoparticle, leading to an increase in U-2932 cells were cultured in RPMI 1640 medium, and nanoparticle internalization and toxin delivery to SUDHL-2 cells in IMDM medium. Culture media for target cancer cells through its specific receptor [12,13]. all cell lines were supplemented with 10% FBS, 1% In our previous work, we have demonstrated that the glutamine and 100 U/mL penicillin-streptomycin targeted protein nanocarrier T22-GFP-H6 avoids renal (Life Technologies). Toledo cells were purchased from filtration (>7 nm) and liver clearance, whereas most of ATCC, U-2932 and SUDHL-6 from DSMZ, and the nanocarrier proteolytic metabolism occurs in SUDHL-2 were kindly provided by Dr L. Pasqualucci tumor tissues [14,15]. In this context, we have (Columbia University, NY, USA). generated a stable nanoparticle, T22-PE24-H6, that Moreover, the Luciferase plasmid (pPK-CMV- incorporates the toxin in a single polypeptide. F3, Promokine) was transfected by electroporation T22-PE24-H6 specifically delivers a de-immunized 24 (Nucleofector TM 2b Device, Lonza) into Toledo cells kDa catalytic domain of Pseudomonas aeruginosa (PE) (Toledo-Luci cells). The expression of the Luciferase exotoxin A to the chemokine receptor CXCR4 gene in transfected cells was detected by the IVIS overexpressing (CXCR4+) cancer cells by interacting Spectrum 200 Imaging System (PerkinElmer). Stable with its T22 ligand [11,16]. These cancer cells are clones were obtained by selection in medium relevant clinical targets since CXCR4 overexpression containing 0.2 mg/mL of geneticin (G418, Life is associated with aggressiveness and dissemination Technologies) for a period of 10 weeks. in many solid and hematological cancers [17–21]. In vitro T22-PE24-H6 cytotoxicity in DLBCL cells Indeed, T22-PE24-H6 could improve treatment was evaluated measuring cell metabolic capacity outcomes in CXCR4+ diffuse large B-cell lymphoma using the colorimetric cell proliferation kit II (XTT, (DLBCL) patients, because of their association with Roche Diagnostics). 30∙104 cells (Toledo, SUDHL-6, poor progression-free as well as overall survival in U-2932 and SUDHL-2) were seeded into 96‐well R-CHOP treated patients; and also because CXCR4+ plates in 100 μL of media and incubated at 37 °C for 24 DLBCL cells are responsible for relapse and resistance h. Then, cells were exposed to different T22-PE24-H6 to R-CHOP [22–24]. Currently, no protein-based concentrations (0.1-5 nM) or carbonate buffer (166 targeted therapeutic nanoparticle has been developed mM NaCO3H pH=8) for 48 h and assessed for to treat disseminated or therapy-resistant DLBCL. viability. The competitive assays were done by Here, we determine the antineoplastic effect of pre-incubating the cells with AMD3100 (ratio the T22-PE24-H6 nanoparticle actively targeting 1T22-PE24-H6:10AMD3100). Fifty μL of the mixture CXCR4+ DLBCL cells to evaluate whether it could http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5171 XTT reagent were added to each well and, after 4 h SuperSignal West Pico Chemiluminescent Substrate incubation, cell viability was quantified by measuring (Thermo Fisher Scientific) and the G:BOX iChemi XT the absorbance at 450 nm wavelength using a spectro- Imaging System (Syngene). photometer (BMG Labtech). Results were expressed Cleaved PARP and active caspase-3 protein as percentage of cell viability in relation to its buffer. levels were quantified using the ImageJ software. The relative density of the protein bands was calculated Western blotting dividing the percent value of each treated sample (7 h, Toledo cells were treated with buffer or 5 nM 15 h, 24 h and 48 h; 5 nM T22-PE24-H6) by the T22-PE24-H6 for different time exposures (7, 15, 24 standard sample (Buffer). To get the normalized and 48 h). Cells were washed twice with PBS and density value, the relative density of each sample lane resuspended in lysis buffer. Then, the suspension was was divided by the relative density of its loading sonicated and rested for 20 min on ice. Cells were control (GAPDH). centrifuged for 10 min at 14000 rpm and 4 ºC. Protein concentration in supernatant was determined using Apoptosis assessment the Bradford protein assay, according to the For the evaluation of JC-1 staining in control and manufacturer’s instructions (BioRad). Cell lysates (50 apoptotic cells, the instruction manual of flow μg) were separated using 12-15% SDS-PAGE and cytometry mitochondrial membrane potential detec- transferred to a nitrocellulose blotting membrane (GE tion kit (BD™ MitoScreen Kit) was followed. Results Healthcare life sciences). Membranes were blocked were analyzed by FACS Calibur flow cytometer and with 5% skim milk in TBST for 2 h at room the Cell Quest Pro software (BD Biosciences). temperature, and then incubated with primary Additionally, the number of cells undergoing antibodies, 1:2000 PARP (556494, BD), 1:1000 apoptosis (early or late) was quantified using the caspase-3 (610322, BD) or 1:10000 GAPDH (MAB374, Annexin V-FITC/propidium iodide (PI) detection kit Millipore). Membranes were washed with TBST and (Merck Millipore), following the manufacturer’s then incubated with the appropriate secondary instructions. Data were analyzed by MACSQuant antibody (1:10000, Jackson Immune Research). analyzer flow cytometry using the MACS Quantify Western blot visualization was performed using the version 2.3 software (Miltenyi Biotec). Figure 1. Graphical image visualizing the highly selective targeting and high cytotoxicity induced by the T22-PE24-H6 nanoparticle on CXCR4+ cancer cells in a disseminated DLBCL mouse model. The image describes critical characteristics of the T22-PE24-H6 polypeptidic nanoparticle that leads to its high CXCR4+ DLBCL-cell uptake within LNs and BM. This nanoparticle reaches the neoplastic tissues without being proteolyzed in the liver or excreted by the kidneys. Once in the affected organ, T22-PE24-H6 interacts with the CXCR4 receptor in lymphoma cells, to induce its internalization by endocytosis and its traffic to Golgi and endoplasmic reticulum (ER). There, the PE24 toxin inactivates EF-2, which inhibits protein synthesis and consequently induces cancer cell death by apoptosis. BM: bone marrow; DLBCL: diffuse-large B-cell lymphoma; EF-2: elongation factor 2; LNs: lymph nodes; PE: Pseudomonas aeruginosa. http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5172 DAPI staining ImageJ software. Area ratio was calculated in relation Lymphoma cells were seeded at 3·105 cells/mL to buffer-treated LNs. (6 mL) in a cell culture dish and incubated at different Histopathology and immunohistochemical times (15, 24 and 48 h) with 5 nM T22-PE24-H6 staining nanoparticle or buffer. Afterwards, cells were All paraffin-embedded organs were stained with centrifuged at 400 g for 5 min, washed with PBS, H&E to perform a complete histopathological analysis centrifuged again, resuspended with 1 mL of 3.7% and a clinical pathologist supervised all samples for paraformaldehyde and fixed for 10 min at -20 ºC. toxicity evaluation. To detect the presence of human Then, cells were washed, resuspended with ~10 µL of B-cells and to quantitate the percentage of organ PBS and placed on a slide. Finally, slides were stained infiltration, we used immunohistochemistry (IHC) with DAPI mounting medium (Thermo Fisher staining with an antibody anti-human CD20cy (clone Scientific) and visualized using a fluorescence micro- L26, Dako). Firstly, 5 and 27 low power fields at 100 scope (Olympus BX53, Olympus). Representative magnifications were taken for LNs and BM, pictures were taken using an Olympus DP73 digital respectively, in each mice group (buffer or T22-PE24- camera and processed with the cellSens Dimension H6). Secondly, the area occupied by B cells in each 1.9 software (Olympus) at 1000 magnifications. organ was selected and the percentage of CD20+ cells In vivo antineoplastic effect in a DLBCL was quantified using the cellSens Dimension 1.9 disseminated mouse model software (Olympus). Four-week-old female NOD-SCID mice were IHC staining was performed in a DAKO obtained from Charles River Laboratories (L-Abreslle, Autostainer Link48 (Agilent) following the France), maintained in SPF conditions and housed manufacturer’s instructions. Representative pictures with sterile food and water ad libitum. In vivo were taken using an Olympus DP73 digital camera experiments were approved by the Hospital de Sant and processed with the cellSens Dimension 1.9 Pau Animal Ethics Committee. After one week in software (Olympus) at 200 or 400 magnifications. quarantine, mice were intravenously injected with Statistical analysis luciferase-transfected Toledo cells (Toledo-Luci cells, All values are expressed as mean ± standard 20·106/200 µL) and divided randomly in two groups. error (SE). Differences between groups were analyzed Three days after cell injection, one group was using the Mann-Whitney U test and were considered administered intravenously with 166 mM NaCO3H statistically significant at p≤0.05. Statistical pH 8 buffer (buffer-treated mice, n=9) and the second calculations were performed using SPSS software v21. group with 5 µg of T22-PE24-H6 (T22-PE24-H6- treated mice, n=9). Both groups were injected three Results times per week for a total of thirteen doses. Lymphoma dissemination was monitored using CXCR4-dependent antineoplastic effect of bioluminescence imaging (BLI, total radiance T22-PE24-H6 in DLBCL cell lines photons) technique in the IVIS Spectrum (Perkin- Levels of CXCR4 membrane expression in the Elmer), once per week. Mice were anesthetized with four tested DLBCL cell lines have been previously 3% isoflurane in oxygen and BLI was captured 5 min reported [15,25]. Toledo cells showed the highest after intraperitoneal injection of firefly D-luciferin levels of CXCR4 expression followed by U-2932, (2.25 mg/mouse, Perkin Elmer). Moreover, mouse SUDHL-6, whereas SUDHL-2 did not express body weight was registered twice a week until the detectable levels of the receptor in their membrane. end of the experiment. Thus, we firstly evaluated the sensitivity of these cell All mice were euthanized when the first mouse lines to the T22-PE24-H6 nanoparticle and its possible presented relevant signs of disease such as poor correlation with their CXCR4 membrane expression. mobility or significant weight loss. At that day, the To that aim, in vitro cytotoxic assays were performed BLI of lymphoma-infiltrated organs, cervical and after 48 h exposure to different nanoparticle concen- renal LNs as well as BM was analyzed ex vivo. Finally, trations (0.1-5 nM). Results showed that T22-PE24-H6 all organs were collected and fixed in 4% therapeutic nanoparticle, at low concentrations, had formaldehyde for further histopathological or an antineoplastic effect in Toledo, SUDHL-6 and immunohistochemical evaluations. U-2932 cells, which express high levels of CXCR4 Cervical LNs size assessment receptor in their membrane (CXCR4+ DLBCL cell lines). In contrast, exposure of SUDHL-2 cells Cervical LNs area was calculated for 18 cervical (CXCR4- DLBCL cell line) to this nanoparticle did not LNs in each group (buffer or T22-PE24-H6) using induce any cytotoxic effect (Figure 2A). http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5173 Figure 2. T22-PE24-H6 induces cytotoxicity in CXCR4+ DLBCL cell lines. (A) In vitro cell viability (%) of different DLBCL cell lines after exposure to T22-PE24-H6 (0.1-5 nM) for 48 h. (B) Competition cell viability assays done by 1 h pre-treatment with 50 nM AMD3100 followed by the addition of 5 nM T22-PE24-H6 during 48 h exposure in DLBCL cell lines. Toledo, SUDHL-6 and U-2932 are CXCR4+ DLBCL cell lines, whereas the SUDHL-2 cell line does not express CXCR4. Experiments were performed in biological triplicates and results expressed as mean ± SE. *p≤0.05. On the other hand, we studied whether the cell at 48 h because of a concomitant increase in late death induced by the T22-PE24-H6 nanoparticle in apoptosis from 22.0±3.8% at 24 h to 62.4±1.3% at 48 h CXCR4+ DLBCL cells was dependent on its (Figure 3C-D). Moreover, we used western blot to internalization through the CXCR4 receptor. In this evaluate the proteolysis of pro-caspase-3 and PARP as context, competition assays were performed by markers of apoptosis induction. The exposure of pre-incubating the cells with AMD3100, a CXCR4 CXCR4+ DLBCL cells to T22-PE24-H6, during the 15 h antagonist, 1 h before the addition of T22-PE24-H6, at to 48 h exposure period, caused a decrease of the a molar ratio of 10:1. Cell exposure with AMD3100 pro-caspase-3 and PARP expression together with the blocked nanoparticle binding to CXCR4 and increase of both cleaved caspase-3 and cleaved PARP produced a complete reversion of cell viability that occurred at the final stage of cell death (Figure (~100%) in all three CXCR4+ DLBCL cell lines (Figure 3E-F). To further confirm apoptosis as the cell death 2B). mechanism, we performed nuclear DAPI staining and Hence, T22-PE24-H6 displays a potent in vitro observed the expected increase of apoptotic bodies antineoplastic effect only in the DLBCL cell lines that over time (Figure 3G). overexpress CXCR4, since endocytosis of the On this basis, the cell death mechanism induced nanoparticle is exclusively mediated through the by T22-PE24-H6 in Toledo CXCR4+ DLBCL cell line is CXCR4 receptor. mediated by apoptosis activation. T22-PE24-H6 induces apoptosis-mediated cell T22-PE24-H6 antineoplastic effect in the death in CXCR4+ DLBCL cells bioluminescent CXCR4+ DLBCL disseminated After demonstrating the cytotoxic effect of the mouse model nanoparticle in CXCR4+ DLBCL cells, we performed Afterwards, we also determined the different assays to identify the mechanism of cell antineoplastic effect of T22-PE24-H6 in a death induction. Firstly, we analyzed whether bioluminescent CXCR4+ DLBCL disseminated mouse depolarization of mitochondria was involved as one model. This animal model, which was generated by of the first events in the cell death mediated by the intravenous injection of Toledo-Luci cells and T22-PE24-H6 nanoparticle. Thus, we monitored described in detail in our previous work, shows mitochondrial membrane potential by JC-1 dye. We CXCR4 overexpression in all organs infiltrated by observed that after 15 h exposure of Toledo cells to 5 lymphoma cells [15]. Here, we started the intravenous nM T22-PE24-H6, 44.6±5.3% of cells showed administration of 5 µg T22-PE24-H6 or buffer (3 depolarization of mitochondrial membrane and this times/week and a total of 13 doses) in this mouse percentage increased up to 65.5±7.1% at 48 h (Figure model 3 days after the injection of Toledo-Luci cells. 3A-B). Secondly, to evaluate whether the nanoparticle Lymphoma dissemination was monitored until the induced cell death through the apoptotic pathway, we end of the experiment by capturing the BLI emitted by performed the AnnexinV-FITC/PI test. We Toledo-Luci cells using the IVIS Spectrum (Figure determined that the percentage of cells in early 4A). We observed a lower total flux of BLI during the apoptosis increased from 29.0±4.4% at 15 h to whole follow-up period in nanoparticle-treated mice 35.5±8.1% at 24 h, decreasing considerably afterwards than in buffer-treated mice. Indeed, this low BLI http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5174 signal in nanoparticle-treated mice indicated a compared to buffer-treated animals (Figure 4B-C). reduction of lymphoma-cell dissemination, which Finally, no significant differences in mouse body was already significant at day 14 after cell injection weight were found between groups (Figure 4D). and became highly significant at days 21, 28 and 33 Figure 3. Apoptosis induction after T22-PE24-H6 exposure in CXCR4+ Toledo DLBCL cell line. (A) Representative dot-plot, showing the percentage of cells undergoing mitochondrial membrane depolarization for this particular replicate (mean percentage is described in the Results text), and (B) quantification of loss of JC-1 red fluorescence (%) in Toledo cells after 15 h, 24 h or 48 h exposure to buffer or 5 nM T22-PE24-H6 nanoparticle. (C) Representative dot-plot, showing percentages related to this replicate, and (D) quantification of viable cells (V), at early or late apoptosis (EA or LA), using Annexin V-FITC/PI test, after 15 h, 24 h and 48 h exposure to 5 nM T22-PE24-H6 or buffer in Toledo cells. (E) Toledo cells treated with buffer or 5 nM T22-PE24-H6 for 7 h, 15 h, 24 h and 48 h, and subjected to Western blot using PARP and caspase-3 antibodies. GAPDH antibody is used as an internal control. (F) Relative protein intensity quantitation of cleaved caspase-3 and cleaved PARP expression normalized to GAPDH. (G) DAPI staining of Toledo cells after exposure to buffer or 5 nM T22-PE24-H6 for 15 h, 24 h and 48 h. Original magnification x1000. Experiments were performed in biological triplicates and results expressed as mean ± SE. *p≤0.05. http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5175 Figure 4. In vivo and ex vivo BLI analysis of Toledo-Luci disseminated mice treated with buffer or T22-PE24-H6 therapeutic nanoparticle. (A) Experimental design followed along the in vivo experiment. (B) Representative images of mice with disseminated lymphoma registered by IVIS Spectrum (BLI signal) at day 21, 28, and 33 after Toledo-Luci cells IV injection in mice treated with buffer or T22-PE24-H6. (C) BLI quantification over the experimental time in mice treated with buffer (n=9) or T22-PE24-H6 (n=9). (D) Body weight of mice treated with buffer or T22-PE24-H6 during the treatment period. (E) Ex vivo BLI quantitation (left) and representative images (right) of lymphoma-infiltrated organs (cervical LNs, renal LNs and BM) for both groups at the end of the experiment. Results are expressed as mean ± SE. BLI: bioluminescence imaging (total flux); BM: bone marrow; IV: intravenous; LNs: lymph nodes; *p≤0.05; **p≤0.01; ***p≤0.005. http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5176 T22-PE24-H6 nanoparticle antineoplastic Lack of toxicity by T22-PE24-H6 in the effect in CXCR4+ DLBCL-infiltrated organs in non-infiltrated-DLBCL organs the disseminated mouse model Finally, once we demonstrated the T22-PE24-H6 At the day of euthanasia (day 34), when the effectiveness to CXCR4+ lymphoma cells we analyzed lymphoma was totally disseminated in buffer-treated its harmlessness in non-affected tissues. We did not mice, we analyzed ex vivo the bioluminescence find any alteration produced by the T22-PE24-H6 emitted by the lymphoma-infiltrated organs in both nanoparticle at the macroscopic level (data not groups (buffer or T22-PE24-H6). We detected a highly shown). At microscopic level (Figure 6), we analyzed significant antineoplastic effect in the cervical LNs in the histopathology (H&E staining) of non-infiltrated T22-PE24-H6 treated mice, observing a 14 fold lymphoma tissues in order to carefully assess the reduction (0.27±0.09E5) compared to the BLI emitted possible toxicity of the nanoparticle. Regarding the in buffer-treated mice (3.76±0.97E5). Moreover, the spleen, we observed the same hematopoietic tissue reduction of lymphoma infiltration was observed also prevalence in buffer-treated than in nanoparticle- in the renal LNs and BM, displaying 5 and 6 times less treated mice. Pancreas and heart tissues were also as BLI total flux in the animals treated with the healthy in nanoparticle-treated as buffer-treated therapeutic nanoparticle (1.07±0.11E4 and 0.38±0.09E6, animals. Furthermore, we did not observe congestion, respectively) compared to the buffer group (5.64± edema or intraalveolar hemorrhage in the lung tissue 1.68E4 and 2.29±0.81E6, respectively) (Figure 4E). in nanoparticle-treated mice. In the liver tissue of Thus, a significant reduction in lymphoma nanoparticle-treated animals, the hepatocytes did not dissemination in T22-PE24-H6-treated mice compared lose their architecture and we did not observe to mice treated with buffer is observed in the organs steatosis or any sign of histological alteration. Finally, infiltrated by CXCR4+ DLBCL cells (LNs and BM). the glomerulus and surrounding renal tubules in nanoparticle-treated mice were clearly visible without T22-PE24-H6 reduces the load of CXCR4+ cytoplasmic vacuolation or eosinophilic protein DLBCL lymphoma cells in affected organs accumulation. Thus, we did not find any Furthermore, we observed macroscopic morphological changes associated with nanoparticle differences in the cervical LNs size between mice administration that could indicate toxicity in treated with nanoparticle or buffer at the end of the non-infiltrated organs. experiment. In this context, the cervical LNs size in To sum up, the therapeutic nanoparticle, at the nanoparticle-treated mice was significantly reduced chosen administration conditions, does not induce more than halfway (47.7±3.2%) compared to the off-target toxicity in the Toledo-Luci disseminated buffer-treated mice (100±4.4%) (Figure 5A). This size mouse model. reduction was related to the lower number of Toledo-Luci cells infiltrating the cervical LNs, as we Discussion demonstrated by IHC using the human B-cell CD20 The administration of the T22-PE24-H6 marker. In buffer-treated mice we found that 94.24± therapeutic nanoparticle at low doses shows a potent 0.47% of cells within cervical LNs were Toledo-Luci antineoplastic effect in a disseminated DLBCL mouse cells (CD20+), whereas the percentage of CD20+ cells model without associated toxicity. Both outcomes are diminished significantly in the nanoparticle-treated achieved by the selective elimination of CXCR4+ mice, reaching only 7.53±9.23% (Figure 5B and E). DLBCL cells, which leads to the control of DLBCL Accordingly, T22-PE24-H6 also induced antineo- growth and dissemination. This is based on the plastic effect in renal LNs, since the percentage of observation of a lower percentage of CD20+ cells lymphoma cells was reduced from 69.89±5.01% in within CXCR4+ DLBCL infiltrated organs in mice treated with buffer to 11.31± 10.83% after nanoparticle-treated animals, while observing no repeated doses of the nanoparticle (Figure 5C and E). histopathological alteration in non-affected normal Regarding the BM, we also observed a significant organs. reduction of the Toledo-Luci cells after T22-PE24-H6 These findings are consistent with our recent treatment. We quantified 71.62±2.50% CD20+ cells demonstration of a high uptake for the T22-GFP-H6 infiltrating the BM in buffer-treated mice, a nanocarrier in subcutaneous (SC) tumors generated percentage that was reduced to 45.13±3.25% after by CXCR4+ lymphoma cells (86.1% of the total ID) and T22-PE24-H6 treatment (Figure 5D-E). its low or negligible biodistribution to non-DLBCL Definitely, and consistently with BLI detection, affected organs (13.9% of the total ID) [15]. This high we demonstrate the antineoplastic effect of T22-PE24- uptake may improve the performance of ADCs, for H6 in CXCR4+ DLBCL-infiltrated organs by mea- which less than 1% ID reaches the tumor, while most suring CD20+ cells in a disseminated mouse model. of the ID is catabolized by the liver and the http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5177 reticuloendothelial system [26,27]. An improved and BM) can profit from the enhanced permeability T22-GFP-H6 uptake was also shown in CXCR4+ and retention (EPR) effect [11,28–30]. Thus, the high lymphoma cells affecting LNs and BM in a T22-PE24-H6 nanoparticle tumor uptake may take disseminated DLBCL mouse model [15]. Indeed, the advantage of the EPR effect, the discontinued endo- T22-PE24-H6 nanoparticle size (60.17 nm) and the thelia in the relevant clinical organs and the active sinusoidal capillaries present in DLBCL niches (LNs targeting to high CXCR4 overexpressing DLBCL cells. Figure 5. Macroscopic and microscopic analysis of T22-PE24-H6 antineoplastic effect in the Toledo-Luci disseminated mouse model. (A) Cervical LNs images in buffer and T22-PE24-H6-treated mice (above) and quantification of cervical LNs area (n=18/group), expressed as area ratio, for both groups (below). (B-D) Representative images of H&E staining and CD20 IHC (human DLBCL cells) of lymphoma-infiltrated organs (cervical LNs, renal LNs and BM) in buffer and T22-PE24-H6-treated mice. Original magnification x200 and insets at x400. (E) Quantitation of the percentage of CD20+ cells in cervical LNs (n=5 fields/group), renal LNs (n=5 fields/group) and BM (n=27 fields/group) tissues for both mouse groups. Results are expressed as mean ± SE. BM: bone marrow; H&E: hematoxylin and eosin; IHC: immunohistochemistry; LNs: lymph nodes; ***p≤0.005. http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5178 Figure 6. Toxicity analysis of T22-PE24-H6 treatment in the Toledo-Luci disseminated mouse model. Histopathology analysis (H&E staining) of non-infiltrated organs (spleen, pancreas, heart, lungs, liver and kidneys) in mice treated with buffer or T22-PE24-H6. Original magnification x200 and insets at x400. H&E: hematoxylin and eosin. We argue that the potent antineoplastic effect by apoptosis [11,37,38]. Interestingly, this mechanism observed in the absence of toxicity in the of action permits the killing of dividing and disseminated mouse model relates also to the non-dividing cancer cells, differing from the cell compactness of the self-assembled nanoparticle cycle-dependent killing induced by genotoxic structure, composed of fusion-protein monomers that chemotherapy [2]. Consistently, using the same incorporate the toxin in a single polypeptide chain, to approach, our group previously proved T22-PE24-H6 prevent leakage in the bloodstream [11,31]. Regarding capacity to selective eliminate CXCR4+ colorectal this issue, we recently determined that the cancer cells in a SC tumor model [11,16]. T22-PE24-H6 protein integrity, and its full-length size Regarding the use of PE exotoxin A in current (29.2 kDa), is maintained in human blood for at least cancer therapies, PE38-based immunotoxins have 10 days [32]. Moreover, this exceptional multimeric been tested in clinical trials against different single polypeptide, which recruits all required hematological cancers [1], being one of them already functional domains, allows nanoparticle purification approved to treat relapsed or refractory HCL patients in a single step, avoiding the need to chemically [3]. However, due to the use of an unstable linkage, conjugate the cytotoxic payload to the carrier while the toxin can be prematurely released from the carrier facilitating its fabrication in endotoxin-free bacterial antibody while is circulating in the bloodstream, systems [33,34]. which together with the immunogenicity induced by Apart from the polypeptidic nanoparticle the toxin itself limit the dosage that can be used to structure, the constitutive overexpression of the treat patients [9,39,40]. For instance, the US CXCR4 receptor in many available DLBCL cell lines FDA-approved moxetumomab pasudotox-tdfk [23,35] and around 30-50% of malignant B-cell immunotoxin causes infusion reactions, edema, fever lymphocytes derived from DLBCL patients [23,36] and anemia, among other symptoms, in relapsed or allows to achieve a highly selective elimination of refractory HCL patients. Some of the patients also aggressive-lymphoma cells by this CXCR4-targeted experience dose-limiting toxicities, mainly hemolytic nanoparticle. Going deeper into the mechanism, uremic syndrome and capillary leak syndrome [3]. T22-PE24-H6 internalizes in CXCR4+ tumor cells by Probably this dosage limitation may underlie the endocytosis. Then, the furin-cleavage sites, inserted observation that PE38-based immunotoxins targeting between the T22 ligand and the functional PE24 toxin, different surface markers or receptors expressed in permit toxin intracellular activation in the cytosol, lymphoma cells (e.g. CD22 or CD25) could not induce inducing protein synthesis inhibition, which finally either partial or complete remissions in non-Hodgkin leads, as we show here, to the induction of cell death lymphoma patients [41,42]. Currently, many PE38 http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5179 derivative-based immunotoxins incorporate sequence 00650, PIE15/00028, PI15/00378 and EU COST Action modifications to reduce their immunogenicity while CA 17140 to R.M., FIS PI17/01246 and RD16/0011/ preserving their efficacy. In this context, PE24 0028 to J.S., and FIS PI15/00272 to E.V.]; Agencia appeared to be less immunogenic than PE38 because Estatal de Investigación (AEI) and Fondo Europeo de of the elimination of several immunogenic B-cell and Desarrollo Regional (FEDER) [grant BIO2016-76063-R, T-cell epitopes. Accordingly, we incorporated the AEI/FEDER, UE to A.V.]; CIBER-BBN [CB06/01/ PE24 toxin to our nanoparticle, as other authors did 1031 and 4NanoMets to R.M., and VENOM4CANCER the same replacement to generate immunotoxins (e.g. to A.V.]; AGAUR [2017 FI_B 00680 to A.F., 2017- PE24-based HA22-LR-8MIT) [7,43]. SGR-865 to R.M., 2017-SGR-1395 to J.S., 2017SGR-229 In conclusion, our novel delivery approach uses to A.V., and 2018FI_B2_00051 to L.S.G.] and a a multimeric single polypeptide-based nanoparticle predoctoral fellowship from the Government of that promises to improve the narrow therapeutic Navarra to N.S.; Josep Carreras Leukemia Research window observed in immunotoxin therapies, Institute [P/AG to R.M.]; a grant from the Cellex especially by reducing the toxicity. Moreover, our Foundation, Barcelona to J.S.; a grant from the results validate CXCR4 overexpressing cells as a Generalitat de Catalunya (PERIS) [SLT002/16/00433 relevant clinical target for treating refractory or to J.S.] and the Generalitat de Catalunya CERCA relapsed DLBCL patients that overexpress CXCR4. Programme [A.V. received an Icrea Academia Ultimately, future experimental work will dictate the Award]. The work was also supported by Grant level of relevance of the multiple factors that PERIS SLT006/17/00093 from the Generalitat de distinguish protein-nanoparticles from immuno- Catalunya to U.U. The bioluminescent follow-up of toxins, a requirement to improve immunotoxin or cancer cells and toxicity studies has been performed nanoparticle targeting selectivity, which for most in the ICTS-141007 Nanbiosis Platform, using its current nanoparticles appear to be still limited [44]. CIBER-BBN Nanotoxicology Unit (http://www. nanbiosis.es/portfolio/u18-nanotoxicology-unit/). Abbreviations Protein production has been partially performed by ADC: antibody-drug conjugate; ATCC: Ameri- the ICTS “NANBIOSIS”, more specifically by the can type culture collection; BLI: bioluminescence Protein Production Platform of CIBER-BBN/ IBB imaging; BM: bone marrow; BPDCN: blastic plasma- (http://www.nanbiosis.es/unit/u1-protein-producti cytoid dendritic cell neoplasm; CTCL: cutaneous on-platform-ppp/). T-cell lymphoma; DAPI: 4′,6-diamidino-2-phenylin- Finally, thanks to the free access Servier Medical dole; DLBCL: diffuse large B-cell lymphoma; DSMZ: Art images (http://smart.servier.com/) that allowed German collection of microorganisms and cell us to create a new composed image (Figure 1). cultures; EPR: enhanced permeability and retention; FACS: fluorescence-activated cell sorting; FBS: fetal Competing Interests bovine serum; FDA: food and drug administration; LSG, NS, UU, IC, RM, EV and AV are GAPDH: glyceraldehyde-3-phosphate dehydrogen- con-inventors of the Patent application PCT/EP2018/ ase; H&E: hematoxylin and eosin; HCL: hairy cell 061732 on the use of Nanostructured proteins. EV, AV leukemia; ID: injected dose; IHC: immunohistochem- and RM are co-founders of Nanoligent. istry; IMDM: Iscove's modified Dulbecco's medium; JC-1: 5,6-dichloro-2-[(E)-3-(5,6-dichloro-1,3-diethyl- References benzimidazol-3-ium-2-yl)prop-2-enylidene]-1,3-dieth 1. Wolf P, Elsässer-Beile U. Pseudomonas exotoxin A: from virulence factor to anti-cancer agent. Int J Med Microbiol. 2009; 299: 161–76. ylbenzimidazole iodide; LNs: lymph nodes; MACS: 2. Alewine C, Hassan R, Pastan I. Advances in anticancer immunotoxin therapy. magnetic-activated cell sorting; PARP: poly(ADP- Oncologist. 2015; 20: 176–85. 3. Nobre CF, Newman MJ, DeLisa A, Newman P. Moxetumomab ribose) polymerase; PBS: phosphate-buffered saline; pasudotox-tdfk for relapsed/refractory hairy cell leukemia: a review of PE: Pseudomonas aeruginosa; R-CHOP: rituximab with clinical considerations. Cancer Chemother Pharmacol. 2019; 84: 255–63. 4. Kreitman RJ, Dearden C, Zinzani PL, et al. Moxetumomab pasudotox in cyclophosphamide, hydroxydaunorubicin, oncovin, relapsed/refractory hairy cell leukemia. Leukemia. 2018; 32: 1768–77. 5. Jen EY, Gao X, Li L, et al. FDA Approval Summary: Tagraxofusp-erzs For prednisone; RPMI: Roswell Park Memorial Institute Treatment of Blastic Plasmacytoid Dendritic Cell Neoplasm. Clin Cancer Res. medium; SC: subcutaneous; SDS-PAGE: sodium 2020; 26: 532–6. 6. Vallera DA, Kreitman RJ. Immunotoxins Targeting B cell dodecyl sulfate polyacrylamide gel electrophoresis; Malignancy—Progress and Problems With Immunogenicity. Biomedicines. SE: standard error; SPF: specific-pathogen-free; TBST: 2019; 7: 1. 7. Kim J-S, Jun S-Y, Kim Y-S. Critical Issues in the Development of Tris-buffered saline Tween 20. Immunotoxins for Anticancer Therapy. J Pharm Sci. 2020; 109: 104–15. 8. Donaghy H. Effects of antibody, drug and linker on the preclinical and clinical Acknowledgments toxicities of antibody-drug conjugates. MAbs. 2016; 8: 659–71. 9. Aruna G. Immunotoxins: A Review of Their Use in Cancer Treatment. J Stem This work was supported by Instituto de Salud Cells Regen Med. 2006; 1: 31–6. 10. Gerber H-P, Senter PD, Grewal IS. Antibody drug-conjugates targeting the Carlos III (ISCIII, Co-funding from FEDER) [PI18/ tumor vasculature. MAbs. 2009; 1: 247–53. http://www.thno.org Theranostics 2020, Vol. 10, Issue 12 5180 11. Sánchez-García L, Serna N, Álamo P, et al. Self-assembling toxin-based 40. Nejadmoghaddam M-R, Minai-Tehrani A, Ghahremanzadeh R, Mahmoudi M, nanoparticles as self-delivered antitumoral drugs. J Control Release. 2018; 274: Dinarvand R, Zarnani A-H. Antibody-Drug Conjugates: Possibilities and 81–92. Challenges. Avicenna J Med Biotechnol. 2019; 11: 3–23. 12. Roncato F, Rruga F, Porcù E, et al. Improvement and extension of anti-EGFR 41. Kreitman RJ, Wilson WH, White JD, et al. Phase I trial of recombinant targeting in breast cancer therapy by integration with the immunotoxin anti-Tac(Fv)-PE38 (LMB-2) in patients with hematologic Avidin-Nucleic-Acid-Nano-Assemblies. Nat Commun. 2018; 9: 1–11. malignancies. J Clin Oncol. 2000; 18: 1622–36. 13. Zhang RX, Li J, Zhang T, et al. Importance of integrating nanotechnology with 42. Kreitman RJ, Squires DR, Stetler-Stevenson M, et al. Phase I trial of pharmacology and physiology for innovative drug delivery and therapy - an recombinant immunotoxin RFB4(dsFv)-PE38 (BL22) in patients with B-cell illustration with firsthand examples. Acta Pharmacol Sin. 2018; 39: 825–44. malignancies. J Clin Oncol. 2005; 23: 6719–29. 14. Unzueta U, Céspedes MV, Ferrer-Miralles N, et al. Intracellular CXCR4+ cell 43. Onda M, Beers R, Xiang L, et al. Recombinant immunotoxin against B-cell targeting with T22-empowered protein-only nanoparticles. Int J malignancies with no immunogenicity in mice by removal of B-cell epitopes. Nanomedicine. 2012; 7: 4533–44. Proc Natl Acad Sci USA. 2011; 108: 5742–7. 15. Falgàs A, Pallarès V, Unzueta U, et al. A CXCR4-targeted nanocarrier achieves 44. Wilhelm S, Tavares AJ, Dai Q, et al. Analysis of nanoparticle delivery to highly selective tumor uptake in diffuse large B-cell lymphoma mouse models. tumours. Nat Rev Mater. 2016; 1: 1–12. Haematologica. 2020; 105: 741–53. 16. Serna N, Sánchez-García L, Unzueta U, et al. Protein-Based Therapeutic Killing for Cancer Therapies. Trends Biotechnol. 2018; 36: 318–35. 17. Kim J, Takeuchi H, Lam ST, et al. Chemokine receptor CXCR4 expression in colorectal cancer patients increases the risk for recurrence and for poor survival. J Clin Oncol. 2005; 23: 2744–53. 18. Chatterjee S, Azad BB, Nimmagadda S. The Intricate Role of CXCR4 in Cancer. Adv Cancer Res. 2014; 124: 31–82. 19. Zhao H, Guo L, Zhao H, Zhao J, Weng H, Zhao B. CXCR4 over-expression and survival in cancer: A system review and meta-analysis. Oncotarget. 2014; 6: 5022–40. 20. Peled A, Tavor S. Role of CXCR4 in the pathogenesis of acute myeloid leukemia. Theranostics. 2013; 3: 34–9. 21. Wald O, Shapira OM, Izhar U. CXCR4/CXCL12 axis in non small cell lung cancer (NSCLC) pathologic roles and therapeutic potential. Theranostics. 2013; 3: 26–33. 22. Laursen MB, Reinholdt L, Schönherz AA, et al. High CXCR4 expression impairs rituximab response and the prognosis of R-CHOP-treated diffuse large B-cell lymphoma patients. Oncotarget. 2019; 10: 717–31. 23. Moreno MJ, Bosch R, Dieguez-Gonzalez R, et al. CXCR4 expression enhances diffuse large B cell lymphoma dissemination and decreases patient survival. J Pathol. 2015; 235: 445–55. 24. Reinholdt L, Laursen MB, Schmitz A, et al. The CXCR4 antagonist plerixafor enhances the effect of rituximab in diffuse large B-cell lymphoma cell lines. Biomark Res. 2016; 4: 12. 25. Recasens-Zorzo C, Cardesa-Salzmann T, Petazzi P, et al. Pharmacological modulation of CXCR4 cooperates with BET bromodomain inhibition in diffuse large B-cell lymphoma. Haematologica. 2019; 104: 778–88. 26. Sun X, Widdison W, Mayo M, et al. Design of Antibody−Maytansinoid Conjugates Allows for Efficient Detoxification via Liver Metabolism. Bioconjug Chem. 2011; 22: 728–35. 27. Erickson HK, Lambert JM. ADME of Antibody–Maytansinoid Conjugates. AAPS J. 2012; 14: 799–805. 28. Vinhas R, Mendes R, Fernandes AR, Baptista PV. Nanoparticles-Emerging Potential for Managing Leukemia and Lymphoma. Front Bioeng Biotechnol. 2017; 5: 79. 29. Garnett MC, Kallinteri P. Nanomedicines and nanotoxicology: some physiological principles. Occup Med (Lond). 2006; 56: 307–11. 30. Deshantri AK, Varela Moreira A, Ecker V, et al. Nanomedicines for the treatment of hematological malignancies. J Control Release. 2018; 287: 194–215. 31. Liu Y, Wu H, Wang H. Protein-based Nanoparticles for Tumor Targeting Theranostics. Nanomaterials for Tumor Targeting Theranostics. China: Co-published with HEP; 2016. 32. Céspedes MV, Cano‐Garrido O, Álamo P, et al. Engineering Secretory Amyloids for Remote and Highly Selective Destruction of Metastatic Foci. Adv Mater. 2020; 32: 1907348. 33. Vazquez E, Mangues R, Villaverde A. Functional recruitment for drug delivery through protein-based nanotechnologies. Nanomedicine (Lond). 2016; 11: 1333–6. 34. Rueda F, Céspedes MV, Sánchez-Chardi A, et al. Structural and functional features of self-assembling protein nanoparticles produced in endotoxin-free Escherichia coli. Microb Cell Fact. 2016; 15: 59. 35. Wester HJ, Keller U, Schottelius M, et al. Disclosing the CXCR4 Expression in Lymphoproliferative Diseases by Targeted Molecular Imaging. Theranostics. 2015; 5: 618–30. 36. Chen J, Xu-Monette ZY, Deng L, et al. Dysregulated CXCR4 expression promotes lymphoma cell survival and independently predicts disease progression in germinal center B-cell-like diffuse large B-cell lymphoma. Oncotarget. 2015; 6: 5597–614. 37. Jendrossek V, Grassmé H, Mueller I, Lang F, Gulbins E. Pseudomonas aeruginosa-Induced Apoptosis Involves Mitochondria and Stress-Activated Protein Kinases. Infect Immun. 2001; 69: 2675–83. 38. Wood SJ, Goldufsky JW, Bello D, Masood S, Shafikhani SH. Pseudomonas aeruginosa ExoT Induces Mitochondrial Apoptosis in Target Host Cells in a Manner That Depends on Its GTPase-activating Protein (GAP) Domain Activity. J Biol Chem. 2015; 290: 29063–73. 39. Masters JC, Nickens DJ, Xuan D, Shazer RL, Amantea M. Clinical toxicity of antibody drug conjugates: a meta-analysis of payloads. Invest New Drugs. 2018; 36: 121–35. http://www.thno.org