A Novel Recombinant Anti-CD22 Immunokinase Delivers

Published OnlineFirst January 29, 2016; DOI: 10.1158/1535-7163.MCT-15-0685

Large Molecule Therapeutics Molecular Cancer Therapeutics A Novel Recombinant Anti-CD22 Immunokinase Delivers Proapoptotic Activity of Death- Associated Protein Kinase (DAPK) and Mediates Cytotoxicity in Neoplastic B Cells Nils Lilienthal1,2, Gregor Lohmann1, Giuliano Crispatzu1, Elena Vasyutina1, Stefan Zittrich3, Petra Mayer1, Carmen Diana Herling4, Mehmet Kemal Tur5, Michael Hallek4, Gabriele Pﬁtzer3, Stefan Barth6,7, and Marco Herling1,4

Abstract

The serine/threonine death-associated protein kinases (DAPK) SGIII against the B-cell–exclusive endocytic glyco-receptor CD22 provide pro-death signals in response to (oncogenic) cellular stres- was created. Its high purity and large-scale recombinant production ses. Lost DAPK expression due to (epi)genetic silencing is found in a provided a stable, selectively binding, and efficiently internalizing broad spectrum of cancers. Within B-cell lymphomas, deficiency of construct with preserved robust catalytic activity. DK1KD-SGIII the prototypic family member DAPK1 represents a predisposing or specifically and efficiently killed CD22-positive cells of lymphoma early tumorigenic lesion and high-frequency promoter methylation lines and primary CLL samples, sparing healthy donor– or CLL marks more aggressive diseases. On the basis of protein studies and patient–derived non-B cells. The mode of cell death was predom- meta-analyzed gene expression profiling data, we show here that inantly PARP-mediated and caspase-dependent conventional apo- within the low-level context of B-lymphocytic DAPK, particularly ptosis as well as triggering of an autophagic program. The notori- CLL cells have lost DAPK1 expression. To target this potential ously high apoptotic threshold of CLL could be overcome by vulnerability, we conceptualized B-cell–specific cytotoxic reconsti- DK1KD-SGIII in vitro also in cases with poor prognostic features, tution of the DAPK1 tumor suppressor in the format of an immu- such as therapy resistance. The manufacturing feasibility of the novel nokinase. After rounds of selections for its most potent cytolytic CD22-targeting DAPK immunokinase and its selective antileukemic moiety and optimal ligand part, a DK1KD-SGIII fusion protein efficiency encourage intensified studies towards specific clinical containing a constitutive DAPK1 mutant, DK1KD, linked to the scFv application. Mol Cancer Ther; 1–14. 2016 AACR.

Introduction death-protective cluster in the presence of cytotoxic signals (1, 2). Knowledge on DAPKs has evolved since then with DAPK1 Functional screens with antisense libraries had enriched established as the prototypic serine-threonine kinase of a family death-associated protein kinases (DAPK) as a prominent of 5 proapoptotic proteins. They encompass DAPK2 (DAPK1- Related Protein 1 ¼ DRP-1), zipper-interacting protein kinase 1 Laboratory of Lymphocyte Signaling and Oncoproteome, Excellence (ZIPK ¼ DAPK3), DAP-related apoptotic kinase (DRAK) 1, and Cluster for Cellular Stress Response and Aging-Associated Diseases – (CECAD), University of Cologne, Koln,€ Germany. 2Federal Institute for DRAK2 (1 4). They share a 50% to 80% homology of their Drugs and Devices (BfArM), Bonn, Germany. 3Institute of Vegetative catalytic domains. Loss-of-function of the DAPK tumor suppres- € 4 Physiology; University of Cologne, Koln, Germany. Department I of sors is implicated as a central tumor-initiating and metastasis Internal Medicine, Center for Integrated Oncology (CIO) Koln-Bonn,€ and CECAD, University of Cologne, Koln,€ Germany. 5Institute of -promoting event in a broad range of cancers, that is, of breast, Pathology, University Hospital, Justus Liebig University Gießen, lung, head and neck, gastrointestinal, and the hematopoetic 6 Gießen, Germany. Department of Experimental Medicine and Immu- system (1, 5, 6). The causes of tumor-associated DAPK deﬁciency notherapy, Institute for Applied Medical Engineering, RWTH Aachen, Aachen, Germany. 7South African Research Chair in Cancer Biotech- are primarily epigenetic mechanisms (promoter hypermethyla- nology, Institute of Infectious Disease and Molecular Medicine (IDM), tion), but also genomic deletions with subsequent LOH are Department of Integrative Biomedical Sciences, Faculty of Health observed(1,7,8). Sciences, University of Cape Town, South Africa. The DAPK proteome of multimeric complexes involves phos- Note: Supplementary data for this article are available at Molecular Cancer phorylated substrates that transmit apoptotic or autophagic sig- Therapeutics Online (http://mct.aacrjournals.org/). nals in response to cellular stresses including those by redox N. Lilienthal and G. Lohmann contributed equally to this article. burden or oncogenes (1, 2, 9). Robust protective DAPK function Corresponding Author: Marco Herling, Laboratory of Lymphocyte Signaling and promotes p19ARF-mediated p53 activity (10). Loss of DAPK Oncoproteome, Department of Internal Medicine I, University Hospital of proﬁciency is tumorigenic in the context of oncogenic stress, that Cologne, Building 15, Level 1, Room 1.010, Kerpener Str. 62, Cologne 50937, is, by Myc- or E2F1, as this attenuates safeguarding p53-conferred Germany. Phone: 49221-478-5194; Fax: 49221-478-88847; E-mail: cell death. Lost DAPK can also no longer retain oncogenic ERK1/ [email protected] 2 in the cytoplasm (11). The operational downstream pathways doi: 10.1158/1535-7163.MCT-15-0685 (caspase- and p53-dependent vs. -independent apoptosis; ref. 1) 2016 American Association for Cancer Research. and cell biologic outcomes [i.e., apoptosis (1, 10) versus

www.aacrjournals.org OF1

Lilienthal et al.

autophagic preservation (1) vs autophagic cell death (12, 13) vs. and 2014 from the DSMZ and from the ATCC. Only original stock necroptosis (14)] appear cell context specific and to be dictated by propagated immediately upon arrival for 2 to 3 passages was the nature of the upstream stressor and the executing DAPK family picked for studies and cultures terminated after the 10th round of member (1, 2, 15). passaging (5–7 weeks). Upon thawing for experimentation Reduced DAPK1 mRNA transcription due to promoter hyper- between 2012 and 2015, all lines were authenticated by flow methylation and aberrant distribution of affected CpG islands cytometry confirming their characteristic immunophenotype, was demonstrated in virtually all cases of sporadic chronic lym- including Ig composition and TCL1 levels (28–30). The same phocytic leukemia (CLL; refs. 6, 16). In addition, a CLL haplotype applied to the FL/DLBCL lines LP, FN, CJ, LR, MS, DBr, DS, EJ with allele-specific imbalances of DAPK1 expression indicates established and provided by Dr. R.J. Ford (UT M.D. Anderson that recurring germline single-nucleotide variants and promoter Cancer Center, Houston, TX). Their identity has been described methylation can be considered as predisposing to (familial) CLL previously (29, 31). Further details are indicated in the Supple- (16, 17). DAPK1 promoter hypermethylation was also shown to mentary Data. Experimentations were done in suspension cul- be associated with more aggressive disease and adverse clinical tures under conditions as instructed and published (28, 29). outcome in diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL; refs. 5, 18). Western blots, IHC, and flow cytometry Giventheircentralroleintumorigenesisandclonalsuste- Protocols and reagents for immunoblots (DAPKs and apopto- nance, DAPKs have been intensely investigated as therapeutic tic proteins), IHC (DAPK1), and flow cytometry are outlined in targets. Strategies employ gene therapy, development of func- the Supplementary Data. tional DAPK activators, or demethylating agents to reestablish DAPK expression (19, 20). This also extents (reversely) to In silico meta-analysis of DAPK family gene expression from DAPK inhibitors in autoinflammatory disease and graft rejec- array-based data tion (15, 20–22). We explored here if reinstating DAPK activity Via manual search of literature and data repositories, we would perturb the high apoptotic threshold in DAKP-deficient obtained available primary data out of 8 reports on array-based neoplastic B cells that are burdened by incurring (oncogenic) gene expression profiling (GEP) of CLL samples with specific stress. We further hypothesized that an efficient immunoli- controls. Details and further meta-analyses on B-cell subsets and gand-based targeting would potentiate that selectivity. Our other B-cell lymphomas are found in the Supplementary Data. All DAPK20-CD30L immunokinase already showed encouraging datasets were separately background corrected and annotated in vitro (23) and in vivo (24) activity in systems of Hodgkin using the BioConductor packages "affy" and "biomaRt" in lymphoma. R-3.1.0. The datasets were quantile-normalized and replicates A novel CD22-targeting construct containing a constitutively were combined by mean. Sample names were assigned from the active DAPK1 variant showed here efficient binding, internaliza- GEO (32) entry to test for differential expression via t test or tion, and cell death induction in malignant B cells. This immu- Wilcoxon/Mann–Whitney test. Only the highest fold change nokinase revealed a favorable in vitro profile of induced apoptosis (FC; P < 0.1) was obtained and visualized in the summarizing and autophagy, including efficacy in high-risk CLL. Its novel heatmap. design and high-yield purification protocols are described with the advantages and feasibility aspects encouraging further explo- fi ration toward a clinical application. This concept of immunoli- Cloning, transfection, protein puri cation, and mass gand-based toxic reconstitution (not mere rescue replenishment) spectrometry of a tumor suppressor function would supplement mAbs, small- The transient transfection of HEK293T cells employed a lipo- fi molecule inhibitors, and T-cell–based therapies (25) whose some-based protocol (33). Puri cation of recombinant protein implementation into the therapeutic approaches in B-cell lym- containing supernatants was done in a three-step procedure. First, fi phomas, particularly CLL, has ushered a new era of potentially an immobilized nickel-nitrilotriacetic (NiNTA) acid metal af nity chemotherapy-free targeted treatments (26). chromatography (IMAC; His-Trap; GE Lifesciences) based fast protein liquid chromatography (FPLC) was performed. After buffer exchange, the eluted fractions were applied on a CM sepharose Materials and Methods column for cation exchange chromatography (CEC). Peptide Cell isolation and cultures analysis was carried out using an ESI Q-Tof-2 MS (Waters Micro- Peripheral blood samples were obtained from 31 CLL patients mass) mass spectrometer (23). Sequences of individual peptides (diagnosed according to iwCLL criteria; ref. 27) that were either were identified using the Mascot algorithm (Matrix Science). treatment-na€ve or had a minimum interval from any last therapy of 4 weeks. After informed written consent, sampling was per- Assays of competitive surface binding and internalization formed in accordance with the Declaration of Helsinki and the Specific binding of DK1KD-SGIII and all other generated pro- guidelines at the University of Cologne (IRB approval #01-163). teins to primary cells and cell lines were determined by flow CLL B cells, healthy donor–derived B cells (purities as per flow cytometry. Competitive binding and internalization assays as well cytometry; for CLL >95%), and healthy donor peripheral blood as confirmatory fluorescence microscopy are described in detail in mononuclear cells (PBMC) were isolated as described previously the Supplementary Data. Briefly, after DK1KD-SGIII incubation (28) or, for tonsillar B cells, followed a MACS protocol (Supple- and washing, cells were labeled with monoclonal Alexa Fluor 488 mentary Data). The cell lines REH, Nalm6, SP53, Mino, Granta, anti-penta histidine antibodies (AHA; Qiagen). The competition BC-1, MEC1, JVM3, BJAB, Namalwa-PNT, Daudi, Raji, Ramos, assays were based on saturation by an APC-labeled anti-human DoHH2, L428, KMS-12-BM, RPMI-8226, HEK293T, and U937 CD22 antibody RFB-4 (Invitrogen). For flow cytometry–based (Supplementary Table S1) were originally acquired between 2012 internalization assays, DK1KD-SGIII and AHA were added to the

OF2 Mol Cancer Ther; 2016 Molecular Cancer Therapeutics

Immunokinase-Based B-Lymphotoxic Delivery of Activated DAPK1

target cells. After 30-minute incubation, surface proteins were contaminating non-B cells might have contributed to the signal removed by addition of proteinase K for 45 minutes (Qiagen) and from this pool of na€ve B cells. Among non-CLL B-cell lympho- levels of internalized AHA-labeled protein were determined. mas, unequivocal DAPK1 expression was most frequently seen in Measurements were performed on FACSCalibur (Beckton Dick- DLBCL lines. For the vast majority of CLL, there was an absence of inson) and Gallios (Beckman Coulter) cytometers. Antibodies DAPK1 and DAPK2 proteins. DAPK1 was not detected in 10 of 11 against human CD5, CD19, IgG2a, and IgG1 were from Biole- cases (91%) and 7 of 9 (78%) cases lacked noticeable DAPK2 gend. Data were analyzed with the Cyflogic software (Cyflo). levels (Fig. 1; Supplementary Table S2). Similarly, the CLL-derived lines MEC1 and JVM-3 expressed no or negligible amounts of Kinase assays DAPK1/2. Beyond initial reports of lost DAPK1 mRNA in CLL (16, In vitro kinase activity of recombinant DAPK1 was determined 17), this pattern had not been shown for DAPK1 and DAPK2 using the PKLight Assay Kit (Lonza) according to the manufac- proteins before. turer's instructions. Instead of protein kinase A (PKA), recombi- The predominant lack of DAPK1 expression in CLL alongside a nant myosin light chain II (MLCII; ProSpec) served as a substrate low-level retention in other B-cell lymphomas and normal B-cell for DK1KD. subsets was confirmed by comparative meta-analyses of DAPK expression from publicly available raw data of array-based gene fi Cell viability and apoptosis expression pro les (GEP). For CLL, we performed this from 9 The colorimetric 3-[4,5-dimethylthiazol-2-yl-2,5-diphenyl] tet- technically comparable datasets comprising comparisons of CLL razolium bromide (MTT) assay (Promega) assessed cell viability and normal B cells from 8 reports (protocol in online Supple- DAPK by measuring metabolic activity in duplicates per sample as mentary Data). Therein, most of the genes revealed a described previously (28). Averaged absorbances of the experi- downregulation in CLL-to-normal comparisons (Fig. 1C), but ZIPK mental conditions were normalized to the control conditions. with often showing an opposite directionality to the rest of Apoptosis rates were measured by flow cytometry on a Gallios the gene family. Data extracted from the EBI/Atlas RDF repository fi fi cytometer (Beckmann Coulter) with 5 105 cells stained with (35) con rmed these ndings (status 07/20/2015; online Sup- þ AnnexinV/7AAD (Becton Dickinson). Preincubations with plementary Data). The overall subtle ( 4to 4) fold changes in 20 mmol/L of the pan-caspase inhibitor ZVAD-fmk (Promega) these CLL versus B-cell comparisons implicated what had been were done for 60 minutes. reported before, namely that normal B cells include sizeable DAPK-silenced populations as compared with non-B cells (36). In CLL, scenarios of treatment seem to be associated with upre- Statistical analyses gulation of DAPK(1) or selection for less DAPK(1)-hypermethy- Parametric Student t test estimations (either paired or non- lated subclones. Additional metacomparisons revealed that the paired) were performed in GraphPad Prism. P values <0.05 were already repressed levels of DAPK1 show very low variation across considered significant. Data are presented as means with indicat- normal B-cell subsets (Supplementary Fig. S1B) and that DAPK1 ed error bars as SEM. levels in most non-CLL B-cell lymphomas exceed those in normal Further information about sources of DNA, vectors, cloning of B cells (Supplementary Fig. S1C). DAPK1 and DAPK2 mutants, cloning of the most suitable scFv, antibodies, transfections, purification, and mass spectrometry as Creation of constitutively active and proapoptotic DAPK1 well as in silico meta-analyses are given in the Supplementary Data. mutants by removal of the autoinhibitory CaM domain The tumor-suppressive function of DAPK1 likely relies on its Results high proapoptotic/-autophagic activity (1). Accordingly, the The reduced expression of the DAPK1 tumor suppressor in observed loss of DAPK expression might contribute to the prom- B lymphocytes and derived tumors is particularly found inent prosurvival phenotype of CLL. Therefore, we postulated that in CLL cells constitutively active DAPK mutants would reconstitute DAPK The scarce literature on DAPK expression in inflammatory cells signaling and may reduce the leukemia-specific apoptotic thresh- implicates T cells and macrophages as the main source (15). old. The function of wild-type DAPK1 is tightly regulated by an þ Alternative splicing creates various DAPK1 isoforms. A kinase- autoinhibitory Ca2 /Calmodulin (CaM) element located next to less truncated s-DAPK1 variant can mediate protein degradation its kinase domain (Fig. 2A). DAPK1 further consists of a cyto- of full-length DAPK1 (34). Here, we evaluated the protein expres- skeleton binding domain and a death domain. Removal of sion of full-length DAPK1 and DAPK2 in CLL patient samples, in the CaM sequence renders this truncated DAPK1 constitutively multiple B-cell tumor lines, and in B cells from healthy donors. We active (23, 37–39). Therefore, we generated different mutants conclude that in the context of a generally low-level DAPK derived from DAPK1 and -2, named DAPK1KDCyt, DK1KD, expression in the B-cell lineage, particularly CLL cells, lack notice- DAPK2DCaM, and DK2KD, all lacking the regulatory CaM ele- able DAPK1. In detail, prominent DAPK1 signals in normal ment (Fig. 2A). Overexpressed in Raji lymphoma and JVM-3 CLL- lymph nodes originated from perivascular macrophages. In reac- like cells (pcDNA3.1 vector), DK1KD was the most efficient of tive tonsils, only a minority of large centroblastic cells of follicular these constructs (including in comparison with full-length germinal centers revealed cytoplasmic signals (Supplementary DAPK1 and -2) in inducing cell death (Supplementary Fig. S2) Fig. S1A). Pan-CD19 isolates from tonsils (N ¼ 3; purities and was hence selected for further analysis. 98%–99%) showed weak DAPK1 expression. The CD19-positive fraction derived from healthy donor peripheral blood samples Conception of an anti-CD22 scFv/DK1KD immunokinase (N ¼ 4) revealed signals below those from DAPK transfectants, fusion protein but above those from CLL (Supplementary Table S2; Fig. 1 and We aimed for CLL-specific delivery of the constitutively active Supplementary Fig. S1). Given their purities of 88% to 89%, DAPK mutant DK1KD via fusion to an immunoglobulin scFv. The

www.aacrjournals.org Mol Cancer Ther; 2016 OF3

Lilienthal et al.

Figure 1. Expression of DAPK1 and DAPK2 is downregulated in the vast majority of CLL, but differential patterns across DAPK family members exist. A and B, representative Western blots, each for DAPK1 (A) and DAPK2 (B). Positive controls: DAPK1 (A) and DAPK2 (B) transfected (t.) HEK293T cells. Pan-CD19 isolates of peripheral- blood (PB)-derived healthy donor B cells (88.7% purity) reveal unequivocal signals. Asterisks, Raji and Ramos cells (Burkitt's lymphoma) reveal a weak signal only after longer exposures, but not CLL samples or the CLL-like cells MEC1 and JVM-3, except for a weak DAPK2 signal in JVM-3 lysates (summaries in Supplementary Tables S1 and S2). GADPH or ß-Actin as loading controls. M, molecular weight marker. For CLL patient #, refer to Supplementary Table S2 (Supplementary Data). C, heatmap generated from summarized array-based DAPK family gene expression data (9 technically comparable publicly available sets). For protocol including batch exclusions, see online Supplementary Data. Data processing included background correction, annotation, replicate combination, quantile normalization, and testing for signiﬁcant (, t-test or enforced Wilcoxon/Mann–Whitney test) differential expression. Color bars for 3 distinct comparisons: (1) CLL versus normal B cells (various subtypes); (2) CLL with IgHV gene unmutated versus mutated status; (3) CLL with post-to-pretreatment and clinical comparisons.

OF4 Mol Cancer Ther; 2016 Molecular Cancer Therapeutics

Immunokinase-Based B-Lymphotoxic Delivery of Activated DAPK1

initial screening for the most suitable binding domains for such for detection of producer cells expressing the fusion protein an immunoligand-based fusion protein entailed preselection of without an eGFP labeled secretion product. antibodies and antibody fragments based on criteria of reason- HEK293T cells were transfected with the pMS versions (Sup- able antigen specificity, antigen density, low immunogenicity, plementary Fig. S3) followed by a novel multi-step up-scalable slow degradation (this excluded CD5; ref. 40), and capacity of purification protocol resulting in an average yield of 2 mg/L. The internalization. Given these criteria and based on published data first step included the protein purification from the supernatants for similar constructs as well as considering the surface marker in a NiNTA-driven IMAC-based FPLC (detailed description in profile of CLL, we chose CD19, CD22, and CD40 (41–43). Several online Supplementary Data). The various fractions were assessed mAb and/or scFv for each of these epitopes as well as a CD40L by Coomassie staining after SDS gel separation (Fig. 2C). In the construct were tested (for sources and details, see Acknowledge- second step, after buffer exchange via dialysis, the protein-con- ments and Supplementary Data). We first selected 3 most suitable taining fractions were purified by cation exchange chromatogra- constructs. phy (CEC). Contaminating proteins eluted from CEC were found The specific antibody anti-CD19 HD37 as well as the scFv's in fractions 3–8 (e.g., fraction 5 in Fig. 2D) while the target protein anti-CD22 SGIII and anti-CD40 G28-5 were cloned into the pMS eluted later and was found in the second elution peak (fractions vector, expressed recombinantly, and tested for binding to Raji 35–38 in Fig. 2D). These fractions revealed only one band lymphoma cells and primary CLL isolates. To exclude unspecific exclusively at the expected 62–70 kDa region by both Coomassie binding, these ligands were also tested on CD19-negative, CD22- stains and anti-His Western blot analysis (Fig. 2E, left). To further negative, and CD40-deficient myeloid/monocytic U937 cells. confirm the identity of the purified protein from this elution Specific binding was detected for CD22 (SGIII) and CD40 fraction, the 62–70 kDa band was excised from the Coomassie- (G28-5) in Raji and CLL cells (Fig. 2B). In CLL, receptor binding stained gel, analyzed via electron spray ionization tandem mass also often mediates prosurvival signaling, rescuing primary CLL spectrometry (ESI-MS/MS), and compared with the predicted cells from their pronounced spontaneous apoptosis in vitro. sequence of the fusion protein in the Mascot protein databank. Hence, survival/viability (AnnexinV/7AAD flow cytometry/MTT Peptide mass/charge profiles identified DK1KD-SGIII with assays) of CLL cells (3 patient samples) was assessed before (0h) sequence coverage of 49.8% and an overall score of 1016.1 and 48 hours after incubation with the binding domain candi- (Fig. 2E, right). dates. It revealed that the anti-CD40 scFv G28-5 decreased spontaneous in vitro apoptosis by almost 20%, whereas CLL cell Purified DK1KD-SGIII demonstrates functional kinase activity survival/viability were not affected by the other candidate ligands, through phosphorylation of MLCII including SGIII. HD37 was also excluded from further studies. Its Catalytic activity of the DAPK1 mutant DK1KD in fusion to binding to the B-cell receptor coreceptor CD19 on positive target SGIII would be necessary to induce cell death in CLL cells upon cells was highly heterogeneous (an example of HD37 binding delivery. Therefore, catalytic activity of DK1KD-SGIII and the failure in a CLL sample is given in Fig. 2B) and misfolding in the respective controls DK1KD and SGIII were investigated by an in HEK293T system was suspected. Consequently, the anti-CD22 vitro kinase activity assay. DAPK10s potent proapoptotic/auto- scFv SGIII was selected for further work as part of the delivery phagic capabilities are mediated by phosphorylation of its approach of constitutive DAPK1. SGIII is a humanized scFv various targets, among which myosin light chain II (MLCII) is derived from the murine anti-CD22 mAb RFB-4 by specificity one of the best established. Consequently, in vitro kinase activity grafting into frameworks based on scaffolds preselected for sta- assays with the exogenous recombinant substrate MLCII were bility from phage display libraries. SGIII exhibits superior antigen performed. Protein kinase A with its substrate kemptide (lucif- binding and stability (44). Recent data, including from clinical erin) served as the positive control system. The underlying trials, confirm the suitability of CD22 as a target of mAb-based principle of this assay is a competition for ATP between luci- therapies (45). We finally screened various B-cell lymphoma lines ferase and the kinase DK1KD-SGIII (DK1KD). Therefore, luci- for CD22 expression with the original RFB-4 clone. It corrobo- ferase activity, measured in relative light units (RLU) by a rated the results of the SGIII binding studies and revealed that luminescence reader, is inversely proportional to DK1KD kinase target receptor densities are largely independent of major histo- activity with higher luciferase quenching reflecting higher genetic differentiation stages except for the known CD22-low/ DK1KD activity (Supplementary Fig. S4). In the presence of 5 absent entities of Hodgkin lymphoma and multiple myeloma nmol/L DK1KD or 5 nmol/L DK1KD-SGIII, bioluminescent (Supplementary Table S1b). signals dropped significantly as compared with the kinase-free control, indicating DK1KD dependent kinase activity. In con- A novel high-yield purification protocol for recombinant trast, 5 nmol/L of SGIII did not affect light emission, hence did DK1KD-SGIII not elicit nonspecific kinase activity (Fig. 2F, left). In addition, no The selected components "DK1KD" and "SGIII" as the best significant difference of catalytic activity was identified between performing DAPK kinase mutant and scFv binding ligand, DK1KD and DK1KD-SGIII demonstrating that the fusion of respectively (above), were cloned into the pMS vector (Sup- DK1KD to SGIII did not affect DK1KD's enzymatic activity. plementary Fig. S3) for expression of a fused recombinant Moreover, kinase activity was dependent exclusively on the con- immunoligand "DK1KD-SGIII". The pMS enables expression centration of DK1KD as part of the DK1KD-SGIII fusion protein. and purification of the target protein through an Igk leader Different concentrations of DK1KD-SGIII (5–50 nmol/L), unlike sequence promoting immediate secretion of the (fusion) pro- respective equimolar concentrations of SGIII, led to dose-depen- tein into the supernatant (46). Therefore, DK1KD did not dent increases in kinase activity (Fig. 2F, right). Taken together, the execute its cytotoxic effects in the HEK293T producers. A SGIII ligand-tethered kinase DK1KD demonstrated specific and 6xHis-tag remains part of the secreted protein and facilitates potent catalytic activity by phosphorylating DAPK10s physiologic its later detection. An IRES-eGFP creates a bicistronic transcript substrate MLCII.

www.aacrjournals.org Mol Cancer Ther; 2016 OF5

Lilienthal et al.

Figure 2. Kinase and ligand selection followed by high-yield puriﬁcation of the DK1KD-SGIII fusion construct. A, schematic illustration of DAPK protein family (top 5), the generated DAPK1 mutant constructs (subsequent 4, italics), and the DK1KD-SGIII immunokinase (bottom). Comparisons to DAPK1 wild-type are as percent homologies of amino acid sequences of the kinase domains. DK1KD-SGIII construct: constitutively active DAPK1 domain (DK1KD) lacking the Ca2þ/Calmodulin-regulatory element (CaM Reg.) linked to a humanized anti-CD22 scFv binding domain (SGIII) plus His-tag; NLS, nuclear localization signals. B, selected constructs: the anti- (a)-CD22 scFv SGIII (blue) and the a-CD40 scFv G28-5 (black) show increased binding in Raji (pos. control) and primary CLL cells, in contrast to an inconsistent performance of the a-CD19 antibody (Ab) HD37 (green; lack of binding illustrated here). Myelomonocytic U937 cells served as a negative control. Incubation for 20 minutes, analysis by ﬂow cytometry. scFv, single chain variable fragment; Control, Isotype IgG1. (Continued on the following page.)

OF6 Mol Cancer Ther; 2016 Molecular Cancer Therapeutics

Immunokinase-Based B-Lymphotoxic Delivery of Activated DAPK1

Recombinant DK1KD-SGIII binds target cells and is from already internalized CD22–DK1KD–SGIII –AHA complexes. internalized via specific engagement of surface CD22 by the This uptake of DK1KD-SGIII occurred within 45 minutes. Confir- SGIII scFv component matory fluorescence microscopy also demonstrated a rapid (1 For DK1KD-SGIII as an immunotherapeutic, highly specific hour) intake of DK1KD-SGIII into primary CLL cells (Fig. 3E). delivery into target cells is necessary. We first analyzed the surface binding of DK1KD, DK1KD-SGIII, and SGIII to primary CLL cells Selective, dose-dependent cell death induction by and various CD22 positive (þ) or negative () cell lines via DK1KD-SGIII in CD22(þ) lymphoma cell line systems and isotype-controlled flow cytometric analysis with Alexa Fluor primary CLL 488 anti-His antibodies (AHA). This demonstrated that SGIII We next addressed the question whether DK1KD-SGIII, mim- mediated selective binding of the construct to human CD22 icking reinstated DAPK1, is able to resemble DAPK10s cytotoxic (þ) cells. We documented strong binding of DK1KD-SGIII and potential. Cell death induction of DK1KD-SGIII in CD22(þ) and SGIII, but not of DK1KD, to B-cell lymphomas, that is, with CD22() cells was determined by AnnexinV/7AAD flow cytome- increased intensities in JVM-3, Raji, MEC1. No binding was try and MTT assays after 48-hour incubation with increasing detected in CD22() U937 cells. The binding of DK1KD-SGIII concentrations of the immune ligands. Treatment of CD22(þ) to CLL cells was uniform and dose dependent (Fig. 3A and CLL-like B-cell lines (JVM3, MEC1) and lymphoma lines (Raji, Supplementary Figs. S5 and S6). Confirmatory fluorescence Ramos, DOHH2) with DK1KD-SGIII resulted in cell death/ microscopy of CLL cells incubated with DK1KD-SGIII recombi- reduced viability in a dose-dependent fashion (Fig. 4A). IC50 nant protein and AHA visualized its strong accumulation on the values of CD22-low JVM-3 cells were approximately 8-fold higher cell surface (Fig. 3B). (extrapolated: 1,300 nmol/L) as compared with CD22-high Epitope specificities of DK1KD-SGIII as a selective CD22- MEC1 or DOHH2 cells (170 nmol/L and 175 nmol/L, respective- scFv–based construct were addressed in more detail in com- ly). Further supporting a CD22/compound–response relation- petitive binding assays. Binding of DK1KD-SGIII to CLL cells ship, both CD22 expression as measured by DK1KD-SGIII binding as detected by AHA was increasingly prevented by preincuba- and IC50 values for DK1KD-SGIII in Raji cells (500 nmol/L) were tion with increments of the APC-coupled a-CD22 mAb RFB-4, in between those of MEC1 and JVM-3 cells (Fig. 4A and Supple- while incubation with nonspecific IgG2a reduced DK1KD- mentary Fig. S5). No cytotoxicity was observed in CD22() U937 SGIII binding only slightly (Fig. 3C). Successful binding of cells. Neither incubation of purified SGIII nor DK1KD with CD22 the preincubated RFB-4-APC was measured simultaneously in (þ) or with CD22() cell lines increased cell death/reduced a different channel. Reciprocally, preincubated SGIII-based viability (Supplementary Fig. S7). Immunoblots for the distal constructs at 20-fold excess largely prevented detection of apoptotic executioner PARP showed cleavage of PARP upon RFB-4-APC signals. treatment with 500 nmol/L DK1KD-SGIII in all CD22(þ) cell The recombinant immune kinase needs to be internalized by the systems, whereas no processed PARP was detected in CD22() targeted CLL cells to exhibit its cytotoxic function. In flow cyto- U937 cells (Fig. 4B). The purine analogue fludarabine (chemo- metry–based internalization assays, cells were either treated with therapeutic backbone of most CLL therapies) induced apoptosis proteinase K, which removed extracellular proteins including (cPARP) irrespective of CD22 status. CD22, followed by incubation with 50 nmol/L DK1KD-SGIII and Importantly, DK1KD-SGIII showed an advantageous high AHA or vice versa. Cells exclusively treated with proteinase K or selectivity as compared with the classical cytostatic fludarabine. DK1KD-SGIII served as negative or positive controls of the spec- The pronounced cell death induction by the complete immuno- ificity of the fluorescent signal, respectively. As expected, preincuba- kinase DK1KD-SGIII in CLL (n ¼ 4 cases), but not by DK1KD or tion of CD22(þ) Ramos lymphoma cells with proteinase K and SGIII alone (Fig. 4C), contrasted the absence of effects of such subsequent incubation with DK1KD-SGIII revealed hardly any antileukemic concentrations of DK1KD-SGIII on healthy donor binding of the CD22-targeted immune conjugate. The protein- PBMC (3 donors; Fig. 4D), while fludarabine was cytotoxic in ase-K–mediated loss of surface CD22 was confirmed by RFB4-APC. both cell systems. Cell survival as means SEM at 100 nmol/L In contrast,strong fluorescence signals weredetected when DK1KD- DK1KD-SGIII: 1.075 0.029 in PBMC versus 0.750 0.100 in SGIII was added prior to proteinase K (Fig. 3D). As proteinase K CLL; at 500 nmol/L: 1.027 0.038 in PBMC versus 0.573 0.076 removed any extracellular receptor–bound protein–AHA com- in CLL, P ¼ 0.005; for fludarabine at 5 mmol/L: 0.733 0.061 in plexes, it had to be concluded that this fluorescent signal originated PBMC versus 0.395 0.139 in CLL; at 25 mmol/L: 0.710 0.061

(Continued.)C,first purification step with NiNTA-based immobilized metal ion affinity chromatography: Coomassie-stained gel indicating removal of nontarget proteins from elution fractions containing DK1KD-SGIII (red frames mark size-validated target bands) S, supernatant; L, loading; W, wash; E, elution fractions; cE, concentrated elution fraction; R, regeneration. D, second purification step: cation exchange chromatography (CEC) on a CM-sepharose column with high-purity yield of DK1KD-SGIII. UV adsorption peaks (blue) indicate elution of targeted protein (2 peaks; representative fractions in E). Green line shows increasing NaCl concentration. Purple line marks injection of the dialyzed NiNTA eluate. Arrows indicate the fractions analyzed by Coomassie stainings and a-his Western blots as well as in subsequent mass spectrometry (MS) of E. E, high relative purity of DK1KD-SGIII in the absence of contaminating bands in fractions 35 and 38 (Coomassie stained gel; left). Corresponding immunoblots proved the identity of the protein detected by an a-His antibody detecting DK1KD-SGIII's C-terminal his-tag (red frame, size confirmed band of interest). Note the higher enrichment in fraction 38 than in 35, corresponding to the tip of the peak versus its slope in the CEC, respectively (see D). M, molecular weight marker. Right, identification of the target protein DK1KD-SGIII by electron spray ionization tandem MS (ESI-MS/MS)-based peptide-mass fingerprinting of fraction 38. Score: 1016.1; sequence coverage: 49.8%; No. of unique peptides: 26 (underlined). F, kinase activity assays by bioluminescence measurement of ATP consumption (cell-free system). DK1KD-dependent phosphorylation of the exogenous substrate myosin light chain II as per quenched relative light units (RLU). Because of the competition of kinase with luciferase for ATP, kinase activity is inversely proportional to RLU. Results normalized to kinase-free buffer. Left, similar high kinase activity of DK1KD-SGIII and DK1KD at 5 nmol/L (mean SEM: 0.52 0.22 and 0.55 0.08 reduction of RLU, respectively) as compared with the scFv SGIII. Right, dose-dependent increase of kinase activity by titrated DK1KD-SGIII. No significant RLU reduction by SGIII.

www.aacrjournals.org Mol Cancer Ther; 2016 OF7

Lilienthal et al.

Figure 3. The DK1KD–SGIII immunoconjugate specifically binds to target cells and is internalized via SGIII-scFv–mediated engagement of the CD22 B-cell epitope. A, flow cytometric analysis after incubation with DK1KD-SGIII (blue), SGIII (purple), and DK1KD (black) in CD22(þ) Raji cells, primary CLL cells, and CD22() U937 cells. Isotype-controlled mean fluorescence intensity (MFI) demonstrates binding of DK1KD-SGIII and SGIII to CD22(þ) cells, whereas the anti-CD22 scFv-less DK1KD shows no binding. Labeling with anti-His6 Alexa Fluor 488 (AHA). Control (red): AHA. B, fluorescence microscopy shows accumulation of AHA-labeled DK1KD-SGIII (200 nmol/L, 15-minute incubation) on the surface of cytospin-prepared CLL cells. C, flow cytometric competitive binding assays. Binding of 50 nmol/L DK1KD-SGIII (blue) to CD22 of CLL cells was inhibited by preincubation with 10-fold molar excess of a-CD22 mAb RFB-4-APC (black). Preincubation with unspecific isotype IgG2a (green) only slightly reduced DK1KD-SGIII binding. Left histogram, channel for detection of AHA labeling of DK1KD-SGIII. Right, channel for APC (allophycocyanin) of anti-CD22 mAb RFB-4-APC. D, internalization (45 minutes) of AHA-labeled DK1KD-SGIII into CD22(þ) Ramos cells (flow cytometry). ProtKþDK1KD-SGIII (yellow): preincubation with Proteinase K (ProtK) leading to RFB-4-APC confirmed loss of CD22 prior to DK1KD-SGIII exposure; DK1KD-SGIIIþProtK (black): in cells treated with Proteinase K (eliminating surface-bound complexes) after DK1KD-SGIII and AHA incubation the signals originate from internalized CD22–DK1KD–SGIII–AHA complexes (no permeabilization-based AHA flow cytoemtry); DK1KD-SGIII alone (blue); ProtK (red): control. E, fluorescence microscopy showing internalization of DK1KD-SGIII labeled by AHA (see B) into CLL cells over time.

in PBMC versus 0.2850 0.124 in CLL (P ¼ 0.04). This was DK1KD-SGIII acts antileukemic via classical apoptosis and further corroborated by exclusive DK1KD-SGIII binding (AHA induction of autophagy in CLL samples irrespective of assay, previous paragraph) to and killing of B cells within healthy associated clinical risk proﬁle þ þ þ donor PBMC (CD5 /CD19 ) and CLL samples (CD5 /CD19 ) Despite highly efﬁcient therapeutic options in CLL (26), þ while sparing T cells (CD5 /CD19 ) or other non B/T leukocytes relapsed/refractory or even upfront resistant disease remains a (Fig. 4E and F). problem. We tested whether DK1KD-SGIII was capable of killing

OF8 Mol Cancer Ther; 2016 Molecular Cancer Therapeutics

Immunokinase-Based B-Lymphotoxic Delivery of Activated DAPK1

Figure 4. DK1KD-SGIII induces cell death in CD22(þ) cell lines and primary CLL cells in a dose-dependent manner. A, MTT assay–based viability (normalized to PBS control) progressively decreases under increasing concentrations of DK1KD-SGIII (48 hours) in the CD22(þ) lymphoma cell lines: Raji (IC50: 445.4 nmol/L), MEC1 (IC50: 165.5 nmol/L), JVM-3 (IC50: 1331 nmol/L), and DOHH2 (IC50: 164.8 nmol/L); IC50 for CD22-negative U937 cells not reached. B, immunoblots for cleaved products of PARP indicating terminal cell death. DK1KD-SGIII shows dose-dependent cytotoxicity in CD22(þ) Raji cells, while CD22() U937 cells are not affected. Fludarabine as positive control in both cell types. C, cell survival by flow cytometry (AnnexinV/7AAD counts) after 48-hour treatment. Significant decrease of survival of CLL cells (n ¼ 3) by treatment with the complete immunokinase DK1KD-SGIII, but not with single domains DK1KD or SGIII as well as compared with healthy donor PBMC (n ¼ 3). D, cell survival by flow cytometry (AnnexinV/7AAD counts) after 48-hour treatment. DK1KD-SGIII mediates death to CLL cells (n ¼ 4 samples) but not to freshly isolated healthy donor PBMC (n ¼ 3), while fludarabine exhibits toxicity to both cell types. E, targeting of CD22(þ) B-cell subpopulations in both healthy donor PBMC and CLL samples as per flow cytometric–binding assays (AHA labeling, Fig. 3C) after þ þ þ þ DK1KD-SGIII treatment for 48 hours. Top, CD5/CD19 gates with B cells (CD5 /CD19 ) in red among PBMC and in CLL (CD5 /CD19 ); T cells (purple, CD5 /CD19 ) can only be detected in healthy donor PBMC; B/T cells (green, CD5 /CD19 ). Bottom, within gated cells only B cells (red) reveal DK1KD-SGIII binding (MFI þ shifted AHA signal) in contrast to T and non-B/T cells. F, selective elimination of CD22(þ) B cells within healthy donor PBMC (CD5 /CD19 , left, n ¼ 3) and within a CLL patient sample (CD5þ/CD19þ, right, n ¼ 1) after 48 hours of incubation with DK1KD-SGIII. Percentages as relative abundances of respective populations per sample.

www.aacrjournals.org Mol Cancer Ther; 2016 OF9

Lilienthal et al.

Table 1. Characteristics of CLL patients with analyzed responses to DK1KD-SGIII b Patient Binet Previous Cytogenetics/P53 IGHV Response to IC50 (nmol/L) age stage therapya mutation status Zap70/CD38 DK1KD-SGIII DK1KD-SGIII CLL 1 48 B Untreated 13q n.d. Zap70þ/CD38þ Yes 400 CLL 2 Unknown A Untreated XX n.d. n.d. Yes 300 CLL 3 48 B BR, FCR 13q U Zap70/CD38 Yes 275 CLL 4 Unknown A n.d. 13q n.d. n.d. Yes n.d. CLL 5 73 C FC, BR, A 11q/13q M Zap70/CD38 Yes 450 CLL 6 73 A CR 11q M Zap70/CD38 Yes 450 CLL 7 59 A BR 13q M n.d. Yes 600 CLL 8 Unknown C FC, BR, CHOP n.d. n.d. n.d. No n.r. CLL 9 64 n.d. Clb, A, L XX M Zap70/CD38 Yes n.d. CLL 10 73 C FCR XX U Zap70/CD38 Yes 875 CLL 11 69 n.d. Untreated 17p U n.d. Yes n.d. CLL 12 55 A Untreated XX M Zap70þ/CD38þ Yes n.d. CLL 13 78 C FCR, A, L n.d. U Zap70/CD38 Yes n.d. CLL 14 72 C Untreated 17p U n.d. No n.a. CLL 15 Unknown A FCR, A n.d. n.d. n.d. Yes n.d. NOTE: A complete Table also including those cases in which only DAPK1/2 expression was evaluated without response evaluations to DK1KD-SGIII (16 additional cases) is given as Supplementary Table S2. Abbreviations: A, alemtuzumab; B, bendamustine; C, cyclophosphamide; CHOP, cyclophosphamide, doxorubicin, vincristine, prednisolone; Clb, Chlorambucil;

F, fludarabine; IC50, inhibitory concentration 50; L, lenalidomide; M, IGHV gene mutated; n.a., not applicable; n.d., not defined; No, no DAPK expression detected; n.r., not reached; R, rituximab; U, IGHV unmutated; unknown, not known from patient file; yes, DAPK expression detected. aAnti-leukemic treatment regimen(s) given over clinical course >1 month before sample was drawn. bIn vitro treatment with 500 nmol/L for 48 hours: yes, >20% viability reduction (above control) after treatment; unknown patient ages as a result of blinded identifiers when enrolled in clinical trials.

CLL cells ex vivo from chemoimmunotherapy-resistant disease. totally abrogated by the pan-caspase inhibitor ZVAD-fmk (Fig. Incubation of freshly isolated CLL cells with either DK1KD-SGIII 5F). In conjunction with the AnnexinV/7AAD data, this implicat- or fludarabine (48 hours) was followed by cell death/viability ed that the DK1KD-SGIII–evoked PARP-mediated cell death was analysis by AnnexinV/7AAD flow cytometry/MTT assays. We to a large part via caspase-dependent classical apoptosis. How- found 13 of 15 CLL samples responding (control-corrected ever, incubation with DK1KD-SGIII also induced the expression cytotoxicity of >20%) to DK1KD-SGIII (Table 1), providing of the autophagy marker LC3B which is known to correlate with further proof-of-concept for the DAPK1 immunoligand the number of autophagosomes (47). It indicates that, as for approach. Correlation with available clinicopathologic data native DAPK1, autophagy might represent a route alternative to including previous treatment of patients revealed similar efficacy conventional apoptosis in mediating DK1KD-SGIII-induced of DK1KD-SGIII across the subsets defined by high-risk viability reduction (Fig. 5F). This is particularly relevant, as strata. Figure 5A–C show representative cases, for example, a this mostly p53-independent program might be responsible for clinically fludarabine-resistant patient (in vitro IC50 not reached) the preserved efficacy of DK1KD-SGIII noted in 2 of the 3 tested in which DK1KD-SGIII (IC50: 868.7 nmol/L) induced cell death CLL cases with cytogenetics (del11q23, del17p) indicating dose dependently (Fig. 5C). Clinical pretreatment (Fig. 5B) or defects in classical p53-mediated apoptosis. Fittingly, the the presence of poor risk cytogenetic aberrations (i.e., p53-mutated MEC1 cells were efficiently killed by DK1KD- del11q23, Fig. 5A) also did not influence the in vitro perfor- SGIII as well (Fig. 4A). mance of DK1KD-SGIII. Genuine DAPK1 regulates the execution of predominantly type I (apoptosis) and type II (autophagy) programmed cell death Discussion (PCD; refs. 4, 14, 39). Therefore, we finally investigated the We present a novel and efficient strategy to reconstitute the mechanisms employed by DK1KD-SGIII to induce B-cell lym- activity of proapoptotic DAPK1 in malignant B cells, which we phoma/leukemia cell death and to overcome the elevated apo- confirm to lack expression of this important tumor suppressor, ptotic threshold of CLL. Analysis of nuclear morphology of especially in the majority of cases of CLL. The concept was an DK1KD-SGIII–treated CLL cells by fluorescence microscopy immunoligand-based delivery of a constitutively active DAPK1 revealed a high prevalence of fragmentation characteristic for kinase domain DK1KD, which is stripped off its autoregulatory apoptotic cells (Fig. 5D). Furthermore, DK1KD-SGIII treatment element. After rounds of receptor selection, DAPK1 was fused to resulted in time-dependent increases in AnnexinVþ/7AAD (12 SGIII, a humanized anti-CD22 scFv. This novel immunokinase hours, 24 hours) and AnnexinVþ/7AADþ (48 hours) cells as a (not a classical immunotoxin) DK1KD-SGIII showed superior typical apoptotic sequel. In contrast, an AnnexinV/7AADþ cell stability, robust binding to CD22, and rapid internalization into population, indicating a necroptotic route, was not detected. This target cells. We established a protocol for the high-yield purifi- largely excluded this type of PCD as a prominent mode of action cation of the fusion protein via FPLC-based IMAC and CEC. of DK1KD-SGIII (Fig. 5E). DK1KD-SGIII selectively induced apoptosis and autophagy in Immunoblot analysis for the apoptotic execution-phase pro- B-cell lymphoma lines and primary CLL, in the latter irrespective tein caspase-3 revealed marked increases in proteolytically proof clinical therapy resistance. cessed (cleaved) caspase-3 products upon exposure of CLL cells to DAPK proteins, as best established for DAPK1, represent DK1KD-SGIII, which (including PARP processing) was (sub) important integrative nodes between apoptosis and autophagic

OF10 Mol Cancer Ther; 2016 Molecular Cancer Therapeutics

Immunokinase-Based B-Lymphotoxic Delivery of Activated DAPK1

Figure 5. DK1KD-SGIII induces cell death via conventional apoptosis and triggers autophagy in CLL samples irrespective of clinical risk categories. A–C, viability as per MTT assay (means; SEM) of freshly isolated CLL cells (patient # and details in Supplementary Table S2) after 48-hour DK1KD-SGIII or fludarabine treatments. A, the presence of defined chromosomal aberrations does not determine differential reduction of viability by DK1KD-SGIII. Shown are representative patients: #18 (IC50:300 nmol/L; normal karyotype), #19 (IC50: 275 nmol/L; isolated del13q14), and #21 (IC50: 450 nmol/L; del11q23/ del13q14). B, in vitro activity of DK1KD- SGIII is independent of previous clinical treatment; shown is therapy- na€ve patient #15 (IC50: 300 nmol/L) compared with CR-treated patient #22 (IC50: 450 nmol/L), BR- pretreated patient #23 (IC50: 600 nmol/L), and FCR-pretreated patient #26 (IC50: 875 nmol/L). B, bendamustine; C, cyclophosphamide; F, fludarabine; R, rituximab. C, DK1KD- SGIII reduces viability (IC50: 875 nmol/L) in the presence of clinical and in vitro fludarabine resistance (IC50: n.r.). D, fluorescence microscopy (DAPI stain) reveals nuclear fragmentation of CLL cells after treatment with 500 nmol/L DK1KD-SGIII for 48 hours. E, AnnexinV/7AAD flow cytometry of freshly isolated CLL cells after treatment with 500 nmol/L DK1KD- SGIII for indicated times. Dead cells (7AADþ) arise through AnnexinV expression, strictly via early apoptosis (12 hours) to late apoptosis (24–48 hours), while no obvious route to AnnexinV/7AADþ stages is discernable. F, immunoblots of two representative CLL patients (Supplementary Table S2), both previously chemoimmunotherapy treated (>4 weeks prior to sampling). Left, patient #22 (CR treated); right, #26 (FCR treated). Cleavage (C) of PARP and caspase-3 after incubation with DK1KD-SGIII are reversible by the pan-caspase inhibitor ZVAD. Only DK1KD-SGIII, but not fludarabine, induces an increased expression of autophagy-specific LC3B-I/II. Reductions of DK1KD-SGIII induced cPARP by ZVAD without affecting LC3B-I/II. Levels of Bcl2 revealed no informative changes. Concentrations: DK1KD-SGIII, 500 nmol/L; fludarabine, 5 mmol/L; 48-hour treatment for both. cell death pathways (1, 2). Mounting evidence implicates their igenetic step in a broad spectrum of cancers (1, 2). We hypoth- dysregulation, mostly via (epigenetic) suppression, as a failed esized that the proleukemogenic role of DAPK1 and its tumor- safeguarding p53-activating mechanism, hence, a central tumor- associated loss would provide an actionable vulnerability and by

www.aacrjournals.org Mol Cancer Ther; 2016 OF11

Lilienthal et al.

that a strong therapeutic rationale for the unique concept of signals from the original anti-CD22 RFB-4 antibody implicates reinstatement of a tumor suppressor. Important evidence for a the reliability of this standard diagnostic read-out in the screening preserved relevance of deficient DAPK at the tumor stage derived for tumors suitable for targeting by DK1KD-SGIII. We consider a from functional antisense-based screens (1). certain degree of toxicity towards normal CD22(þ) B cells Antibody–drug conjugates (ADC) hold great promise towards (Fig. 4F) as an acceptable trade-off. more targeted cancer therapies. These chimeric proteins of ever- The stages of development of this novel immunokinase includ- optimized chemistries are generally composed of a classical ed thorough evaluations of functionality. The DK1KD component cytotoxic moiety and a binding domain connected by a linker showed robust catalytic activity, the DK1KD-SGIII fusion protein (48, 49). The ideally tumor-specificaffinity of ADC results from an was stable, and it revealed specific binding to CD22(þ) cells advantageous tumor-to-normal distribution of the targeted sur- followed by a rapid endocytosis. As expected, on the basis of our face epitope. Stability, affinity, internalization kinetics, and cell- experiments of genetic DK1KD overexpression, the active immu- intrinsic processing are further determinants of ADC perfor- noligand DK1KD-SGIII induced marked cell death in the mance. Prominent examples of successful clinical application 5 CD22(þ) lymphoma cell lines (2 Burkitt, 1 FL/DLBCL, 2 and marketing approval are conventional immunotoxins, that CLL-derived; IC50s ranging from 170 to 1,300 nmol/L) and in is, brentuximab-vedotin (anti-CD30 mAb with monomethyl 13 of 15 freshly isolated primary CLL samples (IC50s ranging from auristatin E, for refractory or relapsed Hodgkin lymphoma and 275 to 875 nmol/L). No cytotoxicity was observed in the non-B systemic anaplastic large cell lymphoma), gemtuzumab-ozoga- cells of healthy donor or CLL samples, most likely as they are micin (anti-CD33 mAb with a calicheamicin derivative, for CD22(). The fusion protein also bound the CD22(þ) B cells of relapsed AML), or trastuzumab-emtansine (anti-HER2 mAb with healthy donors, but further experiments need to show how þ DM1, for inoperable/metastasized HER2 breast cancer; ref. 50). differentially DK1KD-SGIII acts in death induction of normal B Other immunoligands such as those with bispecific antibody cells versus lymphoma/CLL cells. constructs or natural ligands are designed to attract components Seminal work demonstrated that silencing of the tumor sup- of the immune system to harness their antitumor potential (51). pressor DAPK1 is a key feature of familial and sporadic CLL, Our novel immunokinase format (not an immunotoxin) has implicating it in early leukemogenesis (16, 17). We confirm here already shown promising activity in Hodgkin lymphoma xeno- the aberrant expression pattern at the protein level. Our studies of grafts as a DAPK20-CD30L construct (24). protein expression and in silico GEP meta-analysis implicate that In the process of ligand-epitope selection, crucial in determin- DAPK1 levels are reduced in normal B cells as well (Fig. 1 and ing the fate of the whole construct (52), we arrived at anti-CD22 Supplementary Fig. S1; Supplementary Table S1A). This is in based on several of its advantages over other B-cell–specific mAb accordance with previous findings of marked DAPK promoter or scFv with available sequence information. First, the binding of hypermethylation in a higher percentage of normal B cells than in the humanized anti-CD22 scFV SGIII was consistent and stable non-B lineage cells (36). The differential size of DAPK1 hyper- (e.g., in contrast to the anti-CD19 mAb HD37) and it did not methylated subsets within the B-cell fraction of healthy indivi- induce apoptosis as compared with the CD40 antibody fragment duals defined by IgM expression (36) or otherwise (Supplemen- G28-5. Furthermore, unlike the CLL-associated targets for tary Fig. S1) indicates that there might be a predestined DAPK- "naked" antibodies CD20 or CD52, the transmembrane receptor deficient precursor compartment of particular lymphomagenic CD22 is rapidly internalized (42, 45). CD22 is a member of the susceptibility (53). SIGLEC family of lectins and emerges as one of the most intensely In CLL, characterized by a genetically and environmentally pursued B-cell–specific targets in formats of immunoligands (e.g., instructed high apoptotic threshold, it is desirable to therapeuti- inotuzumab ozogamicin) or CAR T-cell–directed therapies (45). cally circumvent conventional und p53-dependent forms of PCD. Finally, we had chosen here an anti-CD22 scFv version over a mAb In fact, an attractive feature of DK1KD-SGIII is that it not only construct for various reasons of immediate consequence or for induces classical caspase- and PARP-mediated apoptosis, but also subsequent clinical application: easier cloning and recombinant triggers the expression of autophagic LC3B in CLL cells. The latter production, facilitated endocytosis as well as lower immunoge- clearly distinguishes the immunokinase from fludarabine (Fig. nicity, and lack of unspecific uptake by Fc receptor carrying cells, 5F). As caspase inhibition was associated with near-complete both due to the missing Fc region. Although CD22 regulates abrogation of DK1KD-SGIII–induced PARP cleavage without apoptotic signaling (45), engagement with our SGIII component affecting LC3B levels, further studies have to investigate whether without the DK1KD part did not noticeably influence survival. the construct induces actual autophagic cell death and whether It remains speculative to which degree the levels of DAPK1 this is PARP independent, or if an autophagic self-preservative expression (its tumor- or B-cell-of-origin associated loss; Fig. 1 program is activated in a subset of CLL cells. This is of particular and Supplementary Fig. S1; Supplementary Tables S1A and S2) interest as fludarabine resistance had been associated with auto- affect the response towards DK1KD-SGIII. Likely, the constitu- phagy addiction (54). tively active DK1KD overrides the regulated (stress-dependent) Although showing low to intermediate efficacy in some cell operation of natural DAPK1; hence (residual), DAPK1 expression lines, further legitimate hope for a clinically meaningful would not pose major restrictions towards DK1KD-SGIII efficacy. advantageous profile of cell death induction by DK1KD-SGIII In fact, the specifics of lymphoma-associated dysregulated sur- is based on our observation that it is equiefficacious across CLL vival pathways and high (oncogenic) stress levels would provide of distinct clinical risk features. Importantly, the immunoki- some selectivity. We postulate that the levels of CD22 expression nase sustained its high activity in chemoimmunotherapy (a key factor for binding and internalization) more profoundly refractory CLL and required a low IC50 in p53 mutated and determine the effect of our construct. Fittingly, there was a trend of fludarabine resistant (55) MEC1 cells. Studies on larger higher CD22 expression with higher sensitivity to DK1KD-SGIII. cohorts of risk-stratified CLL need to solidify these preliminary The near-complete correlation of SGIII scFv binding with the findings.

OF12 Mol Cancer Ther; 2016 Molecular Cancer Therapeutics

Immunokinase-Based B-Lymphotoxic Delivery of Activated DAPK1

Overall, in a proof-of-principle, the strategy of reinstating B-cell Writing, review, and/or revision of the manuscript: N. Lilienthal, G. Lohmann, lymphoma associated lost tumor-suppressive DAPK1 by deliver- E. Vasyutina, S. Zittrich, M. Hallek, S. Barth, M. Herling ing its constitutive kinase domain via a CD22-specific immuno- Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Zittrich, P. Mayer, M.K. Tur, M. Hallek, ligand is highlighted by encouraging biotechnologic feasibility, G. Pfitzer, M. Herling superior selectivity, and notable in vitro efficacy. The cytotoxicity Study supervision: P. Mayer of this prototype will have to be increased in subsequent steps of Other (purification of protein): G. Pfitzer preclinical optimization, for example, including evaluations of valency (up to 10-fold increase; ref. 56), introduction of a trans- location-driving protein domain (up to 20-fold increase; ref. 57) Acknowledgments and removal of the 6xHis-tag, or using SGIII as a diabody (9- to The authors thank R.J. Ford, M. Binder, and L. Frenzel for provision of cell lines and D. Beutner for human tonsils. DNA for the a-CD19 antibody HD37 48-fold increase; ref. 56). DK1KD-SGIII carries high potential for fi a fi was a kind gift from Prof. M. Little (Af med), DNA for the -CD22 scFv SGIII intensi ed preclinical testing. Subject of such additional studies was a kind gift from Dr. M. Arndt (National Center for Tumor Diseases, should be the detailed modes of cell death execution, immuno- University Hospital of Heidelberg, Heidelberg, Germany), and DNA for the genicity, the role of cell-intrinsic mechanisms of immunotoxin a-CD40 scFv G28-5 was a kind gift from Dr. Tanja de Gruijl (Division of resistance (58), or the in vivo performance of species-adapted Immunotherapy, Department Medical Oncology, Vrije Universiteit Medical constructs to assess the penetration to sanctuary sites and milieu Center, Amsterdam, the Netherlands). compartments. We envision such optimized immunokinase variants to broaden the armamentarium of antilymphoma/CLL therapies in notoriously hard-to-treat cases ideally in the context Grant Support of low residual tumor burden. This work was supported by The German Cancer Aid Max-Eder award (to M. Herling), the German Research Foundation (DFG; HE-3553/3-1, to M. fl Herling; and SCHW-1711/1-1, to C.D. Herling) as part of the collaborative Disclosure of Potential Con icts of Interest research group KFO-286, by a grant from the CLL Global Research Foundation No potential conflicts of interest were disclosed. (to M. Herling), the German Jose-Carreras leukemia foundation (DJCLS R 12/ 08), and by the local CECAD initiative (to M. Herling). N. Lilienthal received a Authors' Contributions stipend by a joint pharmacology graduate program between the University of Conception and design: N. Lilienthal, M.K. Tur, S. Barth, M. Herling Cologne and the Bayer Health Care AG. E. Vasyutina was supported by the local € Development of methodology: N. Lilienthal, S. Barth Koln-Fortune program. Acquisition of data (provided animals, acquired and managed patients, The costs of publication of this article were defrayed in part by the payment of advertisement provided facilities, etc.): N. Lilienthal, G. Lohmann, S. Zittrich, P. Mayer, page charges. This article must therefore be hereby marked in C.D. Herling accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): N. Lilienthal, G. Lohmann, G. Crispatzu, E. Vasyutina, Received August 19, 2015; revised January 8, 2016; accepted January 13, 2016; C.D. Herling, M. Herling published OnlineFirst January 29, 2016.

References 1. Gozuacik D, Kimchi A. DAPk protein family and cancer. Autophagy 11. Chen C-H, Wang W-J, Kuo J-C, Tsai H-C, Lin J-R, Chang Z-F, et al. 2006;2:74–9. Bidirectional signals transduced by DAPK-ERK interaction promote the 2. Michie AM, McCaig AM, Nakagawa R, Vukovic M. Death-associated protein apoptotic effect of DAPK. EMBO J 2005;24:294–304. kinase (DAPK) and signal transduction: regulation in cancer. FEBS J 12. Bialik S, Bresnick AR, Kimchi A. DAP-kinase-mediated morphological 2010;277:74–80. changes are localization dependent and involve myosin-II phosphoryla- 3. Kawai T,Matsumoto M, Takeda K,Sanjo H,Akira S.ZIP kinase,a novel serine/ tion. Cell Death Differ 2004;11:631–44. threonine kinase which mediates apoptosis. Mol Cell Biol 1998;18:1642–51. 13. Zalckvar E, Berissi H, Eisenstein M, Kimchi A. Phosphorylation of Beclin 4. Inbal B, Shani G, Cohen O, Kissil JL, Kimchi A. Death-associated protein 1 by DAP-kinase promotes autophagy by weakening its interactions with kinase-related protein 1, a novel serine/threonine kinase involved in Bcl-2 and Bcl-X L. Autophagy 2009;5:720–2. apoptosis. Mol Cell Biol 2000;20:1044–51. 14. Eisenberg-Lerner A, Kimchi A. DAP kinase regulates JNK signaling by 5. Kristensen LS, Asmar F, Dimopoulos K, Nygaard MK, Aslan D, Hansen JW, binding and activating protein kinase D under oxidative stress. Cell Death et al. Hypermethylation of DAPK1 is an independent prognostic factor Differ 2007;14:1908–15. predicting survival in diffuse large B-cell lymphoma. Oncotarget 2014;5: 15. Steinmann S, Scheibe K, Erlenbach-Wuensch K, Neufert C, Schneider-Stock 9798–810. R. Death-associated protein kinase: a molecule with functional antagonis- 6. Chim CS, Fung TK, Wong KF, Lau JS, Liang R. Frequent DAP kinase but not tic duality and a potential role in inflammatory bowel disease (Review). Int p14 or Apaf-1 hypermethylation in B-cell chronic lymphocytic leukemia. J Oncol 2015;47:5–15. J Hum Genet 2006;51:832–8. 16. Raval A, Tanner SM, Byrd JC, Angerman EB, Perko JD, Chen S-S, et al. 7. Li F-F, Yang Y, Wang X-L, Hong Y-Y, Wang N-F, Chen Z-D. Promoter Downregulation of death-associated protein kinase 1 (DAPK1) in chronic methylation of DAPK gene may contribute to the pathogenesis of nonsmall lymphocytic leukemia. Cell 2007;129:879–90. cell lung cancer: a meta-analysis. Tumour Biol 2014;35:6011–20. 17. Wei Q-X, Claus R, Hielscher T, Mertens D, Raval A, Oakes CC, et al. 8. Kalantari M, Osann K, Calleja-Macias IE, Kim S, Yan B, Jordan S, et al. Germline allele-specific expression of DAPK1 in chronic lymphocytic Methylation of human papillomavirus 16, 18, 31, and 45 L2 and L1 genes leukemia. PLoS One 2013;8:e55261. and the cellular DAPK gene: considerations for use as biomarkers of the 18. Giachelia M, Bozzoli V, D'Alo F, Tisi MC, Massini G, Maiolo E, et al. progression of cervical neoplasia. Virology 2014;448:314–21. Quantification of DAPK1 promoter methylation in bone marrow and periph- 9. Schlegel CR, Georgiou ML, Misterek MB, Stocker€ S, Chater ER, Munro CE, eral blood as a follicular lymphoma biomarker. J Mol Diagn 2014;16:467–76. et al. DAPK2 regulates oxidative stress in cancer cells by preserving mito- 19. Pineda de las Infantas MJ, Torres-Rusillo S, Unciti-Broceta JD, Fernandez- chondrial function. Cell Death Dis 2015;6:e1671. Rubio P, Luque-Gonzalez MA, Gallo MA, et al. Synthesis of 6,8,9 poly- 10. Raveh T, Droguett G, Horwitz MS, DePinho RA, Kimchi A. DAP kinase substituted purine analogue libraries as pro-apoptotic inducers of human activates a p19ARF/p53-mediated apoptotic checkpoint to suppress onco- leukemic lymphocytes and DAPK-1 inhibitors. Org Biomol Chem 2015; genic transformation. Nat Cell Biol 2001;3:1–7. 13:5224–34.

www.aacrjournals.org Mol Cancer Ther; 2016 OF13

Lilienthal et al.

20. Huang Y, Chen L, Guo L, Hupp TR, Lin Y. Evaluating DAPK as a therapeutic 39. Inbal B, Bialik S, Sabanay I, Shani G, Kimchi A. DAP kinase and DRP-1 target. Apoptosis 2014;19:371–86. mediate membrane blebbing and the formation of autophagic vesicles 21. Leonczak P, Gao L-J, Ramadori AT, Lescrinier E, Rozenski J, De Jonghe S, during programmed cell death. J Cell Biol 2002;157:455–68. et al. Synthesis and structure-activity relationship studies of 2-(1,3,4- 40. Porro G, Lento P, Marcucci F, Gromo G, Modena D. Different cytotoxic oxadiazole-2(3H)-thione)-3-amino-5-arylthieno[2,3-b]pyridines as inhi- activity and intracellular fate of an anti-CD5-momordin immunotoxin in bitors of DRAK2. ChemMedChem 2014;9:2587–601. normal compared to tumour cells. Cancer Immunol Immunother 1995; 22. Gao L-J, Kovackova S, Sala M, Ramadori AT, De Jonghe S, Herdewijn P. 40:213–8. Discovery of dual death-associated protein related apoptosis inducing 41. Sieber T, Schoeler D, Ringel F, Pascu M, Schriever F. Selective internaliza- protein kinase 1 and 2 inhibitors by a scaffold hopping approach. J Med tion of monoclonal antibodies by B-cell chronic lymphocytic leukaemia Chem 2014;57:7624–43. cells. Br J Haematol 2003;121:458–61. 23. Tur MK, Neef I, Jost E, Galm O, J€ager G, Stocker€ M, et al. Targeted restoration 42. Tu X, LaVallee T, Lechleider R. CD22 as a target for cancer therapy. J Exp of down-regulated DAPK2 tumor suppressor activity induces apoptosis in Ther Oncol 2011;9:241–8. Hodgkin lymphoma cells. J Immunother 2009;32:431–41. 43. Hanna R, Ong GL, Mattes MJ. Processing of antibodies bound to B-cell 24. Tur MK, Neef I, J€ager G, Teubner A, Stocker€ M, Melmer G, et al. Immu- lymphomas and other hematological malignancies. Cancer Res 1996; nokinases, a novel class of immunotherapeutics for targeted cancer ther- 56:3062–8. apy. Curr Pharm Des 2009;15:2693–9. 44. Krauss J, Arndt MAE, Martin ACR, Liu H, Rybak SM. Specificity grafting of 25. Porter DL, Levine BL, Kalos M, Bagg A, June CH. Chimeric antigen receptor- human antibody frameworks selected from a phage display library: gen- modified T cells in chronic lymphoid leukemia. N Engl J Med 2011; eration of a highly stable humanized anti-CD22 single-chain Fv fragment. 365:725–33. Protein Eng 2003;16:753–9. 26. Hallek M. Signaling the end of chronic lymphocytic leukemia: new 45. Sullivan-Chang L, O'Donnell RT, Tuscano JM. Targeting CD22 in B-cell frontline treatment strategies. Blood 2013;122:3723–34. malignancies: current status and clinical outlook. BioDrugs 2013;27: 27. Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero G, Dohner€ 293–304. H, et al. Guidelines for the diagnosis and treatment of chronic lymphocytic 46. Stocker€ M, Tur MK, Sasse S, Krußmann€ A, Barth S, Engert A. Secretion of leukemia: a report from the International Workshop on Chronic Lympho- functional anti-CD30-angiogenin immunotoxins into the supernatant of cytic Leukemia updating the National Cancer Institute-Working Group transfected 293T-cells. Protein Expr Purif 2003;28:211–9. 1996 guidelines. Blood 2008;111:5446–56. 47. Tanida I, Ueno T, Kominami E. LC3 and autophagy. Methods Mol Biol 28. Herling M, Patel KA, Weit N, Lilienthal N, Hallek M, Keating MJ, et al. High 2008;445:77–88. TCL1 levels are a marker of B-cell receptor pathway responsiveness and 48. Kreitman RJ. Immunotoxins for targeted cancer therapy. AAPS J 2006;8: adverse outcome in chronic lymphocytic leukemia. Blood 2009;114: E532–51. 4675–86. 49. Merten H, Brandl F, Pluckthun€ A, Zangemeister-Wittke U. Antibody-drug 29. Herling M, Patel KA, Hsi ED, Chang K-C, Rassidakis GZ, Ford R, et al. TCL1 conjugates for tumor targeting-novel conjugation chemistries and the in B-cell tumors retains its normal b-cell pattern of regulation and is a promise ofnon-IgG binding proteins. Bioconjug Chem 2015;26:2176–85. marker of differentiation stage. Am J Surg Pathol 2007;31:1123–9. 50. Zolot RS, Basu S, Million RP. Antibody-drug conjugates. Nat Rev Drug 30. Lanemo Myhrinder A, Hellqvist E, Bergh A-C, Jansson M, Nilsson K, Discov 2013;12:259–60. Hultman P, et al. Molecular characterization of neoplastic and normal 51. Rothe A, Jachimowicz RD, Borchmann S, Madlener M, Keßler J, Reiners KS, "sister" lymphoblastoid B-cell lines from chronic lymphocytic leukemia. et al. The bispecific immunoligand ULBP2-aCEA redirects natural killer Leuk Lymphoma 2013;54:1769–79. cells to tumor cells and reveals potent anti-tumor activity against colon 31. Ford RJ, Goodacre A, Ramirez I, Mehta SR, Cabanillas F. Establishment and carcinoma. Int J Cancer 2014;134:2829–40. characterization of human B-cell lymphoma cell lines using B-cell growth 52. Dosio F, Brusa P, Cattel L. Immunotoxins and anticancer drug conjugate factor. Blood 1990;75:1311–8. assemblies: the role of the linkage between components. Toxins 2011; 32. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene 3:848–83. expression and hybridization array data repository. Nucleic Acids Res 53. Oakes CC, Seifert M, Assenov Y, Gu L, Przekopowitz M, Ruppert AS, et al. 2002;30:207–10. DNA methylation dynamics during B cell maturation underlie a contin- 33. Hennemann H, Vassen L, Geisen C, Eilers M, Mor€ oy€ T. Identification of a uum of disease phenotypesin chronic lymphocytic leukemia. Nat Genet. novel Kruppel-associated€ box domain protein, Krim-1, that interacts 2016 Jan 18. [Epub ahead of print] with c-Myc and inhibits its oncogenic activity. J Biol Chem 2003;278: 54. Sharma A, Singh K, Mazumder S, Hill BT, Kalaycio M, Almasan A. BECN1 28799–811. and BIM interactions with MCL-1 determine fludarabine resistance in 34. Lin Y, Stevens C, Harrison B, Pathuri S, Amin E, Hupp TR. The alternative leukemic B cells. Cell Death Dis 2013;4:e628. splice variant of DAPK-1, s-DAPK-1, induces proteasome-independent 55. Henrich S, Christopherson RI. Multiple forms of nuclear p53 formed in DAPK-1 destabilization. Mol Cell Biochem 2009;328:101–7. human Raji and MEC1 cells treated with fludarabine. Leukemia 2008; 35. Jupp S, Malone J, Bolleman J, Brandizi M, Davies M, Garcia L, et al. The EBI 22:657–60. RDF platform: linked open data for the life sciences. Bioinformatics 56. Ribbert T, Thepen T, Tur MK, Fischer R, Huhn M, Barth S. Recombinant, 2014;30:1338–9. ETA0-based CD64 immunotoxins: improved efficacy by increased valency, 36. Reddy AN, Jiang WW, Kim M, Benoit N, Taylor R, Clinger J, et al. Death- both in vitro and in vivo in a chronic cutaneous inflammation model in associated protein kinase promoter hypermethylation in normal human human CD64 transgenic mice. Br J Dermatol 2010;163:279–86. lymphocytes. Cancer Res 2003;63:7694–8. 57. Hetzel C, Bachran C, Fischer R, Fuchs H, Barth S, Stocker€ M. Small cleavable 37. Cohen O, Feinstein E, Kimchi A. DAP-kinase is a Ca2þ/calmodulin- adapters enhance the specific cytotoxicity of a humanized immunotoxin dependent, cytoskeletal-associated protein kinase, with cell death-induc- directed against CD64-positive cells. J Immunother 2008;31:370–6. ing functions that depend on its catalytic activity. EMBO J 1997;16: 58. Pasetto M, Antignani A, Ormanoglu P, Buehler E, Guha R, Pastan I, et al. 998. –1008 Whole-genome RNAi screen highlights components of the endoplasmic 38. Cohen O, Kimchi A. DAP-kinase: from functional gene cloning to estab- reticulum/Golgi as a source of resistance to immunotoxin-mediated cyto- lishment of its role in apoptosis and cancer. Cell Death Differ 2001;8:6–15. toxicity. Proc Natl Acad Sci U S A 2015;112:E1135–42.

OF14 Mol Cancer Ther; 2016 Molecular Cancer Therapeutics

A Novel Recombinant Anti-CD22 Immunokinase Delivers Proapoptotic Activity of Death-Associated Protein Kinase (DAPK) and Mediates Cytotoxicity in Neoplastic B Cells

Nils Lilienthal, Gregor Lohmann, Giuliano Crispatzu, et al.

Mol Cancer Ther Published OnlineFirst January 29, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-15-0685

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2016/01/29/1535-7163.MCT-15-0685.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/early/2016/03/25/1535-7163.MCT-15-0685. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.