Volume 24, Issue 7, Pages 1312-1322 (July 2016) Enhanced ADCC and NK Cell Activation of an Anticarcinoma Bispecific Antibody by Genetic Insertion of a Modified IL-15 Cross-linker  Joerg U Schmohl, Martin Felices, Elizabeth Taras, Jeff S Miller, Daniel A Vallera  Molecular Therapy  Volume 24, Issue 7, Pages 1312-1322 (July 2016) DOI: 10.1038/mt.2016.88 Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 1 1615EpCAM TriKE elicits superior purification properties over EpCAM16 BiKE. Diagrams (a) and (b) indicate gene placement of the various BiKE and TriKE moieties in pET expression vectors. (c) The absorbance tracing for 1615EpCAM TriKE eluted from the FFQ ion exchange column as the first phase in drug purification using a 3-step elution protocol. The first peak eluted from the column represents our product. (d) Similar quantity of inclusion bodies were refolded and purified for EpCAM16 BiKE, the absorbance of peak 1 is displayed. (e) The second step purification over a size exclusion column. (f) The SDS-PAGE gel and Coomasie Blue staining shows the purity and size. Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 2 Analysis of EpCAM expression on cancer cell lines of various origins and blocking. (a) EpCAM expression was analyzed using EpCAM-FITC on BT-474, PC-3, UMSCC-11B, Calu-3, and negative control U87 glioma and Daudi lymphoma cell lines. (b) Flow cytometry based blocking assay testing the ability of anti-EpCAM scFv (EpCAM) to block the binding of FITC-labeled 1615EpCAM on EpCAM+ HT-29 cells. CD2219 BiKE was used as a negative control since HT-29 cells do not express CD22 and CD19. Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 3 Evaluation of the activity of the 1615EpCAM TriKE in 51chromium release assays. Freshly isolated peripheral blood mononuclear cells were added to HT-29 target cells with the respective effector:target ratios in a chromium release assay to quantitate NK cell activity as indicated. Donors were chosen with naturally different levels of circulating NK cells; (a) 3.8%, (b) 6.4%, (c) 15%, or (d) > 80% enriched NK cells. Effectors and target cells were treated with no antibody (No Ab), anti-CD16 scFv (CD16), EpCAM16 BiKE, or 1615EpCAM TriKE. (e) Comparison of killing with TriKE amongst donors. (f) Comparison of killing with the EpCAM16 BiKE. Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 4 Testing the presence of CD107a positivity as an indicator for lytic degranulation and the presence of IFN-γ positive cells using flow cytometry. Freshly isolated PBMCs containing NK effectors (E) were added to target cells (T). (a) CD107a and (b) IFN-γ expression on CD56+CD3- NK cells was measured after incubation with or without HT-29 targets and anti-EpCAM scFv, EpCAM16, 1615EpCAM, or EpCAM16 plus IL-15 (n = 3). (c) CD107a expression and (d) IFN-γ on NK cells after incubation with EpCAM expressing HT-29 targets and 1615EpCAM or controls: anti-EpCAM scFv (EpCAM), anti-CD16 scFv (CD16), interleukin-15 (IL-15) cloned in our laboratory (self) or obtained from NCI (NCI), CD22CD19 bispecific scFv (CD2219), or IL-12/IL-18 control without targets. Significance was observed with 1615EpCAM treatment compared to controls (marked with *), (n = 3). (e) CD107a and (f) IFN-γ expression after incubation with EpCAM- HL60 target cells (n = 3). P values were estimated with one-way analysis of variance and presented with SD. Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 5 Lytic degranulation in different target cancer cell lines. 1615EpCAM was studied for its ability to induce lytic degranulation in various EpCAM+ cell lines. Effector peripheral blood mononuclear cells were incubated with different EpCAM bearing target cell lines (n = 3) including BT-474 (a), SK-BR-3 (b), PC-3 (c), DU145 (d), UMSCC-11B (e), NA (f), and SKOV-1 (g). 107a expressing cells were evaluated within the gated CD56+/CD3- NK-cell population. TriKE added to effector (E) and target (T) cells induced significantly higher percentages of CD107a expression compared to controls (Effectors + Targets; E+T+ interleukin-15 (IL-15), E+T+CD16CD133 and E+IL-12/IL-18) (marked with *)). The TriKE showed also enhanced degranulation compared to the BiKE EpCAM16 (marked with # or ##). P values were estimated with one-way analysis of variance and presented with SD. Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 6 Interferon-γ (IFN-γ) production in different cancer cell lines. In the same flow cytometry experiment shown in Figure 5, 1615EpCAM was evaluated for IFN-γ production in different EpCAM bearing target cell lines (n = 3) including BT-474 (a), SK-BR-3 (b), PC-3 (c), DU145 (d), UMSCC-11B (e), NA (f), and SKOV-1 (g). The percentage of IFN-γ expressing cells were enhanced by 1615EpCAM treatment as compared to controls in all cell lines (Effectors + Targets (+) interleukin-15 (IL-15), CD16CD133 and IL-12/IL-18) (marked with *) except of the supraphysiologic IL-12/IL-18 control. 1615EpCAM expression was significantly higher in BT-474 and SK-BR-3 cancer cell lines (marked with #) compared to the BiKE EpCAM16. P-values were estimated with one-way analysis of variance and presented with SD. Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 7 Cytokine profile of treatment with 1615EpCAM TriKE and EpCAM16 BiKE. Freshly isolated peripheral blood mononuclear cells were NK cell enriched from normal donor (n = 6) using magnetic beads. NK cells were cocultured with highly EpCAM positive HT-29 colon carcinoma cells for 24 hours. Supernatant was collected after incubation and assessed for GM-CSF, IL-6, IL-8, and TNF-α using a Luminex based multiplex assay. (a) GM-CSF secretion was significantly higher in the 1615EpCAM TriKE group compared to the EpCAM16 BiKE and to effectors (E) plus targets (T) alone. IL-6 (b) and IL-8 (c) showed no significant differences between BiKE and TriKE. TNF-α levels were significantly higher compared to E plus T and E alone when the BiKE or the TriKE was added. However, no difference between both drugs was obvious. P values were estimated with one-way analysis of variance and presented with SD. Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions

Figure 8 Proliferation capabilities of 1615EpCAM TriKE. (a) Peripheral blood mononuclear cells (PBMCs) were treated with 1615EpCAM TriKE and EpCAM16 BiKE. The discrete peaks in the histograms mark successive generations of NK cells after cell division leading in a repetitive slight reduction of florescence intensity. Whereas NK cells show a typical proliferation pattern, T-cells do not. Representative of five experiments. (b) PBMCs cells were cocultured with the TriKE and the BiKE and NK cell proliferation was evaluated. Representative of five experiments. (c) PBMCs cells were cocultured with the TriKE, the BiKE, anti-CD16scFv (CD16), Interleukin (IL)-15, anti-EpCAM scFv (EpCAM) and DT2219, a targeted toxin comprised of Diphtheria enterotoxin linked to anti-CD22 and anti-CD19 scFv. Evaluation of the NK cell Expansion Index showed a significantly (P < 0.001) enhanced index in the IL-15 comprising drug as well as with IL-15 alone, marked with *, (n = 5). (d) Purified NK cells were exposed to the TriKE and the BiKE. After 7 days, a reactive dye was used to differentiate alive and dead cells. This reactive dye is able to permeate the impaired membranes of dead NK cells, resulting in more intense staining (right peak) whereas no penetration in alive cells led to a weaker staining (left peak). Molecular Therapy 2016 24, 1312-1322DOI: (10.1038/mt.2016.88) Copyright © 2016 American Society of Gene & Cell Therapy Terms and Conditions