Engineering the Human Fc Region Enables Direct Cell Killing by Cancer Glycan–Targeting Antibodies without the Need for Immune Effector Cells or Complement

Fc engineering enhances avidity and direct cell killing of cancer-targeting anti-glycan antibodies to create superior clinical candidates for cancer immunotherapy. Murine IgG3 glycan-targeting mAb often induces direct cell killing in the absence of immune effector cells or complement via a proinflammatory mechanism resembling oncotic necrosis. This cancer cell killing is due to noncovalent association between Fc regions of neighboring antibodies, resulting in enhanced avidity. Human isotypes do not contain the residues underlying this cooperative binding mode; consequently, the direct cell killing of mouse IgG3 mAb is lost upon chimerization or humanization. Using the Lewisa/c/x -targeting 88mAb, we identified the murine IgG3 residues underlying the direct cell killing and increased avidity via a series of constant region shuffling and subdomain swapping approaches to create improved (“i”) chimeric mAb with enhanced tumor killing in vitro and in vivo. Constant region shuffling identified a major CH3 and a minor CH2 contribution, which was further mapped to discontinuous regions among residues 286–306 and 339–378 that, when introduced in 88hIgG1, recapitulated the direct cell killing and avidity of 88mIgG3. Of greater interest was the creation of a sialyl-di-Lewisa–targeting i129G1 mAb via introduction of these selected residues into 129hIgG1, converting it into a direct cell killing mAb with enhanced avidity and significant in vivo tumor control. The human iG1 mAb, termed Avidimabs, retained effector functions, paving the way for the proinflammatory direct cell killing to promote antibody-dependent cellular cytotoxicity and complement-dependent cytotoxicity through relief of immunosuppression. Ultimately, Fc engineering of human glycan-targeting IgG1 mAb confers proinflammatory direct cell killing and enhanced avidity, an approach that could be used to improve the avidity of other mAb with therapeutic potential. Significance: Fc engineering enhances avidity and direct cell killing of cancer-targeting anti-glycan antibodies to create superior clinical candidates for cancer immunotherapy.


Introduction 1
The cancer glycome is a rich source of targets for monoclonal antibody (mAb) development due to 2 the alterations associated with the transformation process, as well as glycans being key co-accessory 3 molecules for cancer cell survival, proliferation, dissemination and immune evasion (1,2). A number of 4 anti-glycan mAbs are in clinical development, as passive or active immunotherapy or reformatted for 5 chimeric antigen receptor (CAR) T cells (3)(4)(5). Additionally, Dinutuximab beta, an anti-GD2 mAb, is 6 currently used for the treatment of neuroblastoma (6).

7
We previously described a panel of cancer glycan targeting mAbs with Lewis a/c/x , Lewis y (7,8) as well 8 as sialyl-di-Lewis a reactivity (9). Intriguingly, some of these glycan-binding mAbs exhibited a direct 9 cytotoxic effect on high-density target expressing cancer cells, independent of the presence of 10 complement or immune effector cells. This direct cytotoxic ability has also been observed for other 11 anti-glycan mAbs and typically involves mAb-induced homotypic cellular adhesion, cytoskeletal 12 rearrangement followed by cell swelling, membrane lesions and eventual cellular demise (7,10-13). In 13 most cases the cell death is a form of non-classical apoptosis, potentially involving the generation of 14 reactive oxygen species (ROS), and most closely resembling oncotic necrosis (14,15). Importantly, 15 akin to immunogenic or inflammatory cell death (ICD), the coinciding release of in inflammatory 16 mediators -damage associated molecular patterns (DAMPs) -has the potential to recruit innate 17 immune cells to the tumor site that may further increase mAb-mediated effector functions (16). Thus, 18 these anti-glycan mAbs can be important tools to remobilise the full potential of the immune system in 19 an otherwise immunosuppressive environment.

20
The direct killing ability of anti-glycan mAbs is mediated by murine (m) IgG3, an isotype that exhibits 21 non-covalent interactions between adjacent Fc regions, thereby increasing avidity, via prolonging 22 target occupancy; a process termed "intermolecular co-operativity" (17,18). In humans, the IgG2 23 isotype can increase avidity via dimerization involving one or more Cys residues in its hinge region 24 (19). However, this inefficient process, combined with poor ADCC and CDC activity render the hIgG2 25 an unattractive clinical candidate.

26
Our panel of mAbs induce strong in vitro and in vivo tumor killing in preclinical mouse models (7,8) 27 and thus are candidates for clinical development. Chimerization of the mIgG3 mAbs onto a human 28 IgG1 backbone coincided with a dramatic reduction in direct cytotoxicity, leading us to hypothesize 29 that this was the result of diminished intermolecular cooperativity. Consequently, the rationale for this 30 study was to identify the key residues within mIgG3 that are responsible for non-covalent Fc 31 interactions and transfer them into hIgG1 in order to recapitulate the mIgG3-observed direct 32 cytotoxicity and avidity, thereby creating a chimeric hIgG1 with superior clinical utility.

33
We report here the identification of discontinuous regions within the mIgG3 CH2 and CH3 domains 34 that endow this isotype with direct cytotoxicity and increased avidity. Transfer of these residues into 35 the hIgG1 isotype, creates an improved 'i'hIgG1 with increased in vitro and in vivo anti-tumor activity. Colorectal cancer cell lines (COLO205 and HCT15) as well as the murine myeloma NS0 cell line were 4 purchased from ATCC (Virginia, USA). All cell lines were authenticated using short tandem repeat 5 profiling and tested monthly for the presence of Mycoplasma. Human serum albumin (HSA)-APD-6 sialyl-Lewis a and HSA-APD-Lewis a were from IsoSepAB (Sweden). Cell lines were maintained in 7 RPMI medium 1640 (Sigma) supplemented with 10% fetal calf serum, L-glutamine (2mM) and sodium 8 bicarbonate-buffered. Parental murine FG88.2 and FG129 mAbs were generated, as previously 9 described (7);(9)). Millipore) and sodium azide added to a final concentration of 0.2% (w/v). mAb was purified on protein 30 G columns (HiTrap ProteinG HP, GE Healthcare) using an AKTA FPLC (GE Healthcare). Columns 31 were washed with PBS/Tris buffer (PBS with 50mM Tris/HCl, pH7.0) before mAb elution with a rapid 32 gradient into 100mM glycine, pH12 (supplemented with 0.05% v/v Tween 20), collecting 2ml fractions.

33
Fractions containing mAb were pooled, neutralized to pH 7.0 (using 1M HCl) and dialyzed against 34 PBS, before concentration determination and storage at -80°C. All transiently expressed mAb 35 constructs were analyzed for cell binding using flow cytometry, as a read-out for correct folding, and 36 compared to the parental 88mIgG3 and 88hIgG1, prior to use in functional assays.

37
Indirect immunofluorescence and flow cytometry 38 Cancer cells (1x10 5 ) were incubated with primary mAbs (at 33.3nmol/L or titrated) for 1h at 4°C, as 39 previously described (7) followed by 1h incubation at 4°C with anti-mouse or anti-human FITC-labelled secondary antibody, and fixing in 0.4% formaldehyde. Stained samples were analyzed on a 1 MACSQuant 10 flow cytometer and analyzed using FlowJo v10.

3
The kinetic parameters of the 88 and 129 mAbs binding to Lewis a -or sialyl-Lewis a -APD-HSA were 4 determined by Surface Plasmon Resonance (SPR, Biacore 3000, GE Healthcare). Increasing 5 concentrations (0.3nmol/L-200nmol/L) of mAb were injected across a CM5 chip and data were fitted 6 to a heterogeneous ligand binding model using BIAevaluation 4.1. The chip contained four cells, two 7 of which, HSA-coated (in-line reference cells), the other two were coated with low (30-80 response 8 units (RU)) and high amounts (360-390 RU) of the respective glycan-APD-HSA.

9
In vitro cytotoxicity 10 Propidium Iodide (PI) uptake and proliferation inhibition were performed to analyze the direct cytotoxic 11 effect of the mAbs. COLO205 or HCT15 cells (5 x 10 4 ) were incubated with mAbs for 2h at 37°C 12 followed by the addition of 1µg of PI for 30min. Cells were resuspended in PBS and run on a 13 Beckman Coulter FC-500 or on a MACSQuant 10 flow cytometer and analyzed with WinMDI 2.9 or 14 FlowJo v10 software, respectively. Proliferation inhibition was assessed by using the water-soluble 15 tetrazolium salt WST-8 (CCK8 kit, Sigma-Aldrich) to measure the activity of cellular hydrogenases 16 which is directly proportional to the number of viable cells. Briefly, after overnight plating of cancer 17 cells (1000 cells/90µl/well), constructs were added at different concentrations in a final volume of 18 10µl/well and the plates were incubated at 37°C, (5%CO2) for 72-96h. WST-8 reagent was then 19 added (10µl/well) and after a further 3h incubation, the plates were read at 450nm (Tecan Infinite F50) 20 and percentage inhibition calculated. EC 50 values were determined using nonlinear regression (curve 21 fit) with GraphPad Prism v 8.0 (GraphPad Inc, La Jolla, CA).

22
Immune effector function determination 23 ADCC and CDC were performed as described previously (7). 51 Cr-labeled target cells (5 × 10 3 ) were 24 co-incubated with 100μL of peripheral blood mononuclear cells (PBMC) from healthy donors (ADCC) 25 or 10% (v/v) autologous serum (CDC) and with mAbs at a range of concentrations; the effector to 26 target ratio was 100:1 (E:T)). Spontaneous and maximum release [counts per minute (cpm)] were 27 evaluated by incubating the labelled cells with medium or with 10% (v/v) Triton X-100, respectively.

30
Scanning electron microscopy 31 HCT15 or COLO205 cells (1 x 10 5 ) were grown on sterile coverslips for 24h prior to mAb (0.2µmol/L) 32 addition for 18h at 37°C. Controls included medium alone and 0.5% (v/v) hydrogen peroxide (H 2 O 2 ) 33 (Sigma). Cells were washed with pre-warmed 0.1 M sodium cacodylate buffer pH7.4 (SDB) and fixed 34 with 12.5% (v/v) glutaraldehyde for 24h. Fixed cells were washed twice with SDB and post-fixed with 35 1% (v/v) osmium tetroxide (pH 7.4) for 45min. After a final wash with H 2 O, the cells were dehydrated 36 in increasing concentrations of ethanol and exposed to critical point drying, before sputtering with 37 gold, prior to SEM analysis (JSM-840 SEM, JEOL).

Recombinant human FcRn binding analysis
The ability of the mAbs to bind to recombinant human (rh) FcRn (R&D Systems) was evaluated using 1 direct ELISA at pH6.0 and pH7.0. Briefly, high-binding ELISA plates were coated with 250ng/well 2 rhFcRn followed by blocking with protein-free blocking buffer (Thermo Fisher Scientific). Primary mAb 3 dilutions (in phosphate buffer pH6.0 or pH7.0) were added (1h at room temperature), followed by 4 washing with respective phosphate buffers containing 0.05% (v/v) Tween 20, and detection of bound 5 mAbs with goat F(ab) 2 anti-human IgG(Fab) 2 HRP antibody (Abcam). The anti-hCTLA4 hIgG1 mAb 6 Ipilimumab (clinical grade) was included as a positive control.

7
Biophysical characterization of the mAbs (size exclusion chromatography with multi-angle 8 light scattering (SEC-MALS) and analytical ultracentrifugation (AUC)) 9 SEC-MALS experiments were performed using a Superose 6 10/300 Increase column (GE 10 Healthcare) on an AktaPure 25 System (GE Healthcare). mAb samples (100μL at 1mg/mL), were 11 loaded and eluted with one column volume (24mL) of buffer, at a flow rate of 0.5mL/min. The eluting 12 protein was monitored using a DAWN HELEOS-II 18-angle light scattering detector (Wyatt 13 Technologies) equipped with a WyattQELS dynamic light scattering module, a U9-M UV/Vis detector 14 (GE Healthcare), and an Optilab T-rEX refractive index monitor (Wyatt Technologies). Data were 15 analyzed by using Astra (Wyatt Technologies) using a refractive index increment value of 0.185mL/g.

16
For AUC characterization, sedimentation velocity scans were recorded for each mAb sample at 17 concentrations of 5.0, 2.5 and 0.5μmol/L. All experiments were performed at 50,000 rpm, using a 18 Beckman Optima analytical ultracentrifuge with an An-50Ti rotor at 20˚C. Data were recorded using 19 the absorbance optical detection system at 280nm. The density and viscosity of the buffer was 20 measured experimentally using a DMA 5000M densitometer equipped with a Lovis 200ME viscometer 21 module. The partial specific volume of the antibodies was calculated using SEDFIT from the amino    were randomly allocated to treatment groups based on their mean tumor volume (~103mm 3 ± 13mm 3 ) 1 on study day 6 and dosed intravenously (i.v.), biweekly, with mAbs (0.1mg) or vehicle (PBS, 100µl) up 2 until week 5. Body weight and tumor volume were assessed three times weekly and reduction in 3 tumor volume analyzed statistically using two-way ANOVA with Bonferroni's post-test at day 35, when 4 all control animals were still in the study (GraphPad Prism v 7.4, GraphPad Inc, La Jolla, CA).

6
The error bars shown in the figures represent the mean ± SD. Titration curves for functional assays 7 (direct cell killing, immune effector functions) were analyzed with two-way ANOVA with the construct 8 factor P values graphed. Functional affinity results as well as fixed-concentration functional assays 9 were analyzed with one-way ANOVA with Dunnett's corrections for multiple comparisons. All 10 analyses were performed with GraphPad Prism v 7.4 (GraphPad Inc, La Jolla, CA), with * P ≤ 0.05, ** 11 P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001.

13
Results 1 m88G3 exhibits avid glycan binding as well as direct cytotoxicity in the absence of 2 complement and immune effector cells, both of which are reduced upon chimerization to 3 88hIgG1 4 We have previously shown that the hybridoma-produced mIgG3 mAb FG88.2 exerts a direct 5 cytotoxic effect on high-binding cancer cell lines, such as COLO205 and HCT15, in the absence of 6 complement or effector cells (7). This direct cytotoxicity involved mAb-induced cellular aggregation, 7 proliferation inhibition as well as irregular pore formation through an oncolytic mechanism. We 8 subsequently created a chimeric, HEK293-expresssed, hIgG1 mAb, 88hIgG1, for clinical exploitation. 9 88hIgG1 maintained equivalent HCT15 and COLO205 cancer cell binding levels (Fig. 1A), compared 10 to the hybridoma-produced FG88.2, as well as the HEK293-expressed 88mIgG3. The latter mAb was 11 generated to rule out expression system related effects such as differential Fc glycosylation, due to 12 the use of murine hybridoma cells versus HEK293 cells. Surprisingly, 88hIgG1, exhibited significantly 13 reduced direct cytotoxicity on COLO205 and HCT15, across two functional assays, PI uptake and 14 proliferation inhibition, compared to 88mIgG3 (Fig. 1, panels B-D). 88mIgG3 also displayed a modest 15 reduction in direct cytotoxicity compared to the hybridoma-produced FG88.2, suggesting that 16 differential glycosylation of the Fc region by the two expression settings (mouse hybridoma versus 17 HEK293 cells) contributed to the effect. Combined, the results indicated that the direct cell killing 18 could be related to the kinetic binding behaviour of the different isotypes. Consequently, the kinetic 19 binding of our isotype-switched mAbs was analyzed on a Lewis a -APD-HSA coated chip using SPR ( 20 Supplementary Table 1). FG88.2 displayed avid Lewis a -APD-HSA binding with fast apparent on-rates 21 (k on ~ 10 4 1/smol/L) and very slow off rates (k off ~ 10 -6 1/s) on the high-density flow cell. The HEK293-22 produced 88mIgG3 exhibited an apparent faster on-rate (k on ~ x 10 5 1/smol/L) and a somewhat faster 23 off-rate (k off ~1 0 -4 1/s) compared to FG88.2, that could explain the slightly reduced cytotoxicity 24 compared to FG88.2. In comparison, 88hIgG1 bound its target with an apparent fast on-rate (k on ~ 25 10 5 1/smol/L), but in contrast to the mIgG3 isotypes displayed a much faster dissociation phase 26 (apparent k off ~ 10 -2 1/s), that is likely to underly its reduced cytotoxic activity upon cancer cell binding.

27
The mAb binding behaviour on the low-density flow cell was largely comparable between the three 28 mAbs, with equilibrium dissociation constants (Kd) of the order of 10 -8 mol/L for all three isotypes.

30
Domain analysis of the mIgG3 constant region indicate a major contribution by the mIgG3 CH3 31 domain with a minor involvement of the CH2 32 Collectively, the results outlined above suggested that the high Lewis a -APD-HSA avidity exhibited 33 by FG88.2 and 88mIgG3, predominantly driven by their slow target dissociation and potentially 34 resulting from the intermolecular cooperativity of the mIgG3 isotype, contributed to their direct 35 cytotoxic effect. We thus set out to engineer a hIgG1 cancer glycan targeting mAb with direct 36 cytotoxic activity, via the transfer of selected mIgG3 constant region residues into 88hIgG1. Firstly, 37 mIgG3 contributing regions were identified through the creation of hybrid 88hIgG1 constructs, 38 containing mIgG3 CH1, CH2 or CH3 domains. Preliminary analyses ascertained that mIgG3 CH1 39 had a negligible contribution to the direct cytotoxicity ability of 88mIgG3, as introducing mIgG3 CH1 into 88hIgG1 (1m1) did not lead to a significant increase in cytotoxicity (Fig. 2, Panels A and B).

1
Conversely, introducing hIgG1 CH1 into 88mIgG3 (3h1), equally, did not instigate a significant 2 reduction in killing activity (Fig. 2, Panels A and B). Next, in a gain-of-function approach, the mIgG3 3 CH2 and CH3 domains, separately, were introduced in 88hIgG1. 88hIgG1 containing murine CH3 4 (1m3) exhibited a significant gain in PI uptake on HCT15, as well a significant increased proliferation 5 inhibition of COLO205 cells, when compared to 88hIgG1 (Fig. 2, Panels C and D). Introducing 6 murine CH2 into 88hIgG1 (1m2) led to small, but not significant, increase in killing activity across both 7 assays (Fig. 2, Panels C and D). As a confirmation of the contributions made by both domains, the 8 reverse strategy was adopted, whereby a loss of cytotoxicity activity was evaluated due to the 9 introduction hIgG1 CH2 or CH3 domains into 88mIgG3. This scenario led to a significant decrease in 10 cytotoxicity for 88mG3 containing hIgG1 CH3 (3h3), corroborating the previous gain-of-function 11 results. Importantly, this strategy also identified a small contribution by the murine CH2, as 88mIgG3 12 containing human CH2 (3h2) exhibited a significant decrease in cytotoxicity activity (Fig. 2, Panels C 13 and D). Next, the kinetic binding behaviour of the hybrid constructs was analyzed. The hybrid 14 construct 1m3 exhibited a modest, but significant increase in avidity (decreased Kd), whilst 3h3, 15 containing human CH3, displayed a significant decrease in avidity (increased Kd, Fig. 2, Panel E), in 16 both cases, mirroring the direct cytotoxicity. Human CH2 in construct 3h2 also led to a modest, but 17 significant drop in avidity. In all cases, the changes in avidity were predominantly driven by changes in 18 the off-rate of the mAbs, with 1m3 showing a significantly decreased off-rate compared to 88hIgG1,

19
whereas 3h3, as well as 3h2, exhibited a significantly increased off-rate compared to 88mIgG3 ( across both assays (Fig. 3, panels A-D), obviating the need for adding additional subdomains. Avidity analysis of the subdomain constructs, compared to 88hIgG1, revealed a striking improvement in 1 avidity for SD286-397, as well as SD339-397, both now matching the 88mIgG3 avidity, with a more 2 modest improvement for SD286-345 (Fig. 3, Panel E). The improved avidity resulted mainly from a 3 dramatically reduced apparent off-rate (~10 -6 1/s) for SD286-397 as well as SD339-397, with the 4 SD286-345 off-rate showing a more modest improvement (~ 10 -3 1/s) (Fig. 3, Panel F). These results 5 add further weight to the cytotoxicity observations and support the notion that creating a mAb with a 6 reduced target dissociation rate upholds direct cytotoxicity.

7
Although SD339-397, with 27 mIgG3 residues, recapitulated up to 90% of the desirable attributes of 8 88mIgG3, notably the slow dissociation and enhanced cytotoxicity, it exhibited a significantly reduced 9 CDC activity compared to 88hIgG1 (Fig. 3, Panel G), but it maintained ADCC activity compared to

33
Reversal of one in silico identified MHCII binding cluster generates the lead candidate, 1 improved 'i' 88G1, with robust cell killing, increased avidity, pore-forming ability and sound 2 immune effector functions

3
We performed an in silico screen of the SD286-306+339-378 sequence, containing 26 mIgG3 4 residues, for MHCII binding epitopes (Immune Epitope Database, IEDB), in order to assess potential 5 immunogenicity. Class II-restricted T helper cells are relevant to the humoral immune response and 6 predicted binding clusters have been shown to be strong indicators of T cell responses (21). Two 7 potential MHCII binding clusters, were identified: cluster 1 (residues 294-315) which would be 8 potentially immunogenic in a wide range of HLA types and cluster 2 (residues 365-393) which would 9 potentially only be weakly immunogenic in HLA-DR*0401 and HLA-DR*01101( Supplementary Fig.1).

10
Reversion of three murine residues, 294 (A to E), 300 (F to Y) and 305 (A to V), within cluster 1, to 11 human residues, produced a human sequence section to which individuals would have been 12 tolerized. Similarly, reversal to human sequence of two residues 351 (I to L) and 371 (N to G), within 13 cluster 2, removed two potential MHCII binding epitopes. Consequently, we created two additional 14 SD286-306+339-378 -based constructs: DI1 and DI2, containing three and two human reverted 15 residues, respectively, and assayed their cytotoxicity and avidity. DI1 maintained significantly 16 improved cytotoxicity compared to 88hIgG1. Additionally, the direct cytotoxicity coincided with a 17 favourable avidity profile, with an apparent off-rate of (~ 10 -4 1/s) and a Kd of 0.5 nmol/L that was 18 similar to 88mIgG3 (Fig. 5, Panel C, Table 1 and Supplementary Fig. 2). In contrast, DI2 showed a 19 small, but consistently decreased activity compared to 88mIgG3 (Fig. 5, Panels A and B) as well as a 20 significantly decreased avidity compared to 88mIgG3 (Fig. 5, Panel C). As this cluster was only 21 potentially weakly immunogenic in two HLA-DR types, these two residues have not been reverted.

28
Earlier work on the parental hybridoma-produced FG88.2 had demonstrated its pore-forming ability, 29 which was surmised to underlie its cytotoxicity (7). We thus set out to analyze the pore-forming ability 30 of i88G1 on HCT15, using SEM. Incubation of HCT15 with i88G1 or 88mIgG3, but not 88hIgG1, 31 resulted in monolayer disruption, cell rounding and clustering. At higher magnification, irregular pore   We recently described the generation of a sialyl-di-Lewis a recognizing mAb (129 mAb) with 1 development potential for cancer immunotherapy (9). The 129 mAb has a more favorable tumor 2 versus normal human tissue distribution compared to the above-described 88 mAb, resulting from 3 wide-ranging tumor tissue binding, combined with very restricted normal tissue reactivity. Neither the 4 hybridoma-produced FG129, a murine IgG1 mAb, nor the chimeric 129hIgG1, exhibited direct 5 cytotoxicity. This led us to test the hypothesis that the introduction of the 23 above-selected mIgG3 6 constant region residues into the Fc region of 129hIgG1 would create an 'i'129G1 with direct 7 cytotoxicity and improved avidity and thus exhibit superior clinical utility.

8
We evaluated the direct cytotoxicity of i129G1 on COLO205, previously shown to be a high-binding 9 cancer cell line for FG129 (9). The i129G1 displayed significantly improved (compared to 129hIgG1), 10 dose-dependent inhibition of proliferation (Fig. 6, Panel A and Table 1), with an EC 50 of 45.6 nmol/L, 11 as well as a significantly improved, but more modest, PI uptake (Fig. 6, Panel B). In comparison, 12 negligible direct cytotoxicity was observed on the low to moderate binding ASPC1 or BXPC3 13 ( Supplementary Fig. 3). Next, we analyzed the avidity of i129G1 using a sialyl Lewis a -APD-HSA-14 coated chip and SPR. The i129G1 mAb exhibited significantly improved avidity compared to 15 129hIgG1 (Fig. 6, panel C, Table 1 and Supplementary Fig. 2), resulting predominantly from an 16 improvement in off-rate by almost two logs (2.6 x 10 -4 s -1 and 5.5 x 10 -6 s -1 for 129hIgG1 and i129G1, 17 respectively). On COLO205, i129G1 maintained ADCC activity in the nanomolar range (EC 50 2.4 18 nmol/L), compared to 1.7 nmol/L for 129hIgG1, but the overall percentage lysis was significantly 19 reduced (Fig. 6, panel D, Table 1). The CDC activity of i129G1, however, was significantly increased 20 compared to the parental 129hIgG1, with EC 50 of 8.2 nmol/L and 75 nmol/L, respectively (Fig. 6,

21
Panel E, Table 1). The direct cytotoxicity as well as improved avidity of i129G1 led us to analyze its 22 pore-forming ability on COLO205. The incubation of COLO205 with i129G1, caused the formation of 23 large cell clumps with uneven surfaces, as well as the appearance of irregular pore-like structures 24 (Fig. 6, Panel F). Incubation with 129hIgG1, at the same concentration, also led to a degree of cell 25 clumping, but smaller and fewer clumps were observed, without evidence of pore formation.

26
The direct cytotoxicity and improved avidity of i129G1 directed us towards analyzing the in vivo anti-27 tumor activity of i129G1 in comparison with the parental 129hIgG1 in a COLO205 xenograft model.

28
The i129G1 mAb instigated a significant reduction in tumor volume compared to vehicle control (two-29 way ANOVA, P <0.0001, Fig.6, Panel G and Supplementary Fig. 4) which remained significant when 30 compared to 129hIgG1, thereby corroborating the in vitro results. No adverse effects on mean body 31 weight were observed (Fig. 6, Panel H).

32
In order to ascertain that our Fc-engineering approach had not impacted on the biopharmaceutical 33 development potential of the i129G1 mAb, we evaluated its in vitro FcRn binding ability, as well as its 34 solution aggregation status using a range of biophysical and biochemical approaches. In vitro binding 35 of i129G1 to rhFcRn at pH6.0 and pH7.0 was compared to the parental 129hIgG1 as well as clinically 36 validated Ipilimumab, also a hIgG1. At pH6.0, 129hIgG1 as well as i129G1 display improved binding 37 compared to Ipilimumab. Furthermore, i129G1 exhibited significantly improved rhFcRn binding 38 compared to 129hIgG1 at the highest concentrations tested (Fig. 6, Panel I). Some rhFcRn binding 39 was observed at pH7.0, mainly at the higher concentrations, with the parental 129hIgG1 displaying higher reactivity compared to i129G1. Next, we evaluated the solution-phase characteristics of 1 i129G1 compared to the parental 129hIgG1 using SEC-MALS and AUC. The SEC-MALS profile of 2 129hIgG1 as well as i129G1 were similar, containing a main peak (16-18mL) consistent with an 3 antibody monomer as well as two minor peaks corresponding to higher molecular weight (MW) 4 species (15mL and 8mL (void volume), respectively) (Fig.6, Panel J, i). AUC profiles of both mAbs 5 across the three concentrations tested, revealed a slight increase in the number of higher MW 6 species for i129v1, the main mAb monomer peak being 80.6% ± 2.6% and 67.6% ± 2.3% of all 7 species detected in the sample for 129hIgG1 and i129v1, respectively (Fig.6, Panel J, ii). The latter 8 analysis prompted us to investigate whether the small increase in higher molecular weight species in 9 the i129v1 sample would lead to complement activation in normal human serum in the absence of

16 1
Whereas unmodified cancer glycan-targeting mAbs often exhibit anti-tumor activity in preclinical 2 animal models, they perform disappointingly in the clinic (3,22-24). One possible explanation is that 3 mIgG3 anti-glycan mAbs exhibited direct cytotoxic activity, which was significantly reduced when 4 chimerized or humanized to hIgG1 (10-13). Similarly, the Lewis a/c/x FG88.2 used in this study, a 5 mIgG3 isotype, exhibited high avidity as well as direct cytotoxicity upon binding to high target-6 expressing cancer cells (7), both of which were significantly reduced on chimerization to hIgG1.

7
It is perhaps not surprising that the direct cytotoxicity of cancer glycan-targeting mIgG3 mAbs was

29
In the current study we describe the creation of hIgG1 anti-glycan mAbs with increased avidity and 30 direct cytotoxic activity through the transfer of selected mIgG3 constant region residues. Candidate 31 residues were identified through screens based on increased direct cytotoxicity and avidity, when 32 introduced into hIgG1 (gain-of-function), and/or decreased direct cytotoxicity and avidity when 33 replaced by the respective hIgG1 residues in mIgG3 (loss-of-function), using the Lewis a/c/x FG88.2.

34
Differences in segmental flexibility between the two mAbs due to the changed CH1 and hinge 35 regions as well as a direct contribution by murine IgG3 CH1 were ruled out, as the introduction of 36 murine IgG3 CH1 into 88hIgG1 did not increase direct cytotoxicity. Neither did the introduction of 37 hIgG1 CH1 into 88mIgG3 decrease direct cytotoxicity. The murine IgG3 hinge region has somewhat 38 greater flexibility, compared to other murine isotypes (50), but an involvement of the hinge region, in isolation, is unlikely to be solely responsible for the observed direct cytotoxicity and improved avidity, 1 as was recently shown for an erythrocyte glycan binding mIgG3 mAb (36).

2
Focusing on the mIgG3 Fc region, a major contribution by CH3 was identified, with effects evident in 3 improved cytotoxicity as well as avidity, the latter mainly the result of a decreased dissociation rate. A 4 minor contribution by CH2 was only evident when screened via the loss-of-function approach,

21
The introduction into 88hIgG1 of a discontinuous section comprising residues 286-306 and 339-378, 22 recapitulated 88mIgG3 cytotoxicity and avidity, whilst maintaining immune effector functions (ADCC 23 and CDC). The likely explanation for the greater than anticipated number of mIgG3 residues required 24 for increased avidity through intermolecular cooperativity is the combined effect of directly interacting 25 as well as conformational residues, the latter potentially creating a permissive framework. A role for 26 charge distribution patterns, notably in CH2, can also not be ruled out, as it has been shown to 27 enhance mIgG3 binding to negatively charged multivalent antigen and is distinct from hIgG1 (36,37).

28
The introduction of 26 mIgG3 in hIgG1 may create MHCII binding epitopes that have the potential to 29 drive HAMA responses in patients. IEDB analysis of the 26 mIgG3 residue-containing hybrid 30 88hIgG1, revealed two clusters (residues 294-315 and 365-378), one containing several potentially 31 high-scoring epitopes. Residues in cluster 1, at positions 294, 300 and 305, were reverted to human 32 sequence with maintained avidity and direct cytotoxicity. On the other hand, reverting residues at 33 positions 351 and 371 (cluster 2, with weaker binding scores) led to a small but significant decreased 34 cytotoxicity, hence were maintained in the final construct. Importantly, this superior 88hIgG1 hybrid 35 mAb, with mIgG3-matching direct cytotoxicity and avidity, induced cellular aggregation, pore formation 36 and cell lysis on high-binding HCT15, suggesting a similar cell killing mechanism compared to the 37 parental FG88.2 (7). The pore formation and eventual cell lysis share similar cellular disintegration features with necroptosis, but cannot be distinguished from necrosis or secondary necrosis (52). The 1 eventual outcome from the released DAMPs -constitutive or induced as a result of activated stress 2 pathwaysduring this inflammatory cell death depends on the cellular environment as well as the 3 underlying signalling cascades, but collectively have the potential to create an inflammatory 4 environment that may further enhance immune effector functions and/or instigate an adaptive immune 5 response through cross-presentation of released tumor antigens (16,53). Advantageously, i88G1 6 maintained immune effector functions with CDC activity being significantly improved, and ADCC 7 activity being somewhat reduced, compared to 88hIgG1.
As the Fc residues involved in 8 FcgammaRIIa/RIIIa binding are predominantly located in the lower hinge and adjacent top of CH2 9 region, it is unlikely that our introduced changes have a direct effect on this interaction, but we cannot 10 rule out an indirect effect (54).

11
Further validation of our approach, came from the introduction of the selected 23 mIgG3 residues 12 into the sialyl-di-Lewis a targeting 129hIgG1, that has a more favorable normal tissue distribution whilst  tumor control that was significantly better than 129hIgG1, the latter exhibiting no significant tumor 30 reduction, further emphasizing the value of direct cytotoxic ability.

31
Additionally, it was important to ascertain that our Fc-engineering had not impacted on the solution 32 self-association of i129G1. Although the biophysical analysis suggested a small increase in the 33 proportion of higher MW species in the i129G1 sample, more apparent from AUC than SEC-MALS, 34 this did not result in a significantly increased C4d generation upon incubation with healthy human 35 donor serum. We did not observe a reduction in rhFcRn by i129G1 binding, suggesting that the 36 pharmacokinetic aspects equally had not been compromised by our Fc-engineering.

37
The creation of improved cancer glycan targeting mAbs, with enhanced avidity as well as direct 38 cytotoxicity, through establishing intermolecular cooperativity binding, may lead to superior clinical 39 utility. Additionally, it is plausible that mAb multimerization upon glycan target engagement through 1 alternative strategies may equally lead to increased avidity and ensuing direct cytotoxicity. Our 2 approach may also have value for mAbs targeting cancer-associated proteins, where longer target 3 residence time may lead to more profound biological effects, but this remains to be validated.

4
Importantly, reinstating the unusual, proinflammatory cell killing mode observed for many glycan-5 targeting mIgG3 mAbs, into the hIgG1 framework, opens the door to combination immunotherapy.