Enhanced Antitumor Activity of an Anti-5T4 Antibody Drug … · 03/11/2015 · Experimental...

Cancer Therapy: Preclinical

Enhanced Antitumor Activity of an Anti-5T4Antibody–Drug Conjugate in Combination withPI3K/mTOR inhibitors or TaxanesBoris Shor, Jennifer Kahler, Maureen Dougher, Jane Xu, Michelle Mack,Ed Rosfjord, Fang Wang, Eugene Melamud, and Puja Sapra

Abstract

Purpose: Targeted treatment of solid or liquid tumors withantibody–drug conjugates (ADCs) can lead to promising clinicalbenefit. The aim of the study is to investigate combination regi-mens of auristatin-based ADCs in preclinical models of cancer.

Experimental Design: An auristatin-based anti-5T4 antibodyconjugate (5T4-ADC) and auristatin payloads were combinedwith the dual PI3K/mTOR catalytic site inhibitor PF-05212384(PF-384) or taxanes in a panel of tumor cell lines. Drug interac-tions in vitro were evaluated using cell viability assays, apoptosisinduction, immunofluorescence, mitotic index, and immuno-blotting. Breast cancer cells treated with auristatin analogue or5T4-ADC were profiled by total- and phospho-proteomics. Anti-tumor efficacy of selected combinations was evaluated in 5T4-positive human breast or lung tumor xenografts in vivo.

Results: In vitro, auristatin-based agents displayed strong syn-ergistic or additive activity when combined with PF-384 or tax-

anes, respectively. Further, treatment of 5T4-ADC plus PF-384resulted in stronger induction of apoptosis and cell line–specificattenuation of pAKT and pGSK. Interestingly, proteomic analysisrevealed unique effects of auristatins on multiple componentsof mRNA translation. Addition of PF-384 further amplifiedeffects of 5T4-ADC on translational components, providing apotential mechanism of synergy between these drugs. In humantumor xenografts, dual targeting with 5T4-ADC/PF-384 or 5T4-ADC/paclitaxel produced substantially greater antitumor effectswith longer average survival as compared with monotherapytreatments.

Conclusions: Our results provide a biologic rationale forcombining 5T4-ADC with either PI3K/mTOR pathway inhibi-tors or taxanes and suggest that mechanisms underlying thesynergy may be attributed to cellular effects of the auristatinpayload. Clin Cancer Res; 1–12. �2015 AACR.

IntroductionAntibody–drug conjugates (ADCs) are poised to become an

important class of cancer therapeutics, as evidenced by the prom-ising response rates when administered as single agents to che-morefractory cancer patients. However, despite significant surviv-al benefit, most patients eventually relapse or show diseaseprogression after treatment with single-agent ADCs (1–4). Onestrategy to improve clinical outcomes for ADCs is through com-binationwith chemotherapy ormolecularly targeted agents.Mostof the known ADCs undergoing clinical testing contain tubulininhibitors, including auristatins and maytansinoids. Auristatinsare fully synthetic water soluble dolastatin analogues that accountfor more than 50% of ADCs in clinical development. Mechanis-tically, auristatins bind the same site on tubulin as the vinca

alkaloids, destabilize microtubules, block microtubule assembly,and arrest cells in the G2–Mphase of the cell cycle resulting in celldeath (5, 6). Currently, two auristatins, monomethyl auristatin-Eand monomethyl auristatin-F (MMAF), are being investigated inthe context of ADCs, with several other auristatin analoguesentering clinical development. We have previously described anADC that targets 5T4, an oncofetal antigen expressed on tumor-initiating cells (7). The 5T4-mcMMAF (5T4-ADC) is comprised ofa humanized anti-5T4 A1 antibody linked to the MMAF via anoncleavable maleimidocaproyl linker and thus, the ADC'smechanism of action is mechanistically linked to the biologicactivity of the auristatin moiety that is liberated upon internal-ization of the ADC into the cancer cell.

The PI3K/mTOR pathway plays an essential role in onco-genesis, and is implicated in the emergence of resistance toseveral therapeutic drugs in clinical settings, including erloti-nib, gefitinib, or trastuzumab in patients with HER2-positivesolid cancers (8–11). Accordingly, agents that normalize PI3K/mTOR pathway activity, including direct inhibitors of PI3K(GDC-0941), AKT (MK-2206), or mTOR (everolimus), haveshown promising activity in trastuzumab-resistant patientswhen combined with trastuzumab both in preclinical and clini-cal investigations (12–15). PF-05212384 (PF-384) is a pan-class I isoform PI3K and mTORC1/2 inhibitor, which is in theearly stages of clinical development for solid tumor indica-tions (16, 17). Unlike PI3K pathway inhibitors, taxanes arecytotoxic agents that are used as a first- or second-line therapy

Oncology Research Unit, Pfizer Worldwide Research and Develop-ment, Pearl River, New York.

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

CorrespondingAuthor: Puja Sapra, Bioconjugates Discovery andDevelopment,Oncology Research Unit, Pfizer Worldwide Research and Development, 401North Middletown Road, Pearl River, NY 10965. Phone: 845-602-3389; Fax:845-602-5557; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-15-1166

�2015 American Association for Cancer Research.

ClinicalCancerResearch

www.aacrjournals.org OF1

Cancer Research. on December 25, 2020. © 2015 American Association forclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst August 28, 2015; DOI: 10.1158/1078-0432.CCR-15-1166

http://clincancerres.aacrjournals.org/

for many malignancies, including breast, ovarian, or lungcancers (18, 19). Paclitaxel and docetaxel are the two mostcommon taxanes, which arrest mitotic division by binding tob-tubulin and stabilizing preexisting microtubules (20, 21).With evidence of single-agent activity in the clinical setting,taxanes have been evaluated extensively, but mostly empiricallyin combination with other chemotherapy agents, includingplatinum compounds, anthracyclines, or radiation (22).

This report identifies a novel therapeutic strategy to increaseefficacy of 5T4-ADCby combining itwith PF-384or paclitaxel andsuggests that a similar approach can be applied to the broad classof auristatin-based drugs.We aimed to evaluate cellular responsesof a subset of 5T4-positive tumor cells to cotreatment with 5T4-ADC/PF-384 or 5T4-ADC/paclitaxel. By applying global proteo-mics approaches, we identify auristatin-specific pathway pertur-bations that could give insight into the mechanisms of synergy.Our results reveal modest but selective changes in the cellularmRNA translation machinery with 5T4-ADC or auristatin, aneffect that is greatly amplified by the addition of PF-384. Whentested in xenograft models of breast and lung cancer, cotargetingwith 5T4-ADC/PF-384 or 5T4-ADC/paclitaxel combinationsresults in enhanced antitumor activity and thus may have apotential to generate more durable responses in 5T4-positivepatients in the clinic.

Materials and MethodsCell lines and reagents

Human tumor cell lines NCI-H1975, Calu-6, NCI-H358,HCC2429, MDA-MB-468, MDA-MB-231, CAOV-3, TOV-112D,OV-90, OVCAR-3, SKOV-3, HT-29, NCI-N87, Raji, and Ramoswere purchased from the ATCC. MDAMB361-DYT2 cells wereobtained fromDr. D. Yang (GeorgetownUniversity, Washington,DC). Cell lines were authenticated annually by short-tandemrepeat analysis (Promega STR profiling service) and routinelytested for mycoplasma contamination (ATCC). MDAMB435/5T4 are cells stably transfected with human 5T4 that weredescribed previously (23). The 37622A1 NSCLC patient-derived

xenograft, and the establishment and characterization of pri-mary serum-free culture TUM622 from 37622A1, were described(24). Each cell line was cultured in its standard medium asrecommended by the ATCC. For in vitro studies, chemotherapeu-tic drugs were obtained from Sigma-Aldrich. PF-05212384(PKI-587), MMAF-Ome, and auristatin 101 were obtained fromPfizer WWMC. Preparation of 5T4-ADC (A1mcMMAF) was de-scribed previously.

Synergy assaysThe effects of drug combinations were evaluated using Chou–

Talalay median effect analysis (25). Cells were treated with eachdrug alone and in combination in two independent 96-well platesin a diagonal matrix format, and proliferation was measured byusing a CellTiter Glo kit (Promega). Results were expressed assurviving fractions (fraction affected), based on the measuredluminescence counts of treated samples, compared with that ofuntreated controls. Seven diagonals representing various dose–effect curves with fixed drug ratios were used to measure thecombination indexes (CI) for each of the combinations withCalcusyn software (Biosoft). In each experiment, CI indexes atED50 levels were averaged for the three dose–effect curves thathad 7 to 8 data points. The CIs from two to three independentexperiments were averaged to generate a single number shownin Figs. 1A and 4A. The Chou–Talalay method was used tocalculate CI, with the CI values of <0.9 considered as evidenceof synergy; 0.9–1.1, additive effects; CI > 1.1, antagonism (25).

Western blottingEqual amounts of proteins were subjected to immunoblotting

analysis using NuPAGE electrophoresis system (Life Technolo-gies). The primary antibodies for P-AKT (S473, S308), AKT, P-GSK(S21/9), GSK, P-eIF4G1 (S1108), eIF4G1, eIF4G2, P-eIF4B(S422), eIF4B, P-eIF2a (S51), eIF2a, eIF3A, eP-eEF2 (T56), eEF2,4E-BP1 (T37/46), 4E-BP1, P-H3, cPARP, GAPDH, P-Aurora A(T288)/B (T232)/C (T198), total Aurora B, and secondary anti-bodieswere obtained fromCell Signaling Technology. Antibodiesfor total 5T4 were from Abcam (EPR5530/ab129058) ordescribed previously (24).

Xenograft efficacy studiesFemale athymic nu/nu mice (18–23 g) were obtained from

Charles River Laboratories. Mice were injected with tumor cellssubcutaneously and animals with staged tumors were administeredintravenously with saline (vehicle), 5T4-ADC, PF-384, paclitaxel,or combinations 5T4-ADC plus PF-384, 5T4-ADC plus paclitaxel.ADCs were administered based on mAb protein content at 2 or3 mg Ab/kg on a q4d (every 4 day) schedule, with 8 to 10 miceper group. 5T4-ADC, PF-384, and paclitaxel were administeredat clinically equivalent doses. All procedures using mice wereapproved by the Pfizer Institutional Animal Care and Use Com-mittee according to established guidelines. Time-to-endpoint(TTE; time to no tumor or rate of tumor tripling) were used forcomparison of drug-treated groups. The t test was used to compareindividual tumor volumes as indicated in the figure legends.

Additional Materials and Methods are available in the Sup-plementary Information.

Results5T4-ADC or auristatins synergize with PI3K/mTOR inhibitor

in vitro

Translational Relevance

Antibody–drug conjugates (ADCs) are emerging as a prom-ising therapeutic modality for the treatment of cancers. How-ever, despite encouraging activity, patients treated with single-agent ADCs show relapse or progression due to the emergenceof resistance. Combination of ADCs with other chemothera-peutic or targeted agents is being evaluated in the clinic toimprove the overall response. In this study, in preclinicalmodels, we systematically evaluated combination partner(s)for an auristatin-based antibody–conjugate (5T4-ADC) anddemonstrate enhanced activity when 5T4-ADCwas combinedwith a PI3K/mTOR inhibitor or with taxanes. Our findingsprovide a rationale to evaluate 5T4-ADC in combination withPI3K/mTOR pathway inhibitors or taxanes in the clinic. Addi-tionally, our data suggest that the mechanistic basis of theobserved synergy may be attributed to the mechanism ofaction of payload (i.e., auristatin). As several auristatin-basedADCs are in clinical trials, we believe that our findings can beapplicable to other ADCs using auristatins.

Shor et al.

Clin Cancer Res; 2015 Clinical Cancer ResearchOF2




We performed a focused combinatorial screen of selectedstandard of care agents or signaling inhibitors combined withauristatin-based drugs in a panel of breast, lung, or ovarian cancercells (unpublished data and Supplementary Table S1). Cell lineswith previously characterized and clinically relevant 5T4 expres-sion levels were also included in the analysis to enable direct

comparison of the drug interactions with 5T4-ADC conjugate tothe unconjugated auristatins (Supplementary Fig. S1, ref. 7).Through this screen, we identified the dual PI3K/mTOR inhibitorPF-384 as showing most consistent interactions with microtu-bule-destabilizing agents or the 5T4-ADC across the panel of 16cell lines (Fig. 1A). Overall, a similar pattern of synergistic or

β

A B

PF384 (50 nmol/L)Control

5T4ADC+PF3845T4ADC (1 µg/mL)

5T4ADC+PF-384

MMAF-OMe+PF-384

Aur101+PF-384

PTX+PF-384

0.0

0.5

1.0

1.5

Combinations with PF-384

CI i

ndex

H1975Calu6H358HCC2429MDA468MDA361-DYT2MDA231MDA435-5T4OV90OVCAR3AA0857SKOV3

HT29N87

CAOV3TOV-112D

MDA-468, 8 days

0.0

0.5

1.0

1.5

Impe

danc

e in

dex

Control PF-384 5T4-ADC 5T4ADC+PF384

***

***

H-1975, 24 h

AKT

P-AKT (T308)

P-AKT (S473)

- 10 50 - - 10 10 50 50 5T4-ADC, µg/mL

- - - 5 50 5 50 5 50 PF-384, nmol/L

P-GSK3 (S21/9)

GSK3

P-H3 (S10)

GAPDH

c-PARP

1 2 3 4 5 6 7 8 9

1 2 3 4 5 6 7 8 90.00

0.25

0.50

0.75

1.00

Rel

ativ

e de

nsito

met

ric in

dex P-AKT473

P-AKT308

1 2 3 4 5 6 7 8 90.00

0.25

0.50

0.75

1.00

Rel

ativ

e de

nsito

met

ric in

dex

P-GSKβ

C D

E

MDA-468 spheroid growth, 7 days

0.01 0.1 1 10 1000

25

50

75

100

125

5T4-ADC (mg/mL)

Sphe

roid

via

bilit

y (%

con

trol

)

5T4-ADC + PF-384 (10)5T4-ADC

PF-384 (10)

0

20

40

60

80

100MDA-468, spheroid growth

Sphe

roid

via

bilit

y (%

con

trol

)

PF-384(10 nmol/L)

5T4-ADC(1mg/mL)

5T4-ADC+PF-384

**

**

01234567

MDA-468, 24 h

5T4-ADC 5T4ADC+PF384

PF-384

Fold

indu

ctio

n (c

aspa

se-3

/7)

* *

0

2

4

6

PF-3845T4-ADC 5T4-ADC+PF-384

Fold

indu

ctio

n (c

aspa

se-3

/7)

**

H-1975, 24 h

Figure 1.Effects of auristatin-based agents combined with PF-384. A, dot plot summary showing the range of CI index values in a panel of cancer cell lines obtainedfollowing analysis of drug combinations with PF-384 or with mTOR-specific inhibitor. CI indexes were determined using the Chou–Talalay method asdescribed in Materials and Methods and are presented at the ED50 level for each of the combinations. Results are the average of at least threeindependent experiments. The CI has been interpreted as follows: very strong synergy (<0.1), strong synergy (0.1–0.3), synergism (0.3–0.7), moderatesynergism (0.7–0.85), slight synergism (0.85–0.9), nearly additive (0.9–1.1), slight antagonism (1.1–1.2), and moderate antagonism (1.2–1.45). Dashed lines areat CI values of 1.1 and 0.7. PTX, paclitaxel; VINO, vinorelbine; Aur101, auristatin-101. B, dynamic monitoring of cell growth in MDA-468 cells with thexCELLigence System. B, Left, concentration- and time-dependent cytotoxic effects of 5T4-ADC, PF-384, or combination in MDA-468 cells. Onerepresentative experiment of the three is shown. B, Right, impedance index values are presented for the 8-day time point and show statistically significantenhancement of cell growth inhibition for the 5T4-ADC plus PF-384 combination. (��, P < 0.001; Student t test). The means and SEM are shown. C,MDA-468 cells grown as 3D spheroids Matrigel were treated with increasing doses of 5T4-ADC, a fixed dose of PF-384 (10 nmol/L), or a combination ofboth the drugs for 7 days. Spheroid viability was measured as described in Materials and Methods. Plotted is the percentage of growth relative to therespective untreated control. Values are means � SEM. Dashed line indicates% viability for the fixed dose of PF-384. B, Right, histogram plots of theselected data points at the indicated drug concentrations. Asterisks represent significant differences from 5T4-ADCþPF-384 (�� , P < 0.01; Student t test).D, enhanced induction of caspase-3/7 in MDA-468 (left) or H-1975 (right) cells treated with 5T4-ADC (10 mg/mL) plus PF-384 (1 mm/L) for24 hours. Fold induction in caspase-3/7 activity was determined as described in Materials and Methods. Means and SEMs of triplicate experiments areshown. � , P < 0.05; �� , P < 0.01; �� , P < 0.001. Asterisks show statistically significant differences between each of the single drugs alone and acombination; Student t test. E, H-1975 cells were treated with indicated concentrations of 5T4-ADC and PF-384, or concurrently with the combination ofboth for 24 hours. Cell lysates were subjected to immunoblotting analysis with antibodies to P-AKT, AKT, P-GSK, GSK, P-H3, cPARP, or with antibodiesto GAPDH as a loading control. E (right) densitometric analyses of E was performed using ImageJ, as described in Supplementary Materials and Methods.Densitometric analyses of data shows protein levels of P-AKT (S473, T308), P-GSK3b (S9).

Combination of 5T4-ADC with PI3K/mTOR Inhibitors or Taxanes

www.aacrjournals.org Clin Cancer Res; 2015 OF3




additive effects detected for 5T4-ADC was also observed for thecell permeable version of parental payload MMAF, MMAF-OMe,or for a recently described auristatin analogue PF-06380101(Aur101; ref. 5). Structurally unrelated microtubule modulatorswith diverse mechanisms of action (both stabilizer paclitaxel anddestabilizer vinorelbine) also showed additive or synergisticrelationships when combined with PF-384 in a smaller subsetof cell lines, suggesting a shared mechanism of drug interactionthat is directly related to the inhibition of microtubules (Fig. 1A).

We also performed real-time monitoring of cell growth usingthe xCelligence system.Over the course of approximately 11 days,we observed that exposure of MDA-468 cells to 5T4-ADC/PF-384and MMAF-OMe/PF-384 combinations reduced proliferationmore than either of the individual agents alone (Fig. 1B; Supple-mentary Fig. S2A). Furthermore, when tested in 3D spheroidgrowth assays, significant enhancement of cytotoxicity wasdetected for the 5T4-ADC/PF-384 andMMAF-OMe/PF-384, com-pared with the single-agent treatments (Fig. 1C; SupplementaryFig. S2B). Similar observations were made for the lung cancerH-1975 spheroid model (data not shown). Thus, these resultsconfirmed our findings above and provided additional insightson the time dependence of synergistic effects in vitro.

Effect of 5T4-ADC/PF-384 on apoptosis, cell-cycle, andPI3K/mTOR markers

To evaluate if synergistic growth inhibition induced by the5T4-ADC and PF-384 combinations is due to apoptosis,we determined proapoptic signal caspase-3/7 activation inH-1975 or MDA-468 cells. 5T4-ADC or PF-384 alone led tomodest activation of caspase-3/7 in both the cell lines, mea-sured 24 hours after incubation (Fig. 1D). However, thecombination treatment showed markedly enhanced inductionof caspase-3/7. Similar results were obtained when the samecells were treated with MMAF-OMe/PF-384, suggesting that theinduction of apoptosis in 5T4-ADC/PF-384 combinationis mechanistically linked to the action of parental payloadMMAF-OMe (Supplementary Fig. S2C). Immunoblot analysisfor cleaved PARP further demonstrated the induction of apo-ptosis by the combination of 5T4-ADC/PF-384 in H-1975(Fig. 1E). Collectively, these findings demonstrate that anti-proliferative effects observed with combinations could beaccounted for, at least in part, by the enhanced apoptoticresponse mediated by caspase-3/7.

As auristatin-containing drugs have been described to impaircell-cycle progression, we first asked if PF-384 can modifymitotic arrest when combined with 5T4-ADC in MDA-468cells. As expected, 5T4-ADC alone markedly induced a signif-icant accumulation of cells in mitosis as evidenced by the levelsof phosphorylated histone H3 (P-H3; Supplementary Fig. S3A).Interestingly, the addition of PF-384 substantially reducedADC-mediated mitotic arrest. The same observations weremade in H-1975 cells when PF-384 was combined with 5T4-ADC or free auristatins (Supplementary Fig. S3B and data notshown). Analysis of the cell cycle in H-1975 cells showed that5T4-ADC/PF-384 consistently reduced ADC-mediated G2–Marrest, with a modest increase in G1 stage compared with5T4-ADC alone (Supplementary Fig. S3C). These results raisedthe possibility that combined treatment with 5T4-ADC/PF-384may suppress the spindle assembly checkpoint function, lead-ing to inappropriate transition out of mitosis. Aurora B inhi-bition has been shown to selectively relax the spindle check-

point invoked by microtubule inhibitors (26, 27). We observedan increase of Aurora A, B, and C phosphorylation by the 5T4-ADC, an effect blocked by the addition of 50 nmol/L PF-384(Supplementary Fig. S3D). Consequently, these results supportthe hypothesis that suboptimal Aurora activity may be respon-sible for accelerated mitotic exit and apoptosis in cells exposedto the combination of both the drugs.

To examine the changes in signaling after treatment withsingle-agent PF-384 or 5T4-ADC, we initially used phospho-kinase antibody arrays. In the H-1975 lung tumor modeltreated with PF-384, we detected expected declines in P-AKT,P-S6K, P-GSK-3b, P-NOS, and P-PRAS40 markers 6 or 16 hoursfollowing drug treatment (Supplementary Fig. S4A). Surpris-ingly, treatment with 5T4-ADC has resulted in the inhibition ofP-AKT and P-GSK-3b 16 hours after drug exposure (Supple-mentary Fig. S4B). To confirm these findings, we analyzed byimmunoblot the downstream effectors of PI3K/mTOR path-ways as well as markers for mitotic arrest and apoptosis.H-1975 (lower 5T4-expresser, L858R/T790M EGFR mutations)and MDA-468 (high 5T4-expresser, PTEN�/�) are cancer celllines that are equally sensitive in vitro to the dual inhibitor ofPI3K/mTOR or to auristatin derivative MMAF-OMe. We treatedH-1975 or MDA-468 cells with the dual PI3K/mTOR inhib-itor PF-384, 5T4-ADC, or a combination of the two. In agree-ment with the previous report (16), PF-384 alone effectivelyreduced phosphorylation of downstream markers reflectingactivation status of the PI3K/mTOR pathway: P-AKT S308,P-AKT S473, and P-GSK-3a/b S21/9 24 hours after drug expo-sure (Fig. 1E; Supplementary Fig. S4C). In H-1975 cells, treat-ment with 50 mg/mL 5T4-ADC led to modest reduction inP-AKT S308, P-AKT S473, and GSK-3b (S9), as quantified bydensitometry. In the same 24-hour experiment, 5T4-ADC/PF-384 treatment showed stronger decline in phosphorylationlevels of PI3K/mTOR downstream effectors than each of thesingle agents, as confirmed through densitometric analysis(Fig. 1E). Increasing concentration of 5T4-ADC alone inducedphosphorylation of H3, but the combination appeared nearlyequivalent or even somewhat lower than 5T4-ADC alone,which is in good agreement with the mitotic index valuesdetermined by flow cytometry. Furthermore, we found that5T4-ADC/PF-384 caused a greater induction of cleaved PARPthan either 5T4-ADC or PF-384 alone. The potentiation ofPF-3840s effect on downstream markers by 5T4-ADC appearsto be cell line specific because experiments on MDA-468showed no further reduction of P-AKT and P-GSK by thecombination treatment as compared with PF-384 alone (Sup-plementary Fig. S4C). In MDA-468, we detected no suppres-sion of PI3K/mTOR pathway markers with a single-agent 5T4-ADC. These findings support the conclusions that enhancedcytotoxicity observed in 5T4-ADC/PF-384 combination islinked, at least in part, to the induction of caspase-3/7 andPARP-dependent apoptosis and correlates with the strongersuppression of PI3K/mTOR pathway biomarkers in H-1975cells.

Auristatin agents cooperate with PF-384 to regulatetranslation

We hypothesized that previously uncharacterized effects ofauristatins on additional intracellular targets may underliethe observed synergy with PF-384. Total proteomics and phos-phoproteomics were independently applied to evaluate changes

Shor et al.





in protein abundance and phosphorylation in response to aur-istatin-based agents in MDA-468, a cell line that had no demon-strable changes in phosphorylation of AKT or GSK upon admin-istration of 5T4-ADC.

Analysis of protein interaction networks for all significantlychanged proteins in total proteome perturbed with 0.5 nmol/Lauristatin showed several clusters of related functional classesof proteins, with a highly distinct group containing compo-nents of mRNA translation and mRNA biogenesis (Fig. 2A;Supplementary Fig. S5A). Specifically, we detected remarkableoverrepresentation of mRNA translation factors in the down-regulated protein set and of ribosomal proteins in the upregu-lated group after a 24-hour cell exposure to either of the twodoses of MMAF-Ome (0.5 and 5 nmol/L; Supplementary Fig.S5B and S5C and Supplementary Table S2A). Label-free phos-

phoproteomics of cells treated with the MMAF-Ome or 5T4-ADC for 6 hours demonstrated unanticipated enrichment of"translation factors," "mRNA processing," and "mRNA splic-ing" for both MMAF-OMe and 5T4-ADC in the list of down-regulated phosphopeptides (Supplementary Fig S5D and S5Eand Supplementary Table S2B). Consistent with the commonmechanism of action between the ADC and free unconjugat-ed payload, there was a substantial overlap in differentiallyexpressed phosphopeptides between MMAF-OMe and 5T4-ADC, with the overrepresentation of translation or mRNAprocessing-related components in the shared datasets (Supple-mentary Fig. S5D). Thus, our results raise the intriguing pos-sibility that protein synthesis is one of the convergence pointsfor cellular action of MMAF-OMe or 5T4-ADC. This notion,together with the well-known role for the PI3K/mTOR signaling

AMDA-468

- - - 1 10 1 1 10 10 5T4-ADC, mg/mL

- 5 50 - - 5 50 5 50 PF-384, nmol/L

eIF4G2

eIF4G1

P-eIF4G1 (S1108)

eIF4B

P-eIF2a (S51)

eIF2a

eIF3A

P-eEF2 (T56)

GAPDH

eEF2

P-4E-BP1 (T37/46)

4E-BP1

P-eIF4B (S422)

6 hrs 16 hrs 24 hrs0.00

0.25

0.50

0.75

1.00

1.25MDA-468-Luc

Luci

fera

se le

vels

(nor

mal

ized

) CntrlPF-3845T4-ADC5T4-ADC/PF-384CHX

B

C

eIF4G2

eIF4G1

- 10 50 - - 10 10 50 50 5T4-ADC mg/mL

- - - 5 50 5 50 5 50 PF-384, nmol/L

eIF4B

P-eIF4G1 (S1108)

eIF2a

P-eIF2a (S51)

eIF3A

P-eEF2 (T56)

eEF2

GAPDH

P-4E-BP1 (T37/46)

4E-BP1

H-1975

P-eIF4B (S422)

Figure 2.Cooperative suppression of protein translation by ADC/PF-384 combination. A, STRING network analysis of all significantly changed proteins in a totalproteomics experiment for MDA-48 cells treated with 0.5 nmol/L MMAF-Ome. Major cluster of interacting proteins includes the translation factors,ribosomal proteins, and ribonucleoproteins that form a densely connected module, which is denoted by a red rectangle. Only connected nodes are shown forsimplicity. STRING network analysis and visualization was performed using the online STRING database 9.1 (45). B, total protein abundance andphosphorylation status of selected proteins involved in mRNA translation. MDA-468 (left) or H-1975 (right) cells were treated with the 5T4-ADC, PF-384 orcombinations of both drugs for 24 hours at the indicated concentrations. Total protein abundance or phosphorylation levels were measured withantibodies indicated and as described in Materials and Methods. C, effect of auristatin-based agents alone or in combination with PF-384 on cap-dependentsynthesis of luciferase reporter. MDA-468 cells stably transduced with monocistronic luciferase reporter were used to assay inhibition of cap-dependenttranslation as described in Materials and Methods. Cells were treated with individual drugs or with combinations for 24 hours. The concentrations ofdrugs were: PF-384 50 nmol/L, 5T4-ADC 10 mg/mL, PTX (paclitaxel) 10 nmol/L, CHX (cycloheximide) 30 mg/mL. 5T4-ADC/PF-384: 5T4-ADC 10 mg/mL andPF-384 50 nmol/L. Results are mean � SE of biologic triplicates from a single experiment representative of the two. 5T4-ADC and PF-384 were statisticallysignificantly different compared with vehicle-treated control by the two-tailed Student t test. 5T4-ADC/PF-384 combination was significantly differentcompared with each of the single drug controls by the two-tailed Student t test.






pathway in the regulation of protein synthesis (28, 29), sup-ports the hypothesis that cellular synergy of 5T4-ADC/PF-384or MMAF-OMe/PF-384 might be at least partially attributed tothe cooperative inactivation of translation.

To substantiate the above findings, we focused on theeffects of 5T4-ADC/PF-384 combination on steady-statelevel and phosphorylation status of key translation factorsinvolved in the PI3K/mTOR pathway. Single-agent 5T4-ADCshowed modest cell type-specific effects, each modulatingexpression and/or phosphorylation of a select subset ofproteins to a different extent (Fig 2B). In both the cell lines,treatment with 5T4-ADC decreased levels of eIF4G1, eIF4G2,eIF4B, and eIF3A, but upregulated P-eIF2a and P-eEF2.More importantly, combination treatments caused coopera-tive changes in a distinct set of translational regulators. Forexample, in MDA-468, a cell model that was used for pro-teome-wide analysis, 5T4-ADC/PF-384 combination led to astronger decline in eIF4G2, eIF4B, eIF2a, and eIF3A levels,which coincided with greater induction of P-eIF2a and P-eEF2 than for each agent alone (Fig 2B). In the H-1975model, the same drug combination caused decreases inthe expression levels of P-eIF4G1, P-eIF4B, eIF4B, P-4E-BP1,and 4E-BP1 with a concomitant increase in P-eIF2a andP-eEF2. To test whether these observations also translate intofunctional impairment of general protein synthesis, we mon-itored the activity of firefly luciferase in MDA-468 cells stablytransduced with a cap-dependent monocistronic reporter.Addition of PF-384 or 5T4-ADC significantly suppressedproduction of luciferase by approximately 40% to 50% after16- or 24-hour incubation with drugs, whereas treatment witha positive control, protein synthesis inhibitor cycloheximide,fully reduced luciferase production in this system. Treat-ment with 5T4-ADC/PF-384 led to a stronger decline inluciferase activity when compared with either single agentalone (Fig. 2C). No reduction of luciferase transcription wasobserved in this experiment in response to drug exposure asmeasured by RT-PCR (data not shown). Overall, our resultssuggest that combining auristatin-based drugs with inhibitorsof the PI3K/mTOR pathway can lead to a specific reprogram-ming of translational factor repertoire at the level of expres-sion and/or phosphorylation, which causes suppression ofprotein synthesis.

In vivo combination therapy with 5T4-ADC and PF-384To investigate whether cooperative action of 5T4-ADC/PF-

384 in vitro could be observed in the in vivo setting, we testedthe efficacy of the respective single agents and of the combi-nation in two previously characterized tumor xenograftmodels with broad range of 5T4 expression levels as shownby flow cytometry and IHC staining (Supplementary Fig. S1B;ref. 7). Treatment of animals bearing MDA-468 breast cancerxenografts with 2 mg/kg 5T4-ADC caused initial robust tumorsuppression followed by stasis, whereas PF-384 at 7.5 mg/kgshowed a very minor inhibition of tumor growth over thevehicle-treated arm (Fig. 3A). In contrast, concurrent admin-istration of both the drugs led to a more complete tumorregressions clearly observed in all tumors treated by the endof the study. Subsequent TTE analysis of time to no tumor(tumor regression) showed that a much shorter time wasneeded to achieve complete tumor regressions in the combi-nation arm versus 5T4-ADC alone (P < 0.0001 by the log-rank

test), with all animals in the 5T4-ADC/PF-384 group becom-ing tumor-free by the day 36 (Fig. 3B). In the H-1975 lungcancer model, treatment with 3 mg/kg 5T4-ADC resultedin tumor stasis followed by regrowth of tumors, whereas7.5 mg/kg PF-384 elicited only nominal antitumor activity(Fig. 3C). The 5T4-ADC/PF-384 combination resulted in amore complete, but still unsustained suppression of tumorgrowth. The percentage of animals with less than three-foldincrease in tumor volume was used as survival endpoint forthe analysis of H-1975 model. TTE analysis indicated a sta-tistically significant delay in tumor tripling rate for thecombination group compared with 5T4-ADC (3 mg/kg; P ¼0.0356, log-rank test) or PF-384 alone (7.5 mg/kg, P < 0.0001,log-rank test, Fig. 3D).

Combination of 5T4-ADC with taxanesThe initial drug interaction screen also identified paclitaxel

as an agent that potentiated MMAF-OMe- or 5T4-ADC growthinhibitory effects. The combination results varied from syner-gism to additivity in most of the cell models tested as measur-ed by the CI index (Fig. 4A). This effect was not unique to pacli-taxel, as docetaxel, a structurally similar taxane, also showedfavorable interactions with MMAF-OMe or 5T4-ADC. More-over, when MMAF-OMe or 5T4-ADC was substituted foranother auristatin analogue Aur101, or an unrelated micro-tubule-depolymerizing agent vinorelbine, we also observedpotentiation of their cytotoxicity by paclitaxel. According toa high-resolution structural data obtained for dolastatin-10(30) or for the new auristatin analogue bound to tubulin(5), auristatins bind at a site adjacent to the vinca binding siteat the interface of two tubulin molecules and in close proximityto the b-tubulin nucleotide exchange (Fig. 4A, Right). Weexamined additional cellular changes after cotreatment withpaclitaxel. Modest enhancement of cytotoxicity in a 3D spher-oid assay and stronger induction of caspase-3/7 was observedin the MDA-468 cells treated with 5T4-ADC plus paclitaxel for48 hours compared with the single drug controls (Fig. 4Band C). A potential explanation for the cooperative actionbetween 5T4-ADC and paclitaxel includes modulation ofcell-cycle progression and altered microtubule dynamics. Asexpected, an M-phase-specific marker P-H3 was markedlyinduced in the MDA-468 cells treated with paclitaxel or 5T4-ADC. The cotreatment with both drugs slightly enhanced themean mitotic increase relative to single agents, but this trenddid not reach statistical significance (Supplementary Fig. S6A).Furthermore, we observed enhanced PARP cleavage andincrease in levels of P-H3 in MDA-468 or H-1975 cells treatedwith 5T4-ADC/paclitaxel combination versus either single drug(Fig. 4D). Interestingly, fluorescent microscopy with an anti-tubulin antibody demonstrated a significant collapse of micro-tubule network around the nucleus and formation of themicrotubule aggregates with 5T4-ADC/paclitaxel combinationin MDA-468 cells (Supplementary Fig. S6B). Unlike the com-bination's effect, the 5T4-ADC alone treated cells showeddisintegration of microtubule bundles with more intense stain-ing at cell periphery and lesser cytoplasmic volume. Collec-tively, these findings demonstrate that antiproliferative effectsobserved with the 5T4-ADC/paclitaxel could be accounted for,at least in part, by the enhanced apoptotic response mediatedby caspase-3/7 and PARP with parallel induction of P-H3 andmore pronounced loss of microtubule integrity.

Shor et al.





Given the favorable interactions observed between 5T4-ADC and taxanes in vitro, we also evaluated the potentialantitumor activity of this combination in vivo. MDA-468xenografts were tested with concurrent combinations of5T4-ADC (2 mg/kg) plus paclitaxel (10 or 22.5 mg/kg;Fig. 5A–D). Paclitaxel monotherapy had no pronouncedeffect on tumor growth at 10 mg/kg, but resulted in strongantitumor activity at the higher dose of 22.5 mg/kg (Fig. 5Aand C). Single drug treatment with 2 mg/kg 5T4-ADC in thismodel led to sustained but incomplete tumor regressionfollowed by the stasis. However, coadministration of 5T4-ADC and paclitaxel at two different doses resulted in profoundand lasting tumor regression for the duration of the study.Importantly, TTE analysis of these combinations revealedsignificantly shorter time needed to achieve complete tumor

regressions for the combination arms compared with singledrugs alone (Fig. 5B and D). In H-1975 human lung cancerxenograft model, paclitaxel treatment at 10 mg/kg resulted intumor growth delay (Fig. 5E). 5T4-ADC at 3 mg/kg inhibitedthe growth of xenografts, with tumor regrowth evident 2 weeksafter treatment was stopped. In contrast, the dual treatmentwith 5T4-ADC/paclitaxel produced marked enhancement inantitumor activity compared with monotherapy treatments.Moreover, log-rank tests showed statistically significant delayin tumor tripling rate for the combination group comparedwith 5T4-ADC or paclitaxel alone treatment arms (Fig. 5F). Insummary, these data suggest that paclitaxel, when used atclinically achieved exposures, can strongly enhance antitumorefficacy of 5T4-ADC in preclinical models of human lung andbreast cancer.

A

C

0 10 20 30 40 50 60 700

25

50

75

100

MDA-468

Time (days)

Perc

ent w

ith tu

mor

s

5T4ADC (2)+PF384 (7.5)5T4ADC (2)PF384 (7.5)

0 10 20 30 400

25

50

75

100

H-1975

Time (days)

% A

nim

als

with

< 3

fold

incr

ease

in tu

mor

vol

ume

VehiclePF384 (7.5)5T4ADC (3)5T4ADC(3)+PF384(7.5)

MDA-468

0 10 20 30 40 500

200

400

600

800

1,000

1,200 Vehicle

5T4ADC(2)5T4ADC(2)+PF384(7.5)

PF384(7.5)

Day

Tum

or v

olum

e (m

m3 )

Tum

or v

olum

e (m

m3 )

H-1975

0 5 10 15 20 25 30 350

500

1,000

1,500

2,000

2,500 Vehicle

5T4ADC(3)5T4ADC(3)+PF384(7.5)

PF384(7.5)

Day

B

D

Figure 3.The 5T4-ADC plus PF-384 combination treatment leads to enhanced therapeutic effects in breast and lung cancer models in vivo. A and B, mice bearingsubcutaneous MDA-468 human breast tumor xenografts were treated with vehicle, 5T4-ADC (i.v. 2 mg/kg, q4d), PF-384 (i.v. 7.5 mg/kg, q4d), or acombination. A, tumor growth curves. Tumor volume was determined at the indicated times after the onset of treatment. Points, mean of values from10 mice/group; bars, SE. B, time to endpoint (TTE) plots for the treatment groups in A show change in percentage of animals with tumors over the time.Endpoint is defined as the time elapsed for animal to become tumor-free. TTE analysis demonstrates significantly enhanced rate of tumor regressionswith combination of 5T4-ADC plus PF-384 compared with the single-agent activity of 5T4-ADC (P < 0.0001, log-rank Mantel–Cox test). PF-384 didnot elicit regressions in this experiment. C and D, mice bearing subcutaneous H-1975 human lung tumor xenografts were treated with 5T4-ADC (i.v. 3 mg/kg,q4d), PF-384 (i.v. 7.5 mg/kg, q4d) or a combination. C, tumor growth curves. Tumor volume was determined at the indicated times after the onset oftreatment. Points, mean of values from 10 mice/group; bars, SE. TTE plots for the treatment groups in C show the percentage of animals with less thanthree-fold increase in tumor volume over time. Endpoint is defined as the time at which tumor volume has tripled. TTE analysis of data demonstratessignificant delay at rate of tumor tripling for the 5T4-ADC plus PF-384 combination compared with the single-agent activity of 5T4-ADC (P¼ 0.0356, log-ranktest) or PF-384 (P < 0.0001, log-rank test) alone. To minimize fluctuations in the tumor growth curve plots and facilitate interpretation of the data,the mean tumor volume for each group was plotted until >10% of the mice in the group were sacrificed.






Discussion

Optimizing the efficacy of ADCs by systematic nonclinicalassessment of combinations remains an important objectivefor ADC development. We hypothesized that major synergisticeffects for the antibody–auristatin conjugates may be mediatedby the pharmacologic action of payload itself and thereforesearched for the common chemotherapeutic agents or signalinginhibitors that could potentiate either free or conjugated aur-istatins, such as 5T4-ADC. Here, we describe novel and previ-ously uncharacterized potentiation of auristatin-based agentsby PF-384 or taxanes in vitro that translates to enhanced anti-

tumor efficacy in tumor xenograft models. The combination ofMMAF-Ome or 5T4-ADC and PF-384 has resulted in consis-tently synergistic drug effect in tumor cell lines of lung, breast,and ovarian cancer origin, the three putative tumor types thatshow broad 5T4 expression (7). Analysis of the in vitro cyto-toxicity data for the cell line panel used in this study revealedthat neither common mutations in these cell lines nor theirtissue lineage or sensitivity of the individual drugs alonecould be used to predict synergistic responses to the combina-tions involving auristatins and other drugs. The favorablepharmacologic interactions were also observed when differentmicrotubule-targeting agents paclitaxel (polymerizing) or

B

A

C

GTPColchicine

GDP

Auristatin

β αα

Taxol

5T4A

DC+PTX

5T4-A

DC+DOCET

MMAF-OMe+

PTX

MMAF-OMe+

DOCET

Aur101+

PTX

Aur101+

DOCET

VINO+PTX

0.0

0.5

1.0

1.5

2.0

CI i

ndex

H1975Calu6H358HCC2429MDA468MDA231OV90OVCAR3SKOV3CAOV-3TOV-112D

MDA-468 spheroid growth, 7 days

0.01 0.1 1 10 1000

20

40

60

80

100

120

5T4-ADC (1 mg/mL)

Sphe

roid

via

bilit

y (%

con

trol

)

5T4-ADC/PTX (1 nmol/L)

5T4-ADC

PTX (1 nmol/L)

0

20

40

60

80

100MDA-468 spheroid growth, 7 days

Sphe

roid

via

bilit

y (%

con

trol

)

PTX(1 nmol/L)

5T4-ADC(1 mg/mL)

5T4-ADC+PTX

***

**

0

1

2

3

4

5

MDA-468, 48 h

PTX5T4-ADC 5T4ADC+PTX

Fold

indu

ctio

n (c

aspa

se-3

/7)

****

MDA-468

c-PARP

- 10 100 - - 10 100 10 100 PTX, nmol/L

- - - 1 10 1 1 10 10 5T4ADC, mg/mL

P-H3 (S10)

GAPDH

c-PARP

P-H3 (S10)

GAPDH

H-1975

D

Combination with taxanes

Figure 4.Effects of auristatin-based agents combined with taxanes. A, summary of CI values in a panel of cancer cell lines. Dot plot showing the range of CIindex values obtained following analysis of drug combinations with microtubule inhibitors. CI indexes were determined using the Chou–Talalay method asdescribed in Materials and Methods and are presented at the ED50 level for each of the combinations. Results are the average of at least threeindependent experiments. The CI has been interpreted as in Fig. 1A. Dashed lines are at CI values of 1.1 and 0.7. B, combination of 5T4-ADC or MMAF-OMewith paclitaxel (PTX) leads to stronger suppression of cell growth in 3D culture. MDA-468 cells were treated with increasing doses of 5T4-ADC, fixeddose of paclitaxel (1 nmol/L), or a combination of both drugs for 7 days. Spheroid viability was measured as described in Materials and Methods. Plotted isthe percentage of growth relative to the respective untreated control. Dashed line indicates% viability for the fixed dose of MMAF-OMe. Right,histogram plots of the selected data points at the indicated drug concentrations. C, induction of caspase-3/7 activity by the combination of 5T4-ADCwith paclitaxel in MDA_468 cells. Left, cells treated with 5T4-ADC (10 mg/mL) plus PF-384 (1 mmol/L); right, cells treated with MMAF-OMe (0.22 nmol/L) pluspaclitaxel (6 nmol/L) (D) for 48 hours. Fold induction in caspase-3/7 activity was determined as described in Materials and Methods. Means and SEMsof triplicate experiments are shown. � , P < 0.05; �� , P < 0.01; �� , P < 0.001. Asterisks show statistically significant differences between each of the singledrugs alone and a combination; Student t test. D, effect of single-agent 5T4-ADC and combinations with paclitaxel on apoptosis and mitotic markermodulation. MDA-468 (top) or H-1975 (bottom) cells were treated with indicated concentrations of 5T4-ADC and paclitaxel, or concurrently with thecombination of both for 24 hours. Cell lysates were subjected to immunoblotting analysis with antibodies to cPARP, P-H3(S10), or with antibodies to GAPDHas a loading control.

Shor et al.





vinorelbine (depolymerizing, unpublished observation) werecombined with PF-384, suggesting that microtubule damage isa global signal that can be potentiated by the suppression of

PI3K/mTOR signaling. Surprisingly, we also found reproduc-ible but cell- and drug-dependent inhibition of AKT and/orGSK3 phosphorylation in response to single-agent 5T4-ADC.

E

A

C

0 7 14 21 28 35 42 49 56 63 70 77 84 91 98 1050

25

50

75

100

MDA-468

Time (days)

Perc

ent w

ith tu

mor

5T4ADC (2)+PTX (10)

5T4ADC (2)PTX (10)

0 7 14 21 28 35 42 49 56 63 70 77 84 91 98 1050

25

50

75

100

MDA-468

Time (days)

Perc

ent w

ith tu

mor

s

5T4ADC (2) + PTX (22.5)

5T4ADC (2)

PTX (22.5)

0 7 14 21 28 35 42 49 56 630

20

40

60

80

100

H1975

Time (days)

% A

nim

als

with

< 3

fold

Incr

ease

in tu

mor

vol

ume Vehicle

PTX (10)5T4ADC (3)

5T4ADC (3)+PTX (10)

MDA-468

0 10 20 30 40 500

200

400

600

800

1,000

1,200

1,400 Vehicle

5T4ADC(2)

PTX(10)

5T4ADC(2)+PTX(10)

Day

MDA-468

0 10 20 30 40 500

200

400

600

800

1,000

1,200

1,400Vehicle

5T4ADC(2)

PTX(22.5)

5T4ADC(2)+PTX(22.5)

Day

Tum

or v

olum

e (m

m3 )

Tum

or v

olum

e (m

m3 )

Tum

or v

olum

e (m

m3 )

H-1975

10 20 30 40 50 600

1,000

2,000

3,000

VehiclePTX (10)5T4ADC (3)5T4ADC(3)+PTX(10)

Days

B

D

F

Figure 5.The 5T4-ADC and paclitaxel (PTX) combination treatment leads to enhanced therapeutic effects in breast and lung cancer models in vivo. A and C, mice bearingsubcutaneous MDA-468 human breast tumor xenografts were treated with 5T4-ADC (i.v. 2 mg/kg, q4d), paclitaxel (p.o. 10 mg/kg, q4d), or a combination.A, tumor growth curves. Tumor volume was determined at the indicated times after the onset of treatment. Points, mean of values from 10 mice/group;bars, SE. B, time to endpoint (TTE) plots for the treatment groups in A show change in the percentage of animals with tumors over the time. Endpoint isdefined as the time elapsed for animal to become tumor-free. TTE analysis of data demonstrates significantly faster rate of complete tumor regressionsachieved with the combination of 5T4-ADC and paclitaxel compared with the single-agent activity of 5T4-ADC (P ¼ 0.0071, log-rank test) or paclitaxel(P ¼ 0.01, log-rank test). C, similar to A, but paclitaxel was used at a dose of 22.5 mg/kg (p.o. q4d). D, TTE analysis of data from C shows significantly fasterrate of complete tumor regressions achieved with the combination of 5T4-ADC plus paclitaxel compared with the single-agent activity of 5T4-ADC(P ¼ 0.00821, log-rank test) or paclitaxel, which has not produced any tumor regressions at this dose. E, mice bearing subcutaneous H-1975 tumors weretreated with 5T4-ADC (i.v. 3 mg/kg, q4d), paclitaxel (p.o. 10 mg/kg, q4d), or a combination. 5T4-ADC combined with paclitaxel is more efficacious thantreatment with single agents. F, TTE analysis of data performed similarly to Fig. 3B, demonstrates significant delay of rate of tumor tripling for the 5T4-ADC pluspaclitaxel combination compared with the single-agent activity of 5T4-ADC (P < 0.0001, log-rank test) or paclitaxel (P ¼ 0.0001, log-rank test) alone.






This finding is unexpected but in agreement with the reportby Asnaghi and colleagues who demonstrated inhibitory effectsof nocodazole on phosphorylation of mTOR at Ser 2448 (31).We, however, could not consistently detect synergistic sup-pression of PI3K/mTOR pathway markers in MDA-468 cells.Unlike H-1975, an MDA-468, a PTEN�/� breast cancer model ismore sensitive to each single agent alone, which may make thedetection of cooperative effects on downstream pathways tech-nically difficult.

We also applied a system-wide approach to explore theadditional cellular action of auristatins that could help explainmolecular mechanisms behind the synergistic interactions withthe PF-384. Both total proteomics with MMAF-OMe and phos-phoproteomics performed with 5T4-ADC and MMAF-OMeuncovered mRNA translation as one of the predominantlyaffected processes. These findings were further corroboratedby Western blot analysis of selected translation initiation orelongation components. More proteins were modulated byMMAF-Ome, 5T4-ADC, or PF-384 alone, with only some show-ing cooperative effects by combined treatment with 5T4-ADC/PF-384 or MMAF-OMe/PF-384. This comports with other large-scale phosphoproteomic studies that identified a number ofinitiation factors differentially phosphorylated in response tothe nocodazole treatment (32, 33). Whereas targeting of trans-lational components by auristatin-based agents is a novelfinding, the effects of PF-384 are quite expected, given theknown role of PI3K and mTOR kinases in the control of proteinsynthesis. One anticipated consequence of ADC-mediatedmodulation of protein synthesis machinery is a decline inglobal translation rates. Stronger reduction of cap-dependentsynthesis of luciferase reporter by the 5T4-ADC/PF-384 com-bination relative to single-agent drugs is consistent with theupregulation of P-eIF2a (S51) and P-eEF2 (T56), changes thatare suggestive of a slowdown in translation. Thus, coopera-tive suppression of protein synthesis can, at least in part, belinked to the observed cellular synergy with PF-384. Notably,structurally unrelated and highly specific mTOR inhibitorWYE-132, which is known to disrupt the cap-dependentmRNA translation and inhibit global protein synthesis, alsoshowed synergy when combined with auristatin derivativein vitro (Supplementary Table S3), This implies that reducingcap-dependent translation by targeting mTOR kinase activityalone may be sufficient to enhance therapeutic effects of theauristatin-based agents.

Given the highly specific effects of MMAF-OMe and 5T4-ADCon microtubules, how these agents perturb mRNA translationand how PF-384 can potentiate this mechanism? One potentialexplanation is based on the substantial evidence of interactionsbetween tubulin cytoskeletal components and ribosomal pro-teins, translation initiation factors and various mRNPs. Inaddition, many mRNAs encoding mitotic regulators or trans-lational components are known to localize to mitotic spindles(34–40). This and the evidence that ongoing translation ismaintained throughout the cell cycle, without substantialdecline during mitosis (35) is consistent with the notion thattubulin-localized protein synthesis maybe especially importantfor the efficient progression through mitosis. Hence, we pro-pose that the entire class of auristatin-containing agents,including ADCs, can disrupt the tubulin-bound pools of trans-lational components, thereby modulating their abundance. Wespeculate that the combined action of ADC and PF-384 disrupts

protein synthesis during mitotic transition, or collectivelyimpacts translation of the key mRNAs required for the survivalduring mitotically arrested state. Furthermore, it is also possiblethat both the drug classes could affect mRNA translation indifferent stages of cell cycle. Interestingly, cotreatment withPF-384 impairs Aurora kinase phosphorylation in our experi-ments, a phenotype that is generally consistent with compro-mised spindle checkpoint and decreased mitotic index in thesecells. These data are also in agreement with what has beenpreviously described for combination of docetaxel and anotherPI3K inhibitor GDC-0941, where decreased time of mitoticarrest was mechanistically linked to the induction of apoptosisin synchronized cells (41). Our results provide one potentialscenario to explain the synergistic activity and it is likely thatmultiple mechanisms may determine favorable pharmacologicoutcomes in tumor cells.

Remarkably, the combination therapy with 5T4-ADC/PF-384in vivo significantly improved antitumor activity and reducedtumor volumes in models of breast and lung cancer as comparedwith the effects of single drug treatment alone. Other importantpreclinical work demonstrated that T-DM1 plus the pan-PI3Kinhibitor GDC-0941 or plus dual PI3K/mTOR combinationsresulted in the enhanced antitumor activity both in vitro andin trastuzumab-resistant or in PIK3CA mutant breast cancerxenograft models in vivo (13). The T-DM1 plus GDC-0941 wastested in a 3þ 3 design dose-escalation phase Ib study in patientswith advanced HER2-positive metastatic breast cancer. Resultshave been reported only on 13 patients, with dose-limingtoxicities of a grade 4 thrombocytopenia and grade 3 fatigueobserved in two initial cohorts. The combination regimen wasbetter tolerated in a third cohort that enrolled at a reduced doseof T-DM1 (3.0 mg/kg) and GDC-0941 (100 mg; ref. 42). Itremains to be seen if combining T-DM1 with PI3K inhibitorsgenerates meaningful clinical activity in these patients.

Several lines of evidence illustrate an important therapeuticpotential for another promising combination between the5T4-ADC and a taxane, paclitaxel. Enhanced antitumor activityof 5T4-ADC plus paclitaxel is intriguing but not counterintu-itive. Whereas both agents act on the microtubules, MMAF-OMe binds to a distinct site than the taxane binding site, in amanner similar to the vinca alkaloids and thus may affectadditional tubulin-dependent functions in a paclitaxel-inde-pendent manner. This may lead to therapeutic synergy whencombined with paclitaxel. There is considerable interest in theADC field in testing the clinical activity and safety of theconjugates with other cytotoxic antimitotic agents, such astaxanes. This hypothesis is currently under investigation inclinical trials evaluating combination therapy of T-DM1 plusdocetaxel in early-stage HERþ breast cancer (43) and in a phaseIII study comparing brentuximab vedotin plus AVD versusABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine)alone (44).

In conclusion, our report provides strong preclinical frame-work and the rationale for combination therapy of 5T4-ADCwith taxanes or 5T4-ADC with PF-384 in clinical trials for thetreatment of lung, breast, or ovarian cancer. The dual-targetingapproach presented here, with both an auristatin-based agentand PI3K/mTOR pathway inhibitor or taxanes, could serve asan important model for enhancing antitumor activity of otherauristatin-based ADCs and overcoming potential drug resis-tance in the clinic.

Shor et al.





Disclosure of Potential Conflicts of InterestP. Sapra and B. Shor have ownership interest in Pfizer and are listed as

co-inventors on a provisional patent, which is owned by Pfizer, directedto subject matter including auristatin combinations. No potential conflictsof interest were disclosed by the other authors.

Authors' ContributionsConception and design: B. Shor, P. SapraDevelopment of methodology: J. Kahler, M. Dougher, E. MelamudAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J. Kahler, M. Dougher, M. Mack, E. Rosfjord, F. Wang,E. MelamudAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): B. Shor, J. Kahler, M. Dougher, M. Mack, E. Rosfjord,F. Wang, E. MelamudWriting, review, and/or revision of the manuscript: B. Shor, J. Kahler,E. Rosfjord, P. SapraAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): M. Dougher, E. RosfjordStudy supervision: B. Shor, E. Rosfjord, P. Sapra

AcknowledgmentsThe authors are grateful to Dr. Fred Immerman for help with statis-

tical analysis; Dr. Shuyan Lu for help with cell impedance analysis;Dr. Jeremy Myers for helpful discussions; members of ADC conjugationteam for preparation and providing ADCs; Drs. Kevin Parris and AndreasMaderna for providing taxol binding model; Worldwide MedicinalChemistry (WWMC) for providing O-Me-MMAF and auristatin-101; JudyLucas and members of the Oncology In vivo Group and vivarium staffin Pearl River for the animal studies. The authors acknowledge SeattleGenetics Inc. and Oxford BioMedica for access to technology and forhelpful discussions.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received May 19, 2015; revised August 11, 2015; accepted August 13, 2015;published OnlineFirst August 28, 2015.

References1. Hurvitz SA, Dirix L, Kocsis J, Bianchi GV, Lu J, Vinholes J, et al.

Phase II randomized study of trastuzumab emtansine versus trastu-zumab plus docetaxel in patients with human epidermal growthfactor receptor 2-positive metastatic breast cancer. J Clin Oncol2013;31:1157–63.

2. Verma S, Miles D, Gianni L, Krop IE, Welslau M, Baselga J, et al. Trastu-zumab emtansine for HER2-positive advanced breast cancer. N Engl J Med2012;367:1783–91.

3. Pro B, Advani R, Brice P, Bartlett NL, Rosenblatt JD, Illidge T, et al.Brentuximab vedotin (SGN-35) in patients with relapsed or refractorysystemic anaplastic large-cell lymphoma: results of a phase II study. J ClinOncol 2012;30:2190–6.

4. Younes A, Gopal AK, Smith SE, Ansell SM, Rosenblatt JD, Savage KJ,et al. Results of a pivotal phase II study of brentuximab vedotin forpatients with relapsed or refractory Hodgkin's lymphoma. J Clin Oncol2012;30:2183–9.

5. Maderna A,DoroskiM, SubramanyamC, Porte A, Leverett CA, Vetelino BC,et al. Discovery of cytotoxic dolastatin 10 analogues with N-terminalmodifications. J Med Chem 2014;57:10527–43.

6. Doronina SO, Toki BE, Torgov MY, Mendelsohn BA, Cerveny CG, ChaceDF, et al. Development of potent monoclonal antibody auristatin con-jugates for cancer therapy. Nat Biotechnol 2003;21:778–84.

7. Sapra P,DamelinM,Dijoseph J,Marquette K,Geles KG,Golas J, et al. Long-term tumor regression induced by an antibody–drug conjugate that targets5T4, an oncofetal antigen expressed on tumor-initiating cells. Mol CancerTher 2013;12:38–47.

8. Fruman DA, Rommel C. PI3K and cancer: lessons, challenges and oppor-tunities. Nat Rev Drug Discov 2014;13:140–56.

9. Chandarlapaty S, Sakr RA, Giri D, Patil S, Heguy A, Morrow M, et al.Frequent mutational activation of the PI3K–AKT pathway in trastuzumab-resistant breast cancer. Clin Cancer Res 2012;18:6784–91.

10. Nagata Y, Lan KH, Zhou X, Tan M, Esteva FJ, Sahin AA, et al. PTENactivation contributes to tumor inhibition by trastuzumab, and loss ofPTEN predicts trastuzumab resistance in patients. Cancer Cell 2004;6:117–27.

11. Jeannot V, Busser B, Brambilla E, Wislez M, Robin B, Cadranel J, et al. ThePI3K/AKT pathway promotes gefitinib resistance in mutant KRAS lungadenocarcinoma by a deacetylase-dependent mechanism. Int J Cancer2014;134:2560–71.

12. Andre F, O'Regan R, Ozguroglu M, Toi M, Xu B, Jerusalem G, et al.Everolimus for women with trastuzumab-resistant, HER2-positive,advanced breast cancer (BOLERO-3): a randomised, double-blind, place-bo-controlled phase 3 trial. Lancet Oncol 2014;15:580–91.

13. Sampath D, Fields C, Li G, Prior W, Parsons K, Friedman L, et al. AbstractS3–6: combination therapy of the novel PI3K inhibitor GDC-0941 anddual PI3K/mTOR inhibitor GDC-0980 with trastuzumab-DM1 antibody

drug conjugate enhances anti-tumor activity in preclinical breast cancermodels in vitro and in vivo. Cancer Res 2011;70:S3–6.

14. Hudis C, Swanton C, Janjigian YY, Lee R, Sutherland S, Lehman R, et al. Aphase 1 study evaluating the combination of an allosteric AKT inhibitor(MK-2206) and trastuzumab in patients with HER2-positive solid tumors.Breast Cancer Res 2013;15:R110.

15. Hurvitz SA, Dalenc F, Campone M, O'Regan RM, Tjan-Heijnen VC,Gligorov J, et al. A phase 2 study of everolimus combined withtrastuzumab and paclitaxel in patients with HER2-overexpressingadvanced breast cancer that progressed during prior trastuzumab andtaxane therapy. Breast Cancer Res Treat 2013;141:437–46.

16. Mallon R, Feldberg LR, Lucas J, Chaudhary I, Dehnhardt C, Santos ED, et al.Antitumor efficacy of PKI-587, a highly potent dual PI3K/mTOR kinaseinhibitor. Clin Cancer Res 2011;17:3193–203.

17. ShapiroGI, Bell-McGuinn KM,Molina JR, Bendell J, Spicer J, Kwak EL, et al.First-in-human study of PF-05212384 (PKI-587), a small-molecule, intra-venous, dual inhibitor of PI3K and mTOR in patients with advancedcancer. Clin Cancer Res 2015;21:1888–95.

18. Yared JA, Tkaczuk KH. Update on taxane development: new analogs andnew formulations. Drug Des Devel Ther 2012;6:371–84.

19. JoshiM, Liu X, Belani CP. Taxanes, past, present, and future impact on non-small cell lung cancer. Anticancer Drugs 2014;25:571–83.

20. Schiff PB, Horwitz SB. Taxol stabilizes microtubules in mouse fibroblastcells. Proc Natl Acad Sci U S A 1980;77:1561–5.

21. Perez EA. Microtubule inhibitors: differentiating tubulin-inhibiting agentsbased on mechanisms of action, clinical activity, and resistance. MolCancer Ther 2009;8:2086–95.

22. Gudena V, Montero AJ, Gluck S. Gemcitabine and taxanes in metastaticbreast cancer: a systematic review. Ther Clin Risk Manag 2008;4:1157–64.

23. Boghaert ER, Sridharan L, Khandke KM, Armellino D, Ryan MG, MyersK, et al. The oncofetal protein, 5T4, is a suitable target for antibody-guided anti-cancer chemotherapy with calicheamicin. Int J Oncol 2008;32:221–34.

24. Damelin M, Geles KG, Follettie MT, Yuan P, Baxter M, Golas J, et al.Delineation of a cellular hierarchy in lung cancer reveals an oncofetalantigen expressed on tumor-initiating cells. Cancer Res 2011;71:4236–46.

25. Chou TC.Drug combination studies and their synergy quantification usingthe Chou–Talalay method. Cancer Res 2010;70:440–6.

26. Ditchfield C, Johnson VL, Tighe A, Ellston R, Haworth C, Johnson T,et al. Aurora B couples chromosome alignment with anaphase bytargeting BubR1, Mad2, and Cenp-E to kinetochores. J Cell Biol 2003;161:267–80.

27. VanderPorten EC, Taverna P, Hogan JN, Ballinger MD, Flanagan WM,Fucini RV. The Aurora kinase inhibitor SNS-314 shows broad therapeuticpotential with chemotherapeutics and synergy with microtubule-targetedagents in a colon carcinoma model. Mol Cancer Ther 2009;8:930–9.






28. ThoreenCC, Chantranupong L, KeysHR,Wang T, GrayNS, Sabatini DM. Aunifying model for mTORC1-mediated regulation of mRNA translation.Nature 2012;485:109–13.

29. Pelletier J, Graff J, Ruggero D, Sonenberg N. Targeting the eIF4F translationinitiation complex: a critical nexus for cancer development. Cancer Res2015;75:250–63.

30. Bai RL, Pettit GR, Hamel E. Binding of dolastatin 10 to tubulin at a distinctsite for peptide antimitotic agents near the exchangeable nucleotide andvinca alkaloid sites. J Biol Chem 1990;265:17141–9.

31. Asnaghi L, Calastretti A, Bevilacqua A, D'Agnano I, Gatti G, Canti G, et al.Bcl-2 phosphorylation and apoptosis activated by damaged microtubulesrequire mTOR and are regulated by Akt. Oncogene 2004;23:5781–91.

32. Franz-Wachtel M, Eisler SA, Krug K, Wahl S, Carpy A, Nordheim A, et al.Global detection of protein kinase D-dependent phosphorylationevents in nocodazole-treated human cells. Mol Cell Proteomics 2012;11:160–70.

33. Nagano K, Shinkawa T, Mutoh H, Kondoh O, Morimoto S, Inomata N,et al. Phosphoproteomic analysis of distinct tumor cell lines in response tonocodazole treatment. Proteomics 2009;9:2861–74.

34. Kim S, Coulombe PA. Emerging role for the cytoskeleton as an organizerand regulator of translation. Nat Rev Mol Cell Biol 2010;11:75–81.

35. Coldwell MJ, Cowan JL, Vlasak M, Mead A, Willett M, Perry LS, et al.Phosphorylation of eIF4GII and 4E-BP1 in response to nocodazole treat-ment: a reappraisal of translation initiation during mitosis. Cell Cycle2013;12:3615–28.

36. Blower MD, Feric E, Weis K, Heald R. Genome-wide analysis demon-strates conserved localization of messenger RNAs to mitotic microtu-bules. J Cell Biol 2007;179:1365–73.

37. Lecuyer E, Yoshida H, Parthasarathy N, Alm C, Babak T, Cerovina T, et al.Global analysis of mRNA localization reveals a prominent role in orga-nizing cellular architecture and function. Cell 2007;131:174–87.

38. Sharp JA, Plant JJ, Ohsumi TK, Borowsky M, Blower MD. Functionalanalysis of the microtubule-interacting transcriptome. Mol Biol Cell2011;22:4312–23.

39. Skop AR, Liu H, Yates J 3rd, Meyer BJ, Heald R. Dissection of themammalian midbody proteome reveals conserved cytokinesis mechan-isms. Science 2004;305:61–6.

40. JangCY, KimHD,ZhangX,Chang JS, Kim J. Ribosomal protein S3 localizeson themitotic spindle and functions as amicrotubule associated protein inmitosis. Biochem Biophys Res Commun 2012;429:57–62.

41. Wallin JJ, Guan J, Prior WW, Lee LB, Berry L, Belmont LD, et al. GDC-0941,a novel class I selective PI3K inhibitor, enhances the efficacy of docetaxel inhuman breast cancer models by increasing cell death in vitro and in vivo.Clin Cancer Res 2012;18:3901–11.

42. Krop I,Wolff A,Winer E,Miller K, Park B,Ware J, et al. Abstract P6–15–02: aphase Ib study evaluating safety, tolerability, pharmacokinetics (PK), andactivity of the phosphoinositide-3 kinase (PI3K) inhibitor GDC-0941 incombination with trastuzumab-MCC-DM1 (T-DM1) in patients withadvanced HER2-positive breast cancer. Cancer Res 2010;70:P6-15-02.

43. MartinM,Dewar J, Albanell J, Limentani S,Chang J, StrasakA, et al. AbstractP4-12-07: Neoadjuvant trastuzumab emtansine and docetaxel, with orwithout pertuzumab, in patients with HER2-positive early-stage breastcancer: Results from a phase 1b/2a study. Cancer Res 2013;73:P4-12-07.

44. Ansell SM, Younes A, Connors JM, Gallamini A, Kim WS, Friedberg JW,et al. Phase 3 study of brentuximab vedotin plus doxorubicin, vinblastine,and dacarbazine (AþAVD) versus doxorubicin, bleomycin, vinblastine,and dacarbazine (ABVD) as front-line treatment for advanced classicalHodgkin lymphoma (HL): Echelon-1 study. ASCO Meeting Abstracts2014;32:TPS8613.

45. Franceschini A, Szklarczyk D, Frankild S, Kuhn M, Simonovic M, Roth A,et al. STRING v9.1: protein-protein interaction networks, with increasedcoverage and integration. Nucleic Acids Res 2013;41:D808–15.


Shor et al.




Published OnlineFirst August 28, 2015.Clin Cancer Res Boris Shor, Jennifer Kahler, Maureen Dougher, et al. TaxanesConjugate in Combination with PI3K/mTOR inhibitors or

Drug−Enhanced Antitumor Activity of an Anti-5T4 Antibody

Updated version

10.1158/1078-0432.CCR-15-1166doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2015/10/02/1078-0432.CCR-15-1166.DC1Access the most recent supplemental material at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://clincancerres.aacrjournals.org/content/early/2015/11/03/1078-0432.CCR-15-1166To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-15-1166

http://clincancerres.aacrjournals.org/content/suppl/2015/10/02/1078-0432.CCR-15-1166.DC1

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/early/2015/11/03/1078-0432.CCR-15-1166


Enhanced Antitumor Activity of an Anti-5T4 Antibody Drug … · 03/11/2015 · Experimental...

Documents

Transcript of Enhanced Antitumor Activity of an Anti-5T4 Antibody Drug … · 03/11/2015 · Experimental...