Uncialamycin as a novel payload for antibody drug conjugate (ADC) based targeted cancer therapy

Article abstract of DOI:10.1016/j.bmcl.2018.12.021

Uncialamycin analogs were evaluated as potential cytotoxic agents in an antibody-drug conjugate (ADC) approach to treating human cancer. These analogs were synthesized using Hauser annulations of substituted phthalides as a key step. A highly potent uncialamycin analog 3c with a valine-citrulline dipeptide linker was conjugated to an anti-mesothelin monoclonal antibody (mAb) through lysines to generate a meso-13 conjugate. This conjugate demonstrated subnanomolar potency (IC50 = 0.88 nM, H226 cell line) in in vitro cytotoxicity experiments with good immunological specificity to mesothelin-positive lung cancer cell lines. The potency and mechanism of action of this uncialamycin class of enediyne antitumor antibiotics make them attractive payloads in ADC-based cancer therapy.

Full text of DOI:10.1016/j.bmcl.2018.12.021

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Uncialamycin as a novel payload for antibody drug conjugate (ADC) based
targeted cancer therapy
⁎
Naidu S. Chowdaria, , Chin Pana, Chetana Raoa, David R. Langleyb, Prasanna Sivaprakasamc,
Bilal Sufia, Daniel Derwina, Yichong Wanga, Eilene Kwoka, David Passmorea,
Vangipuram S. Rangana, Shrikant Deshpandea, Pina Cardarellia, Gregory Vitec, Sanjeev Gangwara
a Bristol-Myers Squibb Research & Development, 700 Bay Road, Redwood City, CA 94063, USA
b Bristol-Myers Squibb Research & Development, 5 Research Parkway, Wallingford, CT 06492, USA
c Bristol-Myers Squibb Research & Development, PO Box 4000, Princeton, NJ 08543, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
Uncialamycin analogs were evaluated as potential cytotoxic agents in an antibody-drug conjugate (ADC) ap-
proach to treating human cancer. These analogs were synthesized using Hauser annulations of substituted
phthalides as a key step. A highly potent uncialamycin analog 3c with a valine-citrulline dipeptide linker was
conjugated to an anti-mesothelin monoclonal antibody (mAb) through lysines to generate a meso-13 conjugate.
This conjugate demonstrated subnanomolar potency (IC50 = 0.88 nM, H226 cell line) in in vitro cytotoxicity
experiments with good immunological specificity to mesothelin-positive lung cancer cell lines. The potency and
mechanism of action of this uncialamycin class of enediyne antitumor antibiotics make them attractive payloads
in ADC-based cancer therapy.
Antibody-drug conjugates
Uncialamycin
Peptide linker
Lung cancer
Antibody–drug conjugates (ADCs) offer the possibility of delivering
a cytotoxic payload to cancer cells by exploiting their cell surface an-
tigens while sparing normal cells that do not express the target anti-
gens.1,2 Interest in ADC technology has expanded with approval of
ADCETRIS® (brentuximab vedotin)3, KADCYLA® (ado-trastuzumab
emtansine),4 MYLOTARG™ (gemtuzumab ozogamicin)5 and BESPO-
NSA® (inotuzumab ozogamicin).6
based on the size of the enediyne core (see Fig 1).
Neocarzinostatin19 and kedarcidin belong to the nine-membered
enediyne group, and calicheamicin, dynemicin, esperamicins, name-
namicin, and uncialamycin belong to ten-membered enediyne
group.20,21 All of these enediynes contain a core consisting of two
acetylenic groups conjugated to a double bond within the 9- or 10-
membered carbocycle. As a consequence of this structural feature, these
compounds share a common mode of action. Bergman cyclization of the
enediyne carbocycle produces a transient benzenoid diradical.18 When
positioned within the minor groove of DNA, this diradical abstracts
hydrogen atoms from the deoxyribose backbone of duplex DNA. These
DNA-centered radicals can then cause interstrand crosslinks and react
with molecular oxygen leading ultimately to DNA double-strand clea-
vage and apoptosis.22 Herein we report our binding model of a transient
benzenoid diradical of uncialamycin in a double-helical B-DNA. The
uncialamycin (and its benzenoid diradical intermediate) models are
based on our previously published dynemicin bound DNA models.23
Uncialamycin intercalates into the DNA placing the enediyne group in
the minor groove while the anthraquinone moiety protrudes into the
major groove. The model is very stable over 1200 nsec molecular dy-
namics (MD) simulation.24,25 Multiple H-bonds observed between un-
cialamycin and DNA anchor uncialamycin into the intercalation cleft.
The development of a new ADC involves the optimization of the
antigen target,7 the antibody,8 the linker,9–11 and the cytotoxic pay-
load. Engineered antibodies that bind selectively to a specific tumor
antigen on the surface of cancer cells provide a wider therapeutic
window for an ADC compared to the parent cytotoxic payload alone.
While there is a large number of ADCs currently in development, cy-
totoxic payloads have generally been limited to a few natural products
(auristatin, maytansinoid and calcheamicin).1 Recently camp-
tothecin,12 tubulysin,13 and benzodiazepine14 based ADCs have ad-
vanced into various stages of clinical trials. Finding new cytotoxic
payloads with novel mechanisms is highly desirable to expand the
scope of ADC targeted cancer therapy.15
Our group recently focused on enediynes16–18 as a payload class for
ADCs due to their higher potency and broader anticancer activity across
cell lines. Enediyne natural products are classified into two subgroups
⁎ Corresponding author.
E-mail address: naidu.chowdari@bms.com (N.S. Chowdari).
https://doi.org/10.1016/j.bmcl.2018.12.021
Received 26 September 2018; Received in revised form 1 December 2018; Accepted 10 December 2018
0960-894X/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Chowdari,N.S.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2018.12.021

N.S. Chowdari et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 1. Structures of payloads used in FDA approved drugs: Calicheamicin (MYLOTARG and BESPONSA), MMAE (ADCETRIS), and maytansine (KADCYLA).
Uncialamycin is a relatively new enediyne that was isolated from a
Streptomyces strain found on the lichen Cladonia uncialis.21 It was re-
ported to exhibit potent antitumor activity. Since the natural abun-
dance of uncialamycin was limited, a synthetic route was developed by
the Nicolaou group, which confirmed uncialamycin’s structure and al-
lowed production of sufficient quantities for biological evaluations.26–28
High potency and structural simplicity of uncialamycin relative to other
enediynes prompted us to consider it as a viable payload. Herein we
describe the synthesis of a highly potent uncialamycin analog capable
of DNA double strand cleavage as a novel payload for use as an ADC.29
A protease-cleavable valine-citrulline dipeptide linker9 with a mal-
eimide group was initially conjugated to the mesothelin antibodies of
interest. Mesothelin is a 40 kDa protein present on normal mesothelial
cells and overexpressed in several human tumors, including mesothe-
lioma, ovarian, non-small cell lung cancer, and pancreatic adenocarci-
noma.
We prepared uncialamycin 3a according to reported procedure.26
Attempts to attach a dipeptide linker to the NH of uncialamycin in-
dicated that this amino group was not reactive under various peptide
coupling conditions. We then introduced an NH2 group onto un-
cialamycin by using the substituted phthalides 9 and 11 in Hauser
annulation to provide compounds 3b and 3c, respectively (Scheme 1).
3b was tested for various cancer cell lines to confirm its activity
(Table 1). 3b has shown excellent activity across the cell lines. More-
over, comparison to the approved ADC payload, monomethyl auristatin
Fig. 2. Binding model of uncialamycin in DNA. The double-helical DNA surface
E (MMAE) indicated improved activity for uncialamycin 3b. It was
is colored by the electrostatic potential and ligand is shown as green carbon
2–60 fold more potent compared to MMAE.
sticks. The two black dots represent diradicals of benzenoid intermediate from
the Bergman cyclization. This figure was prepared using Pymol (The Pymol
Molecular Graphics system, v.1.6.1.0 Schrödinger, LLC).
Cyanophthalides were prepared as shown in Scheme 2. Commer-
cially available aminophthalide 4 was protected with a 2,2,2-tri-
chloroethoxycarbonyl (Troc) group and subsequent bromination and
hydrolysis provided compound 6. Cyanation using potassium cyanide
provided cyanohydrin 7. Intramolecular esterification using dicyclo-
hexyl carbodiimide (DCC) provided ester 8. The Troc protecting group
was removed using zinc in acetic acid to provide aminophthalide 9.
Reductive amination of 9 with sodiumtriacetoxyborohydride in the
presence of the t-butoxycarbonyl (Boc) protected aminoacetaldehyde
provided amine 10a in which the Boc was removed using tri-
fluoroacetic acid. The monomethoxytrityl (MMT) protecting group was
The intermediate diradical is well positioned for hydrogen abstraction
from the sugar moieties on either strand of the duplex DNA. Fig. 2
summarizes the key events from our theoretical mechanistic studies of
uncialamycin binding to DNA.
Although some enediynes taken into the clinic have shown sig-
nificant toxicity due to poor tumor selectivity, a polymer conjugate of
neocarzinostatin has been successfully developed and marketed for use
against hepatoma.19
2

N.S. Chowdari et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Scheme 1. Synthesis of uncialamycins through Hauser annulationa aReagents and conditions: (a) LiHMDS; (b) 3HF.Et3N.
Table 1
conjugate the linker-payload using a Michael reaction through the
maleimide group. These conjugates displayed low levels of aggregation
(< 10%), a common obstacle for more lipophilic ADC payloads, high-
lighting a physiochemical advantage of the uncialamycin class. Linker-
payloads 13 and 15 were also conjugated to a CD70 targeting antibody
(2H5) that served as an isotype non-binding control mAb. CD70 is a
tumor necrosis factor (TNF)-like type II integral membrane protein and
is transiently expressed on activated T- and B-lymphocytes.32 It is
highly expressed in renal cell carcinoma cell lines (786-O) and mini-
mally expressed in H226 cell lines.33 ADCs synthesized from the CD70
In vitro inhibition of proliferation by MMAE and uncialamycin 3b across a
panel of tumor cell lines (IC50 in nM).
Compound A-2780
CCRF-CEM A549
H-2087
HCT-116 MDA-MB231
MMAE
0.199
0.003
0.180
0.014
0.176
0.076
0.071
0.037
0.175
0.007
0.38
3b
0.085
installed to provide compound 11.
The compounds prepared above (3a-c) were tested in a H226 lung
cancer cell line. These compounds had IC50 values of 3.8 nM for 3a,
1.0 nm for 3b, and 0.98 nM for 3c. Their lower potency in the H226 cell
line compared to other cell lines could be due to poor cell membrane
permeability. Although compounds 3b and 3c have similar biological
activity, compound 3b has poor chemical reactivity for appending
linkers. We subsequently chose compound 3c to append linkers for
evaluation as ADCs. We prepared a protease-cleavable dipeptide linker-
payload 13 with a self-immolating spacer by reaction of linker carbo-
nate 12 with the uncialamycin analog 3c. A non-cleavable linker-pay-
load 15 was prepared by reaction of linker-OSu 14 with 3c (Scheme 3).
Linker-payload compounds 13 and 15 were conjugated to a 6A4
antibody which targets the mesothelin antigen.30 Mesothelin is highly
expressed in mesotheliomas, pancreatic, ovarian and lung cancer cells.
Some of the lysines present on the antibody were modified with 2-
iminothiolane to introduce thiol groups,31 which were used to
antibody provided
a better payload drug-to-antibody-ratio (DAR)
compared to the 6A4 antibody. ADCs prepared from linker-payload 13
were subjected to cathepsin B enzyme digestion to release the payload
3c upon self-immolation of spacer (Scheme 4). Cathepsin B is a cysteine
protease found in all mamalian cell lysosomes.34 Meso-13 ADC showed
highly specific cytotoxic activity on H226 (IC50 = 0.88 nM) cell lines.
IC50 of Meso-13 ADC is slightly lower than the corresponding un-
conjugated payload 3c (0.98 nM). This is attributed to the poor per-
meability of unconjugated payload 3c and highly efficient lysosomal
cleavage of ADC. Whereas CD70-13 ADC showed highly specific cyto-
toxic activity on renal cell carcinoma cell lines (786-O)
(IC50 = 0.76 nM). ADC 15 was not active on either cell lines. This
might be due to the non-cleavable nature of the linker employed
(Table 2). This result is contrary to the KADCYLA ADC where the non-
cleavable linker has worked well for a herceptin target.4
In conclusion, the new uncialamycin analogs conjugated to
Scheme 2. Synthesis of phthalides 9 and 11a. aReagents and conditions: (a) Troc-Cl, Pyridine, rt, 58%; (b) NBS, CCl4, reflux; (c) H2O, reflux, 76% over two steps;
(d) KCN, H2O; (e) DCC, Ethyl acetate, rt, 78% over two steps; (f) Zn, AcOH, 66%; (g) Na(OAc)3BH; (h) TFA, 78% over two steps (i) MMT-Cl, TEA, 48%.
3

N.S. Chowdari et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Scheme 3. Synthesis of linker-payload compounds 13 and 15.
antibodies targeting mesothelin were used to generate ADCs of both
cleavable and non-cleavable variations that demonstrated specific ac-
tivity against a mesothelin-expressing cell line in vitro. The on-target
mechanism of action was confirmed both through the use of a control
ADC as well as through the use of cell lines not expressing the target
antigen. Further optimization of ADCs and efficacy evaluation studies in
vivo are underway.
Scheme 4. Uncialamycin ADC and its proteolytic cleavage to release payload 3c.
4

N.S. Chowdari et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 2
References
In vitro inhibition of proliferation by uncialamycin ADCs on H226 and 786-O
cell lines.
1. Goeij BE, Lambert JM. Curr Opin Immnol. 2016;40:14.
2. Polakis P. Pharmcol Rev. 2016;68:3.
ADC
DAR
IC50 (nM)
3. Senter PD, Sievers EL. Nat Biotechnol. 2012;30:631.
4. Verma S, Miles D, Gianni L, et al. N Engl J Med. 2012;367:1783.
5. Amadori S, Venditti A, Voso MT, et al. J Clin Oncol. 2016;34:972.
6. Lamb YN. Drugs. 2017:1.
H226
786-O
Meso-13
CD70-13
Meso-15
CD70-15
1.2
1.4
1.0
4.8
0.88
33.37
0.76
7. Senter PD. Curr Opin Chem Biol. 2009;13:235.
12.37
39.60
28.24
8. Hamblett KJ, Senter PD, Chace DF, et al. Clin Cancer Res. 2004;10:7063.
9. Dubowchik GM, Radia S, Mastalerz H, et al. Bioorg Med Chem Lett. 2002;12:1529.
10. Alley SC, Benjamin DR, Jeffrey SC, et al. Bioconjug Chem. 2008;19:759.
11. Junutula JR, Flagella KM, Graham RA, et al. Clin Cancer Res. 2010;16:4769.
12. Cardillo TM, Sharkey RM, Rossi SL, Arrojo R, Mostafa AA, Goldenberg DM. Clin
Cancer Res. 2017;23:3405.
27.43
19.68
Author contributions
13. Cong, Q, Cheng, H, Gangwar, S., US 8,980,824.
14. Jeffrey SC, Burke PJ, Lyon RP, et al. Bioconjug Chem. 2013;24:1256.
15. Verma VA, Pillow TH, DePalatis L, et al. Bioorg Med Chem Lett. 2015;25:864.
16. Shao R-G, Zhen YS. Med Chem. 2008;8:123.
All authors have given approval to the final version of the manu-
script.
17. Joshi MC, Rawat DS. Chem Biodivers. 2012;9:459.
18. Doyle TW, Borders DB. Enediyne antibiotics as antitumor agents. New York: Marcel-
Dekker; 1995:2.
Notes
The authors declare no competing financial interests.
19. Maeda H, Edo K, Ishida N. Neocarzinostatin: the past, present, and future of an antic-
ancer drug. New York: Springer; 1997.
20. Myers AG, Fraley ME, Tom NJ, Cohen SB, Madar DJ. Chem Biol. 1995;2:33.
21. Davies J, Wang H, Taylor T, Warabi K, Huang X, Andersen RJ. Org Lett. 2005;7:5233.
22. Kennedy DR, Ju J, Shen B, Beerman TA. PNAS. 2007;104:17632.
23. Langley DR, Doyle TW, Beveridge DL. J Am Chem Soc. 1991;113:4395.
24. Phillips JC, Braun R, Wang W, et al. J Comp Chem. 2005;26:1781.
25. Langley DR. J Biomol Struct Dyn. 2012;16:487.
Acknowledgment
26. Nicolaou KC, Zhang H, Chen JS, Crawford JJ, Pasunoori L. Angew Chem.
2007;119:4788.
We thank our colleagues in the lead evaluation group for their as-
sistance in cellular proliferation assays and Catherine Bolger for re-
viewing the manuscript.
27. Nicolaou KC, Chen JS, Zhang H, Montero A. Angew Chem. 2008;47:185.
28. Nicolaou KC, Wang Y, Lu M, et al. J Am Chem Soc. 2016;138:8235.
29. Chowdari, NS, Gangwar, S, Sufi, B, US 9,156,850 and US 8,709,431.
30. Terret, JA, Pogue, SL, Toy, K, Yang, L, Rao, CR, Chen, B. US 8,425,904.
31. Weber C, Reiss S, Langer K. Int J Pharm. 2000;211:67.
32. Grewal IS. Expert Opin Ther Targets. 2008;12:341.
Appendix A. Supplementary data
33. Jacobs J, Deschoolmeester V, Zwaenepoel K, et al. Pharmacol Ther. 2015;155:1.
34. Kirschke H, Barrett AJ, Rawlings ND. Sheterline P, ed. Protein Profile – Proteinases 1:
Lysosomal Cysteine Properties. London: Academic; 1995 1587.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2018.12.021.
5