A Fragmenting Hybrid Approach for Targeted Delivery of Multiple Therapeutic Agents to the Malaria Parasite

Article abstract of DOI:10.1002/cmdc.201100002

Hybrid drugs with a twist: The coupling of iron(II)-promoted trioxolane ring scission with a β-elimination reaction enables the targeted delivery of multiple drug activities to the malaria parasite. A prototypical fragmenting hybrid (shown) comprises an i

Full text of DOI:10.1002/cmdc.201100002

DOI: 10.1002/cmdc.201100002
A Fragmenting Hybrid Approach for Targeted Delivery of Multiple
Therapeutic Agents to the Malaria Parasite
Sumit S. Mahajan,[a] Edgar Deu,[b] Erica M. W. Lauterwasser,[a] Melissa J. Leyva,[c] Jonathan A. Ellman,[c]
Matthew Bogyo,*[b] and Adam R. Renslo*[a]
Artemisinin combination therapies (ACT) represent the current
standard of care in the treatment of uncomplicated malaria.
The widespread adoption of ACT has been motivated by a
desire to minimize the emergence of drug resistance and to
address the problem of recrudescence associated with artemi-
sinin monotherapy.[1–4] We set out to explore a single-molecule
‘fragmenting hybrid’ strategy in which an artemisinin-like per-
oxide is employed to deliver a partner drug, only upon activa-
tion by ferrous iron in the parasite. In principle, iron(II)-depen-
dent drug delivery from a fragmenting hybrid could alleviate
unwanted off-target bioactivities of the partner drug, which
would be inactive in its hybrid form.
Our design for fragmenting hybrids was inspired by earlier
work demonstrating that antimalarial 1,2,4-trioxolanes, includ-
ing the investigational agent arterolane (1, Figure 1),[5] undergo
Scheme 1. a) Mechanism of iron(II)-promoted breakdown of the 1,2,4-trioxo-
lane ring in arterolane (1). b) Proposed unraveling of a fragmenting hybrid
(2) to release an amine-bearing drug species via b-elimination from a retro-
Michael substrate (6).
thought to be initiated by free heme, a byproduct of hemo-
globin degradation in the parasite digestive vacuole (DV). Be-
cause heme concentrations are enormous (estimated at
0.4m)[9] in the parasite DV and significantly lower (~10À16 m)[10]
in human plasma, we reasoned that the presence of this reac-
Figure 1. Structure of the investigational antimalarial arterolane (1) and an
iron(II)-targeted fragmenting hybrid (2).
tive species in parasites could be exploited for selective drug
iron(II)-promoted ring opening to afford both reactive carbon-
centered radical species—the presumed agent responsible for
parasite toxicity—and carbonyl containing byproducts (e.g., 5,
Scheme 1a).[6–8] In parasites, scission of the trioxolane ring is
delivery. Thus, to achieve parasite-targeted hybrid fragmenta-
tion, we embedded a masked retro-Michael linker within the
1,2,4-trioxolane ring system (Figure 1). In the hypothesized ac-
tivation–release sequence, free ferrous iron heme in the para-
site DV mediates opening of the trioxolane ring, which un-
masks the carbonyl function of the retro-Michael linker (6),
leading to release of the second drug species via a b-elimina-
tion reaction (Scheme 1b). The release of free partner drug
from the linker distinguishes this approach from conventional
covalent hybrids,[11] while the introduction of the retro-Michael
linker greatly expands the scope of possible conjugation
chemistry as compared to other peroxidic prodrugs that re-
quire drug conjugation at a carbonyl function.[12–14]
[a] Dr. S. S. Mahajan,+ Dr. E. M. W. Lauterwasser, Prof. Dr. A. R. Renslo
Department of Pharmaceutical Chemistry, Small Molecule Discovery Center
University of California, San Francisco, Mission Bay Campus
1700 4th Street, San Francisco, CA 94158 (USA)
Fax: +(1)415-514-4507
E-mail: adam.renslo@ucsf.edu
[b] Dr. E. Deu,+ Prof. Dr. M. Bogyo
Department of Pathology, Stanford School of Medicine
300 Pasteur Drive, Stanford, CA 94305 (USA)
Fax. (+1)650-725-7424
Conclusive demonstration of dual action for a hybrid drug
species is challenging since simple potency comparisons are
often ambiguous and difficult to interpret. Thus, although frag-
menting hybrids of mefloquine and primaquine could be read-
ily prepared, it was not trivial to establish that the quinoline
species had been released from these hybrids, since the trioxo-
lane moiety itself is a potent antimalarial agent. Therefore, we
turned to studies of fragmenting hybrids containing a protease
inhibitor, reasoning that delivery of such a species could be
confirmed using activity-based probes of protease activity.
E-mail: mbogyo@stanford.edu
[c] M. J. Leyva, Prof. Dr. J. A. Ellman
Department of Chemistry, University of California, Berkeley
Berkeley, CA 94720-1460 (USA)
Fax: +(1)510-642-8369
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201100002.
Re-use of this article is permitted in accordance with the Terms and Condi-
tions set out at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN) 1860–
7179/homepage/2452_onlineopen.html
ChemMedChem 2011, 6, 415 – 419
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
415

MED
Dipeptidyl aminopeptidase 1 (DPAP1)[15] is an essential cysteine
protease involved in the latter stages of hemoglobin degrada-
tion in the parasite DV. Inhibitors of this exo-protease are well
suited for study of fragmenting hybrids since relevant activity-
based probes are available and inhibitor potency is dramatical-
ly attenuated when the N-terminal amino group is acylated or
otherwise blocked.[16,17] Herein, we demonstrate proof of prin-
ciple for the fragmenting hybrid approach by demonstrating
the successful release of the potent and irreversible DPAP1 in-
hibitor ML4118S[15] from fragmenting hybrid 8 (Figure 2). We
Scheme 2. Synthesis of fragmenting hybrid 8 and nonfragmenting control
compound 9. Reagents and conditions: a) MeONH2, pyridine, RT, 48 h, 83%;
b) O3, cyclohexa-1,4-dione, pentane/CH2Cl2 (3:2), 0oC, 1 h, 33%; c) mCPBA,
NaHCO3, CH2Cl2, RT, 48 h, 56%; d) N-(3-aminopropyl)-2-pyrrolidinone, tolu-
ene, 50oC, 5 h, 64%; e) p-NO2PhOC(O)Cl, Et3N, DMAP, CH2Cl2, RT, 16 h, 96%;
f) ML4118S, DMF, DMAP, RT, 16 h, 38%; g) Dess–Martin periodinane, CH2Cl2,
RT, 30 min; h) 1m KMnO4, 5% NaH2PO4, tBuOH, RT, 30 min, 74% (two steps);
i) ML4118S, HATU, HOBt, DMF, DIEA, RT, 2 h, 48%. Abbreviations: 2-(7-Aza-
1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate
(HATU); meta-Chloroperoxybenzoic acid (mCPBA); N,N-Diisopropylethyl-
amine (DIEA); 4-Dimethylaminopyridine (DMAP); N,N-Dimethylformamide
(DMF); 1-Hydroxybenzotriazole (HOBt).
this material could then be joined to ML4118S via a carbamate
linkage to provide hybrid 8 in 38% yield. Similar coupling reac-
tions with the more reactive and less hindered amines present
in the antimalarial drugs primaquine and mefloquine proceed-
ed in yields of >80% and >55%, respectively (unpublished re-
sults). The ability to form fragmenting hybrids from a variety of
primary and secondary amines is significant because most of
the agents currently used in malaria combination therapies
possess such functional handles. Also prepared from 13 was
the crucial amide-linked control compound 9, which can be ac-
tivated by ferrous iron but cannot release free ML4118S be-
cause b-elimination is precluded. Hence, two-step oxidation of
alcohol 13 to the corresponding carboxylic acid (74% overall
yield), followed by HATU-mediated coupling to ML4118S,
afforded conjugate 9 in 48% yield.
Figure 2. Structures of the carbamate-linked fragmenting hybrid 8 and an
amide-linked congener 9. Compound 9 is an important control compound
that can be activated by ferrous iron but cannot release free ML4118S. Both
8 and 9 contain the irreversible DPAP1 inhibitor ML4118S, the a-keto posi-
tion of which is not configurationally stable.[15] R= CH2CH2CH2-N-pyrrolidi-
none.
With hybrid 8 in hand, validation of the proposed iron(II)-
promoted fragmentation chemistry was undertaken, employ-
ing LC/MS for the detection of reaction products (Scheme 1).
Using reaction conditions that are standard in the field,[21]
0.3 mmol of hybrid 8 was treated with 100 equivalents of fer-
rous bromide in a 1:1 mixture of water and acetonitrile. These
conditions afforded rapid fragmentation of the trioxolane ring
and clean formation of the expected products: lactone 14 and
retro-Michael substrate 15 (Scheme 3). The subsequent b-elimi-
nation reaction of 15 to generate 16 and ML4118S was also
observed and was slower than the initial reaction with iron
(see Supporting Information). In control experiments without
ferrous bromide, hybrid 8 remained intact throughout the ex-
perimental time period. These results thus confirm that trioxo-
lane scission and hybrid fragmentation proceeds as predicted,
effecting iron-dependent release of ML4118S.
anticipated that ML4118S in its hybrid form 8 should possess
very weak DPAP1 inhibitory activity since the linkage is made
at the terminal amino function. Therefore, the observation of
DPAP1 inhibition by hybrid 8 in parasites would indicate suc-
cessful delivery of active ML4118S from the hybrid. We could
measure the inhibitory activity in parasite extracts using FY01
(Figure 3a), a fluorescently labeled activity-based probe that
specifically targets DPAPs.[18] Although not a viable drug candi-
date for various reasons, hybrid 8 has proven to be a useful
tool for validating the fragmenting hybrid concept in live para-
sites.
Compounds 8 and 9 were prepared from the key trioxolane
intermediate 13, which was in turn prepared using established
methods of trioxolane synthesis.[19,20] Briefly, Griesbaum co-ozo-
nolysis of 2-adamantanone methyl oxime and cyclohexa-1,4-
dione afforded ketone 11, which following Baeyer–Villiger oxi-
dation provided lactone 12 on a gram scale (Scheme 2). Ring
opening of 12 with a primary amine provided alcohol 13, and
With the fragmentation chemistry validated in vitro, we
turned to the study of 8, 9, 13, and ML4118S in the context of
cultured intra-erythrocytic Plasmodium falciparum parasites.
416
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 415 – 419

Figure 3. Validation of the fragmenting hybrid concept in live parasites. a) Activities of ML4118S and 8, 9, 13 towards DPAP1. Parasite lysates were treated for
30 min in acetate buffer with different concentrations of hybrid 8, trioxolane 13, ML4118S, or the control hybrid 9. Residual DPAP1 activity was labeled with
1 mm of FY01 for 1 h and visualized by fluorescent scan of SDS-PAGE gels. b) Potency against P. falciparum parasites in culture: ML4118S (*); 8 (^); 9 (&);
13 (~). Ring stage parasites were treated with increasing concentrations of the indicated compound and cultured for ~75 h. Parasitemia was quantified by
FACS analysis and fitted to a dose–response curve. The EC50,Pot values are reported in Table 1. c) Kinetics of DPAP1 inhibition in vivo. A synchronous culture of
P. falciparum at trophozoite stage was treated with 50 nm of the indicated compound or DMSO. After 0.5–6 h of treatment, parasites were separated from the
erythrocytes by saponin lysis, and the residual DPAP1 activity was labeled with 1 mm of FY01 in acetate buffer containing 1% nonidet P40.
nor the hybrid control 9 blocked DPAP1 activity, even at the
highest concentration studied (10 mm). Thus, the DPAP1 activi-
ty of ML4118S is blocked in the hybrid form (9). Hybrid 8 par-
tially inhibited DPAP1 at the highest concentration (10 mm),
but was at least 100-fold less potent than ML4118S itself, pre-
sumably due to capping of its free amine function in the
hybrid form (Table 1). An authentic sample of the linker side
product 16 (see scheme S1 in the Supporting Information for
the synthesis) had no effect on parasite viability or DPAP1
inhibition (Table 1; figure S3 in the Supporting Information).
We next evaluated the antimalarial activities of the com-
pounds using intra-erythrocytic P. falciparum parasites (Fig-
ure 3b and Table 1). Trioxolane intermediate 13 and hybrid
control 9 exhibited antimalarial activities in the mid-nanomolar
range (EC50,Pot =29 nm and 52 nm, respectively). The potent ac-
tivity of 9 confirms action via the trioxolane moiety, since no
inhibition of DPAP1 is conferred by this compound in live para-
Scheme 3. Observed reaction products (LC/MS) following treatment of frag-
menting hybrid 8 with excess ferrous bromide. R= CH2CH2CH2-N-pyrrolidi-
none.
We first measured inhibition of DPAP1 by treating parasite ly-
sates for 30 min with increasing concentrations of test com-
pounds followed by labeling of residual DPAP1 activity with
FY01 (Figure 3a). Free ML4118S fully inhibited DPAP1 with an
IC50 value of 70 nm. As expected, neither trioxolane 13 alone
ChemMedChem 2011, 6, 415 – 419
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
417

MED
ple, the slow release of a partner drug in such hybrids could
complement the rapid action of the trioxolane moiety, much
as the ACT strategies seek to combine a longer-acting partner
drug with a rapid-acting artemisinin. As demonstrated with
hybrid 8, intrinsic bioactivities of the partner drug can be
masked in the hybrid form, raising the possibility that unde-
sired on/off-target effects of known drugs might similarly be
attenuated using this approach. Antimalarial agents like prima-
quine and amodiaquine that exhibit systemic toxicities might
be more safely administered in the form of a fragmenting
hybrid, as this would limit systemic exposure to free partner
drug. More speculatively, agents conferring irreversible target
inhibition or polypharmacology might be safely delivered
using a fragmenting hybrid approach in which systemic expo-
sure to free drug is avoided. Currently, we are exploring next-
generation fragmenting hybrids that overcome limitations of
the initial prototypical systems described herein.
Table 1. DPAP1 Inhibitory activities, antimalarial activities, and rates of
hybrid fragmentation in vitro and in parasite cultures.[a]
Compd
IC50[b] [nM]
DPAP1
EC50,Pot[c] [nM]
P. falciparum
t = [h]
1
2
in vitro[d]
in vivo[e]
ML4118S
8
9
13
16
70 (13)
10000
>10000
>10000
>10000
5.2 (0.4)
4.0 (0.2)
52 (7)
n/a
9
n/a
n/a
n/a
n/a
1.5 (0.25)
n/a
n/a
n/a
29 (13)
>10000
[a] n/a: not applicable. The standard deviation for measured values is
shown in parentheses. [b] Half maximal inhibition of DPAP1 in parasite ly-
sates after 30 min treatment with inhibitor. DPAP1 activity was measured
using the FY01 probe. [c] Antimalarial potency measured by treating a
culture of P. falciparum at ring stage with increasing concentrations of
compound. The decrease in parasitemia was quantified by FACS analysis
and fitted to a dose–response curve. [d] Half-life for the release of
ML4118S from hybrid species 8 as measured in vitro by LC/MS spectrom-
etry. [e] Half-life for the release of ML4118S from hybrid 8 in living para-
sites. This value was estimated based on the kinetics of DPAP1 inhibition
observed in culture and the independently determined second-order rate
constant for inhibition of DPAP1 by ML4118S in vitro.
Acknowledgements
ARR acknowledges the support of the Sandler Foundation (USA)
and the Bill and Melinda Gates Foundation (USA). JAE acknowl-
edges the support of the US National Institutes of Health (NIH)
(GM54051). MB also acknowledges the support of the US NIH
(AI078947 and EB005011).
sites (see below). The observation of trioxolane-based activity
in 9 also rules out an alternative decomposition mechanism in-
volving acid-mediated Hock fragmentation. Hock fragmenta-
tion would not produce cytotoxic carbon radicals, and thus the
observation of potent activity by 9 (and 13) suggests action
via the canonical mechanism of trioxolane toxicity (Scheme 1).
Significantly, ML4118S and its hybrid form 8 were both active
at single-digit nanomolar concentrations (EC50,Pot =5.2 nm and
4.0 nm, respectively), approximately tenfold more potent than
9 or 13. The enhanced potency of 8 relative to 9 suggests that
active ML4118S is indeed released from hybrid 8 within para-
sites; that hybrid 8 and ML4118S have similar potencies sug-
gests additive activity rather than synergistic or antagonistic
activity.
Keywords: antiparasitic agents · drug delivery · hybrid drugs ·
malaria · prodrugs
[1] A. M. Dondorp, F. Nosten, P. Yi, D. Das, A. P. Phyo, J. Tarning, K. M. Lwin,
F. Ariey, W. Hanpithakpong, S. J. Lee, P. Ringwald, K. Silamut, M.
Imwong, K. Chotivanich, P. Lim, T. Herdman, S. S. An, S. Yeung, P. Sin-
ghasivanon, N. P. Day, N. Lindegardh, D. Socheat, N. J. White, N. Engl. J.
Med. 2009, 361, 455.
[2] H. Noedl, D. Socheat, W. Satimai, N. Engl. J. Med. 2009, 361, 540.
[3] W. O. Rogers, R. Sem, T. Tero, P. Chim, P. Lim, S. Muth, D. Socheat, F.
Ariey, C. Wongsrichanalai, Malar. J. 2009, 8, 10.
To further investigate the release of ML4118S from hybrid 8
in parasites, we measured the kinetics of DPAP1 inhibition
using the FY01 probe (Figure 3c). No inhibition of DPAP1 activ-
ity was observed upon exposure of parasites to compound 13
for 0.5–6 h, which indicates that the toxic effects of the trioxo-
lane moiety do not alter the levels of DPAP1 activity for at
least 6 h. As expected, ML4118S completely inhibited DPAP1
activity in parasites at all time points while control hybrid 9
was unable to inhibit DPAP1 even after 6 h. Hybrid 8, on the
other hand, inhibited DPAP1 in a time-dependent fashion, with
complete inhibition observed after 3 h. These results are con-
sistent with a slow release of ML4118S from fragmenting
hybrid 8. Based on the inhibition rate constant measured for
[4] P. Olliaro, T. N. Wells, Clin. Pharmacol. Ther. 2009, 85, 584.
[5] J. L. Vennerstrom, S. Arbe-Barnes, R. Brun, S. A. Charman, F. C. K. Chiu, J.
Chollet, Y. Dong, A. Dorn, D. Hunziker, H. Matile, K. McIntosh, M. Padma-
nilayam, J. Santo Tomas, C. Scheurer, B. Scorneaux, Y. Tang, H. Urwyler,
S. Wittlin, W. N. Charman, Nature 2004, 430, 900.
[6] Y. Tang, Y. Dong, X. Wang, K. Sriraghavan, J. K. Wood, J. L. Vennerstrom,
J. Org. Chem. 2005, 70, 5103.
[7] D. J. Creek, W. N. Charman, F. C. Chiu, R. J. Prankerd, K. J. McCullough, Y.
Dong, J. L. Vennerstrom, S. A. Charman, J. Pharm. Sci. 2007, 96, 2945.
[8] D. J. Creek, W. N. Charman, F. C. Chiu, R. J. Prankerd, Y. Dong, J. L. Ven-
nerstrom, S. A. Charman, Antimicrob. Agents Chemother. 2008, 52, 1291.
[9] D. J. Sullivan, Jr., I. Y. Gluzman, D. G. Russell, D. E. Goldberg, Proc. Natl.
Acad. Sci. USA 1996, 93, 11865.
[10] A. Robert, F. Benoit-Vical, B. Meunier, Coord. Chem. Rev. 2005, 249, 1927.
[11] B. Meunier, Acc. Chem. Res. 2008, 41, 69.
[12] P. M. O’Neill, P. A. Stocks, M. D. Pugh, N. C. Araujo, E. E. Korshin, J. F. Bick-
ley, S. A. Ward, P. G. Bray, E. Pasini, J. Davies, E. Verissimo, M. D. Bachi,
Angew. Chem. 2004, 116, 4289; Angew. Chem. Int. Ed. 2004, 43, 4193.
[13] P. M. O’Neill, A. Mukhtar, S. A. Ward, J. F. Bickley, J. Davies, M. D. Bachi,
P. A. Stocks, Org. Lett. 2004, 6, 3035.
[14] P. Gibbons, E. Verissimo, V. Barton, G. L. Nixon, R. K. Amewu, J. Chad-
wick, P. A. Stocks, G. A. Biagini, A. Srivastava, P. J. Rosenthal, J. Gut, R. C.
Guedes, R. Moreira, R. Sharma, N. Berry, M. L. Cristiano, A. E. Shone, S. A.
Ward, P. M. O’Neill, J. Med. Chem. 2010, 53, 8202.
ML4118S in vitro (ki =10200mÀ1
s
À1; figure S2a in the Support-
1
ing Information), we estimate the t = of ML4118S release from
2
hybrid 8 in parasites to be ~1.5 h (kr =0.00013 sÀ1; figure S2b
in the Supporting Information), which is sufficiently rapid to be
useful in the context of antimalarial therapy.
In summary, we have shown that a prototypical fragmenting
hybrid delivers multiple antimalarial activities in a targeted
fashion to the intra-erythrocytic P. falciparum parasite. In princi-
[15] E. Deu, M. J. Leyva, V. E. Albrow, M. J. Rice, J. A. Ellman, M. Bogyo, Chem.
Biol. 2010, 17, 808.
418
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 415 – 419

[16] J. G. Olsen, A. Kadziola, C. Lauritzen, J. Pedersen, S. Larsen, S. W. Dahl,
FEBS Lett. 2001, 506, 201.
Dorn, X. Wang, J. M. Karle, Y. Tang, S. Wittlin, R. Brun, J. L. Vennerstrom,
J. Med. Chem. 2005, 48, 4953.
[20] Y. Tang, Y. Dong, J. M. Karle, C. A. DiTusa, J. L. Vennerstrom, J. Org. Chem.
2004, 69, 6470.
[21] D. J. Creek, F. C. Chiu, R. J. Prankerd, S. A. Charman, W. N. Charman, J.
Pharm. Sci. 2005, 94, 1820.
[17] D. Turk, V. Janjic, I. Stern, M. Podobnik, D. Lamba, S. W. Dahl, C. Laurit-
zen, J. Pedersen, V. Turk, B. Turk, EMBO J. 2001, 20, 6570.
[18] S. Arastu-Kapur, E. L. Ponder, U. P. Fonovic, S. Yeoh, F. Yuan, M. Fonovic,
M. Grainger, C. I. Phillips, J. C. Powers, M. Bogyo, Nat. Chem. Biol. 2008,
4, 203.
[19] Y. Dong, J. Chollet, H. Matile, S. A. Charman, F. C. K. Chiu, W. N. Charman,
B. Scorneaux, H. Urwyler, J. Santo Tomas, C. Scheurer, C. Snyder, A.
Received: November 12, 2010
Published online on January 24, 2011
ChemMedChem 2011, 6, 415 – 419
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
419