Photoactivatable Prodrug of Doxazolidine Targeting Exosomes

Article abstract of DOI:10.1021/acs.jmedchem.8b01508

Natural lipid nanocarriers, exosomes, carry cell-signaling materials such as DNA and RNA for intercellular communications. Exosomes derived from cancer cells contribute to the progression and metastasis of cancer cells by transferring oncogenic signaling molecules to neighboring and remote premetastatic sites. Therefore, applying the unique properties of exosomes for cancer therapy has been expected in science, medicine, and drug discovery fields. Herein, we report that an exosome-targeting prodrug system, designated MARCKS-ED-photodoxaz, could spatiotemporally control the activation of an exquisitely cytotoxic agent, doxazolidine (doxaz), with UV light. The MARCKS-ED peptide enters a cell by forming a complex with the exosomes in situ at its plasma membrane and in the media. MARCKS-ED-photodoxaz releases doxaz under near-UV irradiation to inhibit cell growth with low nanomolar IC₅₀ values. The MARCKS-ED-photodoxaz system targeting exosomes and utilizing photochemistry will potentially provide a new approach for the treatment of cancer, especially for highly progressive and invasive metastatic cancers.

Full text of DOI:10.1021/acs.jmedchem.8b01508

Subscriber access provided by the Henry Madden Library | California State University, Fresno
Article
Photoactivatable Prodrug of Doxazolidine Targeting Exosomes
Ryo Tamura, Alla Balabanova, Samuel A Frakes, Austin
Bargmann, Jan Grimm, Tad H Koch, and Hang Hubert Yin
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01508 • Publication Date (Web): 31 Jan 2019
Downloaded from http://pubs.acs.org on February 1, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
Photoactivatable Prodrug of Doxazolidine Targeting
Exosomes
7
8
9
Ryo Tamura1,2,3, Alla Balabanova1, Samuel A. Frakes1, Austin Bargmann1, Jan
Grimm3, Tad H. Koch1, Hang Yin1,2,4*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1Department of Chemistry and Biochemistry and 2BioFrontiers Institute,
University of Colorado Boulder, Boulder, CO 80309, USA
3Molecular Pharmacology Program, Memorial Sloan Kettering Cancer Center,
New York, NY 10065, USA.
4School of Pharmaceutical Sciences, Tsinghua University-Peking University Joint
Center for Life Sciences, Tsinghua University, Beijing 100082, China
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 42
1
2
3
4
Abstract
5
6
7
8
9
Natural lipid nanocarriers, exosomes, carry cell signaling materials such as DNA and RNA for
intercellular communications. Exosomes derived from cancer cells contribute the progression
and metastasis of cancer cells by transferring oncogenic signaling molecules to neighbor and
remote premetastatic sites. Therefore, applying the unique properties of exosomes for cancer
therapy has been expected in science, medicine and drug discovery fields. Herein, we report that
an exosome-targeting prodrug system, designated MARCKS-ED-Photodoxaz, could
spatiotemporally control the activation of an exquisitely cytotoxic agent, doxazolidine (doxaz),
with UV light. MARCKS-ED peptide enters a cell by forming a complex with exosomes in situ
at its plasma membrane and in the media. MARCKS-ED-Photodoxaz releases doxaz under near
UV irradiation to inhibit cell growth with low nanomolar IC50 values. The MARCKS-ED-
Photodoxaz system targeting exosomes and utilizing photochemistry will potentially provide a
new approach for the treatment of cancer, especially for highly progressive and invasive
metastatic cancers.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

Page 3 of 42
Journal of Medicinal Chemistry
1
2
3
4
Introduction
5
6
7
8
9
Extracellular vesicles (EVs) are significant nanocarriers in cell-to-cell communications to
deliver bioactive molecules including DNA, RNA and proteins, which otherwise could be easily
degraded in plasma without EVs. EVs consist of exosomes with diameter of 30-150 nm and
microvesicles with diameter of 100-1000 nm. Exosomes are released by exocytosis after being
formed in endosome, and microvesicles are produced by direct budding from the plasma
membrane.1–4 EVs have been intensely studied in biomedical research including cancer growth
and metastasis, immune system modulation, and drug delivery system development.5–8
Exosomes derived from cancer cells deliver oncogenic molecules to tumor microenvironments
and sites of future metastases. For instance, exosomes from cancer cells prepare the environment
for incoming cancer cells in premetastatic sites.9,10 Furthermore, cancer cell-derived exosomes
transfer signaling molecules to the immune system that modulate the immune response against
cancer cells.7,11,12 Exosomes have been used for delivering therapeutic cargos such as small
molecules, proteins and nucleic acids to disease sites in pre-clinical and clinical studies.6,8,13–15
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We have previously reported an exosome-sensing peptide called MARCKS-ED, derived
from the effector domain of myristoylated alanine-rich C kinase substrate (MARCKS) protein.16
MARCKS modulates phospholipase C signaling by segregating phosphatidyl inositol-
bisphosphate (PIP2).17,18 MARCKS-ED comprises 25 amino acids, out of which the 13 basic
amino acid residues and 5 phenyl alanine residues contribute most of the binding to negatively
charged and highly curved lipid membrane (Table 1). MARCKS-ED binds exosomes with higher
affinity than larger vesicles because of their characteristic externalized phosphatidylserine (PS)
and a highly curved membrane.16,19–21 Exosomes in nature deliver bioactive molecules for cell-
to-cell communications, otherwise degraded in plasma. Therefore, exosomes are attractive drug
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 42
1
2
3
4
5
6
7
8
9
delivery vehicles. The recent methods for applications of exosomes to drug delivery are based on
isolating exosomes from cell cultures, followed by drug loading by electroporation, incubation or
sonication.6,8,14,22 However, since cancer cells continuously release exosomes in tumor
microenvironments that are eventually circulating through the vascular system, utilizing those
exosomes minimizes potential unwanted side effects from heterogenous origins and does not
require large scale production and isolation of pure exosomes.23–25 Here, we report a novel drug
delivery technology with an exosome sensing peptide, MARCKS-ED, which binds exosomes in
situ and delivers a cargo into cancer cells.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We use an exquisitely cytotoxic agent derived from doxorubicin, doxazolidine (doxaz),
as the anticancer therapeutic cargo (Fig. 1A). Doxorubicin (dox) is an FDA approved anticancer
drug. However, dose dependent cardiotoxicity and acquisition of multidrug-resistance phenotype
including elevated expression of p-170 glycoprotein (P-gp) hamper clinical efficacy.26–28 Doxaz
is a dox-formaldehyde conjugate at the 3’-NH2 and 4’-OH groups. The IC50 values of doxaz to
cancer cells are 1 to 4 orders of magnitude lower than dox and its potency is not affected by P-
170gp because doxaz is an uncharged DNA alkylating agent at sites of 5’-GC-3’ while dox is a
cation at physiological pH that inhibits the function of topoisomerase II.29,30 However, the half-
life of doxaz is approximately 3 min with respect to hydrolysis to dox under physiological
conditions due to the unstable oxazolidine ring.29,31,32 Therefore, prodrugs with stable
carbamylated amino groups, p-aminobenzyloxycarbonyl (PABC) groups, have been investigated
for therapeutic applications of doxaz. We have reported two types of enzymatically cleavable
prodrugs of doxaz: Pentyl PABC-Doxaz (PPD) and GaFK-PABC-Doxaz (see abbreviations for
definitions of acronyms).33–35 PPD was designed to be activated by carboxylesterase 2 (CES2)
expressed in liver, non-small-cell lung, colon, pancreatic, renal, and thyroid cancer cells. PPD is
4
ACS Paragon Plus Environment

Page 5 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
effectively cleaved by CES2 and inhibited the tumor growth in xenografted mouse models of
liver and non-small-cell lung cancer cell lines.34 GaFK-PABC-Doxaz is the prodrug of doxaz,
which is activated by proteolytic enzymes strongly associated with cancer progression and poor
prognosis.35 PPD and GaFK-PABC-Doxaz are efficiently activated by enzymes, but those
enzymes are also present to some extent in normal cells. Therefore, we utilized a
photoactivatable linker to achieve a potentially selective prodrug activation with light inside
cancer cells.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results
Design and Synthesis
Doxaz is readily synthesized from dox in the presence of excess paraformaldehyde to form
an oxazolidine ring at 3’-NH2 and 4’-OH of the daunosamine sugar ring. Additional
formaldehyde links two doxaz molecules to yield doxoform (doxf). Doxf hydrolyzes to doxaz
which hydrolyzes to dox in aqueous medium. The half-life of doxf under physiological
conditions is approximately 1 minute and the half-life of doxaz is approximately 3 minutes.31
Previous studies showed that carbamoylation of doxaz at the oxazolidine ring extends the half-
life and keeps the drug inactive.33–35 Therefore, we applied a similar prodrug strategy to protect
active oxazolidine by carbamoylating the secondary amine. Doxaz is a powerful cytotoxic agent
but can hydrolyze to much less toxic dox rapidly. Hence, unwanted side effects can be
minimized after the prodrug is activated inside cancer cells such as escape to normal cells. The
final construct consists of three parts: a very strong anticancer cargo (doxaz), an o-nitrobenzyl
photolabile linker to achieve spatiotemporal prodrug activation, and MARCKS-ED peptide that
binds exosomes in tumor microenvironments (Fig. 1B). We used a photolabile linker with a
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 42
1
2
3
4
5
6
7
8
9
small PEG spacer to improve the water solubility and accessibility to the peptide. MARCKS-ED
peptide has been shown to be a sensor of negatively charged phosphatidylserine (PS) and
membrane curvature by both electrostatic interactions and the vertical insertion of phenylalanine
into the lipid bilayer. Exosomes are natural vesicles with PS enriched, highly curved membranes,
which are correlated with tumor progression and metastasis.2,36
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the first step to prepare the photolabile linker, intermediate 5 (Scheme 1) was
synthesized using previously reported methods.37,38 Its carbonyl group was reduced with sodium
borohydride to generate the alcohol 6 in good yield. Then the resulting alcohol was activated to
intermediate 7 with 4-nitrophenyl chloroformate and DMAP. To prepare doxaz, dox free base
was extracted from a clinical sample of dox hydrochloride and transformed to doxf in the
presence of excess paraformaldehyde. Doxf was coupled to the linker 7 with a 51% yield using
HOBt in wet DMSO. The trace of water in wet DMSO hydrolyzes doxf to doxaz, and the
secondary amine of doxaz couples to the activated linker. The modest yield was presumably due
to steric effects at the bulky secondary amine of doxaz. Photodoxaz-PEG-azide 8 shows two
peaks on RP-HPLC because it is a mixture of two diastereomers (Fig. S1). Interestingly, normal
phase column chromatography did not work to purify 8; therefore, reverse phase column
chromatography or preparative HPLC was utilized.
MARCKS-ED peptide and other peptides in this study were synthesized using a standard
Fmoc solid phase peptide synthesis (SPPS) protocol.39 We previously reported that both L- and
D- configurations of MARCKS-ED have PS- and curvature-sensing capabilities, but the isomer
with D- residues is more resistant to proteolytic degradation than the isomer with L-
residues.16,19,20,40 For comparison here, we synthesized both L- and D- configurations of
MARCKS-ED peptide, designated L-MARCKS-ED and D-MARCKS-ED, respectively. In
6
ACS Paragon Plus Environment

Page 7 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
addition, we also synthesized the scrambled sequence of MARCKS-ED peptide (MARCKS-Scr),
a non-binding control peptide (C2BL3L), and a mutant with the critical lysines to aspartic acids
and phenylalanines to alanines (MARCKS-Mut) (Table 1).16 Our initial attempt to conjugate
the peptide to the linker 8 used a copper catalyzed click reaction by installing 4-pentynoic acid to
the N-terminus, but this approach failed, perhaps because the metal catalyst was sequestered by
the highly charged MARCKS-ED or because the relatively bulky peptide renders steric
hinderance.41 Furthermore, the anthracycline and linker are sensitive to acidic and basic
conditions as well as metal catalyzed reactions. Transition metals are known to form a complex
with anthracyclines.42 Therefore, we decided to use a thiol-maleimide Michael addition for this
conjugation since it is a well-established bioconjugation method such as the method used in
antibody drug conjugate syntheses.43,44 To install a maleimide, DBCO-maleimide was reacted
with the intermediate 8. This completely transformed 8 to photdoxaz-PEG-maleimide 9 (Scheme
2). Analytical RP-HPLC showed four peaks since 9 was formed as two sets of two
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
diastereomers, generated by the non-regioselective click reaction (Fig. S1). The thiol was
installed to the N-terminus of the peptides by coupling with S-trityl-3-mercaptopropionic acid
using HATU, and the peptides were each released by TFA global deprotection and cleavage
from the resin. Intermediate 9 was conjugated to the thiol group of the peptides in PBS or
phosphate buffer under a nitrogen atmosphere. The final products were then purified by reverse
phase column chromatography. The RP-HPLC showed one broad peak containing 8 isomers, two
sets of regioisomeric diastereomers with four diastereomers per set, generated by the Michael
addition (Fig. S1). To compare the efficacy of doxaz conjugate to the corresponding dox
conjugate, dox was also reacted with intermediate 7 followed by the same coupling method to L-
MARCKS-ED. We succeeded in the syntheses of the compounds 1, 2, 3 and 4 (Schemes 2 and
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 42
1
2
3
4
5
6
7
8
9
S1, and Fig. S2). However, the Michael additions of MARCKS-Scr and -Mut were not
successful. Therefore, we synthesized Cy5 labeled peptides to explore the cellular uptake of
these peptides. All peptides and peptide conjugates were completely water soluble. Interestingly,
nanoparticle tracking analysis (NTA) showed that MARCKS-ED conjugates 1 and 2, and Cy5-
labeled L- and D-MARCKS-ED, form nanoparticles with a diameter of approximately 100 nm
(Fig. S3). This nanoparticle formation potentially prevents proteolytic degradations and
unwanted prodrug activations, and improves water solubility, a reasonable circulation half-life in
blood, and tumor accumulation.45
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Photolysis
The long-term goal of our approach is that the prodrug can be spatiotemporally activated
by internal light after the prodrug is taken up by cancer cells; thereby, unwanted side effects can
be minimized. As a proof-of-principle of this approach, we first measured the relative rate of
photolysis of photodoxaz-PEG-azide 8 using a monochromatic laser. We used intermediate 8 for
the quantum yield measurements instead of the final products 1, 2, 3, and 4 because of its ease of
preparation and better solubility in the solution of the measurement. Intermediate 8 in a solution
of 65% 15 mM phosphate buffer and 35% acetonitrile was irradiated with an Omnichrome HeCd
monochromatic laser (12 mW, 325 nm) for 30 min. Stirring was achieved by continuous
pipetting, and the optical density at 325 nm assured total absorption through 10% conversion.
The amounts of dox released was monitored by RP-HPLC (Fig. S4, and S5). The quantum yield
(Φ) was calculated using the time at which approximately 10% of dox was released. Figure 2
shows plots of dox released by UV light irradiation as a function of irradiation time. More than
95% of the prodrug was activated by 30 min of UV irradiation with a quantum yield (Φ) of
8
ACS Paragon Plus Environment

Page 9 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
0.025 (Table S1), agreeing well with previously reported quantum yields for analogous
photocleavage reactions.46 Further, we irradiated the peptide conjugates, 1, 2, 3 and 4, for 15 min
with a UV black lamp apparatus, which emits 0.8 mW/cm2 of UV light at the surface of a 96
well plate in the region 320-400 nm with maximum intensity at 350 nm as measured with a
power meter (Fig. S6). The change in solvent from 35% acetonitrile to aqueous culture medium
as well as the change from the azide intermediate to the final prodrug could affect the quantum
yield. We suspect the effect is not dramatic and not so relevant to the prodrug development at
this stage. RP-HPLC analyses showed approximately 50% of prodrug was activated (Fig.S7 A, B
and C). L-MARCKS-Photodox 4 showed 70% photolysis (Fig. S7 D). This difference might
derive from the bulky tertiary carbamate of doxaz prodrugs that potentially decreases the
efficiency of the photoreaction. Unlike dox carbamates, doxaz carbamates are known to exist as
a mixture of slowly interconverting chair/twist boat conformations of the amino sugar ring.47
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Growth Inhibition of MDA-MB-231 Breast Cancer Cells
We successfully showed the photolysis efficacy of our reagents. Next, we investigated
the growth inhibition potency to cancer cells under UV irradiation. For the growth inhibition
assessments, we used the same UV light apparatus used for photolysis experiments described
above (Fig. S6). A 96-well plate was placed 12 cm away from the light source. The power of the
light at this position was approximately 0.8 mW/cm2 (max. wavelength at 350 nm) over the area
of the portion of the 96-well plate used as measured by a Scientech power meter. To test the
phototoxicity of the UV light, we measured its effect on cellular growth of MDA-MD-231 cells.
We found that 15 min of UV irradiation showed minimal effects on cellular growth (Fig. S8).
Next, we investigated the cytotoxicities of the prodrugs, 1, 2, 3 and 4, under 15 min of
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 42
1
2
3
4
5
6
7
8
9
near UV irradiation to MDA-MB-231 cells in culture. The prodrugs were incubated for 6 h at
various concentrations in the media with 10% FBS. After cells were washed with PBS twice,
they were irradiated with UV light for 15 min. Following incubating at room temperature for 15
min, PBS washing twice and addition of fresh media, the cells were grown until the untreated
cells reach 80% confluency or 5 days. For the negative control without UV light, the 96-well
plate containing the cells was wrapped with aluminum foil to prevent the activation by any
ambient UV light. The cell densities were measured by crystal violet staining. The growth
inhibition curves and IC50 values are reported in Fig. 3 and Table 2, respectively. L-MARCKS-
ED-photodoxaz 1 showed significant cytotoxicity with UV black light even though it was
inactive without UV black light (IC50: with UV = 0.015 µM, without UV = 2.5 µM). The D-
configuration of the MARCKS-ED conjugate 2 (IC50: with UV = 0.042 µM, without UV = 2.0
µM) also showed a comparable result, but the IC50 with UV light was slightly higher than with
the L-configuration. By contrast, C2BL3L conjugate 3 (IC50: with UV = 1.3 µM, without UV =
11 µM) was 87-fold less toxic than 1 under UV light. Interestingly, L-MARCKS-ED-photodox 4
(IC50: with UV = 1.2 µM, without UV = 6.5 µM) was 80-fold less toxic than 1 under UV light
even though they differ by only a single carbon. This result supports the oxazolidine ring of
doxaz being essential for high cellular growth inhibition.48 The IC50 values of 1 and 2 are
comparable to PPD and GaFK-PABC-Doxaz with enzymatic activation.33–35
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cellular Uptake
Both L- and D- configurations of MARCKS-ED-photodoxaz, 1 and 2, showed good
potency against cancer cell growth under UV black light irradiation but were essentially inactive
in the absence UV black light irradiation. Subsequently, we explored the cellular uptake
10
ACS Paragon Plus Environment

Page 11 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
properties of the MARCKS peptides. First, the compounds 1, 2, 3 and 4, were incubated with
MDA-MB-231 cells for 6 h, which is the same condition for the growth inhibition assay above.
Following PBS washing, trypsinization and centrifugation, the cellular uptake of the compounds
was analyzed by flow cytometry (Fig. 4). Interestingly, D-MARCKS-ED-photodoxaz 2 had
approximately 2- to 3-fold higher cellular uptake than the L-counterparts, 1 and 4. C2BL3L
conjugate 3 had the lowest cellular uptake. Confocal microscopy also demonstrated cellular
uptake of 1 and 2 (Fig. 5). Since 10 µM of 1 and 2 for 6 h incubation was too toxic, we decreased
the exposure to 5 µM for 3 h. The images clearly showed the uptake of the peptides, and some
peptides appeared on the plasma membrane (Fig. 5 A and B).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Since the fluorescence intensity from the anthracycline is not sufficient for in depth
analysis of cellular uptakes, we subsequently explored uptake using Cy5-labeled peptides.
Although Dox is fluorescent, its fluorescence intensity was insufficient for the experiment.
Hence, Cy5, which has a higher molar absorptivity and fluorescence quantum yield, was
substituted for Dox. All peptides in Table 1 were labeled with Cy5-carboxylic acid using the
standard HATU coupling. 49 MDA-MB-231 and Human Mammary Epithelial Cells (HuMEC),
both cell lines known to uptake exosomes, were incubated with 1 µM of the peptides for 1, 3, or
6 h. 50,51 The fluorescence intensities from cells were then measured by confocal microscopy and
quantified by ImageJ software. Both MDA-MB-231 and HuMEC cells showed a similar trend
with respect to the uptake of photodoxaz peptide conjugates (Fig. S9 A and B). The uptake of D-
MARCKS-ED was significantly greater than L-MARCKS-ED. By contrast, C2BL3L and
MARCKS-Mut did not show cell uptake at all. Interestingly, MARCKS-Scr showed a similar
uptake with the wild type peptide. To see the correlation with exosomes, MDA-MB-231 cells
were incubated with 1 µM L- or D-MARCKS-ED peptide with the exosomes labeled with
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 42
1
2
3
4
5
6
7
8
9
fluorescence dye for 6 h (Fig. S10). Confocal microscopic analyses showed co-localization of the
peptide and exosomes for both L- and D- peptides (Fig. S11 and S12 A and B). The higher
magnification 3D images showed that both the peptide and exosomes are co-localized outside of
the nucleus (Fig. S13). The time course experiment with continuous imaging for 1 h after
addition of the peptide and exosomes showed that they are gradually associating with the plasma
membrane, and some aggregations consisting of MARCKS-ED and exosomes appear on it upon
cell entry (Fig. 6 and S14 and supporting videos). The complexation of the peptide and exosomes
may also take place in media because we previously reported that MARCKS-ED peptides bind
exosomes and liposomes in a biological solution.16 Further, quantification of cellular uptakes of
Cy5-D-MARCKS-ED and cell-mask labeled exosomes showed the decrease in fluorescence
intensity from exosomes when it was co-incubated with the peptide while the cellular uptake of
peptide is not affected by exosomes. This result is presumably due to energy transfer from the
cell-mask dye on exosomes to the peptide that supports the co-localization of the peptides and
exosomes in the cells (Fig. S15).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Discussion
In this work, we developed a strategy of using photoactivatable prodrugs, Doxaz-
MARCKS-ED conjugates, to target selectively cancerous cells. The final constructs 1, 2, 3 and 4
were successfully synthesized as shown in Scheme 2 and Fig. S2. Interestingly, the compounds 1
and 2 form nanoparticles in PBS, presumably due to self-assembly of hydrophobic caged-doxaz
and hydrophilic highly charged peptide. This is potentially beneficial for the stability in plasma
by protecting the construct from unwanted activation and/or metabolic reduction of the nitro
group.52 Photolysis experiments showed the quantum yield of cleavage of photodoxaz-PEG-
12
ACS Paragon Plus Environment

Page 13 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
azide is 2.5%. This low yield may be advantageous for protecting the prodrug from unnecessary
photoactivation by ambient near UV light. However, since UV light permeability in tissue is
limited, further modification of linker to improve photolysis efficiency is probably needed.53
Some photoactivable linkers with high efficiency have been reported during the last decades.54
For this application, photocleavage efficiency is defined as the product of the quantum yield of
reaction and molar absorptivity at the wavelengths of light available.55 For example, Menge and
Heckel reported that a coumarin protecting group is 17 times more effectively uncaged than 2-(o-
nitrophenyl)-propyl caging group due to higher molar absorptivity at comparable wavelengths.56
Therefore, improving the photocleavage efficiency is the focus of our efforts for the success of
future in vivo experiments. Furthermore, a light delivery method will also be important for in
vivo applications because external UV light cannot permeate human skin and tissue permeability
is limited.53,57 Ibsen et al. utilized a LED fiber-optic system to deliver UV light to deep tumor
tissue in mouse experiments and showed successful prodrug activation by inserting a fiber optic
to the tumor.58,59 Less invasive methods were also reported using radioactive species. For
instance, Kotagiri et al. reported an innovative way to generate near UV light inside tumors with
Cerenkov radiation (CR) generated by clinically relevant radionuclides. They demonstrated that
titanium dioxide (TiO2) nanoparticles are excited by CR and induce photodynamic therapy. 60 In
addition to the TiO2 nanoparticles, other nanoparticles and photosensitizers were reported for CR
induced photodynamic therapy. 61–63 Further, Ran et al. applied CR to activate o-nitrobenzyl
ether-caged luciferin in vivo.64 All of these previous studies pointed to the direction that the
photoactivation of a prodrug may be a viable method for clinical applications.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In our current studies, we showed that both L- and D-MARCKS-ED-photodoxaz, 1 and
2, were successfully activated by UV light with low ~nM IC50 values while toxicity was
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 42
1
2
3
4
5
6
7
8
9
significantly lower without UV light irradiation. On the other hand, L-MARCKS-ED-Photodox 4
was 80-fold less toxic than 1 under UV irradiation even though the difference is only one carbon.
These results clearly showed doxaz is more effective than dox. As a comparison, the C2BL3L
control 3 was 87-fold less toxic than 1 likely because it did not get into cells well as we
anticipated. Interestingly, D-MARCKS-ED 2 has more cell penetrating capability than L-
MARCKS-ED 1 but it is slightly less toxic to cancer cells. We have previously reported that D-
MARCKS-ED peptide is roughly 6 times more resistant to proteolytic hydrolysis than L-
MARCKS-ED-peptide.40 We speculated that the stability of the peptide might contribute to the
higher cell uptake of D-MARCKS-ED. Furthermore, A more stable nanoparticle formulation of
the prodrug may be beneficial if photoactivation to doxaz occurs in the nanoparticle and if the
more hydrophobic environment of the nanoparticle slows the rate of hydrolysis of doxaz to dox
at sites remote from its DNA target.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We designed the prodrug to target the exosomes secreted from cancer cells and deliver
the therapeutic cargo into the cells. Since cancer cell-derived exosomes correlate with cancer
progression and metastasis, our approach will potentially inhibit cancer growth, progression and
metastasis caused by exosomes. Nakase et al. reported that cationic lipids or peptides could
enhance exosome delivery to cytosol because polycationic peptides or polymers have cell
membrane penetrating capability that help exosomes enter cells.65,66 We hypothesized that
MARCKS-ED peptide binds the PS enriched highly curved membrane of exosomes and forms
an exosome-peptide complex in situ that localizes on the plasma membrane and enters the cell.
Co-incubation of the peptides with fluorescence labeled exosomes clearly showed co-localization
in cells. Furthermore, the time-course imaging experiments showed that MARCKS-ED peptide
and exosomes co-localize on the plasma membrane and subsequently enter the cell. These
14
ACS Paragon Plus Environment

Page 15 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
experiments imply that they were taken up together by the same mechanism. However, we
cannot rule out other possible mechanisms of peptide uptake. For example, cell-penetrating
peptides (CPPs) such as the TAT peptide derived from the human immunodeficiency virus type
1 (HIV-1) and the penetratin peptide derived from the amphiphilic Drosophila Antennapedia
homeodomain possess unique cell penetrating properties though the exact mechanism of uptake
has not been completely understood. In general, CPPs are polycationic peptides and believed to
enter the cells by direct penetration of plasma membrane or endocytosis pathways.67 As
MARCKS-ED peptide is polycationic, it could enter the cells by a similar manner with CPPs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Exosome targeting is a new strategy for drug delivery. Some exosome-sensing probes
have been reported that potentially support our approach. For example, Carney et al. developed a
cyclic peptide, LXY30, that targets α3β1 integrin on exosomes and demonstrated that LXY30
detects exosomes and even differentiate cancer cell-derived exosomes from noncancerous
exosomes. Furthermore, the exosome uptake to cancer cells was inhibited by LXY30 in a dose
dependent manner.68 Similarly, Ghosh et al. demonstrated that the venceremin peptide, targeting
heat shock proteins on exosomes, captures and precipitates exosomes. 69 Although those studies
were aimed for detecting exosomes, the exosome-sensing probes could be used to modify the
surface of exosomes as well as to deliver therapeutic cargos likewise as we demonstrated in this
study.
Conclusion
In summary, we successfully developed the photoactivatable prodrug of an exquisitely
cytotoxic anticancer agent, doxaz, targeted with a novel cell-penetrating-peptide derived from
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 42
1
2
3
4
5
6
7
8
9
the effecter domain of MARCKS protein. MARCKS-ED-photodoxaz easily enters cancer cells
and with UV irradiation releases the exquisitely toxic but short-lived DNA cross-linking agent,
doxaz, to inhibit cancer cell growth. MARCKS-ED peptide co-localizes with exosomes on the
plasma membrane to enter the cell. Since exosomes are highly correlated with cancer progression
and metastasis, our method may provide a new approach for development of targeted anticancer
therapies.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental
1. General
All chemicals were purchased from Sigma Aldrich (St. Louis, MO), unless otherwise noted.
Clinical samples of doxorubicin hydrochloride were gifted by FeRx, Inc. (Aurora, CO). All
reagents and amino acids for solid phase peptide synthesis were purchased from ChemPep Inc.
(Wellington, FL). All NMR spectra were taken at 400 MHz on a Varian Inova (Palo Alto, CA) or
400 MHz Bruker Avance-III (Billerica, MA) spectrometer in deuterated solvents purchased from
Cambridge Isotope Laboratories, Inc. (Tewksbury, MA). Chemical shifts are reported in δ values
of ppm and were standardized by the residual solvent peak in MestReNova NMR software
(Mestrelab Research, Santiago de Compostela, Spain). Thin layer chromatography (TLC) was
performed on Merck Kieselgel 60Å F254 plates and visualized with a UV lamp at 254 nm. All
peptides were synthesized using a CEM Liberty Microwave Peptide Synthesizer via solid phase
Fmoc peptide chemistry (Matthews, NC). The resin used was a H-Rink-Amide-Chem matrix
with a loading capacity of 0.45 mmol/g, purchased from PCAS BioMatrix Inc (Saint-Jean-sur-
Richelieu, Quebec Canada). UV-vis spectroscopy was performed with a Beckman Coulter DU
730 Life Science UV/Vis Spectrophotometer (Brea, CA). High resolution mass spectrometry was
performed on a Waters Synapt G2 HDMS (q-TOF) instrument (Milford, MA). Low resolution
16
ACS Paragon Plus Environment

Page 17 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
LC-MS was performed on an Agilent 6120 (Q-MS) instrument (Santa Clara, CA). Analytical and
semi-preparative reverse phase HPLC was performed on an Agilent 1200 series instrument
(Santa Clara, CA). Normal phase and reverse phase column chromatographies were performed
on a Biotage Isolera Flash System (Charlotte, NC) using Sorbent Technologies Silica gel 60 Å
(230-240 mesh) (Norcross, GA) and C18 column, respectively. The purity of compounds was
evaluated via HPLC and 1H NMR. All tested compounds are ≥95% purity, unless otherwise
noted. The molar extinction coefficients were dox (480 nm) 11,500 M-1cm-1 and Cy5 (646 nm)
250,000 M-1cm-1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2. Synthesis
Dox free base preparation
A clinical sample of doxorubicin-HCl (40 mg) was dissolved in 10 mL of MeOH in a separatory
funnel. After 10 mL of saturated NaHCO3 (Aq.) was added to adjust the pH to 8.5, the dox free
base was extracted with 20 mL of chloroform three times. The combined organic layers were
dried with sodium sulfate and chloroform was removed via rotatory evaporator. The compound
was dried in vacuo and stored in a -20 °C freezer.
Doxaz and DoxF Synthesis
Dox free base was dissolved in deuterated chloroform, followed by addition of 10 equiv of
prilled paraformaldehyde. The reaction was monitored by 1H NMR. After 2 or 3 days stirring, all
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 42
1
2
3
4
5
6
doxorubicin was converted to doxoform. After the reaction mixture was filtered, the solvent was
removed by rotatory evaporator and the doxoform was dried in vacuo. 35
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N-(2-{2-[2-(2-Azidoethoxy)ethoxy]ethoxy}ethyl)-2-[4-(1-hydroxyethyl)-2-methoxy-5-
nitrophenoxy]acetamide (6).
Compound 5 (1.61 g, 3.43 mmol) was dissolved in MeOH (15 mL). To this solution was added
sodium borohydride (0.519 g, 13.7 mmol) slowly at 0 °C. The resulting solution was warmed to
room temperature and stirred for 5 h. The reaction was quenched by adding saturated NH4Cl
(Aq.). The product was extracted with ethyl acetate three times. The combined organic phase
was washed with brine and dried over sodium sulfate, and the solvent was removed by rotatory
evaporator. The crude product was purified by normal phase silica gel flash column
chromatography eluting with 5% MeOH in DCM (v/v) to give product 6 (1.2 g, 72% yield) with
the following spectral properties: 1H NMR (400 MHz, DMSO-d6) δ 8.09 (t, J = 5.7 Hz, 1H), 7.52
(s, 1H), 7.39 (s, 1H), 5.53 (d, J = 4.4 Hz, 1H), 5.32 – 5.20 (m, 1H), 4.60 (s, 2H), 3.93 (s, 3H),
3.61 – 3.47 (m, 10H), 3.44 (t, J = 5.8 Hz, 2H), 3.41 – 3.35 (m, 3H), 3.29 (q, J = 5.8 Hz, 2H), 1.37
ppm (d, J = 6.2 Hz, 3H); 13C NMR (101 MHz, DMSO-d6) δ 167.06, 153.49, 145.40, 138.97,
138.53, 109.65, 109.28, 69.71, 69.67, 69.58, 69.49, 69.17, 68.80, 67.92, 63.86, 56.04, 49.88,
40.02, 39.81, 39.60, 39.39, 39.18, 38.97, 38.76, 38.24, 25.04 ppm. ESI-MS: [M+H]+ calcd
472.2043, found 472.2043: [M+Li]+ calcd 478.2126, found 478.2123. NMR assignments appear
in Figure S16.
1-(4-{[(2-{2-[2-(2-Azidoethoxy)ethoxy]ethoxy}ethyl)carbamoyl]methoxy}-5-methoxy-2-
nitrophenyl)ethyl 4-nitrophenyl carbonate (7).
18
ACS Paragon Plus Environment

Page 19 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 6 (0.282 g, 0.562 mmol) was dissolved in THF (10 mL) with activated 4Å molecular
sieves. 4-Nitrophenyl chloroformate (0.170 g, 0.843 mmol), DMAP (0.0068 g, 0.056 mmol) and
triethylamine (0.114 g, 1.124 mmol) were added. The resulting mixture was stirred at room
temperature for 16 h. Ethyl acetate and water were poured into the reaction mixture, followed by
washing with saturated NaHCO3 (Aq) five times. The organic phase was washed with brine and
dried with sodium sulfate, and the solvent was evaporated. The crude mixture was purified by
normal phase silica gel flash column chromatography eluting with 5% MeOH in DCM (v/v). The
fractions were collected, evaporated and dried to give 7 (0.322 g, 90% yield) with the following
spectral properties: 1H NMR (400 MHz, DMSO-d6) δ 8.29 (d, J = 9.3 Hz, 2H), 8.12 (t, J = 6.0
Hz, 1H), 7.57 (s, 1H), 7.53 (d, J = 9.3 Hz, 2H), 7.23 (s, 1H), 6.28 (q, J = 6.4 Hz, 1H), 4.66 (s,
2H), 3.99 (s, 3H), 3.61 – 3.47 (m, 10H), 3.44 (s, 2H), 3.35 (s, 4H), 3.28 (d, J = 5.7 Hz, 2H), 1.73
ppm (d, J = 6.5 Hz, 3H); 13C NMR (101 MHz, DMSO) δ 166.88, 154.98, 153.68, 151.13, 146.51,
145.12, 139.40, 131.09, 125.32, 122.54, 109.55, 108.93, 72.92, 69.70, 69.66, 69.57, 69.49, 69.16,
68.80, 67.69, 56.40, 49.87, 40.03, 39.82, 39.61, 39.40, 39.19, 38.98, 38.77, 38.27, 21.11 ppm;
ESI-MS: [M+H]+ calcd 637.2106, found 637.2111. NMR assignments appear in Figure S17.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Photodoxazolidine-PEG-Azide (8).
Compound 7 (0.159 g, 0.250 mmol) and HOBt hydrate (0.045 g, 0.294 mmol) were dissolved in
DMSO (6 mL). After the mixture was stirred at room temperature for 10 min, doxaz (0.0817 g,
0.147 mmol) in DMSO (6 mL) was added. The resulting mixture was stirred at room temperature
for 2 days. The reaction was monitored by analytical HPLC. The unreacted doxaz was
hydrolyzed to dox by adding 1X phosphate buffer (10 mL) and the unreacted 7 was quenched by
adding by 1 equiv of D-glucosamine and DIEA, and 2 equiv of HOBt. Following overnight
stirring, DCM (20 mL) was added and the organic phase was washed with water three times. The
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 42
1
2
3
4
5
6
7
8
9
crude mixture was purified by reverse phase column chromatography using a C18 column
eluting with H2O and ACN with 0.1 % TFA (v/v) to give 8 (0.08 g, 51% yield) with the
following spectral properties: NMR spectra with assignments appear in Figure S18. ESI-MS:
[M-H]- calcd 1051.3499, found 1051.3470.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Photodoxazolidine-PEG-Maleimide (9).
Compound 8 (0.0398 g, 37.8 μmol) and DBCO-maleimide (0.0194 g, 58.5 μmol) were combined
in DMSO (4.5 mL). The resulting mixture was stirred at room temperature for 22 h. To the
reaction mixture were added DCM and water, followed by washing with water three times and
brine, drying with sodium sulfate, and evaporation of solvent. The crude mixture was purified by
normal phase silica gel column chromatography eluting with 90% (v/v) DCM and 10% (v/v)
MeOH to give 9 (0.045 g, 81% yield) with the following spectral properties: NMR spectra with
assignments appear in Figure S19; ESI-MS: [M+H]+ calcd 1480.5109, found 1480.5135.
Photodoxorubicin-PEG-azide (10).
Compound 6 (0.035 g, 0.055 mmol) and HOBt hydrate (0.011 g, 0.074 mmol) were dissolved in
DMSO (2 mL). To this mixture was added dox free base (0.02 g, 0.037 mmol) in DMSO (2 mL).
The resulting mixture was stirred at room temperature overnight and monitored by analytical
HPLC (method 1). To the reaction mixture was added 20 mL of DCM. The resulting solution
was washed with water three times and dried over anhydrous sodium sulfate. The crude mixture
was purified by reverse phase column chromatography using a C18 column eluting with H2O and
ACN with 0.1 % (v/v) TFA to give 10 (0.026 g, 69% yield) with the following spectral
20
ACS Paragon Plus Environment

Page 21 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
properties: NMR spectra with assignments appear in Figure S20; ESI-MS: [M+Na]+ calcd
1063.3396, found 1063.3369.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Photodoxorubicin-PEG-Maleimide (11).
Compound 10 (0.0130 g, 0.0124 mmol) and DBCO-maleimide (0.0064 g, 0.015 mmol) were
dissolved in DMSO (1.5 mL). The resulting mixture was stirred at room temperature for 22 h.
The reaction was monitored by analytical HPLC (method 1). After chloroform was added to this
mixture, it was washed with water three times and dried over anhydrous sodium sulfate. The
crude mixture was purified by normal phase silica gel column chromatography eluting with 95%
(v/v) DCM and 5% (v/v) MeOH to give 11 (7.2 mg, 39 % yield) with the following spectral
properties: NMR spectra with assignments appear in Figure S21; ESI-MS [M+H]+ calcd.
1468.5109, found 1468.5056.
General Method for peptide-thiol conjugation.
L-MARCKS-ED-SH. (L-configuration)
L-MARCKS-ED peptide (sequence: KKKKKRFSFKKSFKLSGFSFKKNKK) (0.1 mmol) was
synthesized by standard Fmoc solid phase peptide synthesis. The peptide on resin (0.1 mmol), s-
trityl-3-mercaptopropionic acid (0.069 g, 0.2 mmol) and HATU (0.077 g, 0.2 mmol) were
combined and dissolved in DMF (8 mL). After the solution reached a homogeneous appearance,
DIEA (0.0516 g, 0.4 mmol) was added. The resulting mixture was stirred at room temperature
overnight. The peptide cleavage from the resin and universal deprotection were carried out with
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 42
1
2
3
4
5
6
7
8
9
TFA cocktail: TIPS (100 μL), H2O (250 μL), EDT (250 μL), and TFA (9.4 mL). The reaction
was left at room temperature for 3 h. After filtering out resin, cold diethyl ether was added and
the precipitated peptides were collected by centrifugation (4500 xg). The crude peptide was
purified by semipreparative RP-HPLC using method 2. The pure product showed the following
spectral properties; ESI-MS: calcd. [M+3H]3+ 1056.6, [M+4H]4+ 792.7, [M+5H]5+ 634.4,
[M+6H]6+ 528.8, [M+7H]7+ 453.4, [M+8H]8+ 396.9, found [M+3H]3+ 1056.6, [M+4H]4+ 792.7,
[M+5H]5+ 634.4, [M+6H]6+ 528.8, [M+7H]7+ 453.4, [M+8H]8+ 396.9.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
D-MARCKS-ED-SH. (D-configuration)
The peptide with each amino acid residue having the D-configuration was similarly prepared. It
showed the following mass spectral properties: ESI-MS: calcd. [M+2H]2+ 1584.4513, [M+3H]3+
1056.6368, [M+4H]4+ 792.7296, found [M+2H]2+ 1584.4550, [M+3H]3+ 1056.6370, [M+4H]4+
792.7276.
C2BL3L-SH
C2BL3L-SH peptide was similarly prepared with the following mass spectral properties. ESI-
MS: calcd. [M+1H]+ 1339.6, [M+2H]2+ 670.3, found [M+1H]+ 1339.6, [M+2H]2+ 670.3.
General method for thiol-maleimide Michael addition for synthesis of 1, 2, 3 and 4.
The peptide-thiol conjugate (0.050 g, 0.016 mmol) was dissolved in degassed 1X PBS (pH=7.4,
10 mL) under a nitrogen atmosphere, to which photodoxaz-maleimide (0.028 g, 0.019 mmol) in
22
ACS Paragon Plus Environment

Page 23 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DMF (1 mL) was added dropwise. The reaction was left in the dark with a nitrogen atmosphere
at room temperature overnight. The crude product was purified by reverse phase column
chromatography using a C18 column eluting with 60% (v/v) H2O with 0.1% (v/v) TFA and 40%
(v/v) ACN with 0.1% (v/v) TFA to give the pure product. The fractions containing the product
were lyophilized. Purity was more than 95% (RP-HPLC) eluting with method 1 and monitoring
at 210, 254, 280 and 480 nm (Figures S35, S36, S37 and S38).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
L-MARCKS-ED-Photodoxazolidine (1). 7% yield. ESI-MS:C223H325N51O56S, calcd. [M+3H]3+
1550.1388, [M+4H]4+ 1162.8561, [M+5H]5+ 930.4864, [M+6H]6+ 775.5733, [M+7H]7+
664.9211, [M+8H]8+ 581.9319, found [M+3H]3+ 1550.1439, [M+4H]4+ 1162.8579, [M+5H]5+
930.4872, [M+6H]6+ 775.5684, [M+7H]7+ 664.9202, [M+8H]8+ 581.9130.
Frag. 1: C194H297N50O43S, calcd. [M+4H]5+ 810.7760, [M+5H]6+ 675.8133, [M+6H]7+ 579.4114,
[M+7H]8+ 507.1100, [M+8H]9+ 450.8756, found [M+4H]5+ 810.6526, [M+5H]6+ 675.7117,
[M+6H]7+ 579.3247, [M+7H]8+ 507.0339, [M+8H]9+ 450.5086.
Frag. 2: C202H308N51O47S, [M+4H]5+, calcd. 847.8120, found 847.6667.
Frag. 3: C202H309N51O48S, [M+5H]5+, calcd. 851.4120, found 851.2692.
D-MARCKS-ED-Photodoxazolidine (2). 25% yield. ESI-MS: calcd. [M+3H]3+ = 1550.1388,
[M+4H]4+ = 1162.8561, [M+5H]5+ = 930.4864, [M+6H]6+ = 775.5733, [M+7H]7+ = 664.9211,
found [M+3H]3+ = 1550.1199, [M+4H]4+ = 1162.8431, [M+5H]5+ = 930.4782, [M+6H]6+ =
775.5732, [M+7H]7+ = 664.9173.
Frag. 1: C194H297N50O43S, calcd. [M+5H]6+ 675.8133, [M+6H]7+ 579.4114, [M+7H]8+ 507.1100,
[M+8H]9+ 450.8756, found [M+5H]6+ 675.7092, [M+6H]7+ 579.1815, [M+7H]8+ 507.0352,
[M+8H]9+ 450.8104.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 42
1
2
3
4
5
6
Frag. 2: C202H308N51O47S, [M+5H]6+, calcd. 706.6767, found 706.5542.
Frag. 3: C202H309N51O48S, [M+6H]6+, calcd. 709.6767, found 709.5551.
7
8
9
C2BL3L-Photodoxazolidine (3). 34.6 % yield. ESI-MS: C128H163N25O46S, calcd. [M+2H]2+
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1410.5546, [M+3H]3+ 940.7056, found [M+2H]2+ 1410.5590, [M+3H]3+ 940.7059.
Frag. 1: C99H135N24O33S, [M+2H]3+ calcd. 740.6446, found 740.9850.
Frag. 2: C107H146N25O37S, calcd. [M+1H]2+ 1203.7700, [M+2H]3+ 802.8467, found [M+1H]2+
1203.509, [M+2H]3+ 802.6745.
L-MARCKS-ED-Photodoxorubicin (4). 26.6% yield. ESI-MS: C222H325N51O56S, calcd.
[M+3H]3+ 1546.1388, [M+4H]4+ 1159.8561, [M+5H]5+ 928.0864, [M+6H]6+ 773.5733, found
[M+3H]3+ 1546.1353, [M+4H]4+ 1159.8542, [M+5H]5+ 928.0833, [M+6H]6+ 773.5701.
Frag. 1: C194H297N50O43S, [M+5H]6+ 675.8133, [M+6H]7+ 579.4114, [M+7H]8+ 507.1100, found
[M+5H]6+ 675.7104, [M+6H]7+ 579.3233, [M+7H]8+ 507.0322.
Frag. 2: C201H308N51O47S, [M+5H]6+, calcd. 704.6733, found 704.5548.
Frag. 3: C201H309N51O48S, [M+6H]6+, calcd. 707.6750, found 707.5566.
Associated Content
Supporting information
The supporting Information is available free of charge on the ACS Publications website at DOI:
Molecular Formula Strings (CSV).
24
ACS Paragon Plus Environment

Page 25 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
Supporting figures, tables and videos, experimental, assignments of proton NMR signals on
spectra, mass spectra, HPLC chromatograms, and sets of NMR spectra (PDF).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Author Information
Corresponding Author:
Dr. Hang Yin
School of Pharmaceutical Sciences, Tsinghua University, Beijing 100082, China
E-mail: yin_hang@tsinghua.edu.cn
Acknowledgements
This work was funded by the National Key R&D Program of China (No. 2017YFA0505200),
National Institutes of Health (NIH R01CA183953), and National Natural Science Foundation of
China (Nos. 21825702 and 21572114). The imaging work was performed at the
BioFrontiers Institute Advanced Light Microscopy Core. Spinning disc confocal
microscopy was performed on Nikon Ti-E microscope supported by the BioFrontiers
Institute and the Howard Hughes Medical Institute. We would like to thank Dr. Benjamin
L Barthel, Mr. Thomas Price Kirby and Dr. Peter Cogan for helpful discussions.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 42
1
2
3
4
Abbreviations
5
6
7
8
9
ACN = acetonitrile, DBCO = dibenzocyclooctyne, DCM = Dichloromethane, DIEA = N,N-
diisopropylethylamine, DMF = dimethylformamide, DMSO = dimethyl sulfoxide, Dox =
doxorubicin, DMAP = 4-(dimethylamino)pyridine, Doxaz = doxazolidine, DoxF = doxoform,
ED = effector domain, EDT = Ethane ditiol, EM = exact mass, ESI-MS = Electrospray
Ionization Mass Spectrometry, EtOH = ethanol, Fmoc = fluoronylmethyloxycarbonyl, FBS =
fetal bovine serum, HATU = 1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-
b]pyridinium 3-oxid hexafluorophosphate, HOBt = hydroxybenzotriazole, HPLC = High
Performance Liquid Chromatography, IC50 = half maximal inhibitory concentration, MALDI-
TOF-MS = Matrix Assisted Laser Desorption/Ionization Time of Flight Mass Spectrometry,
MARCKS = myristoylated alanine-rich C kinase substrate protein, MeOH = methanol, NMR =
Nuclear Magnetic Resonance, NTA = Nanoparticle Tracking Analysis, PBS = phosphate
buffered saline, Photodoxaz = Photodoxazolidine, Photodox = Photodoxorubicin, PS =
phosphatidylserine, RT = room temperature, SDS = Sodium dodecyl sulfate, TFA =
Trifluoroacetic acid, TIPS = triisopropylsilane, PPD = Pentyl PABC-Doxaz , GaFK-PABC-
Doxaz = N-(N-(N-acetylglycyl-D- alanyl-L-phenylalanyl-L-lysyl)-p-aminobenzyloxycarbonyl)-
doxazolidine.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
References
(1)
Tamura, R.; Yin, H. Rationally Designed Peptide Probes for Extracellular Vesicles. In
Advances in Clinical Chemistry; Makowski, G. S., Ed.; Academic Press: Burlington,
2017; Vol. 79, pp 25–41.
(2)
(3)
Kastelowitz, N.; Yin, H. Exosomes and Microvesicles: Identification and Targeting By
Particle Size and Lipid Chemical Probes. ChemBioChem 2014, 15, 923–928.
Lane, R. E.; Korbie, D.; Anderson, W.; Vaidyanathan, R.; Trau, M. Analysis of Exosome
26
ACS Paragon Plus Environment

Page 27 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
Purification Methods Using a Model Liposome System and Tunable-Resistive Pulse
Sensing. Sci. Rep. 2015, 5, 7639.
6
7
8
9
(4)
Minciacchi, V. R.; Freeman, M. R.; Vizio, D. Di. Extracellular Vesicles in Cancer:
Exosomes, Microvesicles and the Emerging Role of Large Oncosomes. Semin. Cell Dev.
Biol. 2015, 40, 41–51.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(5)
(6)
Vader, P.; Breakefield, X. O.; Wood, M. J. A. Extracellular Vesicles: Emerging Targets
for Cancer Therapy. Trends Mol. Med. 2014, 20, 385–393.
Alvarez-Erviti, L.; Seow, Y.; Yin, H.; Betts, C.; Lakhal, S.; Wood, M. J. A. Delivery of
siRNA to the Mouse Brain by Systemic Injection of Targeted Exosomes. Nat. Biotechnol.
2011, 29, 341–345.
(7)
(8)
Greening, D. W.; Gopal, S. K.; Xu, R.; Simpson, R. J.; Chen, W. Exosomes and Their
Roles in Immune Regulation and Cancer. Semin. Cell Dev. Biol. 2015, 40, 72–81.
Tian, Y.; Li, S.; Song, J.; Ji, T.; Zhu, M.; Anderson, G. J.; Wei, J.; Nie, G. A Doxorubicin
Delivery Platform Using Engineered Natural Membrane Vesicle Exosomes for Targeted
Tumor Therapy. Biomaterials 2014, 35, 2383–2390.
(9)
Peinado, H.; Alečković, M.; Lavotshkin, S.; Matei, I.; Costa-Silva, B.; Moreno-Bueno, G.;
Hergueta-Redondo, M.; Williams, C.; García-Santos, G.; Ghajar, C.; Nitadori-Hoshino,
A.; Hoffman, C.; Badal, K.; Garcia, B. A.; Callahan, M. K.; Yuan, J.; Martins, V. R.;
Skog, J.; Kaplan, R. N.; Brady, M. S.; Wolchok, J. D.; Chapman, P. B.; Kang, Y.;
Bromberg, J.; Lyden, D. Melanoma Exosomes Educate Bone Marrow Progenitor Cells
toward a pro-Metastatic Phenotype through MET. Nat. Med. 2012, 18, 883–891.
(10) Sceneay, J.; Smyth, M. J.; Möller, A. The Pre-Metastatic Niche: Finding Common
Ground. Cancer Metastasis Rev. 2013, 32, 449–464.
(11) Robbins, P. D.; Morelli, A. E. Regulation of Immune Responses by Extracellular
Vesicules. Nat. Immunol. 2014, 14, 195–208.
(12) György, B.; Hung, M. E.; Breakefield, X. O.; Leonard, J. N. Therapeutic Applications of
Extracellular Vesicles: Clinical Promise and Open Questions. Annu. Rev. Pharmacol.
Toxicol. 2015, 55, 439–464.
(13) Wahlgren, J.; Karlson, T. D. L.; Brisslert, M.; Vaziri Sani, F.; Telemo, E.; Sunnerhagen,
P.; Valadi, H. Plasma Exosomes Can Deliver Exogenous Short Interfering RNA to
Monocytes and Lymphocytes. Nucleic Acids Res. 2012, 40, e130.
(14) Haney, M. J.; Klyachko, N. L.; Zhao, Y.; Gupta, R.; Plotnikova, E. G.; He, Z.; Patel, T.;
Piroyan, A.; Sokolsky, M.; Kabanov, A. V.; Batrakova, E. V. Exosomes as Drug Delivery
Vehicles for Parkinson’s Disease Therapy. J. Control. Release 2015, 207, 18–30.
(15) Wang, J.; Zheng, Y.; Zhao, M. Exosome-Based Cancer Therapy: Implication for
Targeting Cancer Stem Cells. Front. Pharmacol. 2017, 7, 533.
(16) Morton, L. a; Yang, H.; Saludes, J. P.; Fiorini, Z.; Beninson, L.; Chapman, E. R.;
Fleshner, M.; Xue, D.; Yin, H. MARCKS-ED Peptide as a Curvature and Lipid Sensor.
ACS Chem. Biol. 2013, 8, 218–225.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 42
1
2
3
4
5
(17) Aderem, A. The Marcks Brothers: A Family of Protein Kinase C Substrates. Cell 1992,
71, 713–716.
6
7
8
(18) Arbuzova, A.; Schmitz, A. A. P.; Vergères, G. Cross-Talk Unfolded: MARCKS Proteins.
Biochem. J. 2002, 362, 1–12.
9
(19) Morton, L. A.; Tamura, R.; De Jesus, A. J.; Espinoza, A.; Yin, H. Biophysical
Investigations with MARCKS-ED: Dissecting the Molecular Mechanism of Its Curvature
Sensing Behaviors. Biochim. Biophys. Acta - Biomembr. 2014, 1838, 3137–3144.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(20) Yan, L.; de Jesus, A. J.; Tamura, R.; Li, V.; Cheng, K.; Yin, H. Curvature Sensing
MARCKS-ED Peptides Bind to Membranes in a Stereo-Independent Manner. J. Pept. Sci.
2015, 21, 577–585.
(21) Glaser, M.; Wanaski, S.; Buser, C. A.; Boguslavsky, V.; Rashidzada, W.; Morris, A.;
Rebecchi, M.; Scarlata, S. F.; Runnels, L. W.; Prestwich, G. D.; Chen, J.; Aderem, A.;
Ahn, J.; McLaughlin, S. Myristoylated Alanine-Rich C Kinase Substrate (MARCKS)
Produces Reversible Inhibition of Phospholipase C by Sequestering Phosphatidylinositol
4,5-Bisphosphate in Lateral Domains. J. Biol. Chem. 1996, 271, 26187–26193.
(22) El-Andaloussi, S.; Lee, Y.; Lakhal-Littleton, S.; Li, J.; Seow, Y.; Gardiner, C.; Alvarez-
Erviti, L.; Sargent, I. L.; Wood, M. J. A. Exosome-Mediated Delivery of siRNA in Vitro
and in Vivo. Nat. Protoc. 2012, 7, 2112–2126.
(23) Taylor, D. D.; Shah, S. Methods of Isolating Extracellular Vesicles Impact down-Stream
Analyses of Their Cargoes. Methods 2015, 87, 3–10.
(24) Ha, D.; Yang, N.; Nadithe, V. Exosomes as Therapeutic Drug Carriers and Delivery
Vehicles across Biological Membranes: Current Perspectives and Future Challenges. Acta
Pharm. Sin. B 2016, 6, 287–296.
(25) Aryani, A.; Denecke, B. Exosomes as a Nanodelivery System: A Key to the Future of
Neuromedicine? Mol. Neurobiol. 2016, 53, 818–834.
(26) Staltiel, E.; McGuire, W. Doxorubicin ( Adriamycin ) Cardiomyopathy A Critical Review.
West. J. Med. 1983, 139, 332–341.
(27) Takemura, G.; Fujiwara, H. Doxorubicin-Induced Cardiomyopathy. From the Cardiotoxic
Mechanisms to Management. Prog. Cardiovasc. Dis. 2007, 49, 330–352.
(28) Nielsen, D.; Maare, C.; Skovsgaard, T. Cellular Resistance to Anthracyclines. Gen.
Pharmacol. 1996, 27, 251–255.
(29) Kalet, B. T.; Mcbryde, M. B.; Espinosa, J. M.; Koch, T. H. Doxazolidine Induction of
Apoptosis by a Topoisomerase II Independent Mechanism. J. Med. Chem. 2007, 50,
4493–4500.
(30) Nitiss, J. L. Targeting DNA Topoisomerase II in Cancer Chemotherapy. Nat. Rev. Cancer
2009, 9, 338–350.
(31) Post, G. C.; Barthel, B. L.; Burkhart, D. J.; Hagadorn, J. R.; Koch, T. H. Doxazolidine, a
Proposed Active Metabolite of Doxorubicin That Cross-Links DNA. J. Med. Chem. 2005,
28
ACS Paragon Plus Environment

Page 29 of 42
Journal of Medicinal Chemistry
1
2
3
4
48, 7648–7657.
5
6
7
8
(32) Fenick, D. J.; Taatjes, D. J.; Koch, T. H. Doxoform and Daunoformꢀ: Anthracycline -
Formaldehyde Conjugates Toxic to Resistant Tumor Cells. J. Med. Chem. 1997, 40,
2452–2461.
9
(33) Barthel, B. L.; Torres, R. C.; Hyatt, J. L.; Edwards, C. C.; Hatfield, M. J.; Potter, P. M.;
Koch, T. H. Identification of Human Intestinal Carboxylesterase as the Primary Enzyme
for Activation of a Doxazolidine Carbamate Prodrug. J. Med. Chem. 2008, 51, 298–304.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(34) Barthel, B. L.; Zhang, Z.; Rudnicki, D. L.; Coldren, C. D.; Polinkovsky, M.; Sun, H.;
Koch, G. G.; Chan, D. C. F.; Koch, T. H. Preclinical Efficacy of a Carboxylesterase 2-
Activated Prodrug of Doxazolidine. J. Med. Chem. 2009, 52, 7678–7688.
(35) Barthel, B. L.; Rudnicki, D. L.; Kirby, T. P.; Colvin, S. M.; Burkhart, D. J.; Koch, T. H.
Synthesis and Biological Characterization of Protease-Activated Prodrugs of
Doxazolidine. J. Med. Chem. 2012, 55, 6595–6607.
(36) Sharma, R.; Huang, X.; Brekken, R. A.; Schroit, A. J. Detection of Phosphatidylserine-
Positive Exosomes for the Diagnosis of Early-Stage Malignancies. Br. J. Cancer 2017, 1–
8.
(37) Jain, P. K.; Karunakaran, D.; Friedman, S. H. Construction of a Photoactivated Insulin
Depot. Angew. Chem. Int. Ed. Engl. 2013, 52, 1404–1409.
(38) Holmes, C. P. Model Studies for New O-Nitrobenzyl Photolabile Linkers: Substituent
Effects on the Rates of Photochemical Cleavage. J. Org. Chem. 1997, 62, 2370–2380.
(39) Hansen, P. R.; Oddo, A. Fmoc Solid-Phase Peptide Synthesis. In Peptide Antibodies.
Methods in Molecular Biology; Houen, G., Ed.; Humana Press: New York, 2015; Vol.
1348, pp 33–50.
(40) Kastelowitz, N.; Tamura, R.; Onasoga, A.; Stalker, T. J.; Ormacinda, R.; Brown, P. N.;
Brodsky, G. L.; Brass, L. F.; Branchford, B. R.; Paola, J. Di; Yin, H. Peptides Derived
from MARCKS Block Coagulation Complex Assembly on Phosphatidylserine. Sci. Rep.
2017, 7, 4275.
(41) Presolski, S. I.; Hong, V. P.; Finn, M. G. Copper-Catalyzed Azide–Alkyne Click
Chemistry for Bioconjugation. Curr. Protoc. Chem. Biol. 2011, 3, 153–162.
(42) Fiallo, M. M. L.; Garnier-Suillerot, A. Metal Anthracycline Complexes as a New Class of
Anthracycline Derivatives . Pd ( II ) -Adriamycin and Pd ( II ) -Daunorubicin Complexesꢀ:
Physicochemical Characteristics and Antitumor Activity. Biochemistry 1986, 25, 924–
930.
(43) Chari, R. V. J.; Miller, M. L.; Widdison, W. C. Antibody-Drug Conjugates: An Emerging
Concept in Cancer Therapy. Angew. Chem. Int. Ed. Engl. 2014, 53, 3796–3827.
(44) Ducry, L.; Stump, B. Antibody - Drug Conjugatesꢀ: Linking Cytotoxic Payloads to
Monoclonal Antibodies. Bioconjug. Chem. 2010, 21, 5–13.
(45) Shi, J.; Kantoff, P. W.; Wooster, R.; Farokhzad, O. C. Cancer Nanomedicineꢀ: Progress ,
29
ACS Paragon Plus Environment