Identification of BP16 as a non-toxic cell-penetrating peptide with highly efficient drug delivery properties

Article abstract of DOI:10.1039/c3ob42422g

Antimicrobial peptides are an interesting source of non-cytotoxic drug delivery vectors. Herein, we report on the identification of a new cell-penetrating peptide (KKLFKKILKKL-NH₂, BP16) from a set of antimicrobial peptides selected from a library of cecropin-melittin hybrids (CECMEL11) previously designed to be used in plant protection. This set of peptides was screened for their cytotoxicity against breast adenocarcinoma MCF-7, pancreas adenocarcinoma CAPAN-1 and mouse embryonic fibroblast 3T3 cell lines. BP16 resulted to be non-toxic against both malignant and non-malignant cells at concentrations up to 200 μM. We demonstrated by flow cytometry and confocal microscopy that BP16 is mainly internalized in the cells through a clathrin dependent endocytosis and that it efficiently accumulates in the cell cytoplasm. We confirmed that the cell-penetrating properties of BP16 are retained after conjugating it to the breast tumor homing peptide CREKA. Furthermore, we assessed the potential of BP16 as a drug delivery vector by conjugating the anticancer drug chlorambucil to BP16 and to a CREKA-BP16 conjugate. The efficacy of the drug increased between 6 and 9 times when conjugated to BP16 and between 2 and 4.5 times when attached to the CREKA-BP16 derivative. The low toxicity and the excellent cell-penetrating properties clearly suggest that BP16 is a suitable vector for the delivery of therapeutic agents into cells. This journal is The Royal Society of Chemistry 2014.

Full text of DOI:10.1039/c3ob42422g

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Identification of BP16 as a non-toxic cell-
penetrating peptide with highly efficient drug
delivery properties†
Cite this: Org. Biomol. Chem., 2014,
12, 1652
Marta Soler,‡a,b Marta González-Bártulos,‡a,c David Soriano-Castell,d Xavi Ribas,a
Miquel Costas,a Francesc Tebar,d Anna Massaguer,*c Lidia Feliu*b and Marta Planas*b
Antimicrobial peptides are an interesting source of non-cytotoxic drug delivery vectors. Herein, we report
on the identification of a new cell-penetrating peptide (KKLFKKILKKL-NH2, BP16) from a set of antimicro-
bial peptides selected from a library of cecropin-melittin hybrids (CECMEL11) previously designed to be
used in plant protection. This set of peptides was screened for their cytotoxicity against breast adeno-
carcinoma MCF-7, pancreas adenocarcinoma CAPAN-1 and mouse embryonic fibroblast 3T3 cell lines.
BP16 resulted to be non-toxic against both malignant and non-malignant cells at concentrations up to
200 μM. We demonstrated by flow cytometry and confocal microscopy that BP16 is mainly internalized in
the cells through a clathrin dependent endocytosis and that it efficiently accumulates in the cell cyto-
plasm. We confirmed that the cell-penetrating properties of BP16 are retained after conjugating it to the
breast tumor homing peptide CREKA. Furthermore, we assessed the potential of BP16 as a drug delivery
vector by conjugating the anticancer drug chlorambucil to BP16 and to a CREKA-BP16 conjugate. The
efficacy of the drug increased between 6 and 9 times when conjugated to BP16 and between 2 and 4.5
times when attached to the CREKA-BP16 derivative. The low toxicity and the excellent cell-penetrating
properties clearly suggest that BP16 is a suitable vector for the delivery of therapeutic agents into cells.
Received 4th December 2013,
Accepted 10th January 2014
DOI: 10.1039/c3ob42422g
www.rsc.org/obc
and may be derived from long peptides or proteins present in
nature (i.e., Tat peptide or penetratin) or de novo designed pep-
Introduction
The efficacy of anticancer treatments remains a key challenge tides (i.e., transportan, polyarginine peptides, β-peptides, pep-
in medicine. Current therapies are hindered by the appearance toids, oligocarbamates or polyproline helices).9–12 The
of toxic side effects and the development of multi-drug resist- mechanism of action of CPPs initially involves their binding to
ance by cancer cells.1,2 Moreover, the use of therapeutic agents the negatively charged head groups of lipids or proteins in the
is limited by their low permeability through biological mem- plasma membrane which is then followed by their internaliz-
branes. Therefore, there is an actual interest in developing new ation. The exact mode of internalization is still poorly under-
approaches to cancer treatment.3 A promising strategy in this stood but it has been shown to depend on several factors such
area is the use of cell-penetrating peptides (CPPs) as drug as the structure and concentration of the CPP as well as the
delivery systems.4–9 CPPs are usually short cationic sequences cargo to be transported and the specific cell line.7,8,12–14 It is
believed that endocytosis is the major mode of uptake for CPPs;
however, direct translocation can also occur. Due to their ability
to cross biomembranes in a non-disruptive way, CPPs offer an
opportunity to increase the bioavailability of drugs, enhancing
their activity and reducing their dosage. In this sense, CPPs
have been described as efficient carriers of nucleic acids, pro-
teins, small molecule drugs and imaging agents.6–8,10,11,15–18
Focusing on the idea of obtaining new CPPs with very low
toxicity, cationic antimicrobial peptides (AMPs) have emerged
as good candidates.12,19–23 Like CPPs, AMPs are short
sequences, containing between 9 and 30 amino acids, and
most of them are cationic and amphipathic. These peptides
show a broad spectrum of activity against bacteria, fungi,
enveloped viruses, parasites, and tumor cells, while exhibiting
aQBIS-CAT Research Group, Institut de Química Computacional i Catàlisi (IQCC)
and Departament de Química, Universitat de Girona, Campus Montilivi,
E-17071 Girona, Catalonia, Spain
bLIPPSO, Departament de Química, Universitat de Girona, Campus Montilivi,
E-17071 Girona, Catalonia, Spain. E-mail: marta.planas@udg.edu,
lidia.feliu@udg.edu; Fax: +34 972418150; Tel: +34 972418274
cDepartament de Biologia, Universitat de Girona, Campus Montilivi, E-17071
Girona, Catalonia, Spain. E-mail: anna.massaguer@udg.edu; Fax: +34 972418150;
Tel: +34 972418370
dDepartament de Biologia Cel·lular, Immunologia i Neurociències, Universitat de
Barcelona, E-08036 Barcelona, Catalonia, Spain
†Electronic supplementary information (ESI) available: HPLC, ESI-MS and
HRMS of peptides. See DOI: 10.1039/c3ob42422g
‡These authors contributed equally to this work.
1652 | Org. Biomol. Chem., 2014, 12, 1652–1663
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
low eukaryotic cytotoxicity.24–31 Several AMPs are being studied fact, one member of this library, BP100, has been described as
for the treatment of human diseases, and some of them have an efficient CPP in plant tobacco cells.54 This background
already entered pre-clinical and clinical trials.26,27 The mecha- prompted us to study the suitability of peptides from the
nism of action of AMPs also includes their electrostatic inter- CECMEL11 library as drug delivery vectors in cancer cells. We
action with the negatively charged phospholipid membranes first selected several members of this library displaying
causing morphological changes such as pore formation or cell different antimicrobial and hemolytic activity profiles. Some
lysis; however, their translocation into the cytoplasm is not analogues containing arginine residues were also synthesized
uncommon.32–35 Due to the latter property, the use of AMPs as and included in the study. The cytotoxicity of these peptides
CPPs is a field of interest to develop non-cytotoxic delivery was evaluated in cancer and healthy cell lines. The cellular
vectors. In fact, AMPs, such as LL-37, SynB, melittin and uptake properties of the peptide with the optimal cytotoxic
bLFcin6, are able to translocate across human plasma mem- profile as CPP were then evaluated. We also examined the
branes and to act as drug transporters.36–39
potential of this CPP candidate as a drug carrier by conjugat-
Binding of a low-molecular-weight drug to a CPP results in ing it to CLB. In addition, this candidate was conjugated to the
an improved uptake, which is associated with a decrease of the homing peptide CREKA to obtain a selective delivery vector for
dose required to achieve a significant therapeutic effect.7 In CLB in cancer cells and to confirm that the cell penetrating
particular, several efforts have been made to conjugate the properties were retained after the incorporation of CREKA and
cytotoxic agent chlorambucil (CLB) to CPPs in order to CLB. Thus, we evaluated the cytotoxicity of the CLB-CRE-
improve its efficacy.40–43 Chlorambucil (CLB) is a well-known KA-CPP conjugate against cancer and healthy cells and its cel-
nitrogen mustard typically used to treat leukemia, but also lular uptake properties.
some breast, lung and ovarian cancers. It is believed that CLB
is taken up by passive diffusion and alkylates DNA bringing
about its cross-linking.44 Although this nitrogen mustard is
highly effective against certain cancers, its use is limited by
Results and discussion
Peptide design and synthesis
the lack of selectivity and the low permeability through cell
membranes. To overcome these limitations, CLB has been con- AMPs represent a source of CPPs because both families share
jugated to CPPs such as pVEC or sC18.40,41,43
similar structural characteristics. With the aim of identifying
Despite the high efficiency of CPPs in mediating cellular new CPPs, we focused our attention on a 125-member library
uptake of pharmacologically active molecules, their use for tar- of AMPs (CECMEL11) previously described by Badosa and co-
geted therapy is limited by their low level of selectivity. Several workers.53 The general structure of this library is
approaches aimed at improving the specificity of CPPs towards R-X1KLFKKILKX10L-NH2, where X1 and X10 correspond to
tumor cells have been developed.45 One of these strategies amino acids with various degrees of hydrophobicity and hydro-
relies on the conjugation of a CPP with a homing peptide. philicity (Leu, Lys, Phe, Trp, Tyr, Val) and R includes different
These peptides bind selectively to overexpressed receptors in N-terminal derivatizations (H, Ac, Ts, Bz, Bn). Thus, peptides
human tumor cells providing a means of targeting CPPs of this library are highly cationic and their amphipathic char-
towards desired cells or tissues.46–49 However, most homing acter becomes evident when they are represented by means of
peptides have no internalization properties and only can an Edmunson wheel plot. These structural features have been
deliver cargos to the cell surface. Thus, homing peptide-CPP described as crucial for their antimicrobial activity and also
conjugates may act as efficient vectors for drug delivery into a may confer on them cell-penetrating properties.32–35 In fact,
specific cell target, improving drug bioavailability, and BP100, a member of the CECMEL11 library, has been reported
decreasing side effects and toxicity in healthy cells. An as an efficient agent to transport cargoes into plant cells.54
example of a homing peptide is CREKA (Cys-Arg-Glu-Lys-Ala)
In the present study, we examined the ability of sequences
which was identified by in vivo screening of phage-displayed of this CECMEL11 library to internalize cancer cells and to
peptide libraries in breast tumors of MMTV-PyMT transgenic transport a drug. Two sets of peptides were considered: (i)
mice. CREKA specifically homes to tumors by binding to fibrin undecapeptides BP16, BP76, BP81, BP100 and BP105, and the
and fibrin-associated clotted plasma proteins present in the Arg-containing peptides BP307 and BP308; and (ii) peptide
vessels and the interstitial stroma of tumors.47,50–52 Since conjugates containing in their structure: the undecapeptide
CREKA is not able to internalize cancer cells, it is a good can- BP16 and the DNA alkylating agent CLB (BP325), BP16 and the
didate to design peptide-mediated systems for targeted cell homing peptide CREKA (BP327), and BP16, CLB and CREKA
delivery of drugs.41,42
This study is focused on the search for new CPP candidates BP327, and BP329 were labeled with 5(6)-carboxyfluorescein
and was based on library of linear undecapeptides (CF).
(CECMEL11) previously described by Badosa and co-workers to The five peptides of the first set (BP16, BP76, BP81, BP100
(BP329). To analyze the cellular internalization, BP16, BP325,
a
be used for plant protection.53 This library included sequences and BP105) (Table 1) were selected based on their distinct anti-
highly active against phytopathogenic bacteria and fungi, and bacterial activity against the plant pathogens Erwinia amylo-
low hemolytic activity. Similarly to CPPs, these peptides are vora, Pseudomonas syringae and Xanthomonas vesicatoria, and
cationic and are able to adopt an amphipathic structure. In their different hemolysis percentage at 150 μM.53 BP100 and
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 1652–1663 | 1653

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Peptide sequences, retention times and purities on HPLC and HRMS data
Peptide
Sequencea
tR b (min)
Purityc (%)
HRMS
BP16
BP76
BP81
BP100
BP105
BP307
BP308
KKLFKKILKKL
KKLFKKILKFL
LKLFKKILKFL
KKLFKKILKYL
LKLFKKILKYL
RRLFRRILRYL
RRLFRRILRRL
CREKA
CREKA-KKLFKKILKKL
CLB-KKLFKKILKKL
CLB-CREKA
CLB-CREKA-KKLFKKILKKL
CF-KKLFKKILKKL
CF-CREKA
CF-CREKA-KKLFKKILKKL
CLB-KKLFKKILK(CF)KL
CLB-CREKA-KKLFKKILK(CF)KL
5.98
6.40
6.80
7.40
6.65
7.83
6.08
6.63
6.99
7.71
6.90
6.92
6.76, 6.80d
6.25
7.89
7.56
7.53
87
93
92
100
90
100
99
98
95
91
90
92
93
95
97
81
91
347.2575 [M + 4H]4+, 462.6733 [M + 3H]3+
352.0004 [M + 4H]4+, 468.9983 [M + 3H]3+
348.2477 [M + 4H]4+, 463.9946 [M + 3H]3+
355.9977 [M + 4H]4+, 474.3282 [M + 3H]3+
352.2490 [M + 4H]4+, 469.3289 [M + 3H]3+
391.0091 [M + 4H]4+, 521.0091 [M + 3H]3+
311.6168 [M + 5H]5+, 389.2674 [M + 4H]4+
303.1636 [M + 2H]2+, 605.3194 [M + H]+
395.4636 [M + 5H]5+, 494.0771 [M + 4H]4+
418.5260 [M + 4H]4+, 558.3644 [M + 3H]3+
445.6966 [M + 2H]2+, 890.3832 [M + H]+
452.4770 [M + 5H]5+, 565.3442 [M + 4H]4+
437.0217 [M + 4H]4+, 582.3589 [M + 3H]3+
482.1862 [M + 2H]2+, 963.3620 [M + H]+
583.5907 [M + 4H]4+, 777.7843 [M + 3H]3+
677.0473 [M + 3H]3+, 1015.0638 [M + 2H]2+
654.8565 [M + 4H]4+, 872.8045 [M + 3H]3+
BP327
BP325
BP329
CF-BP16
BP328
BP326
BP330
a All peptides are C-terminal amides. b HPLC retention time. c Percentage determined by HPLC at 220 nm from the crude reaction mixture.
d Retention time corresponding to the two isomers of the 5(6)-carboxyfluorescein (CF) labeled peptide.
BP76 are highly active against these phytopathogens (MIC of for most other homing peptides. In addition, with the aim of
2.5 to 7.5 μM) and low hemolytic (22–34%), BP81 is highly improving the specificity of the cytotoxic agent CLB, BP329
active (MIC of <2.5 to 5 μM) and moderately hemolytic (65%), was synthesized by attaching CLB to CREKA-BP16. Control
BP105 is highly active (MIC of 2.5 to 7.5 μM) and highly hemo- peptides CREKA and CLB-CREKA were also prepared (Table 1).
lytic (91%), and BP16 is poorly active (MIC > 7.5 μM) and non-
These peptides were manually synthesized on a Fmoc-Rink-
hemolytic (0%).
MBHA resin or on an aminomethyl ChemMatrix resin follow-
The Arg-containing peptides BP307 and BP308 were derived ing a standard Fmoc/tBu strategy and were obtained as C-term-
from BP100 and BP16, respectively (Table 1). They were inal amides. Couplings of the conveniently protected Fmoc-
included in this study because most common CPPs, such as amino acids and of CLB were mediated by ethyl 2-cyano-
Tat or penetratin, are Arg-rich peptides.55 This residue has 2-(hydroxyimino)acetate (Oxyma) and N,N′-diisopropylcarbodii-
been shown to play a key role in peptide internalization due to mide (DIPCDI) in N,N-dimethylformamide (DMF) or N-methyl-
the hydrogen-bond formation of the guanidino moiety with 2-pyrrolidinone (NMP), depending on the length of the
phosphates, sulfates and carboxylates on cellular components. sequence. Peptides were cleaved from the resin by acidolytic
BP307 and BP308 were tested for their antibacterial activity treatment and were obtained in excellent purities (87–100%,
against the above pathogens and for their hemolysis. The Table 1), as determined by analytical HPLC. Their identity was
replacement of Lys for Arg in BP16 and BP100 significantly confirmed by ESI-MS and HRMS.
influenced the antibacterial activity. While BP307 resulted to
On the other hand, BP16 and BP327 (CREKA-BP16) were
be less active (MIC of 6.2 to 12.5 μM) than BP100, BP308 dis- labeled with 5(6)-carboxyfluorescein by coupling this fluo-
played higher antibacterial activity (MIC of 3.1 to 12.5 μM) rescent label to their N-terminus leading to CF-BP16 and
than BP16. In contrast, the hemolysis was not affected and BP328, respectively (Table 1). 5(6)-Carboxyfluorescein labeled
BP307 and BP308 exhibited a similar hemolytic activity (28% CREKA was also prepared (CF-CREKA). Moreover, BP325
and 1%, respectively) as the corresponding parent peptide.
(CLB-BP16) and BP329 (CLB-CREKA-BP16), incorporating CLB
Peptide conjugates included in the second set were at the N-terminus, were labeled at the side-chain of Lys9 of the
designed based on BP16 (Table 1), which showed the best BP16 fragment, affording BP326 and BP330, respectively. 5(6)-
profile as a drug delivery vector among all the above tested Carboxyfluorescein was introduced using DIPCDI and Oxyma,
undecapeptides (see below). Peptide BP325 was prepared by followed by piperidine washes before cleavage of the peptide
coupling CLB to the N-terminus of BP16 via an amide bond. from the resin. These washes served to remove overincorpo-
The use of CLB is hampered by its low stability in aqueous rated carboxyfluorescein moieties.56 For the synthesis of pep-
environments and its low permeability through biomem- tides BP326 (CLB-BP16(CF)) and BP330 (CLB-CREKA-BP16-
branes, limitations that have been overcome by its conjugation (CF)), the lysine residue to be labeled was incorporated as
to a CPP.41–44 Moreover, since the use of homing peptides has Fmoc-Lys(Dde)-OH. The N-[1-(4,4-dimethyl-2,6-dioxocyclohex-
been described to increase the cell-type specificity of CPPs, 1-ylidene)ethyl] (Dde) group was selectively removed by treat-
BP327 was designed by conjugating the homing peptide ment with hydrazine prior to the coupling of the fluorescent
CREKA to the N-terminus of BP16. CREKA was selected for its label. Acidolytic cleavage afforded the labeled peptides in
excellent targeting ability to breast tumors and because its excellent purities (81–97%), and they were characterized by
linear structure avoids the additional cyclization step required ESI-MS and HRMS.
1654 | Org. Biomol. Chem., 2014, 12, 1652–1663
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Cell cytotoxicity of the CPP candidates
account for the selective binding of AMPs to cancer cells
involves the higher net negative charge and membrane fluidity
of cancer cells as compared to normal cells. Therefore, due to
their cationic nature, the undecapeptides of this study may
preferentially bind to cancer cell membranes by electrostatic
interaction and subsequently enter cells.
A good CPP to be used as a delivery vector of an anticancer
drug must display no toxicity against cancer as well as healthy
cells. Hence, the cytotoxicity of undecapeptides BP16, BP76,
BP81, BP100, BP105, BP307 and BP308 was screened in breast
adenocarcinoma MCF-7 and pancreas adenocarcinoma
CAPAN-1 cell lines. The IC50 was determined by the 3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)
assay after 48 h of peptide exposure. As shown in Table 2,
except for BP100, the undecapeptides were more active against
CAPAN-1 than against MCF-7 cells. The cytotoxic activity of
BP76 and BP105 was significant against these two cell lines,
displaying IC50 values of 23.5 and 22.7 µM against CAPAN-1,
respectively, and of 26.3 and 29.2 µM against MCF-7 cells,
respectively. BP81, BP100 and its Arg-containing analogue
BP307 were considerably active against one cell line with IC50
values ranging from 24.2 to 35.7 µM. BP16 and BP308 were the
least cytotoxic undecapeptides. BP308 showed an IC50 of 148.2
and 97.0 µM against MCF-7 and CAPAN-1 cells, respectively,
and BP16 displayed no cytotoxic effects against any of the
cancer cell lines tested (IC50 > 200 µM).
These results indicate that these undecapeptides exhibit
lower cytotoxicity against cancer cells in comparison with their
antibacterial activity. However, the structural features that
govern the anticancer activity of the CECMEL11 library
sequences, BP16, BP76, BP81, BP100 and BP105, correlate with
the general trend for the antibacterial activity of this library.53
Thus, peptides with a net charge of +5 or +6 displayed high
cytotoxic activity, whereas BP16 that has a net charge of +7 was
inactive. Moreover, BP16, BP76, and BP100 that only differ in
the amino acid at position 10 displayed a very distinct cytotoxic
activity. The same effect was observed for BP307 and BP308.
This result confirms previous data on how subtle changes in a
peptide sequence influence the biological activity.57–59 On the
other hand, the replacement of Lys with Arg in BP16 and
BP100 caused a different effect on the cytotoxicity. While this
replacement slightly influenced the cytotoxicity of BP100, the
Arg-containing sequence BP308 showed an increased antiproli-
ferative activity compared to BP16.
The analysis of the activity of undecapeptides on non-
malignant mouse embryonic fibroblasts 3T3 revealed that the
antiproliferative activity of BP76 and BP81 was high, with IC50
values of 13.5 and 15.5 µM, respectively (Table 2). BP100,
BP105, BP307 and BP308 were moderately active (IC50 of 39.0
to 68.0 µM). Notably, BP16 was non-toxic against this cell line
exhibiting an IC50 > 200 µM. Furthermore, we also used the
hemolysis assay to assess the toxicity of these peptides.53 As
mentioned above, except for BP105 and BP81, undecapeptides
displayed low hemolysis at 150 μM (0–34%). Interestingly,
BP16 and its Arg analogue BP308 were nonhemolytic even at
375 μM (data not shown). Even though these undecapeptides
did not show selectivity between the malignant and non-malig-
nant cell lines tested, some sequences (BP76, BP100, and
BP307) did not show significant hemolytic activity at concen-
trations much higher than the IC50 values against the cancer
cell lines tested.
Moreover, the stability in fetal bovine serum of BP16 was
also evaluated and compared to that of Tat49. After exposure to
10% serum at different time intervals, the presence of peptide
was analyzed by HPLC. Results showed that both peptides
exhibited a similar stability, with 70% degradation after
45 min of incubation.
Taken together, these results allowed the identification of
BP16, a short and highly cationic peptide with a suitable
activity profile to be considered as an excellent CPP candidate.
In particular, in contrast to most common CPPs that exhibited
cytotoxicity even at low concentrations,60–63 BP16 was non-
toxic to both malignant and non-malignant cell lines at con-
centrations up to 200 μM and, therefore, was selected for
further studies.
Cellular uptake of BP16
One of the major drawbacks of drug delivery applications of
CPPs is their toxicity to normal cells. In this sense, AMPs are
interesting CPP candidates because it has been reported that
some AMPs with anticancer activity do not show significant
cytotoxicity against normal cells at peptide concentrations that
are able to kill cancer cells.19,24,26,28 Parameters that would
To characterize the capacity of internalization of BP16 into
cancer and non-malignant cells, MCF-7 and 3T3 cells were
incubated at 37 °C with 5(6)-carboxyfluorescein labeled BP16
(CF-BP16) at different concentrations (0, 5, 25 and 50 μM) for
different times (1, 3 and 6 h) (Fig. 1). The mean fluorescence
of the cells, corresponding to the peptide uptake, was
Table 2 Cytotoxicity of the set of undecapeptides in 3T3, MCF-7 and CAPAN-1 cells and their hemolytic activity
Cell line
BP16
BP76
BP81
BP100
BP105
BP307
BP308
3T3
MCF-7
CAPAN-1
>200
>200
>200
0
13.5 0.7
26.3 7.5
23.5 7.0
34 2.1
15.5 0.7
40.0 3.4
24.2 7.2
65 1.5
62.5 6.4
34.3 4.0
57.7 13.3
22 2.8
39.0 8.5
29.2 4.2
22.7 1.5
91 6.2
40.0 7.1
64.7 12.5
35.7 1.5
28 3.2
68.0 7.6
148.2 8.8
97.0 15.4
IC50 a (μM)
Hemolysisb (%)
1
0.1
a The IC50 values were determined by the MTT assay after 48 h of peptide exposure. Data represent the mean SD of at least three independent
experiments performed in triplicate. b Percent hemolysis at 150 μM. The confidence interval for the mean is included.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 1652–1663 | 1655

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 2 Uptake of 5(6)-carboxyfluorescein labeled peptides CF-BP16
and CF-Tat49 into MCF-7 cells. Cells were exposed to the peptide at
25 µM at 37 °C for 1, 3 and 6 h. Each column in the graph represents the
mean fluorescence intensity of the cells determined in three indepen-
dent experiments SD.
the peptide uptake by the cells at concentrations higher than
25 μM.
We compared the capacity of internalization of BP16 with
that of the well-known CPP Tat49.10 With this aim, MCF-7 cells
were incubated at 37 °C with CF-BP16 and CF-Tat49 at 25 μM
for 1, 3 and 6 h (Fig. 2). Interestingly, CF-BP16 and CF-Tat49
showed the same mean fluorescence intensity values indicat-
ing that BP16 exhibits an excellent cellular uptake.
Fig. 1 Kinetics of the cellular uptake of BP16. 3T3 and MCF-7 cells
were exposed to different concentrations of 5(6)-carboxyfluorescein-
labeled BP16 (CF-BP16) (0, 5, 25 and 50 μM) at 37 °C for 1, 3 and 6 h.
The fluorescence intensity of the cells, corresponding to the intracellular
uptake of the peptide, was determined by flow cytometry. Each point in
the graphs represents the mean intracellular fluorescence intensity of
three independent experiments + SE.
To gain more insight into the internalization and intracellu-
lar distribution of BP16, the cellular uptake of CF-BP16 was
further analyzed by confocal microscopy. For this purpose,
MCF-7 cells were incubated with 25 μM CF-BP16 at 4 °C for
30 min and at 37 °C for 10, 30, 60, 120 and 180 min (Fig. 3A).
No fluorescence was observed after 30 min of incubation at
4 °C. However, when the cells were treated at 37 °C, a very faint
punctuate fluorescent staining was observed inside the cells
after 10 min, revealing an incipient cellular uptake of the
peptide. The number of fluorescent particles gradually
increased over time and, after 180 min of incubation, there
was a prominent fluorescent staining inside the cells, demon-
strating intense internalization of the peptide during this
period of time. Higher magnification (1000×) images from
cells exposed to CF-BP16 for 180 min (Fig. 3B) revealed that
the fluorescent particles were located throughout the whole
cytoplasm, with significant clustering at the periphery of the
cell nucleus. An optical sectioning indicated that no fluo-
rescence particles were placed inside the cell nucleus (Fig. 3C).
In addition, to find out whether CF-BP16 cellular entry was
dependent on specific interaction with the plasma membrane,
MCF-7 cells were pre-incubated with 25 µM CF-BP16 at 4 °C
for 30 min and, after extensive washing with saline phosphate
buffer, were incubated at 37 °C for 25 min. CF-BP16 internali-
zation was also compared with the Alexa647-conjugated trans-
ferrin (Tf-A647) which is a well-known model of receptor
dependent internalized ligands. Confocal images revealed that
while transferrin-A647 labelled endocytic structures inside the
cell as expected, no intracellular CF-BP16 was observed
quantified by flow cytometry. Cells were harvested by trypsini-
zation, which also prevented non-specific plasma membrane
binding of the peptide. As represented in Fig. 1, CF-BP16 was
efficiently internalized by the cells in a time and concen-
tration-dependent manner. The mean fluorescence of the cells
increased over time with parallel internalization kinetics in
both cell lines. The cellular uptake of CF-BP16 was intense
during the first 3 h of incubation, particularly when the cells
were treated with the peptide at 25 and 50 μM. After treatment
with 50 μM CF-BP16, the mean fluorescence of 3T3 cells
increased from 5 2 (0 h) to 1829 335 (1 h), 4895 464 (3 h)
and 6953 536 (6 h), while the mean fluorescence of MCF-7
cells rose from 4 1 (0 h) to 1431 307 (1 h), 3997 720 (3 h)
and 5830 410 (6 h). These results point out that the internal-
ization of CF-BP16 is more elevated in the non-malignant 3T3
cells than in MCF-7 cells, indicating the lack of selectivity of
BP16 for cancer cells.
A good correlation between peptide concentration and its
uptake by cells was observed, especially when they were treated
with 5 and 25 μM CF-BP16. At the different incubation times,
when CF-BP16 concentration was increased from 5 to 25 μM,
the mean cellular fluorescence increased by 5 to 6-fold in 3T3
cells and by 8- to 13-fold in MCF-7 cells. In contrast, slight
differences in the mean cellular fluorescence (less than 1.7-
fold increase) were determined when both cell lines were incu-
bated with CF-BP16 at 25 or 50 μM, revealing a saturation of
1656 | Org. Biomol. Chem., 2014, 12, 1652–1663
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
has been associated with hydrophobic peptides while hydro-
philic and amphipathic sequences can be internalized by both
mechanisms.64 In addition, the endocytosis process is energy-
dependent and is assumed to be inhibited by peptide incu-
bation at low temperature. Altogether, the results observed for
the incubation of CF-BP16 in MCF-7 cells at 4 and 37 °C
suggest that the endocytic pathway may play a major role in
BP16 internalization.
To further characterize the BP16 endocytosis, the role of
dynamin, which is involved in vesicle scission from plasma
membrane, was analysed by means of dynamin inhibition
using the specific inhibitor dynasore (Fig. 4).65 The BP16 endo-
cytosis was also compared with the transferrin protein as a
specific marker of a dynamin dependent endocytosis. MCF-7
cells were pre-incubated with 100 µM of dynasore for
15 min at 37 °C and then CF-BP16 and Tf-A647 were added
and further incubated for 60 min. Confocal microscopy
images showed an accumulation of Tf-A647 and CF-BP16 in
the same structures at the plasma membrane after dynasore
treatment, likely unscissored clathrin coated pits (CCPs)
(Fig. 4A). On the other hand, in control cells, CF-BP16 and
Fig. 3 Confocal microscopic imaging of the internalization of 5(6)-car-
boxyfluorescein-labeled BP16 (CF-BP16) into MCF-7 cells. (A) MCF-7
cells were exposed for 30 min at 4 °C to 25 µM CF-BP16 and then incu-
bated for the indicated times at 37 °C. The localization of CF-BP16 is
indicated by the green fluorescence. The cell nuclei were stained with
Hoeschst (blue). (B) Higher-magnification (1000×) image of a MCF-7 cell
after 180 min of treatment. (C) Three slices of a merged xz recon-
struction of the image stack (slices at 0.3 microns on the z-axis) are rep-
resented. Arrows indicate the perinuclear localization of CF-BP16. (D)
MCF-7 cells were pre-incubated for 30 min at 4 °C with 25 µM CF-BP16
and transferrin-A647 (Tf-A467, 60 µg ml−1) and then incubated for
25 min at 37 °C with or without a previous saline phosphate buffer wash
as indicated. Confocal images were acquired through the Alexa647 (red)
and CF (green) channels.
(Fig. 3D). This result may indicate the possibility that CF-BP16
is weakly membrane associated and removed after washes or
that CF-BP16 could be receptor independent and preferentially
fluid phase internalized.
These findings highlight a plausible mechanism for the
internalization of BP16. The mechanism of CPP uptake is con-
troversial and still under debate. Two main pathways have
been suggested: endocytosis and direct translocation across
the membrane bilayer.7,8,12–14 These cell uptake phenomena
can be related to the hydrophobicity of the CPP. Translocation
Fig. 4 CF-BP16 is dynamin dependent internalized, and follows the
endocytic degradation pathway. (A) MCF-7 cells were pre-incubated for
15 min at 37 °C with or without 100 µM dynasore and then CF-BP16
(25 µM) and Alexa647-labelled transferrin (Tf-A647, 60 µg ml−1) were
added and further incubated for 60 min. Confocal images were acquired
through the Alexa647 (red) and CF (green) channels. (B) MCF-7 cells
expressing RFP-Lamp1 (red) were incubated with CF-BP16 (green) and
Tf-A647 (blue) for 60 min at 37 °C. Insets show magnified images and
arrows indicate CF-BP16/RFP-Lamp1 positive vesicles (bars are 10 µm).
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 1652–1663 | 1657

View Article Online
Paper
Organic & Biomolecular Chemistry
Tf-A647 were found mostly in different endocytic structures Tf-A647, CF-BP16 and CF-Tat49 entry was strongly inhibited in
inside the cell (Fig. 4A). Indeed, a high degree of co-locali- those Cherry-DPF overexpressing cells compared to non-expres-
zation between CF-BP16 and the ectopically expressed sing cells. In order to quantify such inhibition, internalization
RFP-Lamp1,66 a marker of late endosomes and lysosomes, was of Tf-A647 and CF-BP16 or CF-Tat49 was measured by flow cyto-
clearly detected (Fig. 4B). These results suggest that CCPs are metry and compared to dextran-FITC, which enters cells by
an important port of entry for BP16. However, BP16 and trans- fluid phase and therefore via clathrin dependent and indepen-
ferrin follow different endocytic pathways inside the cell, a dent pathways. Flow cytometry analysis revealed that transfer-
known recycling pathway for transferrin and a degradation rin uptake was ∼55% inhibited in highly cherry-DPF
route for BP16.
expressing cells compared to non-transfected cells, while BP16
Next, in order to ascertain the importance of clathrin and dextran were inhibited by ∼47% and ∼33%, respectively
dependent endocytosis (CDE), the role of this dynamin depen- (Fig. 5C).
dent pathway in CF-BP16 uptake was examined in cells over-
Given that transferrin endocytosis, which is totally depen-
expressing the Cherry-DPF fragment of Eps15 protein, which dent on CDE, was ∼55% inhibited and that inhibition of BP16
by sequestering AP-2 inhibits CDE (Fig. 5).67 The CF-BP16 was close to transferrin and higher than dextran (∼47% vs.
endocytosis was also compared with the related CF-Tat49 ∼33%), it can be reasoned that CDE is playing a major role in
peptide and the Tf-A647, which is specifically internalized via BP16 internalization. Moreover, flow cytometry inhibition data
CDE. Confocal images in Fig. 5 (panels A and B) showed that obtained from Tat49 uptake (∼19%) compared to dextran
(Fig. 5D) suggests that CDE is less important for Tat49 internal-
ization than for BP16. These results indicate that both pep-
tides, BP16 and Tat49, are differentially internalized.
In summary, BP16 internalization is dynamin dependent
and it is mainly internalised via clathrin dependent endocyto-
sis possibly through a weak plasma membrane interaction or
by fluid phase, which is also in agreement with the BP16 local-
ization in late endosomes after 60 min of endocytosis in
MCF-7 cells.
Cell cytotoxicity of peptide conjugates
The potential use of BP16 as a vector for the delivery of the
DNA alkylating agent CLB was assessed by exposing MCF-7,
CAPAN-1 and 3T3 cells to CLB and the CLB-BP16 conjugate
(BP325). The IC50 was determined by the MTT assay after 48 h
of peptide exposure. As shown in Table 3, CLB alone exhibited
low cytotoxicity against the three cell lines (IC50 of 73.7 to
152.5 μM). In contrast, BP325 displayed high activity, with IC50
values of 12.0, 13.7 and 20.6 μM against MCF-7, CAPAN-1 and
3T3 cells, respectively. Thus, the conjugation of CLB to BP16
increased the cell cytotoxicity of this nitrogen mustard by
6-fold in MCF-7 cells, by 7-fold in 3T3 cells and by 9-fold in
CAPAN-1 cells. Since BP16 is non-toxic at these concentrations
(IC50 > 200 μM), this toxicity must be attributed to the CLB
moiety. This increase of activity clearly demonstrates that BP16
contributes to the internalization of CLB through the cell
membrane, enhancing its efficacy. Therefore, in agreement
with earlier reports, the mechanism of action of the CLB-BP16
Fig. 5 Clathrin dependent endocytosis is playing a key role in CF-BP16 conjugate seems to be much more efficient than the passive
diffusion mechanism suggested for CLB alone.41–44
uptake. (A, B) Confocal images of MCF-7 cells expressing Cherry-DPF
and incubated with Tf-A647 (60 µg ml−1) and CF-BP16 (25 µM) (A) or
Homing peptides are employed to enhance the selectivity of
CF-Tat49 (25 µM) (B) for 60 min at 37 °C (bars are 20 µm). (C, D) Internal-
ization of FITC-dextran (2 mg ml−1), Tf-A647 and CF-BP16 (C) or
CPPs towards malignant cells.46–49 Thus, we next investigated
whether the cell-penetrating properties of BP16 were retained
CF-Tat49 (D) after 60 min at 37 °C was quantified by flow cytometry in
Cherry-DPF expressing cells as described in the Experimental section. after conjugation with the breast tumor homing peptide
Histograms show the percentage of the indicated molecules internalized
in highly Cherry-DPF expressing cells versus non-expressing cells. Each
column represents the mean fluorescence intensity of the cells deter-
CREKA and also whether this homing peptide-CPP conjugate
could provide selective internalization of CLB into cancer cells.
With this aim, the cytotoxicity of the conjugate CLB-CRE-
mined in four independent experiments
SD. Statistical significances
KA-BP16 (BP329) was evaluated in MCF-7, CAPAN-1 and 3T3
cells. CREKA, CLB-CREKA and the conjugate CREKA-BP16
between different molecules were determined using the Student’s
t-test, **p < 0.01 and ***p < 0.001 vs. Tf-A647; ##p < 0.01 vs. CF-BP16.
1658 | Org. Biomol. Chem., 2014, 12, 1652–1663
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 3 Cytotoxicity of BP16, CREKA, chlorambucil (CLB), CLB-CREKA, and peptide conjugates in 3T3, MCF-7 and CAPAN-1 cells
Cell line
BP16
CREKA
BP327
CLB
CLB-CREKA
BP325
BP329
3T3
MCF-7
CAPAN-1
>200
>200
>200
>200
>200
>200
74.5 4.1
74.7 8.8
76.2 0.8
152.5 5.6
73.7 4.5
129.0 35.5
>200
>200
>200
20.6 3.3
12.0 2.7
13.7 2.4
33.7 1.9
35.2 1.8
33.0 1.0
IC50 a (μM)
a The IC50 values were determined by the MTT assay after 48 h of peptide exposure. Data represent the mean SD of at least three independent
experiments performed in triplicates.
(BP327) were also assayed for comparison purposes. CREKA
and CLB-CREKA were non-toxic (IC50 > 200 μM) and BP327 dis-
played low cytotoxicity (IC50 of 74.5 to 76.2 μM). In contrast,
the conjugate CLB-CREKA-BP16 (BP329) was significantly
active, with IC50 values ranging from 33.0 to 35.2 μM. These
results showed that the incorporation of CREKA decreased the
activity of the CLB-BP16 conjugate BP325; however, they con-
firmed that BP16 was able to internalize both CREKA and CLB
since the cytotoxicity of BP329 (CLB-CREKA-BP16) increased
2- to 4.5-fold the one of CLB alone. Furthermore, the lack of
selectivity observed for BP329 could be attributed to the fact
that CREKA was originally identified by in vivo phage display
and recognized fibrin-associated clotted plasma proteins in
the tumor stroma.47,50–52 On this basis, the behavior of CREKA
Fig. 6 Uptake of 5(6)-carboxyfluorescein labeled peptides CF-BP16,
in vitro could differ from the one previously described in vivo.
However, these results show that BP329 displayed a higher
activity than CLB alone proving the validity of BP16 as CPP for
cell delivery of therapeutically useful molecules.
CF-CREKA and BP328 (CF-CREKA-BP16) into 3T3 and MCF-7 cells. Cells
were exposed to the peptides at 25 µM for 3 h at 37 °C. Each column in
the graph represents the mean fluorescence intensity of the cells deter-
mined in three independent experiments SD. *p < 0.05 vs. 3T3 cells;
#p < 0.05 vs. CF-BP16 treated cells; $p < 0.05 vs. CF-CREKA.
The internalization ability of the CLB peptide conjugates
BP326 (CLB-BP16(CF)) and BP330 (CLB-CREKA-BP16(CF)) at
25 μM was evaluated by flow cytometry in MCF-7 cells after 3 h
of incubation at 37 °C. To our surprise, while the fluorescence
of BP326 (4272 453) was not significantly different from that
of CF-BP16 (3035 643), BP330 showed a considerably higher
intracellular fluorescence (3966 217) than BP328 (CF-CRE-
KA-BP16) (1789 221). These results are noteworthy since they
prove that BP16 is able to efficiently internalize both CLB and
CREKA.
These findings also reveal that the incorporation of either
CREKA and/or CLB has a strong influence on the cellular
uptake of BP16, reflecting that any modification in the mole-
cular structure of this peptide can lead to relevant changes in
its cell penetrating properties. In fact, it has been reported that
the cargo plays an important role in the internalization mecha-
nism of CPPs.8,11 Interestingly, despite these modifications,
BP16 remains an efficient drug delivery vector.
Cellular uptake of CLB peptide conjugates
In order to determine whether the differences in cytotoxic
activity of the CLB peptide conjugates BP325 (CLB-BP16) and
BP329 (CLB-CREKA-BP16) were related to their internalization
properties, these conjugates were labeled with 5(6)-carboxy-
fluorescein (CF) affording BP326 and BP330, respectively, and
analyzed by flow cytometry. In addition, the homing peptide
CREKA and the CREKA-BP16 conjugate (BP327) were also
labeled (CF-CREKA and BP328, respectively) and included in
the study.
The cellular uptake of CF-CREKA and BP328 at 25 μM was
determined by flow cytometry after 3 h of incubation in 3T3
and MCF-7 cells at 37 °C. As represented in Fig. 6, very low
intracellular fluorescence levels were detected in both cell
lines when treated with CF-CREKA, indicating that CREKA
alone was unable to internalize into the cells. These results are
in accordance with previous studies describing that CREKA
displays no cell-penetrating capacity in cultured breast cancer
cells.42 When comparing the fluorescence of CF-BP16 and
BP328 (CF-CREKA-BP16), it was observed that the cell-penetrat-
ing properties of BP16 were significantly reduced when conju-
gated to the homing peptide, probably because the coupling of
Conclusions
CREKA restricts the interaction of BP16 with the cell mem- In the present study, we have identified BP16 as a new CPP
brane. Notably, as shown by the intracellular fluorescence with high cellular uptake in vitro. In contrast to other CPPs
levels, BP328 retains significant internalization ability which is previously reported, BP16 displays no cytotoxicity against
required to deliver cytotoxic agents inside the cells.
malignant and non-malignant cells and, moreover, shows no
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 1652–1663 | 1659

View Article Online
Paper
Organic & Biomolecular Chemistry
hemolytic activity. We have demonstrated that BP16 exhibits at a flow rate of 100 μL min−1) was delivered by a 1100 Series
highly efficient penetration through endocytic mechanisms, HPLC pump (Agilent). Nitrogen was employed as both the
accumulating in the cell cytoplasm at short time periods. In drying and nebulizing gas.
addition, the conjugation of the DNA alkylating agent CLB to
High-resolution mass spectra (HRMS) were recorded under
BP16 dramatically increases the cytotoxicity of this drug conditions of ESI with a Bruker MicrOTOF-Q IITM instrument
between 6- and 9-fold. We have also shown that, in conju- using a hybrid quadrupole time-of-flight mass spectrometer
gation with the homing peptide CREKA, BP16 is able to (University of Girona). Samples were introduced into the mass
improve the cytotoxic activity of CLB from 2- to 4.5-fold. Taken spectrometer ion source by direct infusion through a syringe
together, these results confirm that BP16 is an excellent non- pump and were externally calibrated using sodium formate.
toxic delivery vector suitable for the effective transport of The instrument was operated in the positive ESI(+) ion mode.
drugs. Further studies on applications of this CPP are under-
Cell lines
way in our laboratory.
The human breast cancer cell line MCF-7, the human pancreas
cancer cell line CAPAN-1 and the mouse fibroblast cell line
3T3 were obtained from the American Tissue Culture Collec-
Experimental section
Materials and methods
tion (ATCC, Rockville, MD, USA). Cells were maintained in
DMEM supplemented with 10% FBS and 100 U ml−1 penicil-
Unless otherwise stated, common chemicals and solvents lin–streptomycin at 37 °C under a humidified atmosphere con-
(HPLC-grade or reagent-grade quality) were purchased from taining 5% CO2. Cells were passaged two times per week.
commercial sources and used without further purification.
Peptide synthesis
The 9-fluorenylmethoxycarbonyl (Fmoc) derivatives and Fmoc-
Rink-4-methylbenzhydrylamine (MBHA) resin (0.56 mmol g−1
)
General method for solid-phase peptide synthesis. Peptides
were obtained from Senn Chemicals International (Gentilly, from Table 1, Tat49 and CF-Tat49, were synthesized manually
France), NovaBiochem (Schwalbach, Germany) or from IRIS by the solid-phase method using Fmoc-type chemistry and the
Biotech GmbH (Marktredwitz, Germany). Aminomethyl Chem- following side-chain protecting groups: tert-butyloxycarbonyl
Matrix resin (0.66 mmol g−1) was obtained from Matrix Inno- (Boc) for Lys, tBu for Tyr and Glu, trityl (Tr) for Cys, and
vation Inc (St-Hubert, Canada). Ethyl 2-cyano-2-(hydroxyimino) 2,2,5,7,8-pentamethylchroman-6-sulphonyl (Pmc) for Arg. A
acetate (Oxyma) was purchased from Novabiochem (Notting- Fmoc-Rink-MBHA resin (0.56 mmol g−1) or an aminomethyl
ham, UK). Trifluoroacetic acid (TFA), triisopropylsilane (TIS), ChemMatrix resin (0.66 mmol g−1) were used as a solid
dimethyl sulfoxide (DMSO), D,L-dithiothreitol (DTT), N,N′-diiso- support to obtain peptide amides. Coupling of Fmoc-Rink
propylcarbodiimide (DIPCDI), chlorambucil (CLB), 5(6)-carboxy- (4 equiv.) onto the aminomethyl ChemMatrix resin was
fluorescein (CF), fluorescein isothiocyanate dextran mol. wt mediated by DIPCDI (4 equiv.) and Oxyma (4 equiv.) in DMF at
10 000 (FITC-dextran) and dynasore hydrate were from Sigma- room temperature overnight. Couplings of the Fmoc-amino
Aldrich (St. Louis, MO, USA). Transferrin Alexa Fluor® 647 con- acids (4 equiv.) were performed using DIPCDI (4 equiv.) and
jugate was from Molecular Probes (Invitrogen, Life Techno- Oxyma (4 equiv.) in DMF under stirring at room temperature
logies, Carlsbad, CA). Piperidine was purchased from Fluka for 2 h, and monitored by the Kaiser test.68 For sequences con-
(Buchs, Switzerland). N-Methyl-2-pyrrolidinone (NMP), N,N-di- taining up to eleven residues, the Fmoc group was removed by
methylformamide (DMF), CH3OH, CH2Cl2, diethyl ether and treating the resin with a mixture of piperidine–DMF (3 : 7, 1 ×
solvents for high performance liquid chromatography (HPLC) 2 min + 1 × 10 min). For longer sequences, Fmoc group
were obtained from Scharlau (Sentmenat, Spain).
removal was carried out with piperidine–DMF (3 : 7, 1 × 3 min
3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide + 3 × 10 min). After each coupling and deprotection step, the
(MTT), paraformaldehyde, and bisbenzimide trihydrochloro- resin was washed with DMF (6 × 1 min) and CH2Cl2 (6 ×
acetic acid (Hoechst 33258) were purchased from Sigma- 1 min), and air-dried. After the coupling of the eleventh
Aldrich (St. Louis, MO, USA). Dulbecco’s modified Eagle’s residue, NMP was used instead of DMF. Peptide elongation
medium (DMEM), phosphate buffered saline (PBS), fetal was performed by repeated cycles of Fmoc group removal,
bovine serum (FBS), penicillin–streptomycin and trypsin were coupling and washings.
obtained from GIBCO BRL (Grand Island, NY, USA). The
RFP-Lamp1 plasmid was kindly provided by Walther Mothes jected to the N-terminal Fmoc group removal. Then, peptides
(Addgene plasmid 1817).66
were cleaved or the peptidyl resins were derivatized with CLB
Once the synthesis was completed, peptidyl resins were sub-
Electrospray ionization mass spectrometry (ESI-MS) ana- and/or with 5(6)-carboxyfluorescein. Cleavage of peptides from
lyses were performed with an Esquire 6000 ESI ion trap LC/MS the resin was performed by treatment with TFA–TIS–H2O
(Bruker Daltonics) instrument equipped with an electrospray (95 : 2.5 : 2.5) for 3 h at room temperature. Peptides containing
ion source. The instrument was operated in the positive ESI(+)
a cysteine residue were cleaved with TFA–TIS–H2O–DTT
ion mode (University of Girona). Samples (5 μL) were intro- (92.5 : 2.5 : 2.5 : 2.5) for 3 h at room temperature. Both pro-
duced into the mass spectrometer ion source directly through cedures were followed by TFA evaporation by bubbling N2 into
an HPLC autosampler. The mobile phase (80 : 20 CH3CN–H2O the solution. Crude peptides were precipitated by adding cold
1660 | Org. Biomol. Chem., 2014, 12, 1652–1663
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
diethyl ether (−20 °C) and collected by centrifugation. This plates 24 h prior to the treatments. Then, cells were treated for
procedure was repeated twice. Finally, peptides were dissolved 48 h with the corresponding compound at concentrations
in H2O–CH3CN (50 : 50 v/v containing 0.1% TFA), lyophilized ranging from 0 to 200 µM. After removal of the treatment, cells
and tested for purity by HPLC. Analysis was carried out with a were washed with PBS and incubated for an additional 2 h in
Kromasil C18 reverse-phase column (4.6 × 40 mm; 3.5 µm par- the dark with fresh culture medium (100 µL) with MTT
ticle size) with a 2–100% B linear gradient over 7 min at a flow (10 µL). The medium was discarded and DMSO (100 µL) was
rate of 1.0 ml min−1. Solvent A was 0.1% aqueous TFA and added to each well to dissolve the purple formazan crystals.
solvent B was 0.1% TFA in CH3CN. Detection was performed at Plates were agitated at room temperature for 10 min and the
220 nm. ESI-MS and HRMS (ESI) were used to confirm peptide absorbance of each well was determined with an absorbance
identity. Peptides from Table 1, Tat49 and CF-Tat49, were microplate reader (ELx800, BioTek, Winooski, USA) at a wave-
obtained in purities ranging from 81 to 100%.
length of 570 nm. Three replicates for each compound were
Synthesis of CLB-peptide conjugates. Derivatization with used, and all treatments were tested at least in three indepen-
CLB was performed by treating the corresponding peptidyl dent experiments. For each treatment, the cell viability was
resin with CLB (5 equiv.), DIPCDI (5 equiv.) and Oxyma determined as a percentage of the control untreated cells by
(5 equiv.) in DMF or NMP under stirring at room temperature dividing the mean absorbance of each treatment by the mean
for 5 h. The completion of the reaction was checked by the absorbance of the untreated cells. The concentration that
Kaiser test.68 The resin was then washed with NMP (6 × reduces by 50% the cell viability (IC50) was established for
1 min), CH3OH (6 × 1 min), and CH2Cl2 (6 × 1 min), and air each compound.
dried.
Hemolysis
Synthesis of N-terminal 5(6)-carboxyfluorescein-labeled pep-
tides. For the N-terminal derivatization with 5(6)-carboxyfluo- The data corresponding to the hemolytic activity of peptides
rescein, this fluorophore (2.5 equiv.) was first pre-activated were previously reported by Badosa et al.53 It was evaluated at
with Oxyma (2.5 equiv.) and DIPCDI (2.5 equiv.) in CH2Cl2– 150 μM by determining the hemoglobin release from erythro-
NMP (1 : 9) for 10 min. The mixture was added to the corres- cyte suspensions of fresh human blood (5% v/v) using absor-
ponding N-terminal deprotected peptidyl resin and reacted bance at 540 nm.
overnight at room temperature protected from light by cover-
ing it with aluminum foil due to the light sensitivity of 5(6)-
Stability of BP16 and Tat49 in serum
carboxyfluorescein. Completeness of the coupling was con- The stability of BP16 and Tat49 was evaluated in fetal bovine
firmed using the Kaiser test.68 The resin was then washed with serum. Each peptide (1 mg) was exposed to 10% aqueous fil-
NMP (1 × 5 min), piperidine–NMP (1 : 5, 1 × 15 min), NMP (6 × tered fetal bovine serum (2 ml) at 37 °C. After 5, 10, 15, 30 and
1 min), CH2Cl2 (6 × 1 min), CH3OH (6 × 1 min), and CH2Cl2 45 min exposure, aliquots (200 μL) were removed and proteins
(6 × 1 min) and air dried.56
were precipitated with CH3CN (200 μL). Samples were cooled
Synthesis of 5(6)-carboxyfluorescein-labeled CLB-peptide to 0 °C for 15 min and centrifuged (11 000 rpm, 5 min). The
conjugates. These conjugates were prepared from a peptidyl supernatant was analyzed by HPLC. The digestion was esti-
resin incorporating the lysine residue to be labeled protected mated as the percentage of degraded peptide calculated from
with
N-[1-(4,4-dimethyl-2,6-dioxocyclohex-1-ylidene)ethyl] the decrease of the HPLC peak area of the native peptide.
(Dde) at the Nε-amino group. After CLB coupling, the Dde
group was removed by treatment with hydrazine–NMP (2 : 98,
Flow cytometry
5 × 20 min) under stirring at room temperature and the depro- The uptake efficiency of CF-BP16, CF-CREKA, and BP328 by
tection progress was monitored by the Kaiser test.68 Then, the 3T3 and MCF-7 cells and of CF-Tat49, BP326, and BP330 by
resin was washed with NMP (6 × 1 min) and CH2Cl2 (6 × MCF-7 cells was quantified by flow cytometry. Aliquots of
1 min) and air dried. Next, fluorophore labeling with 5(6)-car- 50 000 cells were seeded in 24 well plates and allowed to attach
boxyfluorescein was carried out as described for the N-term- for 24 h. Next, the cells were incubated with CF-BP16 at 5, 25
inal carboxyfluorescein labeled peptides.
and 50 µM for 1, 3 and 6 h, with CF-CREKA, BP328, BP326 and
BP330 at 25 µM for 3 h or with CF-BP16 and CF-Tat49 at 25 µM
for 1, 3 and 6 h at 37 °C. The cells were harvested by trypsiniza-
Cytotoxicity assays
Cytotoxicity of peptides BP16, BP76, BP81, BP100, BP105, tion and gently washed with 2% FBS in cold PBS. The fluo-
BP307, BP308, CREKA and BP327, CLB-peptide conjugates rescence of the cells, corresponding to the cellular uptake of
CLB-CREKA, BP325 and BP329, and CLB in 3T3, MCF-7 and the carboxyfluorescein labeled peptides, was analyzed using a
CAPAN-1 cells was determined by the MTT assay. Peptides and FACSCalibur (Becton Dickinson Immunocytometry Systems,
CLB-peptide conjugates were diluted in Milli-Q water to obtain San Jose, CA) equipped with the CellQuestTM software (Becton
2 mM stock solutions. CLB was dissolved in DMSO to provide Dickinson). The mean fluorescence intensity was represented
a 75 mM stock solution.69 Appropriate aliquots of these solu- on a four orders of magnitude log scale (1–10 000). The effect
tions were diluted in the cell culture medium to obtain the of Cherry-DPF expression on CF-BP16, CF-Tat49, transferrin-
final working concentrations. Aliquots of 4000 3T3 cells, 6000 A647 and FITC-Dextran uptake in MCF-7 cells was also ana-
MCF-7 cells or 10 000 CAPAN-1 cells were seeded on 96-well lysed by flow cytometry. MCF-7 cells seeded in 6 well plates
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 1652–1663 | 1661

View Article Online
Paper
Organic & Biomolecular Chemistry
were transfected with Cherry-DPF plasmid using Effectene
(Qiagen, Hiden, Germany) according to the manufacturer’s
specifications. To generate the Cherry-DPF plasmid, the DPF
fragment (aas: 501–874) from human Eps15 was obtained by
polymerase chain reaction and cloned into pEGFP-C1 vector
(Clontech) using XhoI and Pst1 restriction sites and after that
GFP was replaced with the mCherry fluorescent protein. After
24 h of protein expression, cells were incubated with transfer-
rin-A647 (60 µg ml−1) combined with CF-BP16 (25 µM),
CF-Tat49 (25 µM) or FITC-Dextran (2 mg ml−1) for 60 min at
37 °C and washed two times with cold PBS. The remaining
fluorescence at the cell surface was removed by a 3 min cold
acid wash (0.2 M sodium acetate, 0.5 M sodium chloride, pH
4.5) and the intracellular fluorescence was quantified using a
LSRFortessa (Becton Dickinson) equipped with the Cell-
QuestTM software (Becton Dickinson). Ten thousand cells
were analyzed in each experiment.
Notes and references
1 D. Hanahan and R. A. Weinberg, Cell, 2011, 144, 646–674.
2 C. H. Takimoto and E. Calvo, in Cancer Management: A Mul-
tidisciplinary Approach, ed. R. Pazdur, L. D. Wagman,
K. A. Camphausen, W. J. Hoskins, UBM Medica, London,
2008, pp. 42–58.
3 E. Raschi, V. Vasina, M. G. Ursino, G. Boriani, A. Martoni
and F. De Ponti, Pharmacol. Ther., 2010, 125, 196–218.
4 E. L. Snyder and S. F. Dowdy, Pharm. Res., 2004, 21, 389–
393.
5 E. Vivès, J. Schmidt and A. Pèlegrin, Biochim. Biophys. Acta,
2008, 1786, 126–138.
6 K. M. Stewart, K. L. Horton and S. O. Kelly, Org. Biomol.
Chem., 2008, 6, 2242–2255.
7 S. B. Fonseca, M. P. Pereira and S. O. Kelley, Adv. Drug Deliv-
ery Rev., 2009, 61, 953–964.
8 E. Koren and V. P. Torchilin, Trends Mol. Med., 2012, 18,
385–393.
Confocal microscopy
9 F. Milleti, Drug Discovery Today, 2012, 17, 850–860.
10 J. S. Wadia and S. F. Dowdy, Adv. Drug Delivery Rev., 2005,
57, 579–596.
11 V. Kersemans, K. Kersemans and B. Cornelissen, Curr.
Pharm. Des., 2008, 14, 2415–2427.
12 R. Fischer, M. Fotin-Mleczek, H. Hufnagel and R. Brock,
ChemBioChem, 2005, 6, 2126–2142.
13 S. Pujals, J. Fernández-Carneado, C. López-Iglesias,
M. J. Kogan and E. Giralt, Biochim. Biophys. Acta, 2006,
1758, 264–279.
14 P. Lundberg and U. Langel, J. Mol. Recognit., 2003, 16, 227–
233.
15 M. Gooding, L. P. Browne, F. M. Quinteiro and
D. L. Selwood, Chem. Biol. Drug Des., 2012, 80, 787–809.
16 S. A. Nasrollahi, C. Taghibiglou, E. Azizi and E. S. Farboud,
Chem. Biol. Drug Des., 2012, 80, 639–646.
17 T. Lehto, K. Kurrikoff and U. Langel, Expert Opin. Drug
Deliv., 2012, 9, 823–836.
18 I. Nakase, H. Akita, K. Kogure, A. Gräslund, U. Langel,
H. Harashima and S. Futaki, Acc. Chem. Res., 2012, 45,
1132–1139.
MCF-7 cells grown on coverslips and incubated with different
fluorescent-labeled molecules were washed with cold PBS and
fixed with freshly prepared 4% paraformaldehyde in PBS for
15 min at 4 °C. After washing twice with PBS, the coverslips
were mounted in Mowiol (Calbiochem, Merck, KGaA, Darmstadt,
Germany) or using a fluorescence mounting medium (Dako,
Carpinteria, CA, USA). The images were acquired using a Leica
TCS SP5 laser scanning confocal microscope (Leica Microsys-
tems Heidelberg GmbH) equipped with a DMI6000 inverted
microscope, blue diode (405 nm), argon (458/476/488/514),
diode pumped solid state (561 nm) and HeNe (633) lasers.
Final imaging was performed using ImageJ software.
Statistical analysis
The statistical analysis was performed with the SPSS statistical
software for Windows (version 15.0; SPSS Inc., Chicago, IL,
USA). Quantitative variables were expressed as mean and stan-
dard deviation (SD). The normality of the data was tested using
the Shapiro–Wilk test. The differences between data with
normal distribution and homogeneous variances were ana-
lyzed using the parametric Student’s t test. A value of p < 0.05
was considered significant.
19 F. Schweizer, Eur. J. Pharmacol., 2009, 625, 190–194.
20 Y. Li, Q. Xiang, Q. Zhang, Y. Huang and Z. Su, Peptides,
2012, 37, 207–215.
21 K. Takeshima, A. Chikushi, K.-K. Lee, S. Yonehara and
K. Matsuzako, J. Biol. Chem., 2003, 278, 1310–1315.
22 S. T. Henriques, M. N. Melo and M. A. R. B. Castanho,
Biochem. J., 2006, 399, 1–7.
Acknowledgements
This work was supported by Consolider Ingenio CSD/CSD2010-
00065 from MICINN of Spain and by grant BFU2012-38259 23 J. Fernández-Carneado, M. J. Kogan, S. Pujals and E. Giralt,
from Ministerio de Economia y Competitividad of Spain to F.T. Biopolymers, 2004, 76, 196–203.
We also thank the Catalan DIUE of the Generalitat de Cata- 24 C. Leuschner and W. Hansel, Curr. Pharm. Des., 2004, 10,
lunya (2009SGR637). X.R. thanks financial support from 2299–2310.
INNPLANTA project INP-2011-0059-PCT-420000-ACT1. M.C. 25 C. D. Fjell, J. A. Hiss, R. E. W. Hancock and G. Schneider,
and X.R. thank ICREA Academia Awards. D.S.-C. is a recipient Nat. Rev. Drug Discovery, 2012, 11, 37–51.
of FI fellowship (Generalitat de Catalunya). We also acknowl- 26 D. W. Hoskin and A. Ramanoorthy, Biochim. Biophys. Acta,
edge the Serveis Tècnics de Recerca of the University of Girona
for technical support.
2008, 1778, 357–375.
27 E. Randal, Future Microbiol., 2011, 6, 635–666.
1662 | Org. Biomol. Chem., 2014, 12, 1652–1663
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
28 J. S. Mader and D. W. Hoskin, Expert Opin. Invest. Drugs,
2006, 15, 933–946.
M. J. Sailor and E. Ruoslahti, Proc. Natl. Acad. Sci. U. S. A.,
2007, 104, 932–936.
29 P. Bulet, R. Stöcklin and L. Menin, Immunol. Rev., 2004, 51 E. H. M. Lempens, M. Merkx, M. Tirrel and E. W. Meijer,
198, 169–184. Bioconjugate Chem., 2011, 22, 397–405.
30 H. Jenssen, P. Hamill and R. E. W. Hancock, Clin. Micro- 52 W. Qu, W. H. Chen, Y. Kuang, X. Zeng, S. X. Cheng,
biol. Rev., 2006, 19, 491–511.
31 M. Zasloff, Nature, 2002, 415, 389–395.
X. Zhou, R. X. Zhuo and X. Z. Zhang, Mol. Pharmaceutics,
2013, 10, 261–269.
32 B. Bechinger and K. Lohner, Biochim. Biophys. Acta, 2006, 53 E. Badosa, R. Ferre, M. Planas, L. Feliu, E. Besalú,
1758, 1529–1539.
33 R. E. W. Hancock and H. G. Sahl, Nat. Biotechnol., 2006, 24,
1551–1557.
34 J. F. Marcos and M. Gandía, Expert Opin. Drug Discov.,
2009, 4, 659–671.
35 H. W. Huang, Biochim. Biophys. Acta, 2006, 1758, 1292–
1302.
36 B. Fang, H. Y. Guo, M. Zhang, L. Jiang and F. Z. Ren, FEBS
J., 2013, 280, 1007–1017.
37 G. Drin, S. Cottin, E. Blanc, A. R. Rees and J. Temsamani,
J. Biol. Chem., 2003, 278, 31192–31201.
J. Cabrefiga, E. Bardají and E. Montesinos, Peptides, 2007,
28, 2276–2285.
54 K. Eggenberger, C. Mink, P. Wadhwani, A. S. Ulrich and
P. Nick, ChemBioChem, 2011, 12, 132–137.
55 K. Melikov and L. V. Chernomordik, Cell. Mol. Life Sci.,
2005, 62, 2739–2749.
56 R. Fischer, O. Mader, G. Jung and R. Brock, Bioconjugate
Chem., 2003, 14, 653–660.
57 I. Güell, J. Cabrefiga, E. Badosa, R. Ferre, M. Talleda,
E. Bardají, M. Planas, L. Feliu and E. Montesinos, Appl.
Environ. Microbiol., 2011, 77, 2667–2675.
38 K. K. Hou, H. Pan, G. M. Lanza and S. A. Wickline, Bioma- 58 S. Monroc, E. Badosa, E. Besalú, M. Planas, E. Bardají,
terials, 2013, 34, 3110–3119. E. Montesinos and L. Feliu, Peptides, 2006, 27, 2575–2584.
39 S. Sandgren, A. Wittrup, F. Cheng, M. Jönsson, E. Eklund, 59 E. Badosa, R. Ferre, J. Francés, E. Bardají, L. Feliu,
S. Busch and M. Belting, J. Biol. Chem., 2004, 279, 17951–
17956.
M. Planas and E. Montesinos, Appl. Environ. Microbiol.,
2009, 75, 5563–5569.
40 H. Myrberg, L. Zhang, M. Mäe and U. Langel, Bioconjugate 60 S. Aroui, S. Brahim, M. De Waard and A. Kenani, Biochem.
Chem., 2008, 19, 70–75. Biophys. Res. Commun., 2010, 391, 419–425.
41 M. Mäe, H. Myrberg, S. El-Andaloussi and U. Langel, 61 H. Yang, S. Liu, H. Cai, L. Wan, S. Li, Y. Li, J. Cheng and
Int. J. Pept. Res. Ther., 2009, 15, 11–15. X. Lu, J. Biol. Chem., 2010, 285, 25666–25676.
42 S. B. Fonseca and S. O. Kelley, ACS Med. Chem. Lett., 2011, 62 J. Song, M. Kai, W. Zhang, J. Zhang, L. Liu, B. Zhang, X. Liu
2, 419–423. and R. Wang, Peptides, 2011, 32, 1934–1941.
43 J. Hoyer, U. Schatzschneider, M. Schulz-Siegmund and 63 S. Jones and J. Howl, Bioconjugate Chem., 2012, 23, 47–56.
I. Neundorf, Beilstein J. Org. Chem., 2012, 8, 1788–1797.
44 S. R. Rajski and R. M. Williams, Chem. Rev., 1998, 98,
2723–2795.
45 I. Martín, M. Teixidó and E. Giralt, Pharmaceuticals, 2010,
3, 1456–1490.
46 S. Majumdar and T. J. Siahaan, Med. Res. Rev., 2012, 32,
637–658.
47 N. Svensen, J. G. A. Walton and M. Bradley, Trends Pharma-
col. Sci., 2012, 33, 186–192.
48 P. Laakkonen and K. Vuorinen, Integr. Biol., 2010, 2, 326–
337.
49 O. H. Aina, R. Liu, J. L. Sutcliffe, J. Marik, C.-X. Pan and
K. S. Lam, Mol. Pharmaceutics, 2007, 4, 631–651.
50 D. Simberg, T. Duza, J. H. Park, M. Essler, J. Pilch,
L. Zhang, A. M. Derfus, M. Yang, R. M. Hoffman, S. Bhatia,
64 J. Farrera-Sinfreu, E. Giralt, S. Castel, A. Albericio and
M. Royo, J. Am. Chem. Soc., 2005, 127, 9459–9468.
65 E. Macia, M. Ehrlich, R. Massol, E. Boucrot, C. Brunner
and T. Kirchhausen, Dev. Cell., 2006, 10, 839–850.
66 N. M. Sherer, M. J. Lehmann, L. F. Jimenez-Soto,
A. Ingmundson, S. M. Horner, G. Cicchetti, P. G. Allen,
M. Pypaert, J. M. Cunningham and W. Mothes, Traffic,
2003, 4, 785–801.
67 A. Benmerah, C. Lamaze, B. Bègue, S. L. Schmid, A. Dautry-
Varsat and N. Cerf-Bensussan, J. Cell Biol., 1998, 140, 1055–1062.
68 E. Kaiser, R. L. Colescott, C. D. Bossinger and P. Cook,
Anal. Biochem., 1970, 34, 595–598.
69 A. Guaragna, A. Chiaviello, C. Paolella, D. D’Alonzo,
G. Palumbo and G. Palumbo, Bioconjugate Chem., 2012, 23,
84–96.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 1652–1663 | 1663