Cathepsin D activity and selectivity in the acidic conditions of a tumor microenvironment: Utilization in the development of a novel Cathepsin D substrate for simultaneous cancer diagnosis and therapy

Article abstract of DOI:10.1016/j.biochi.2013.07.010

Pro-Cathepsin D (pCD) is an aspartyl endopeptidase which is over expressed in many cancers. This over expression generally led to its secretion into the extracellular culture medium of cancer cells. Moreover, pCD can auto activate and cleave its substrates at an acidic pH compatible with that found in tumor microenvironments (TME). Thus, exploiting these two pathological characteristics of TME offers the opportunity to develop new protease-activated vector on the basis of their specific substrate structures. The aim of this study was to validate new pCD substrates in the extracellular pH conditions of TME. As a first step, we investigated the effect of pH on the catalytic activity and selectivity of mature Cathepsin D (CD). It was found that the increase in the pH of the media led to a decrease in the reaction rate. However, the specificity of mature CD was not affected by a variation in pH. In the second step, the effect of the substrate structure was studied. We demonstrated that the substrate structure had a significant effect on the catalytic activity of CD. In fact, some modifications in peptide structure induced a change in the catalytic behavior that involved a substrate activation phenomenon. We suggest that this activation may be related to the amphiphilic nature of the modified peptide that may induce an interfacial activation mechanism. Finally, pCD, which is the major form found in the extracellular culture medium of cancer cells, was used. We demonstrated that the proform of CD cleave the modified peptide 5 at pH 6.5 with the same cleavage selectivity obtained with the mature form of the protease. These data provide a better understanding of CD behavior in tumor microenvironment conditions and this knowledge can be used to develop more specific tools for diagnosis and drug delivery.

Full text of DOI:10.1016/j.biochi.2013.07.010

Biochimie 95 (2013) 2010e2017
Contents lists available at ScienceDirect
Biochimie
journal homepage: www.elsevier.com/locate/biochi
Research paper
Cathepsin D activity and selectivity in the acidic conditions of a tumor
microenvironment: Utilization in the development of a novel
Cathepsin D substrate for simultaneous cancer diagnosis and therapy
Oussama Achour, Nicolas Bridiau, Meriem Kacem, Régis Delatouche,
Stéphanie Bordenave-Juchereau, Frédéric Sannier, Valérie Thiéry, Jean-Marie Piot,
*
Thierry Maugard, Ingrid Arnaudin
Université de La Rochelle, UMR CNRS 7266, LIENSs, Equipe Approches Moléculaires, Environnement-Santé, Département de Biotechnologies, Avenue Michel
Crépeau, 17042 La Rochelle, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Pro-Cathepsin D (pCD) is an aspartyl endopeptidase which is over expressed in many cancers. This over
expression generally led to its secretion into the extracellular culture medium of cancer cells. Moreover,
pCD can auto activate and cleave its substrates at an acidic pH compatible with that found in tumor
microenvironments (TME). Thus, exploiting these two pathological characteristics of TME offers the
opportunity to develop new protease-activated vector on the basis of their specific substrate structures.
The aim of this study was to validate new pCD substrates in the extracellular pH conditions of TME. As a
first step, we investigated the effect of pH on the catalytic activity and selectivity of mature Cathepsin D
(CD). It was found that the increase in the pH of the media led to a decrease in the reaction rate. However,
the specificity of mature CD was not affected by a variation in pH. In the second step, the effect of the
substrate structure was studied. We demonstrated that the substrate structure had a significant effect on
the catalytic activity of CD. In fact, some modifications in peptide structure induced a change in the
catalytic behavior that involved a substrate activation phenomenon. We suggest that this activation may
be related to the amphiphilic nature of the modified peptide that may induce an interfacial activation
mechanism. Finally, pCD, which is the major form found in the extracellular culture medium of cancer
cells, was used. We demonstrated that the proform of CD cleave the modified peptide 5 at pH 6.5 with
the same cleavage selectivity obtained with the mature form of the protease. These data provide a better
understanding of CD behavior in tumor microenvironment conditions and this knowledge can be used to
develop more specific tools for diagnosis and drug delivery.
Received 27 February 2013
Accepted 8 July 2013
Available online 16 July 2013
Keywords:
Cathepsin D
Pro-Cathepsin D
Tumor microenvironment
Substrate structure
Interfacial activation
Ó 2013 Published by Elsevier Masson SAS.
1. Introduction
of normal tissues because most tumors develop a pathophysio-
logical microenvironment characterized by low oxygen levels
The tumor microenvironment (TME) consists of cells, soluble
factors, signaling molecules, extracellular matrix proteins, and
mechanical cues that surround and feed a tumor cell [1]. TME can
promote malignant transformation, and support tumor growth and
invasion [2]. TME conditions are considerably different from those
(hypoxia) [3], low glucose concentration [4] and micro-acidic
conditions [5]. In addition, a variety of hydrolytic enzymes are
over expressed by tumor cells, activated during the different stages
of tumor progression and often hypersecreted, forming the liquid
milieu of the TME [6]. Based on the understanding of chemical and
functional particularities of the TME, new specific molecular targets
have been identified that enhance the effective development of
new diagnostic tools [7] that are more selective for cancer cells,
leading to personalized treatment. For example, one chemical
characteristic that is common to the majority of solid tumors is
their acidic microenvironment, with an extracellular pH (pHe)
ranging from 5.3 to 7.2 [8]. Whatever the mechanism for acidifi-
cation, the result is a pericellular environment that may favor the
Abbreviations: pCD, pro-Cathepsin D; TME, tumor microenvironment; CD,
Cathepsin D; ppCD, pre-pro-Cathepsin D; MCA, 7-methoxycoumarin-4-acetyl; DNP,
dinitrophenyl; SA, sodium acetate; MES, 2-(N-morpholino)ethanesulfonic acid;
MOPS, 3-(N-morpholino)propanesulfonic acid.
* Corresponding author. Tel.: þ33 546 45 85 62; fax: þ33 546 45 82 65.
E-mail addresses: ingrid.fruitier@univ-lr.fr, ifruitie@univ-lr.fr, ingrid.arnaudin@
orange.fr (I. Arnaudin).
0300-9084/$ e see front matter Ó 2013 Published by Elsevier Masson SAS.
http://dx.doi.org/10.1016/j.biochi.2013.07.010

O. Achour et al. / Biochimie 95 (2013) 2010e2017
2011
activity of proteases such as cysteine or aspartic cathepsins that
2. Materials and methods
have slightly acidic pH optima. In addition, an acidic pHe induces a
redistribution of lysosomes to the tumor cell surface and secretion
of their proteases [9].
2.1. Materials
Most of the research in our laboratory has focused on one
lysosomal enzyme, aspartic CD (CD, EC 3.4.23.5), as an entry point
to innovative cancer treatments [10,11]. CD is a lysosomal aspartyl
endopeptidase present in all cells and tissues and associated with
tumor progression and serve as autocrine growth factor for several
cancer cell types. CD may stimulate cancer growth via its enzymatic
activity by digesting various chemokines and may therefore
attenuate the anti-tumoral immune response [12,13]. Pre-pro-
Cathepsin D (ppCD) is synthesized in the rough endoplasmic re-
ticulum. After the signal peptide is cleaved, the 52 kDa pro-
Cathepsin D (pCD) is targeted to intracellular vesicular structures
like lysosomes, endosomes or phagosomes. In a tumor, pCD escapes
normal targeting mechanisms and can be hypersecreted into the
extracellular space. At slightly acid pH, secreted pCD undergoes
partial maturation and becomes active [14,15]. Therefore, an in-
crease in pCD level is observed in several human neoplastic tissues
like breast [16], thyroid [17] and prostate sarcoma [18].
The catalytic site of CD consists of two aspartic acid residues
Asp33 and Asp231, located in the triad sequences of Asp33eThr34e
Gly35 and Asp231eThr232eGly23 [19]. The reaction mechanism
that has been generally accepted for aspartic protease involves
amide bond hydrolysis through an active site water molecule [20].
In this mechanism, the hydrolysis of the peptide bonds catalyzed by
CD occurs in two stages. In the first stage, two simultaneous proton
transfer between the water molecule and the carboxyl ion Asp33
and between the carboxyl group of Asp231 and the oxygen atom of
the carboxyl group in the substrate lead to the formation of the
enzymeesubstrate complex (EeS), in which the components are
bound by ionic and hydrogen bonds. In the next stage, a mechanism
of double transfer leads to the decomposition of the indirect
product, in which a proton of the hydroxyl group is transferred onto
the Asp33, whereas the Asp231 is transported onto the nitrogen
atom, resulting in the hydrolysis of the peptide bond in the sub-
strate. Concerning substrate selectivity, CD preferentially cleaves
peptide bonds found within the polypeptide chain formed by hy-
drophobic amino acid residues like aromatic trp, tyr and phe or
long-chain aliphatic amino acids such as leu and ile [21]. Moreover,
peptide sequences containing less than five amino acids are resis-
tant to CD hydrolysis [22].
All reagents, unless specified otherwise, were purchased from
Sigma AldrichÒ (St Quentin Falavier, France).
A mature form of bovine CD (34 þ 14 kDa) was also provided by
Sigma AldrichÒ and had an activity of 5.0 units/mg of protein. One
unit, as defined by the supplier, will produce an increase in A280 of
1.0 per min per mL at pH 3.0 and at 37 ꢀC, measured as TCA-soluble
products using hemoglobin as a substrate (1 cm light path).
Recombinant pCD (52 kDa) was purchased from R&D systemsÒ
(USA) with a purity higher than 95%.
CD activity was measured using fluorogenic peptide 1 (R[K-DNP]
LRFFLIPK[G-MCA]) supplied by Sigma AldrichÒ (MCA: 7-Methoxyc-
oumarin-4-Acetyl and DNP: dinitrophenyl). Peptides 3 (LLVVFF)
was prepared by standard solid phase peptide synthesis with a
Fmoc strategy using a Rink amide resin (0.8e1 mmol/g loading) on
a 0.1 mmol scale. The synthesis was realized using an automated
microwave peptide synthesizer CEM Liberty 1 with coupling time
of 15e30 min depending on the amino acid with microwave
heating at 40 ꢀC. The others peptides (Peptides 2, 4 and 5, see
Table 3 for structures) used in this study were synthesized by
Genosphere BiotechnologyÒ (France) with purity greater than 85%.
2.2. Kinetics assays of CD using the fluorogenic substrate
Hydrolysis of fluorogenic peptide 1 was performed in white 96
half-well plates (CorningÒ #3693) using a BMG Labtech Fluostar
Omega spectrofluorometer. The fluorescence was measured at
lem ¼ 390 nm and lex ¼ 330 nm with an interval of 20 s. The re-
action was conducted at 37 ꢀC in a final volume of 100
mL with a
concentration of 8.3 ng/mL of mature CD. Three different buffers
were used to adjust the reaction pH: Sodium acetate buffer (for pH
3.7, pH 4.5 and pH 5.6), 2-(N-morpholino)ethanesulfonic acid
(MES) buffer (for pH 5.6, pH 6 and pH 6.5) and; 3-(N-morpholino)
propanesulfonic acid (MOPS) buffer (for pH 6.5 and pH 6.8). Sub-
strate concentrations were varied from 5
20 M, 25 M, 30 M).
mM to 30 mM (5 mM,10 mM,
m
m
m
2.3. Kinetics assays of CD using the synthetic model peptide
substrates
This study is a part of a project that aims to design a new
intelligent CD cleavable vector for tracking and imaging the TME of
breast cancer based on the response to unique pathological pat-
terns of TMEs (micro-acidic extracellular pH associated with
hypersecreted pCD). The ability to evaluate specific enzyme activity
in vivo would thus have considerable clinical applications like the
improvement of early detection of diseases and targeted drug de-
livery in localized areas or tissues. In the last few decades, some
research teams have demonstrated the feasibility of such an
approach by targeting proteases such as Cathepsin K or CD [23,24].
In this hostile environment, pCD is found in pH conditions that are
far from their known optimal i.e. 3.5. We therefore started by
investigating the effect of pH on the catalytic activity of the mature
form of CD (34 þ 14 kDa) using a commercial fluorogenic substrate.
The effect of pH on CD specificity was also studied using a model
synthetic substrate. As the objective is the synthesis of a CD-
sensitive substrate, the effect of substrate structure on both cata-
lytic activity and specificity of the protease was investigated using
peptides selected according to the specification mentioned below.
Finally, the results were validated using the pCD form that is
considered to be the major form which is over expressed in the
extracellular microenvironment of breast cancer.
Peptides 2, 3, 4 and 5 (Table 3) were first dissolved in pure DMSO
at a concentration of 4 mM. Hydrolysis of the peptides was per-
formed in a final volume of 500 mL in the presence of 8.3 ng/mL of
mature CD and the reaction was conducted at 37 ꢀC in a RadleysÒ
reactor. Three different buffers were used to adjust the reaction pH:
Sodium acetate buffer (for pH 3.7, pH 4.5 and pH 5.6), Sodium MES
Buffer (for pH 5.6, pH 6 and pH 6.5) and Sodium MOPS (for pH 6.5
and pH 6.8). Substrate concentrations were varied from 10
400 M (10 M, 35 M, 70 M, 140 M, 200 M, 400 M). A zero
time aliquot (50 L) was removed from the solution prior to
mM to
m
m
m
m
m
m
m
m
enzyme addition. Aliquots were taken over time (0 h, 0.5 h, 1 h,
1.5 h, 2 h, 4 h, 24 h, 28 h, 34 h, 48 h, 72 h,120 h) and the reaction was
stopped by boiling for 5 min followed by a 5 min centrifugation at
10,000ꢁ g. Samples were stored at ꢂ20 ꢀC for at most 48 h prior to
their analysis by LC/MS-ESI (cf. Section 2.6).
2.4. Phospholipid micelle preparation
Phosphatidylcholin and cardiolipin were diluted in ethanol
(1 mL) at a concentration of 0.12 mM and 0.05 mM, respectively.
Ethanol was dried at 37 ꢀC and phospholipid solutes were
dissolved in water (1 mL). This aqueous solution was sonicated for

2012
O. Achour et al. / Biochimie 95 (2013) 2010e2017
10 min using a probe-type UP50H ultrasonicator provided by
Hielscher (Germany) in order to form micelles. The ultrasonicator is
equipped with a probe (micro tip MS3) with a diameter of 3 mm
vi ¼ Vmax$½Sꢃ=Km þ ½Sꢃ
(1)
(2)
n
n
vi ¼ Vmax$½Sꢃ =Km þ ½Sꢃ
that provides an amplitude of 180
mm. The processor generates
To allow visualization of fitting quality, the experimental data
points are presented as plots along with the theoretical lines fitted
by “Enzyme Kinetics 2004 1.3” software.
mechanical longitudinal vibrations with a frequency of 30 kHz
produced by electrical excitation.
2.5. Kinetics assays of pCD using a synthetic model peptide
substrate
3. Results and discussion
3.1. Effect of pH on CD activity using a fluorescent substrate
Recombinant pCD (52 kDa) was autoactivated by acidic condi-
tions for 1 h at 37 ꢀC in 50
mL of 0.1 M sodium acetate buffer (pH 3.5)
The effect of pH on mature CD activity was investigated using
fluorescence labeled peptide 1 as a substrate. The kinetics param-
eters were determined in different pH conditions, from the opti-
mum pH of the enzyme (3.7) to the pH value of TME (6.8). The
fluorescence intensity reflecting product apparition was followed
over time. Fig. 1 shows the effect of pH on the products appearing
containing 0.2 M NaCl. The reaction media was neutralized with
1 M Tris buffer, pH 8.8 [25]. Peptide 5 ([aoa-K]LLVVFF[D-PEG2]) was
first dissolved in pure DMSO at a concentration of 4 mM. Hydrolysis
of the peptide was performed in a final volume of 150 mL in the
presence of 8.3 mg/mL of pCD and the reaction was conducted at
37 ꢀC in a RadleysÒ reactor. Two different buffers were used to
adjust the reaction pH to 3.7 (optimum condition with sodium
acetate buffer) and 6.5 (tumoral pH condition with sodium MES
over time from a starting substrate concentration of 30
different pH conditions.
mM for the
As expected the reaction was slower at higher pH. Nevertheless,
CD was still active even at pH 6.8, which is far from its optimal pH
but representative of the pH condition of the TME.
Buffer). Substrate concentrations were fixed at 200
tion was performed with and without 24 L of phospholipid mi-
celles prepared as mentioned below. A zero time aliquot (20 L)
mM. The reac-
m
m
We also observed that the products appear after a latency time
that is longer at a higher pH. This latency time was very short for pH
3.7 and increased when the reaction pH moved away from the
optimal pH. For example, at pH 6.8 the products appeared only
10 min after the reaction started while they appeared immediately
after the reaction started at pH 3.7. We note that no product
appeared under the same conditions in the absence of enzyme even
after 24 h. This latency time corresponds to the pre-steady-state
period that occurs before the enzymeesubstrate complex (EeS) is
formed at the same rate that it decomposes. The formation of the
(EeS) complex is probably retarded by the fact that the solubility of
the peptide is reduced when the pH increases. Moreover, in the
mechanism of peptide hydrolysis by CD described here, the (EeS)
complex is highly dependent on the protonation state of the active
site. In acid pH, the first proton transfer that leads to the formation
of the (EeS) complex is certainly faster and easier than when this
transfer takes place in less acidic solvent [20].
was removed from the solution prior to the addition of the enzyme.
Aliquots were taken over time (0 h, 0.5 h, 1 h, 2 h, 4 h, 24 h, 120 h)
and the reaction was stopped by boiling for 5 min followed by a
5 min centrifugation at 10,000ꢁ g. Samples were stored at ꢂ20 ꢀC
for at most 48 h prior to their analysis by LC/MS-ESI (cf. Section 2.6).
2.6. Quantitative analysis of reaction products by LC-MS analysis
Structural and quantitative analysis of the pure and reaction
products were conducted using an LC/MS-ESI system from Agilent
(1200 Infinity LC/MSD Quadrupole mass spectrometer) with a
reverse phase C18 phenomenex column (Luna 5
m C18 100A
150 ꢁ 2 mm). The column was maintained at room temperature
(25 ꢀC) and the elution was performed at 0.3 mL/min using a
gradient that increased from 15 to 35% of eluant B over a 28 min
period after an initial isocratic step (15% B) lasting 2 min. (Eluant A:
water with 0.1% acetic acid and Eluant B: acetonitrile with 0.1%
acetic acid). For peptide 5, the gradient went from 10% to 55% over a
28 min period after an initial isocratic step (10% B) lasting 2 min.
Mass spectrometer analyses were performed in positive mode
with a capillary voltage of 3000 V in a mass range of 50e1600 m/z
and with a fragmentor value of 10. Drying gas flow, nebulizer
pressure and drying gas temperature were set at 13 mL/min, 50 psi
and 200 ꢀC, respectively.
The initial rates were measured as the slope of the linear
segment of the kinetics plot and were fitted with all the single
substrate kinetics models provided by the Enzyme Kinetics Module
of Sigma Plot. For peptide 1, the best-fit results were obtained with
the Hill models (Equation (2)). The kinetics parameters for different
pH conditions were calculated using nonlinear least-squares
The products were detected and quantified at a wavelength of
210 nm based on a standard curve obtained with pure products.
The products that were eluted at similar times were quantified by
mass spectrometry after the extraction of the most abundant ion
using a standard curve obtained with pure products. Chemstation
off-line software was used for processing.
10
pH 3,7 SA
pH 4,5 SA
pH 5,6 SA
pH 5,6 MES
pH 6 MES
pH 6,5 MES
pH 6,5 MOPS
pH 6,8 MOPS
8
6
4
2
0
2.7. Calculation of kinetics parameters
All initial rate data were fitted to all the single substrate kinetics
models provided by the Enzyme Kinetics Module of Sigma Plot,
“Enzyme Kinetics 2004 1.3” (Systat Software Inc., San Jose, USA),
using nonlinear least-squares regression analysis.
0
20
40
60
80
Reaction time (Minutes)
Initial rate data corresponding to the hydrolysis kinetics of
peptides 2 and 3 were best fitted to the Michaelis Menten Equation
(1) [26]. Initial rate data corresponding to the hydrolysis kinetics of
peptides 1 and 5 were best fitted to the Hill Equation (2) [26].
Fig. 1. Effect of pH on the kinetics of product apparition from an initial substrate
concentration of 30 M. The data represent the averages of at least triplicate runs
conducted with the substrate R[K-DNP]LRFYFLIPK[G-MCA]. Y: Cleaved peptide bond.
m
SA: sodium acetate buffer; MES: sodium MES buffer; MOPS: sodium MOPS buffer.

O. Achour et al. / Biochimie 95 (2013) 2010e2017
2013
8e+5
regression analysis based on the model equation. The effect of pH
on Vmax is presented in Fig. 2.
Sample taken before reaction
Sample taken after 24 hours
The maximum rate of the hydrolysis of peptide 1 by CD
decreased as the pH of the medium increased. Indeed, at pH 3.7 the
LLVVYPWTQRFF
6e+5
4e+5
2e+5
0
maximum rate was 1.9 nmol minꢂ1 m ꢂ1, which is 13 times higher
g
LLVVYPWTQRF
&
than the maximum rate observed at pH 4.5. At pH that are found in
the tumor microenvironment (between 6 and 6.8), the maximum
VVYPWTQRFF
rate was between 5 ꢁ 10ꢂ2 and 1 ꢁ 10ꢂ3 nmol minꢂ1 mgꢂ1
.
3.2. Effect of pH on CD specificity using a model dodecapeptide
(Peptide 2)
LVVYPWTQRF
VVYPWTQRF
LVVYPWTQRFF
In order to investigate the effect of pH on CD specificity, a model
dodecapeptide (peptide 2) LLVVYPWTQRFF, belonging to the
hemorphin class, was used as a substrate. The choice of this peptide
was motivated by its stable and specific release by CD from he-
moglobin [27]. Moreover, this peptide contains in its primary
structure the main bonds that are recognized and cleaved by CD
[21,28].
10
15
20
25
30
Time (Minutes)
Fig. 3. Liquid chromatography analysis of a sample taken before peptide 2 hydrolysis
by CD and after 24 h at pH 3.7 in sodium acetate. Separation was performed on a
Peptide 2 was subjected to CD hydrolysis in different pH con-
ditions and at different concentrations. Fig. 3 presents an example
of a chromatographic analysis of a sample taken before the reaction
and after 24 h of hydrolysis by CD at pH 3.7. The identified ions and
their corresponding sequences are presented in Table 1. All ions
presented in Table 1 were found in all pH conditions. The identi-
fication of the product structure indicates that CD cleaves peptide 2
at the C-terminus FYF and the N-terminus LYL, LYV and LLYV (Y:
cleaved peptide bond). As we have already demonstrated [27], the
nonapeptide VVYPWTQRF (product 2-e) is resistant to hydrolysis
and accumulated in the reaction media. These data confirm that CD
preferentially cleaves hydrophobic amino acids [21] and demon-
strates an exolytic activity in addition to the endolytic activity
generally described in the literature [22].
reverse phase C18 phenomenex column (Luna 5
m
C18 100A 150 ꢁ 2 mm). The column
was maintained at room temperature (25 ꢀC) and elution was performed at 0.3 mL/min
using a gradient rising from 15 to 35% of eluant B over a 28 min period after an initial
isocratic step (15% B) lasting 2 min. (Eluant A: water with 0.1% acetic acid and Eluant B:
acetonitrile with 0.1% acetic acid).
Despite the general slowdown in reaction rate, the cleavage
profile remained unchanged at higher pH (Table 1). Based on these
results, we concluded that the specificity of CD does not seem to be
affected by the pH of the media.
3.3. Effect of substrate structure on mature CD catalytic activity and
selectivity
The substrate and products of hydrolysis were quantified over
time. Fig. 4 presents the results obtained for the reaction at pH 4.5
3.3.1. Effect of substrate structure on CD specificity
with a starting substrate concentration of 140
mM. Only two major
In order to study the effect of the substrate structure on the
selectivity of CD cleavage, several peptides with different sequences
were used. The choice of these peptides was made based on the
cleavage sites identified in the previous section and in the objective
of the incorporation of the more selective peptide into a tumor
targeting vector. First, a shorter hexapeptide (peptide 3) (LLVVFF)
containing previously identified cleavage sites was assayed. For
latter functionalization, a lysyl and an aspartyl residue were added
to peptide 3 on the N- and the C-terminus sides, respectively, to
form the octapeptide 4 (KLLVVFFD). Finally, peptide 4 was func-
tionalized by coupling an aminooxy-acetyl group (AOA) onto the N-
terminus lysyl residue and a polyethyleneglycol with 2 monomer
units (PEG2) onto the C-terminus aspartyl residue to give octa-
peptide 5 ([aoa-K]LLVVFF[D-PEG2]). These two changes will allow
products appeared at the first stage of the hydrolysis reaction.
These products correspond to those obtained after the cleavage of
the amino bonds between the LYL, LYV and LLYV at the N-terminus
side. Product 2-c (LLVVYPWTQRF), which is derived from substrate
hydrolysis after cleavage between FYF at the C-terminus side, also
appeared during the first stage of the reaction, albeit at a low
concentration because it is rapidly consumed, leading to the
apparition of smaller peptides (product 2-d and product 2-e) after
the cleavage of the peptide bond between LYL, LYV and LLYV at the
N-terminus side.
2,00
Sodium acetate buffer
MES buffer
MOPS buffer
Table 1
1,90
Mass spectrometry identification of the hydrolysis products of peptide 2 catalyzed
by mature CD at all tested pH.
1,80
0,20
Peptide sequence
Calculated
mass
Elution
time
(min)
Observed ion
(g molꢂ1
)
0,15
0,10
0,05
0,00
Peptide 2
LLVVYPWTQRFF
LVVYPWTQRFF
1568.88
23
20
[Mþ2H]2þ ¼ 785.1
[Mþ3H]3þ ¼ 523.8
[Mþ2H]2þ ¼ 728.5
[Mþ3H]3þ ¼ 486.1
[Mþ2H]2þ ¼ 672
[Mþ2H]2þ ¼ 711
[Mþ2H]2þ ¼ 655
[Mþ2H]2þ ¼ 598.4
[M þ H]þ ¼ 245.5
Product 2-a
1455.72
Product 2-b
Product 2-c
Product 2-d
Product 2-e
Product 2-f
VVYPWTQRFF
LLVVYPWTQRF
LVVYPWTQRF
VVYPWTQRF
LL
1342.56
1421.71
1308.55
1195.39
244.34
18
17.7
14
11
1.5
3,5
4,0
4,5
5,0
5,5
6,0
6,5
7,0
pH
Fig. 2. Effect of pH on the Vmax of mature CD hydrolysis of fluorescence labeled peptide
1. The data represent the averages of triplicate runs at least.

2014
160
O. Achour et al. / Biochimie 95 (2013) 2010e2017
the aminooxyacetic acid, was resistant to CD hydrolysis. The [aoa-
LLVVYPWTQRFF
LVVYPWTQRFF
VVYPWTQRFF
LLVVYPWTQRF
LVVYPWTQRF
VVYPWTQRF
K]LLVVF product was not detected in mass spectroscopy either
because it was rapidly used up by CD either because the FF bond is
cleaved only after the cleavage of the LV bonds.
140
120
100
80
3.3.2. Effect of substrate structure on CD catalytic activity
Fig. 5 presents the effect of substrate concentration on the initial
rate of hydrolysis of peptides 1, 2, 3 and 5 at pH 3.7.
60
40
These plots present the theoretical lines obtained with the best-
fit results along with the experimental data points. In all cases, the
initial rate increased with the increase in substrate concentration up
to the maximum rate (Vmax) obtained for a saturating concentration
of substrate, which corresponds to the initial rate. The velocity
curves obtained for peptide 2 and peptide 3 have a hyperbolic shape,
reflecting a Michaelian behavior. Moreover, the best-fit results were
obtained with the MichaeliseMenten equation (Equation (1)).
The kinetics data for the hydrolysis of modified peptides 1 and 5
present the best-fit with the theoretical data calculated using the
Hill equation (Equation (2)). The kinetics of CD seemed to be
affected by substrate structure. This model describes the behavior
of an enzyme that has multiple substrate fixation sites. In fact, the
velocity curve obtained for peptide 1 and peptide 5 has a sigmoid
shape, which may reflect an allosteric behavior. Allosteric enzymes
display altered kinetics properties that may occur when enzymes
have multiple substrate fixation sites [26]. However, the crystal
structure of CD [29] and previous kinetics studies [20] indicate that
CD has only one substrate fixation site.
Moreover, this sigmoid shape may also reflect an interfacial
activation phenomenon [30]. This phenomenon is well known for
lipase, whose catalytic activity increases at the waterelipid inter-
face. This activation is associated with a conformational change. In
aqueous homogenic solvent, the enzyme has an inactive confor-
mation and the active site is inaccessible for the substrate since it is
not exposed to the aqueous solvent. However, in the presence of an
interface (heterogenic solvent), the enzyme switches to an active
conformation, rendering the active site accessible to the substrate.
Some data found in the literature support the hypothesis that CD is
activated by an interfacial mechanism, like lipase. In fact, some
20
0
0
20
40
60
80
100
120
140
Reaction time (hours)
Fig. 4. Substrate extinction and product appearance over time at pH 4.5 (SA buffer)
with an initial substrate concentration of 140
triplicate runs at least.
mM. The data represent the averages of
the incorporation of the latter peptide into the adapted vector
through covalent bonds.
These peptides were subjected to CD hydrolysis in different pH
conditions and at different concentrations. Samples that were
taken over time were analyzed by liquid chromatography on a C18
column and the products were identified by mass spectrometry.
The ions identified and their corresponding sequences are pre-
sented in Table 2. All ions presented in Table 2 were found in all pH
conditions.
These results show that peptide 3 was cleaved at the same site
as that identified for peptide 2, which is twice as long. Both the N-
terminus and the C-terminus sides were recognized and cleaved by
CD at the LYL, LYV, LLYV and FYF bonds. Unlike peptide 3, peptide 4
was resistant to CD hydrolysis and no product was detected in the
media. These results pointed to the importance of the amino acids
that surround the cleaved site. Indeed, the lysyl and aspartyl that
were added at the N- and the C-terminus sides, for latter func-
tionalization, were certainly too hydrophilic for hydrolysis. Many
studies have shown that the presence of hydrophobic amino acids
around the cleavage site is important for substrate recognition by
CD [21]. Peptide 5 ([aoa-K]LLVVFF[D-PEG2]) was recognized and
cleaved by CD but only at two different positions i.e. the FYF bonds
and the LYV bonds. The LL bond that was cleaved on peptides 2 and
3 was no longer cleaved by CD. The addition of the PEG2 group
reestablished the enzymatic recognition at the C-terminus side. The
amino bond that is located at the N-terminus side, which contains
studies show that the proform of CD interacts with the
the low-density lipoprotein (LDL) receptor-related protein-1
(LRP1 ), that directs pCD to the lipid rafts at the cell membrane
b chain of
b
[31]. Another study pointed out the activator effect of some
amphiphilic molecules that form micelles such as phospholipids
and phosphosphingolipids [32]. The authors of this study noted
that cardiolipin micelles were able to activate the CD lysosomal
fraction by increasing its activity up to 40 times. According to the
authors, the lysosomal membranes, which is similar to micelles,
participate in the activation of the mature form of CD.
Table 2
Mass spectrometry identification of the hydrolysis products of peptide 3 and peptide
5 catalyzed by CD at all tested pH.
Peptide sequence Calculated Elution Observed ion
In our case, both substrates (peptides 1 R[K-DNP]LRFFLIPK[G-
MCA] and 5 [aoa-K]LLVVFF[D-PEG2]) that showed a sigmoid
shape are characterized by the addition of a hydrophilic structure
like the aminooxyacetic acid for peptide 5 or the methylcoumar-
inacetic acid molecule for peptide 1. Added to the hydrophobic
amino acid sequence that is needed for the peptide recognition by
CD, the molecule formed may have an amphiphilic behavior that
contributes to the interfacial activation of CD.
The kinetics parameters of the CD hydrolysis of each peptide
were calculated by non-linear regression using the best-fit model
and the enzyme kinetics sigma plot modules. Table 3 presents the
kinetics parameters obtained for all peptides tested.
mass
time
(min)
(g molꢂ1
)
Peptide 3
LLVVFF
736.95
623.79
510.63
589.78
476.62
363.46
244.34
980.22
23
17
10
18
11
5
[M þ H]þ ¼ 738.1
[M þ H]þ ¼ 624.6
[M þ H]þ ¼ 511.5
[M þ H]þ ¼ 590
Product 3-a LVVFF
Product 3-b VVFF
Product 3-c LLVVF
Product 3-d LVVF
Product 3-e VVF
Product 3-f LL
[M þ H]þ ¼ 476.9
[M þ H]þ ¼ 363.9
[M þ H]þ ¼ 245.5
[M þ H]þ ¼ 981.1
[Mþ2H]2þ ¼ 491.1
[M þ K]þ ¼ 1238.8
[Mþ2K]2þ ¼ 639.22
[M þ K þ H]2þ ¼ 620
[Mþ2H]2þ ¼ 599.6
[M þ H]þ ¼ 771.7
[M þ K]þ ¼ 486.0
[M þ H]þ ¼ 363.9
[M þ H]þ ¼ 426.6
1.5
16
Peptide 4
KLLVVFFD
Peptide 5
[aoa-K]LLVVFF
[D-PEG2]
1198.45
22
The maximum rate of hydrolysis of peptide 3 (LLVVFF) was
45.8 nmol minꢂ1 ꢂ1, which is 10 fold less than the maximum
mg
Product 5-a VVFF[D-PEG2]
Product 5-b [aoa-K] LL
Product 5-c VVF
770.90
445.57
363.46
425.45
15
13.7
5
rate of hydrolysis of peptide 2 (LLVVYPWTQRFF). However, the
Michaelis Menten constant (Km), which reflects the affinity of an
enzyme for its substrate, was twice as high for peptide 2 compared
Product 5-d F[D-PEG2]
3

O. Achour et al. / Biochimie 95 (2013) 2010e2017
2015
1,0
0,8
0,6
0,4
0,2
0,0
300
Peptide 1
Peptide 2
250
200
150
100
50
0
0
5
10
15
20
25
30
35
0
100
200
300
400
500
Substrate concentration (µM)
Substrate concentration (µM)
30
25
20
15
10
5
5
4
3
2
1
0
Peptide 5
Peptide 3
0
0
50
100
150
200
250
0
50
100
150
200
250
Substrate concentration (µM)
Substrate concentration (µM)
Fig. 5. Effect of substrate concentration on the initial rate of CD hydrolysis of peptides 1 (R[K-DNP]LRFFLIPK[G-MCA]), 2 (LLVVYPWTQRFF), 3 (LLVVFF) and 5 ([aoa-K]LLVVFF[D-
PEG2]) at pH 3.7. The data represent the averages of triplicate runs at least.
with peptide 3, indicating a higher affinity of peptide 3 for CD. This
improved affinity can be explained by the small size of the peptide,
which facilitates access to the active site of CD. On the other hand,
long peptides and proteins, which may be less accessible to the
active site, are cleaved faster than small peptides. Therefore, the
resulting catalytic efficiency was higher for peptide 2 than for
peptide 3.
fact, the trigger peptide will be cleaved by the extracellular forms of
CD only after it accumulates in the targeted region.
3.4. Cleavage of peptide 5 by pCD in tumor microenvironment
conditions
The pCD is the major form of CD that is secreted in the extra-
cellular matrix. Thus, the verification of the cleavage of peptide 5
by the proform in tumor microenvironment conditions is neces-
sary as a first step for its validation as recognizable substrate
in vivo. In order to verify if the proform of Cathepsin D is able to
Both modified peptides 1 and 5 showed a lower maximal rate
than the two other peptides (peptides 2 and 3). The maximal rate of
peptide 1 was equal to 1.9 nmol minꢂ1 m ꢂ1, which is 2-fold lower
g
than peptide 5, which had a maximal rate of 3.4 nmol minꢂ1 mgꢂ1
.
Peptide 5, which was designed to be a part of an in vivo vector, was
recognized and cleaved by CD. It turned out that the addition of a
PEG2 group on the C-terminal and of the aminooxyacetic group on
the N-terminal does not obstruct the cleavage of the substrate by
CD. This peptide has a sequence of only 8 amino acids, which is
interesting in terms of the simplicity and cost of synthesis. This
small number of amino acids may also be an advantage in the
design of the final new vector if we consider that we have reduced
the risk of aliasing, which can make the cleavage sites inaccessible
to the enzyme. Finally, the low hydrolysis rate and the substrate
activation phenomenon observed with peptide 5 can be considered
as advantages that can enhance the efficiency of the new vector. In
cleave peptide 5, 200
and pH 6.5 in the presence and in the absence of phospholipid
micelles with 8.3 ng/ L of pCD. Liquid chromatography analysis
mM was exposed to pCD hydrolysis at pH 3.7
m
and mass spectroscopy identification demonstrated that pCD
cleaved peptide 5 at the same sites that were found to be cleaved
by the mature CD form. Fig. 6 presents the initial rate of hydrolysis
of peptide 5 by pCD at pH 3.7 and pH 6.5 with and without
phospholipid micelles.
At pH 3.7 and in the presence of phospholipids, pCD cleaved the
peptide at a rate of 14 ꢁ 10ꢂ3 nmol minꢂ1
mg
ꢂ1. This was about 8
fold higher than the rate obtained at the same pH without phos-
pholipids micelles (1.62 ꢁ 10ꢂ3 nmol minꢂ1
mg
ꢂ1). This data
Table 3
Kinetics parameters for the different peptide sequences used in this study.
Peptide
Vmax (nmol minꢂ1
m
gꢂ1
)
Km
(m
M)
Catalytic efficiency minꢂ1 gꢂ1
m L
1 (R[K-DNP]LRFFLIPK[G-MCA])
2 (LLVVYPWTQRFF)
3 (LLVVFF)
4 (KLLVVFFD)
5 ([aoa-K]LLVVFF[D-PEG2])
1.9 (ꢄ0.23)
516 (ꢄ97.2)
45.8(ꢄ10.7)
21.1
394
193.3
9 ꢁ 10ꢂ5
1.3 ꢁ 10ꢂ3
2.4 ꢁ 10ꢂ4
Resist to hydrolysis
3.4 (ꢄ0.39)
74.1
4.6 ꢁ 10ꢂ5
The data represent the averages of triplicate runs at least.

2016
O. Achour et al. / Biochimie 95 (2013) 2010e2017
mainly hypersecreted in the TME, is able to cleave the peptide with
the same cleavage profiles as the mature form. We also confirmed
the activator effect of phospholipid micelles.
0,016
0,014
0,012
0,010
0,008
0,006
0,004
0,002
0,000
With PLM
Without PLM
Although some details concerning the present findings remain to
be elucidated, in particular the activation mechanism of the
different forms of CD, these results enhance our knowledge on CD
behavior in the heterogenic conditions of a tumor microenviron-
ment. This work provides new arguments in favor of the role of CD
in the proteolysis of the degradation of extracellular matrix proteins
of the tumor microenvironment. It also provides new insights for
the development of a vector that will react specifically to the over
expression of enzymes like pCD in the tumor microenvironment.
Acknowledgments
pH 3.7
pH 6.5
This study was supported by a PhD grant from the Ligue Contre le
Cancer of Charente-Maritime for O. Achour (2011e2014). The Ligue
Contre le Cancer of Charente-Maritime and the PRES (Pôle de
Recherche et d’Enseignement Supérieur) of Limousin Poitou-Char-
entes are acknowledged for financial support through the research
project “Conception et validations de nanovecteurs intelligents à
double fonctionalité pour la thérapeutique et le diagnostic des
tumeurs”. Authors thank the CPER-Poitou-Charentes (2007–2013) for
financial support.
Fig. 6. Initial rate of hydrolysis of 200 mM of peptide 5 by 8.3 ng/mL of pCD at pH 3.7
and pH 6.5 with and without phospholipids micelles (PLM). The data represent the
averages of triplicate runs at least.
confirm the effect of phospholipids micelles, previously described
in the literature, as activators of the lysosomal CD responsible for
protein degradation [32]. Under TME conditions i.e. pH 6.5, pCD
was unable to cleave the peptide in the absence of phospholipids
after 120 h of reaction. However, the addition of phospholipid
micelles activated pCD and allowed the cleavage of peptide 5 with
an initial rate of 1.52 ꢁ 10ꢂ3 nmol minꢂ1 m
g
ꢂ1, which is practically
References
equal to the rate obtained at pH 3.7. The effect of phospholipid
micelles on the pCD form strengthens the hypothesis of an inter-
facial activation of this enzyme. This hypothesis certainly needs
more advanced investigations to be confirmed. In fact, extracellular
media are structured as heterogenic compartments that are sepa-
rated by many membranes and interfaces. This complex environ-
ment can be considered as a non-conventional media for enzymes
that has a significant effect on their activity, selectivity and stability.
The cleavage of peptide 5 by the pCD form can be considered as
an encouraging result for the use of this peptide as the trigger part
of a new vector that will be specifically activated in areas where
pCD is over expressed, such as a tumor microenvironment. Work is
ongoing to associate peptide 5 with others parts of the vector prior
to finer optimization of the peptide sequence, taking into account
the entire vector.
[1] M.A. Swartz, N. Iida, E.W. Roberts, S. Sangaletti, M.H. Wong, F.E. Yull,
L.M. Coussens, Y.A. DeClerck, Tumor microenvironment complexity: emerging
roles in cancer therapy, Cancer Res. 72 (2012) 2473e2480.
[2] I.P. Witz, O. Levy-Nissenbaum, The tumor microenvironment in the post-
PAGET era, Cancer Lett. 242 (2006) 1e10.
[3] B. Wouters, S. Weppler, M. Koritzinsky, W. Landuyt, S. Nuyts, J. Theys, R. Chiu,
P. Lambin, Hypoxia as a target for combined modality treatments, Eur. J.
Cancer 38 (2002) 240e257.
[4] J. Li, R. Ayene, K.M. Ward, E. Dayanandam, I.S. Ayene, Glucose deprivation
increases nuclear DNA repair protein Ku and resistance to radiation induced
oxidative stress in human cancer cells, Cell Biochem. Funct. 27 (2009) 93e101.
[5] M.F. McCarty, J. Whitaker, Manipulating tumor acidification as
treatment strategy, Altern. Med. Rev. 15 (2010) 264e272.
a cancer
[6] F. Fan, A. Schimming, D. Jaeger, K. Podar, Targeting the tumor microenviron-
ment: focus on angiogenesis, J. Oncol. 2012 (2012) 261e281.
[7] A. Razgulin, N. Ma, J. Rao, Strategies for in vivo imaging of enzyme activity: an
overview and recent advances, Chem. Soc. Rev. 40 (2011) 4186e4216.
[8] C. Ward, S.P. Langdon, P. Mullen, A.L. Harris, D.J. Harrison, C.T. Supuran,
I.H. Kunkler, New strategies for targeting the hypoxic tumour microenviron-
ment in breast cancer, Cancer Treat. Rev. 39 (2013) 171e179.
[9] J. Rozhin, M. Sameni, G. Ziegler, B.F. Sloane, Pericellular pH affects distri-
4. Conclusions
bution and secretion of cathepsin
B in malignant cells, Cancer Res. 54
(1994) 6517e6525.
[10] M. Cohen, I. Fruitier-Arnaudin, R. Sauvan, D. Birnbaum, J.M. Piot, Serum levels
of hemorphin-7 peptides in patients with breast cancer, Clin. Chim. Acta 337
(2003) 59e67.
[11] M. Cohen, I. Fruitier-Arnaudin, I. Garreau-Balandier, J.M. Piot, Microplate assay
for determination of cathepsin D activity based on quantification of a specific
and stable peptide released from hemoglobin: VV-hemorphin-7, Anal. Chim.
Acta 486 (2003) 21e29.
[12] O. Masson, A.-S. Bach, D. Derocq, C. Prébois, V. Laurent-Matha, S. Pattingre, et
E. Liaudet-Coopman, Pathophysiological functions of cathepsin D: targeting its
catalytic activity versus its protein binding activity? Biochimie 92 (2010)
1635e1643.
In this work, we investigated CD catalytic activity and specificity
in tumor microenvironment conditions. First of all, the effect of pH
on the CD catalytic activity and specificity was studied. We
conclude that, in the absence of an activator, mature CD is weakly
active at pH values that are found in the tumor microenvironment.
Moreover, we demonstrated that the specificity of CD was not
affected by a variation in pH. We then investigated the effect of
substrate structure on CD activity and specificity and our results
showed that there was a significant effect. The addition of hydro-
philic amino acids around the cleaved site led to the interruption of
substrate cleavage by CD. Interestingly, the functionalization of the
uncleaved peptide by PEG2 groups reestablished enzyme recogni-
tion. Moreover, we observed a significant change in the behavior of
CD that depended on the substrate structure. A substrate activation
phenomenon was observed for the modified substrate. We suggest
that this activation may be related to the amphiphilic nature of the
modified peptide, which may induce an interfacial activation
mechanism such as the one reported for other hydrolases like
lipase. In the next step, we used the pCD for the hydrolysis of the
modified peptide 5. This demonstrated that this proform, which is
[13] M. Wolf, I. Clark-Lewis, C. Buri, H. Langen, M. Lis, L. Mazzucchelli, Cathepsin D
specifically cleaves the chemokines macrophage inflammatory protein-1
macrophage inflammatory protein-1 , and SLC that are expressed in human
breast cancer, Am. J. Pathol. 162 (2003) 1183e1190.
a,
b
[14] G. Nicotra, R. Castino, C. Follo, C. Peracchio, G. Valente, C. Isidoro, The
dilemma: does tissue expression of cathepsin D reflect tumor malignancy?
the question: does the assay truly mirror cathepsin D mis-function in the
tumor? Cancer Biomarkers 7 (2010) 47e64.
[15] M.M. Gopalakrishnan, H.-W. Grosch, S. Locatelli-Hoops, N. Werth,
E. Smolenová, M. Nettersheim, K. Sandhoff, A. Hasilik, Purified recombinant
human prosaposin forms oligomers that bind procathepsin D and affect its
autoactivation, Biochem. J. 383 (2004) 507e515.
[16] S. Aziz, S. Pervez, S. Khan, N. Kayani, M. Rahbar, Immunohistochemical
cathepsin-D expression in breast cancer: correlation with established path-
ological parameters and survival, Pathol. Res. Pract. 197 (2001) 551e557.

O. Achour et al. / Biochimie 95 (2013) 2010e2017
2017
[17] J.-L. Kraimps, T. Métayé, C. Millet, D. Margerit, P. Ingrand, J.-M. Goujon,
P. Levillain, P. Babin, F. Begon, et J. Barbier, Cathepsin D in normal and
neoplastic thyroid tissues, Surgery 118 (1995) 1036e1040.
M. Abrahamson, G. Lalmanach, C.M. Overall, et E. Liaudet-Coopman, Proteol-
ysis of cystatin C by cathepsin D in the breast cancer microenvironment,
FASEB J. 26 (2012) 5172e5181.
[18] V. Vetvicka, J. Vetvickova, M. Fusek, Effect of procathepsin D and its activation
peptide on prostate cancer cells, Cancer Lett. 129 (1998) 55e59.
[19] P. Benes, V. Vetvicka, M. Fusek, Cathepsin Ddmany functions of one aspartic
protease, Crit. Rev. Oncol. Hematol. 68 (2008) 12e28.
[20] S. Bjelic, J. Aqvist, Catalysis and linear free energy relationships in aspartic
proteases, Biochemistry 45 (2006) 7709e7723.
[21] J.M. van Noort, A.C. van der Drift, The selectivity of cathepsin D suggests an
involvement of the enzyme in the generation of T-cell epitopes, J. Biol. Chem.
264 (1989) 14159e14164.
[26] I.H. Segel, Enzyme Kinetics: Behavior and Analysis of Rapid Equilibrium and
Steady State Enzyme Systems, Wiley, 1993.
[27] I. Fruitier, I. Garreau, J.M. Piot, Cathepsin D is a good candidate for the specific
release of a stable hemorphin from hemoglobin in vivo: VV-hemorphin-7,
Biochem. Biophys. Res. Commun. 246 (1998) 719e724.
[28] D.C. Pimenta, A. Oliveira, M.A. Juliano, L. Juliano, Substrate specificity of
human cathepsin D using internally quenched fluorescent peptides derived
from reactive site loop of kallistatin, Biochim. Biophys. Acta 1544 (2001)
113e122.
[22] A. Minarowska, M. Gacko, A. Karwowska, q. Minarowski, Human cathepsin D,
Folia Histochem. Cytobiol. 46 (2008) 23e38.
[23] F.A. Jaffer, D.E. Kim, L. Quinti, C.-H. Tung, E. Aikawa, A.N. Pande, R.H. Kohler,
G.P. Shi, P. Libby, R. Weissleder, Optical visualization of cathepsin K activity in
atherosclerosis with a novel, protease-activatable fluorescence sensor, Cir-
culation 115 (2007) 2292e2298.
[24] C.H. Tung, U. Mahmood, S. Bredow, R. Weissleder, In vivo imaging of pro-
teolytic enzyme activity using a novel molecular reporter, Cancer Res. 60
(2000) 4953e4958.
[25] V. Laurent-Matha, P.F. Huesgen, O. Masson, D. Derocq, C. Prébois, M. Gary-
Bobo, F. Lecaille, B. Rebière, G. Meurice, C. Oréar, R.E. Hollingsworth,
[29] P. Metcalf, M. Fusek, Two crystal structures for cathepsin D: the lysosomal
targeting signal and active site, EMBO J. 12 (1993) 1293e1302.
[30] R. Verger, ‘Interfacial activation’ of lipases: facts and artifacts, Trends Bio-
technol. 15 (1997) 32e38.
[31] M. Beaujouin, C. Prébois, D. Derocq, V. Laurent-Matha, O. Masson, S. Pattingre,
P. Coopman, N. Bettache, J. Grossfield, R.E. Hollingsworth, H. Zhang, Z. Yao,
B.T. Hyman, P. van der Geer, G.K. Smith, E. Liaudet-Coopman, Pro-cathepsin D
interacts with the extracellular domain of the
b chain of LRP1 and promotes
LRP1-dependent fibroblast outgrowth, J. Cell Sci. 123 (2010) 3336e3346.
[32] S. Watabe, N. Yago, Phospholipids activate cathepsin D, Biochem. Biophys. Res.
Commun. 110 (1983) 934e939.