A Suite of Activity-Based Probes to Dissect the KLK Activome in Drug-Resistant Prostate Cancer

Article abstract of DOI:10.1021/jacs.1c03950

Kallikrein-related peptidases (KLKs) are a family of secreted serine proteases, which form a network (the KLK activome) with an important role in proteolysis and signaling. In prostate cancer (PCa), increased KLK activity promotes tumor growth and metastasis through multiple biochemical pathways, and specific quantification and tracking of changes in the KLK activome could contribute to validation of KLKs as potential drug targets. Herein we report a technology platform based on novel activity-based probes (ABPs) and inhibitors enabling simultaneous orthogonal analysis of KLK2, KLK3, and KLK14 activity in hormone-responsive PCa cell lines and tumor homogenates. Importantly, we identifed a significant decoupling of KLK activity and abundance and suggest that KLK proteolysis should be considered as an additional parameter, along with the PSA blood test, for accurate PCa diagnosis and monitoring. Using selective inhibitors and multiplexed fluorescent activity-based protein profiling (ABPP), we dissect the KLK activome in PCa cells and show that increased KLK14 activity leads to a migratory phenotype. Furthermore, using biotinylated ABPs, we show that active KLK molecules are secreted into the bone microenvironment by PCa cells following stimulation by osteoblasts suggesting KLK-mediated signaling mechanisms could contribute to PCa metastasis to bone. Together our findings show that ABPP is a powerful approach to dissect dysregulation of the KLK activome as a promising and previously underappreciated therapeutic target in advanced PCa.

Full text of DOI:10.1021/jacs.1c03950

pubs.acs.org/JACS
Article
A Suite of Activity-Based Probes To Dissect the KLK Activome in
Drug-Resistant Prostate Cancer
Scott Lovell, Leran Zhang, Thomas Kryza, Anna Neodo, Nathalie Bock, Elena De Vita,
Elizabeth D. Williams, Elisabeth Engelsberger, Congyi Xu, Alexander T. Bakker, Maria Maneiro,
Reiko J. Tanaka, Charlotte L. Bevan, Judith A. Clements, and Edward W. Tate*
Cite This: J. Am. Chem. Soc. 2021, 143, 8911−8924
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Kallikrein-related peptidases (KLKs) are a family of
secreted serine proteases, which form a network (the KLK
activome) with an important role in proteolysis and signaling. In
prostate cancer (PCa), increased KLK activity promotes tumor
growth and metastasis through multiple biochemical pathways, and
specific quantification and tracking of changes in the KLK
activome could contribute to validation of KLKs as potential
drug targets. Herein we report a technology platform based on
novel activity-based probes (ABPs) and inhibitors enabling
simultaneous orthogonal analysis of KLK2, KLK3, and KLK14
activity in hormone-responsive PCa cell lines and tumor
homogenates. Importantly, we identifed a significant decoupling
of KLK activity and abundance and suggest that KLK proteolysis
should be considered as an additional parameter, along with the PSA blood test, for accurate PCa diagnosis and monitoring. Using
selective inhibitors and multiplexed fluorescent activity-based protein profiling (ABPP), we dissect the KLK activome in PCa cells
and show that increased KLK14 activity leads to a migratory phenotype. Furthermore, using biotinylated ABPs, we show that active
KLK molecules are secreted into the bone microenvironment by PCa cells following stimulation by osteoblasts suggesting KLK-
mediated signaling mechanisms could contribute to PCa metastasis to bone. Together our findings show that ABPP is a powerful
approach to dissect dysregulation of the KLK activome as a promising and previously underappreciated therapeutic target in
advanced PCa.
dihydrotestosterone (DHT), where it induces expression of a
variety of genes important for PCa cell proliferation and
survival (Figure 1A).6 Among these genes are specific serine
proteases from the 15-member human kallikrein-related
peptidase (KLK) family, which have versatile and crucial
roles in extracellular proteolysis and signaling.7,8 For example,
increased expression and subsequent leakage of KLK2 and
KLK3, also known as prostate-specific antigen (PSA), into the
vasculature are used in PCa diagnosis and progression
monitoring.9,10 KLK2 and KLK3 also have functional roles in
PCa and contribute to disease progression. KLK2 can activate
protease-activated receptors (PARs) on the surface of
neighboring fibroblasts, which in turn release cytokines that
stimulate PCa cell proliferation,11 while in the bone micro-
INTRODUCTION
■
Prostate cancer (PCa) is the most frequently diagnosed cancer
among men in industrialized nations and remains a leading
cause of cancer-related deaths.1 Localized malignancies are
treated with surgery and radiation therapy, and the 5-year
survival rate of patients is close to 100%; however, post-
operative recurrence often progresses to advanced PCa. Initial
treatment for advanced tumors exploits the dependence of PCa
cells on androgens by combining androgen-deprivation therapy
(ADT) with direct targeting of the androgen receptor (AR),2,3
but many patients stop responding and develop castrate-
resistant prostate cancer (CRPC). In CRPC, PCa cells evolve
resistance to androgen-targeting therapies by restoring AR
signaling through diverse mechanisms, and the majority of
patients present with bone metastases with increased risk of
morbidity and mortality due to alterations in skeletal integrity.4
Mapping critical pathways involved in establishing CRPC may
enable identification of novel therapeutic targets which can
ultimately reduce disease recurrence.5
Received: April 15, 2021
Published: June 4, 2021
AR is a transcription factor that dimerizes and translocates
to the nucleus upon binding of androgens such as
© 2021 The Authors. Published by
American Chemical Society
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924
8911

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 1. Human kallikrein-related peptidases in prostate cancer. (A) KLK synthesis: KLK2 and KLK3 expression is induced by the transcriptional
activity of the AR. Conversely, KLK14 expression is negatively regulated by AR signaling and is upregulated upon treatment with enzalutamide, a
competitive AR inhibitor. (B) KLK activation: KLKs are secreted to the tumor microenvironment as proenzymes and mostly require trypsin-like
proteolytic processing to become catalytically active. In prostate cancer a regulatory cascade termed the “KLK activome” has been proposed
whereby the trypsin-like proteases KLK2 and KLK14 autoactivate prior to cross-activating other pro-KLK molecules. Red = trypsin-like activity,
blue = chymotrypsin-like activity, yellow outline = autoactivation. KLK functional roles: In addition to degrading components of the ECM, active
KLK molecules are involved in key stages of PCa progression including cell proliferation, cell migration, and bone metastasis. (C) Schematic
representation of the proposed ABPP workflow to enable quantification of KLK activity in PCa cell conditioned media and tumor homogenate.
ABPs are incubated with PCa cells. Conditioned medium is collected and concentrated using a MWCO spin filter. Protein samples are subjected to
SDS−PAGE, and labeling is then visualized by in-gel fluorescence or streptavidin blotting.
environment KLK3 promotes osteoprogenitor cell prolifer-
ation and osteoblast differentiation and establishment of bone
metastases.12,13 More recently, KLK14 has also been
implicated in CRPC development; while KLK14 is normally
suppressed by AR signaling (Figure 1A), treatment with AR-
targeted drugs may increase KLK14 expression to promote
PCa cell migration and bone matrix colonization.14,15
Despite recent progress in determining the pathophysio-
logical roles of individual KLKs in PCa, the potential of KLK
inhibitors for therapeutic intervention remains to be
determined. Once secreted by PCa cells, KLKs do not work
in isolation but as components of a complex network called the
KLK activome (Figure 1B) that is tightly regulated by
proteolytic KLK autoactivation, cross-activation or deactiva-
tion, and inhibition by endogenous serine protease inhib-
itors.16−19 A method to quantify the activity levels of each KLK
directly in a complex biological system would enable the
dynamic response of the KLK activome to be determined in
relevant PCa models and drug-resistant tumors and during
drug/hormone-mediated perturbations and may lead to
validation of specific active KLKs as drug targets or novel
biomarkers.
Here we present a technology platform that enables
simultaneous orthogonal readout of the activity of KLK2,
KLK3, and KLK14 in androgen-responsive PCa cell lines and
patient-derived xenografts (PDX), based on a chemical toolbox
of first-in-class selective activity-based probes (ABPs) and
inhibitors for activity-based protein profiling (ABPP, Figure
8912
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 2. Development and optimization of a KLK3 ABP using competitive ABPP. (A) Structures of reported KLK3 inhibitors 1 and 2 and first-
generation fluorescent ABP 3. (B) In-gel fluorescence shows selective, dose-dependent labeling of KLK3 after treatment of LNCaP conditioned
media with 3 for 1 h. (C−F) Inhibition of KLK3 protease activity by morpholine-capped peptidyl-DPP analogues. LNCaP conditioned medium
was first incubated with the analogue for 1 h and then with 3 for 1 h. Protein samples were separated by SDS−PAGE, and residual KLK3 activity
was calculated by densitometry. See Figures S4 and S5 for fluorescent gels. (G) Chemical structures of optimized KLK3 inhibitor 15 and ABP 20.
(H) Streptavidin blotting shows potent and selective labeling of KLK3 after treatment of LNCaP conditioned media with 20 for 1 h.
1C). We use this platform to show that active KLK molecules
are secreted into the bone microenvironment by PCa cells
following stimulation by osteoblasts, supporting a double
paracrine signaling mechanism that may contribute to the
establishment of bone metastases mediated by KLK activity.
Furthermore, using selective inhibitors and multiplexed
fluorescent ABPP, we dissect the KLK activome directly in
PCa cells and show that KLK14 drives PCa cell migration, a
key step in the formation of distant metastases. Together these
findings demonstrate that ABPP can provide a unique window
on KLK activity and inhibition in PCa and suggest that KLK-
activome dysregulation is a promising and previously under-
appreciated therapeutic target in CRPC.
RESULTS
■
Development of a Selective Inhibitor and Activity-
Based Probe for KLK3. Proof-of-concept for the first
selective ABP targeting the PCa KLK activome started with
KLK3, which exhibits chymotrypsin-like specificity, in contrast
8913
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 3. Comparative analysis of the S1−S4 subsite preferences of KLK2 and KLK14 using a positional scanning library approach. (A) Structure
of the peptidyl-coumarin scaffold used, exemplified with the P2 sublibrary. P2 = any defined natural or non-natural amino acid. X = isokinetic
mixture of 19 natural amino acids, excluding cysteine. (B−D) Scatter plots showing relative initial rates of hydrolysis of each amino acid for KLK2
(x-axis) and KLK14 (y-axis). The optimal amino acid residue is set to 100%. Amino acids that are close to the dotted diagonal line are equally
tolerated by both KLKs. The preferred natural amino acid for each protease is shown in red, and the amino acid that was used in the final optimized
ABP is shown in green. The position of each amino acid represents the mean value of three replicates. See Figures S11−S14 for complete data. (E)
Structures of KLK2 and KLK14 ABPs before and after P2−P4 optimization.
to KLK2 and KLK14 which are trypsin-like proteases.20 An
subsites was envisaged for the ABP specificity element, grafted
optimized tetrapeptide designed to occupy the KLK S1−S4
onto a peptidyl-diphenyl phosphonate (DPP) warhead which
8914
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
reacts specifically and exclusively with the Ser195 residue in
the KLK3 active site.21 DPP 1 with a tyrosine-mimicking
phenol in the first position (P1) was previously identified as a
modestly potent inhibitor of KLK3,22 while peptidyl-boronic
acid 2 is a potent and selective KLK3 inhibitor (Figure 2A).23
Both DPP and boronic acid inhibitors covalently modify
Ser195, thereby mimicking the tetrahedral intermediate during
amidolysis; however boronic acids form reversible covalent
complexes and are generally less effective as ABPs. We
therefore replaced the boronic acid in 2 with the DPP moiety
in 1 and capped the N-terminus with a tetramethyl rhodamine
(TAMRA) fluorophore to afford 3, which served as a
prototype ABP for KLK3. A mixed solid-phase and solution-
phase approach was used for the synthesis of peptidyl-DPP
compounds, as described previously (Figures S1 and S2).24
With this approach the peptidyl-DPP compounds are obtained
as an equimolar mixture of two diastereoisomers with only the
(R) configuration engaging the target KLK. KLKs are typically
activated after their secretion from the cell, so we first
examined the labeling profile of 3 in conditioned media (CM)
obtained from LNCaP cells treated with androgen (10 nM
DHT). LNCaP is an androgen-responsive human prostate
adenocarcinoma cell line widely used to model key features of
clinical disease, including secretion of KLK2 and KLK3
following androgen stimulation.8 LNCaP CM was treated
with different concentrations of 3 for 1 h, and proteins were
separated by SDS−PAGE. In-gel fluorescence revealed a single
major target at ∼32 kDa, labeled in a concentration-dependent
manner (Figure 2B), and selective covalent modification of
active KLK3 was confirmed by immunoprecipitation (Figure
S3).
Despite exquisite selectivity, maximal KLK3 labeling was
achieved only after treatment with 20 μM 3 for 1 h. We
therefore optimized probe potency by systematically modifying
each amino acid side chain in the P1−P4 positions of a
morpholine-capped peptidyl-DPP inhibitor scaffold and
determined the potency of each analogue using competitive-
ABPP against 3.25,26 LNCaP CM was first incubated with a
peptidyl-DPP analogue for 1 h, followed by 1 h treatment with
20 μM 3 to assess the degree of KLK3 inhibition by
densitometry (Figures S4 and S5 for fluorescent gels and
Figure S6 for structures of each morpholine-capped peptidyl-
DPP analogue). Out of 16 analogues tested, compound 15 was
identified as a potent KLK3 inhibitor (Figure 2C−G). Despite
its preference for Tyr in P1, KLK3 can also process substrates
with P1 Gln,27 and a “hybrid” benzamide side chain in P1
resulted in a significant increase in potency (Figure 2C). The
P2−P4 specificity of KLK3 has been mapped in detail
previously and only minor changes were required to further
optimize potency, including substitution of P2 Leu for
norvaline (Nva) and of P4 Ser for Asn.28 Second generation
biotinylated 20 and fluorescent 21 ABPs (Figures 2G and S7)
based on this scaffold retained exquisite selectivity for KLK3 in
LNCaP CM, but in contrast to first-generation ABP 3 achieved
maximum labeling at only 1 μM and readily detectable labeling
down to 130 nM (Figures 2H and S7). Importantly, 20 and 21
showed no inhibition of KLK2 and KLK14 even at
concentrations as high as 50 μM (Figure S8). Consequently,
these probes are the most selective covalent inhibitors of KLK3
reported to date.
anticipated that development of selective ABPs would require
optimization of the P2−P4 positions to differentiate between
unique preferences of the S2−S4 subsites. Building on prior
work,29−32 we generated a positional scanning substrate library
derived from 19 natural amino acids (excluding Cys) and 86
structurally diverse non-natural amino acids to provide a
detailed analysis of active site preferences at each protease
subsite. Three sublibraries consisting of 105 mixtures of 361
fluorogenic peptidyl coumarins were generated and screened
against KLK2 and KLK14 (Figure 3A; see Figures S9 and S10
for library synthesis and structures of non-natural amino acids),
and scatter plots of relative initial rates of hydrolysis for each
amino acid revealed specificity preferences at the P2, P3, and
P4 positions (Figure 3B−D and Figures S11−S14). Despite a
general preference for aromatic residues at P2, KLK2
processed substrates with a P2 cyclohexyl alanine (Cha) at
the highest rate, whereas KLK14 demonstrated dual specificity
at P2, processing both aliphatic residues such as aminobutyric
acid (Abu) and aromatic residues such as benzyl histidine
(His(Bzl)) (Figure 3B). Both proteases favored basic amino
acids at P3, with KLK14 preferring Lys while KLK2 preferred
the shorter chain diaminobutyric acid (Dab) (Figure 3C). At
P4, KLK2 processed aliphatic (e.g., Abu) and aromatic
residues with a flexible linker such as benzyl-serine (Ser(Bzl)),
and KLK14 preferred medium and large aromatic residues
such as 4-bromophenylalanine (Phe(4-Br)) and benzothiazol-
2-yl alanine (Ala(Bth)) (Figure 3D).
On the basis of these data, we hypothesized that the
divergent specificity preferences of KLK2 and KLK14 could be
exploited to develop selective chemical probes. The preferred
amino acid for each subsite was incorporated into a peptidyl-
DPP scaffold to afford prototype probes 22 and 32 directed
toward KLK2 and KLK14, respectively (Figure 3E), using a
phenylguanidine Arg mimic in P1 to ease synthesis.33 Kinetic
analyses (Figure S15) revealed that although 22 is a potent
inhibitor of KLK2 (kinact/Ki = 3274
89 M−1 s−1), it has
significant cross-reactivity with KLK14 (kinact/Ki = 511 18
M−1 s−1), and we therefore further optimized the KLK2
inhibitor scaffold by systematically altering P2, P3, and P4
residues with other hits from the substrate library screen. From
nine peptidyl-DPP analogues, compound 31 was most optimal
(Table 1, Figure S16) with similar potency toward KLK2
(kinact/Ki = 3076
87 M−1 s−1) delivered by a P2 Tyr
substitution and with >30-fold selectivity over KLK14 (kinact/Ki
= 94 7 M−1 s−1) because of increased steric bulk at P4 (Tle
in place of Abu). We took a similar approach to optimize
compound 32 against KLK14 (kinact/Ki = 523 46 M−1 s−1),
identifying compound 39 (Table 1, Figure S17) with 80-fold
higher potency toward KLK14 (44474 1976 M−1 s−1) and
220-fold selectivity over KLK2 (kinact/Ki = 204 13 M−1 s−1).
To gain further insights into site preference, we performed
molecular modeling studies using previously determined KLK2
and KLK3 crystal structures34,35 in the Molecular Operating
Environment (MOE) software package36 to propose binding
determinants of 31 and 20. For the purpose of modeling the
specificity elements of each probe class, the biotin in 31 was
replaced with an acetyl group, and morpholino-capped
derivative 15 was used to model 20. We found that 31 and
20 likely bind to their cognate KLK active sites in a
conventional extended conformation, with the amino acid
side chains of P1−P4 occupying the S1−S4 pockets (Figure
S18A,B). In both poses the phosphonate warhead is located in
close proximity to the catalytic serine, with the PO moiety
Development of Selective Inhibitors and Activity-
Based Probes for KLK2 and KLK14. Both KLK2 and
KLK14 display a strong preference for Arg in P1,20 and we
8915
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
establishing an H-bonding interaction with the NH group on
the backbone of Gly193, which would activate the
phosphonate for nucleophilic attack (Figure S18C,D). Our
modeling highlighted the differing properties of the S1 and S4
pockets of KLK2 and KLK3 as key determinants of probe
specificity. For example, the P1 phenylguanidine of 31 forms a
salt bridge with the carboxylate group of Asp189 at the bottom
of the S1 pocket of KLK2. In contrast, for KLK3 the residue at
the bottom of the S1 pocket is serine, which is also flanked by
the polar side chains of Ser227 and Thr190. Consequently, the
KLK3 S1 pocket is polar at the bottom and hydrophobic on
the sides and has a preference for medium hydrophobic side
chains with polar neutral head groups. In agreement with this,
the P1 benzamide group of 20 forms key hydrogen bonds with
Ser227 and Thr190. Furthermore, our model suggests that the
P4 Asn in 20 forms a hydrogen bond with Gln174, which is
positioned in the small, polar S4 pocket of KLK3. In contrast,
the bulky P4 Tle residue of 31 is accommodated by the S4
pocket of KLK2, which has significant hydrophobic character
imparted from several residues of the kallikrein loop.37
A
crystal structure for KLK14 is yet to be solved, but a homology
model by de Veer et al. provides a potential explanation for the
selectivity of our KLK2 and KLK14 probes.38 The S2 pocket of
KLK14 is narrow owing to the flanking side chain of His99,
and thus small aliphatic residues, such as the P2 Abu of 39, are
preferred. Conversely, the S4 pocket of KLK14 is large with
Trp215 positioned at the base, which is predicted to form π-
stacking interactions with large aromatic residues such as the
P4 2-Nal of 39.38 The selectivity for 31 is likely achieved due
to the combination of suboptimal interactions of the S2 and S4
pockets of KLK14 with P2 Tyr and P4 Tle, respectively.
Conversely, the selectivity of 39 is likely derived from the fact
that the S4 pocket of KLK2 cannot accept the large aromatic 2-
Nal side chain present at P4.
With ABPs for KLK2, KLK3, and KLK14 in hand, we next
addressed selective readouts of KLK activities in a complex
biological system. Compounds 20, 31, and 39 were incubated
with CM obtained from LNCaP-K14 cells, in which KLK14
expression has been placed under the control of a doxycycline-
inducible promoter, stimulated with 10 nM DHT and 200 nM
doxycycline (Dox).15 Streptavidin blotting revealed a single
target between 25 and 32 kDa for each compound, labeled in a
concentration-dependent manner (Figure 4A), showing that
the only detectable targets of these probes are their cognate
proteases to the exclusion of other off-targets, emphasizing
their impressive selectivity. In line with previous reports
KLK14 appears as a double band due to protein
glycosylation.15,39 Selective covalent modification of KLK2,
KLK3, and KLK14 by 31, 20, and 39, respectively, was
confirmed by streptavidin enrichment and immunoblotting
(Figures S19 and S20). In the following experiments,
biotinylated ABPs (bABPs) 31, 20, and 39 are named
KLK2_bABP, KLK3_bABP, and KLK14_bABP, respectively.
Taken together, detailed analyses of specificity preferences
of KLK2, KLK3, and KLK14 led to the design and discovery of
a toolbox of selective covalent inhibitors and probes.
Profiling KLK Activity in Hormone-Dependent Pros-
tate Cancer. As noted above, KLKs 2, 3, and 14 have been
proposed to play roles at different stages of PCa progression
including facilitating initial prostate tumor expansion and
invasion, and establishment of distant metastases.8 Given that
AR activity is crucial in the development of CRPC and for
continued tumor growth, we next assessed how KLK activity
8916
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 4. KLK activity profiling in PCa cells and PDX homogenates. (A) Streptavidin blotting shows potent and selective labeling of KLK2, KLK3,
and KLK14 by 31, 20, and 39, respectively, after treatment of LNCaP-K14 conditioned media for 1 h. (B) KLK activity and expression in the
LNCaP cell line after treatment with 10 nM DHT 10 μM Enz. KLK activity was assessed by streptavidin enrichment of probe-labeled KLK
molecules following treatment of conditioned media with 1 μM of either 31, 20, or 39 for 1 h. See Figure S21 for blot quantification. (C) KLK
activity and expression in six different PDX homogenates. KLK activity was assessed using the same method as in (B). The tissue origin and AR
expression levels of the six PDX samples are shown. The + sign indicates the staining intensity of PDX samples by an AR antibody. Abbreviation:
IHC = immunohistochemistry, TURP = transurethral resection of the prostate. See Figure S22 for blot quantification.
changes in response to specific AR signaling perturbations. In
the LNCaP-K14 cells used in previous experiments the
expression of KLK14 is increased by activation of a
doxycycline-inducible promoter, and to assess AR-mediated
changes in KLK14 activity, we reverted to WT LNCaP cells,
which are known to have low basal KLK14 expression.15 To
enable detection of each KLK in WT LNCaP cells, we
incubated LNCaP conditioned media with 1 μM of each KLK
ABP, enriched labeled proteins on streptavidin beads, and
visualized activity levels by immunoblotting with the
appropriate KLK antibody (see Figure S19 for workflow). As
expected, the expression and activity of KLK2 and KLK3 in
LNCaP cells were upregulated in response to treatment with
DHT, and this increase was nullified by co-treatment with
enzalutamide (Enz), a competitive AR inhibitor employed in
the clinic to treat CRPC. In contrast, KLK14 expression and
activity was decreased upon treatment with DHT but restored
by co-incubation with Enz (Figure 4B and Figure S21 for blot
quantification).
To assess KLK activity levels at different disease stages, we
profiled six PCa patient-derived xenograft (PDX) tumors
isolated from diverse tissues including prostate and metastases
in lymph node, liver, and bone.40 Fresh-frozen PDX samples
were homogenized in 1% Triton in PBS and incubated with 1
μM KLK ABP, followed by enrichment and immunoblotting.
KLK2 was expressed in all samples, and KLK3 was expressed in
all but one; however, expression and activity were substantially
decoupled, with KLK2 and KLK3 activity seen only in PDXs
8917
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 5. KLK activity profiling in coculture models of LNCaP cells and primary human osteoblasts. (A) Primary human osteoprogenitor cells were
seeded and then osteogenically differentiated for 6 weeks resulting in the deposition of a dense, mineralized ECM. In the direct coculture model
LNCaP cells were seeded onto the matrix in androgen-depleted growth medium, and after 48 h the conditioned medium was collected. In the
indirect coculture model osteoblasts were cultured in androgen-deprived growth medium for 48 h, and the conditioned medium was collected and
then added to LNCaP cells. The conditioned medium was then collected after 48 h. (B) KLK activity and expression in the coculture model of
LNCaP cells and osteoblasts. KLK activity was assessed using the same method as described in Figure 4B. See Figure S24 for blot quantification.
(C) Proposed role of KLK2 and KLK3 in PCa growth in the bone microenvironment. Osteoblasts secrete soluble factors such as IL-6 resulting in
ligand-independent activation of the AR. AR signaling results in PCa cell proliferation and secretion of KLK2 and KLK3 into the bone
microenvironment. KLK2 and KLK3 stimulate proliferation of osteoprogenitor cells and osteoblast differentiation via a TGFβ-dependent
mechanism. This results in further ligand-independent AR activation and establishment of bone metastases.
generated from metastases including lymph node, liver, and
bone. LuCaP 35 PDX, which lacked KLK3 expression and had
decreased KLK2 expression, has also previously been shown to
have low AR expression.40 KLK14 activity decoupled from
expression was evident in all samples, with higher KLK14
activity in LuCaP 35, BM18, and LuCaP 105 (see Figure 4C
and Figure S22 for blot quantification).
Strikingly, we found that simultaneous activation of all three
KLKs was observed only in PDX cells from bone metastases
(BM18 and LuCaP 105). Advanced PCa has a propensity to
metastasize to bone and dysregulate bone resorption and bone
formation through complex paracrine signaling events with
osteoblasts and osteoclasts (cells that mediate bone generation
or absorption, respectively). Treatment options for PCa bone
metastases remain inadequate and generally palliative, and
there is an urgent need to identify novel therapeutic targets.41
Osteoblasts secrete growth factors including IL-6 which can
induce AR signaling in PCa cells even in the absence of
androgens, as seen during ADT,42−44 and we hypothesized that
osteoblasts might therefore also induce PCa cell secretion of
active KLK molecules into the bone microenvironment under
androgen-deprived conditions. To test this hypothesis, primary
human osteoprogenitor cells were isolated from bone tissue
and cultured in osteogenic media (1 M β-glycerophosphate,
200 mM ascorbate-2-phosphate, and 0.1 M dexamethasone)
for 6 weeks as described previously,45,46 forming a dense
mineralized collagen-rich bone-like matrix (Figure S23).
LNCaP cells were then either directly cocultured with
osteoblasts or treated with conditioned media from osteoblasts
(here termed “indirect coculture”) in androgen-depleted media
(Figure 5A), and CM was obtained from indirect or direct
coculture treated with 1 μM of each KLK ABP, followed by
enrichment and immunoblotting. In agreement with our
hypothesis, both direct and indirect coculture of LNCaP
cells with osteoblasts resulted in an increase of total and active
KLK2 and KLK3. In line with the AR-dependent and prostate-
restricted expression of these proteases, neither KLK was
detected in osteoblast CM alone, and very low expression was
seen in CM obtained from androgen deprived LNCaP cells
(Figure 5B and Figure S24 for blot quantification). The
concentration of active and total KLK14 also increased in
coculture compared to LNCaP cells alone, but in contrast to
KLK2 and KLK3 this was due to secretion by osteoblasts,
8918
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 6. Dissecting the KLK activome in LNCaP cells. (A) Structures of fluorescent ABPs 42, 20, and 43 for KLK2, KLK3, and KLK14,
respectively. (B) KLK activity profiling by in-gel fluorescence of conditioned media obtained from LNCaP-WT, LNCaP-mK14, and LNCaP-K14
cells following treatment with DHT (10 nM), Dox (100 ng/mL), and the appropriate biotinylated ABP using fluorescent ABPs 20, 42 and 43. See
Figure S26 for blot and gel quantification. (C) Assessment of KLK activity inhibition on LNCaP cell migration. LNCaP-K14 cells were cultured in
serum-free RPMI media supplemented with 10 nM DHT Dox (100 ng/mL) for 24 h. Cells were added to transwell inserts (100 000 cells in 100
μL of serum-free RPMI), and after 4 h the appropriate ABP was added. Cultures were incubated for a further 24 h before the number of migrating
cells was quantified.
suggesting that osteoblasts may contribute to KLK14 activity
in bone metastases.
phenotypic response in a particular disease setting. Experi-
ments with purified proteins show that KLK14 can activate
pro-KLK2 and pro-KLK3 by proteolytic cleavage of their N-
terminal propeptide sequence,17 and it has been suggested that
KLK2 can also activate pro-KLK3.53 To enable dissection of
the PCa KLK activome directly in LNCaP cells, we further
expanded the KLK probe toolbox by synthesizing KLK2_fABP
(42) and KLK14_fABP (43) (Figure 6A and Figure S2),
fluorescent ABP (fABP) analogues of KLK2_bABP and
KLK14_bABP bearing dyes fluorescing at discrete wave-
lengths, which retain excellent potency and selectivity for
Dissecting the Role of the KLK Activome in Prostate
Cancer Progression. The data presented above provide
evidence that KLK activity in PCa is regulated by the AR at a
level beyond simple changes in expression and raise the
possibility that KLK activities may be actionable targets for
therapeutic intervention at different disease stages, in particular
during PCa metastasis to bone. However, KLKs are predicted
to form complex proteolytic networks and it is often not clear
which protease(s) should be inhibited for maximum
8919
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
KLK2 and KLK14, respectively (Figure S25). KLK2_fABP and
KLK14_fABP were combined with KLK3-selective fABP 21
(KLK3_fABP) in an “fABP cocktail” to a final concentration of
1 μM each and incubated for 20 min with CM obtained from
LNCaP-WT, LNCaP-mK14, or LNCaP-K14 cells following 48
h treatment with 10 nM DHT to enable simultaneous
multicolor readout of KLK2, KLK3, and KLK14 activity.
Both LNCaP-WT and Dox-treated LNCaP-mK14 cells
expressing catalytically inactive mutant KLK14[S195A] under
the control of a Dox-inducible promoter15 had low levels of
active KLK3, while KLK2 and KLK14 activity was below the
detection limit (Figure 6B). However, upon Dox-induced
KLK14 expression in DHT-treated LNCaP-K14 cells a
significant increase in the activity of all three KLKs was
evident (see Figure 6B and Figure S26 for gel quantification).
Importantly, the total abundance of secreted KLK2 and KLK3
remained constant upon treatment with Dox, demonstrating an
increase in the ratio of active to inactive KLK independent
from protein expression. To enable selective inhibition of
active KLK molecules, LNCaP-K14 cells were co-incubated
with 1 μM KLK2_bABP, KLK3_bABP, or KLK14_bABP, 10
nM DHT, and 200 nM Dox for 48 h. Residual KLK activity in
CM was then assessed using the fABP cocktail (Figure 6B).
These data demonstrate that inhibition of either KLK2 or
KLK3 activity does not affect the activity of other KLKs,
suggesting that KLK2 and KLK3 do not cross-activate the
other PCa KLKs. However, inhibition of KLK14 resulted in a
significant decrease in the activity levels of both KLK2 and
KLK3, suggesting that KLK14 cross-activates the proforms of
both these proteases.
Prior to the establishment of bone metastases circulating
PCa cells must first enter the bone microenvironment by
migrating across the sinusoidal wall.54 We have previously
shown that increased KLK14 activity drives PCa cell migration
and colonization of mineralized bone matrices.15 However, the
specific contributions of the KLK activome (KLK2, KLK3, and
KLK14) toward PCa cell migration are yet to be elucidated.
We first confirmed that treatment of LNCaP-K14 cells with 10
μM KLK2_bABP, KLK3_bABP, and KLK14_bABP resulted
in no cytotoxicity over a 96 h time period (Figure S27). We
then assessed the migratory capacity of LNCaP-K14 cells using
a transwell assay (see Figure S28 for standard curve).
Induction of KLK14 expression in LNCaP-K14 cells (200
nM Dox) resulted in increased migration, as expected, which
was nullified by co-treatment with 1 μM KLK14_bABP
(Figure 6C). Conversely, treatment with 1 μM KLK2_bABP
or KLK3_bABP had no significant effect on the number of
migrating LNCaP-K14 cells. Simultaneous treatment of
LNCaP-K14 cells with 1 μM each KLK2_bABP, KLK3_bABP,
and KLK14_bABP resulted in a significant decrease in cell
migration, suggesting that KLK14 activity is a key driver of
PCa cell migration, while KLK2 and KLK3 activities
individually make minor contributions.
disease progression through amplified proteolysis.56−58 How-
ever, delineating the overlapping, synergistic and opposing
activities of individual prostatic KLKs in complex environ-
ments remains very challenging. Inspired by multiplexed ABP
toolkits previously reported for the proteasome,26,59 neutrophil
serine proteases,24 and caspases,32 we developed the first probe
set enabling simultaneous assessment of the activity of each
PCa-relevant KLK in PCa samples. Peptidyl-DPP ABPs were
established for each protease, and potency and specificity were
further optimized in PCa supernatants by competitive ABPP,
an approach that enables assessment of the properties of each
covalent inhibitor directly in a complex biological environ-
ment. Development of selective ABPs for KLK2 and KLK14
was complicated by their similar trypsin-like activity; indeed,
proteases with the same primary (P1) specificity are often
coexpressed in tissues, and to generate selective tools, it is
necessary to explore more complex specificity determinants.60
Here we applied positional scanning libraries to identify
mutually exclusive preferences in the S2−S4 pockets of KLK2
and KLK14; however this approach does not account for
potential cooperativity between protease subsites in driving
selectivity. Further optimization across ABPs incorporating
different combinations of preferred amino acids illustrated the
importance of exploiting cooperativity and in the case of
KLK14 identified an ABP with optimal selectivity which
features none of the top amino acid hits identified for each
individual subsite in the library screen. Interestingly, both
KLK2 and KLK14 demonstrated dual specificity in certain
subpockets, suggesting a role for these subsites in tuning
substrate profiles.61 We note that while the present set of
probes has been optimized for PCa, each tissue type expresses
a different set of secreted proteases, and thus the specificity
sequence required to obtain probe selectivity may depend on
the physiological context.
Recent development of KLK-knockout and transgenic
mice18,62 and selective inhibitors39 have revealed key
associations between individual KLK activities and the onset
or progression of diverse diseases. However, whether these
associations are due to the activity of an individual KLK or
because of cross-activation of other proteases, including other
KLKs, is poorly understood. The KLK activome has to date
been modeled only using purified proteins, and it remains very
challenging to deconvolute the biological relevance of a specific
cross-activation event in more complex systems.16,17 We
leveraged our chemical toolkit to dissect the KLK activome
in PCa cells by inhibiting one KLK and assessing the change in
activity of the remaining KLKs, demonstrating that KLK14
cross-activates KLK2 and KLK3. We suggest that our approach
could be used in principle to dissect any disease-related or
tissue KLK activome and reveal which KLKs represent
potential targets for therapeutic intervention in a specific
context.
Our data using LNCaP cells demonstrate that KLK activity
in the tumor microenvironment is likely regulated by AR
signaling (Figure 4B). However, while KLK2 and KLK3
activity is upregulated by AR activation, KLK14 activity
increases upon AR inhibition. The latter observation suggests a
role for active KLK14 in the development of resistance to AR
competitive inhibitors. Despite the divergent response to AR
signaling, we show that KLK2, KLK3, and KLK14 are
coexpressed and active in hormone responsive cells from
different sites of PCa metastasis, including lymph nodes, liver,
and bone. An explanation for this paradox lies in the
Overall, these data dissect the KLK activome in LNCaP cells
and show that KLK14 activity drives PCa cell migration.
DISCUSSION
■
Despite recent progress in determining the pathophysiological
roles of individual KLKs at different stages of PCa, the
potential of selectively inhibiting these proteases for
therapeutic intervention is yet to be realized.55 It has become
clear that KLKs do not work alone but instead are individual
components of complex networks that when deregulated drive
8920
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
heterogeneity of prostate tumors, which develop resistance to
AR inhibition through diverse mechanisms including upregu-
lation of AR-V7 (an AR splice-variant with constitutive
transcriptional activity that lacks a ligand-binding domain)
and glucocorticoid receptor (GR).63−65 Both AR-V7 and GR
signaling drive increased KLK2 and KLK3 expression even in
the face of Enz treatment. Similarly, a subset of AR-null PCa
cells have high KLK14 expression, which is not perturbed by
DHT treatment.15 We hypothesize that these diverse PCa cell
populations contribute to a buildup of KLK activities in the
tumor microenvironment. In this study we found that
androgen-independent activation of AR in PCa cells by
osteoblasts results in an increase of KLK2 and KLK3 activity
in the bone microenvironment. Previous studies have shown
that KLK3 increases bone volume and osteoblast numbers in
vivo via a TGFβ-dependent mechanism.47−50 It has also been
demonstrated in vitro that KLK2 and KLK14 can activate
TGFβ-1 and TGFβ-2.66 Furthermore, our data show that
osteoblasts, as well as PCa cells, secrete active KLK14 into the
bone microenvironment and that KLK14 can cross-activate
KLK2 and KLK3, thus providing a proteolytic cascade that
may amplify osteoblast proliferation and differentiation and
result in a double paracrine signaling event that drives both
tumor growth and woven bone formation (Figure 5C).51,52 We
propose that the KLK activome warrants further study as an
actionable therapeutic target in bone metastatic PCa, either
alone or in combination with enzalutamide.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.1c03950.
■
sı
Detailed synthetic methods describing probe synthesis
and positional scanning library synthesis, biochemical
methods including cell culture, tumor homogenate
preparation, enzyme assays, and cell migration assays,
Molecular Operating Environment (MOE) docking
studies, synthesis schemes and probe structures, probe
labeling gels, enzyme inhibition assays, raw positional
scanning library data, probe docking poses, blot
quantifications, cell proliferation assays, uncropped
Western blots, NMR spectra, HPLC spectra, and MS
spectra (PDF)
AUTHOR INFORMATION
Corresponding Author
■
Edward W. Tate − Department of Chemistry, Molecular
Sciences Research Hub, Imperial College London, London
W12 0BZ, U.K.; The Francis Crick Institute, London NW1
1AT, U.K.; orcid.org/0000-0003-2213-5814;
Email: e.tate@imperial.ac.uk
Authors
Scott Lovell − Department of Chemistry, Molecular Sciences
Research Hub, Imperial College London, London W12 0BZ,
U.K.
Finally, the chemical tools developed in this study may allow
KLK activity to be explored as a novel biomarker in PCa.
Population-based screening with the PSA test has decreased
PCa mortality; however, due to the test’s relatively poor
specificity, it has also increased the detection of indolent
cancers and resulted in overtreatment of patients. Identification
of novel biomarkers with increased specificity for aggressive
PCa detection may aid in risk stratification and the appropriate
identification of men for prostate biopsy.9 An early hallmark of
PCa is the downregulation of zinc transporter proteins, which
results in a decrease in the concentration of Zn2+ ions in the
prostate.67 Zinc ions allosterically regulate the activity of KLKs,
and thus there is an increase in KLK activity in prostatic fluid
obtained from patients with PCa.19 Measurement of KLK
activity in first void urine samples, which contain prostatic
fluid, may be worth exploring in the quest for a more accurate
diagnostic tool.68−70 To this end, we note that biotinylated
ABPs have previously been integrated into ELISA platforms to
enable highly sensitive and high throughput assessment of
protease activity.71 Similarly, as KLK activity is higher in PCa
tissue compared to neighboring healthy tissue, KLK probes
could in future be used as fluorescent guides for surgeons
striving to obtain negative margins during tumor resection,72
potentially facilitated by the development of quenched-
fluorescent derivatives.73,74
Leran Zhang − Department of Chemistry, Molecular Sciences
Research Hub, Imperial College London, London W12 0BZ,
U.K.
Thomas Kryza − Mater Research InstituteThe University of
Queensland, Translational Research Institute,
Woolloongabba, QLD 4102, Australia; Australian Prostate
Cancer Research Centre-Queensland (APCRC-Q), Institute
of Health & Biomedical Innovation and School of Biomedical
Sciences, Faculty of Health, Queensland University of
Technology, Translational Research Institute,
Woolloongabba, QLD 4102, Australia
Anna Neodo − Department of Chemistry, Molecular Sciences
Research Hub, Imperial College London, London W12 0BZ,
U.K.
Nathalie Bock − Australian Prostate Cancer Research Centre-
Queensland (APCRC-Q), Institute of Health & Biomedical
Innovation and School of Biomedical Sciences, Faculty of
Health, Queensland University of Technology, Translational
Research Institute, Woolloongabba, QLD 4102, Australia
Elena De Vita − Department of Chemistry, Molecular Sciences
Research Hub, Imperial College London, London W12 0BZ,
U.K.
Elizabeth D. Williams − Australian Prostate Cancer Research
Centre-Queensland (APCRC-Q), Institute of Health &
Biomedical Innovation and School of Biomedical Sciences,
Faculty of Health, Queensland University of Technology,
Translational Research Institute, Woolloongabba, QLD
4102, Australia; orcid.org/0000-0002-3364-6655
Elisabeth Engelsberger − Department of Chemistry,
Molecular Sciences Research Hub, Imperial College London,
London W12 0BZ, U.K.
CONCLUSION
■
This study describes the development of a versatile chemical
probe platform to enable dissection of KLK activome activity
in PCa, leading support for the hypothesis that the KLK
activome drives PCa progression and holds promise as an
actionable therapeutic target in CRPC.
Congyi Xu − Department of Chemistry, Molecular Sciences
Research Hub, Imperial College London, London W12 0BZ,
U.K.
8921
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(7) Prassas, I.; Eissa, A.; Poda, G.; Diamandis, E. P. Unleashing the
therapeutic potential of human kallikrein-related serine proteases. Nat.
Rev. Drug Discovery 2015, 14, 183−202.
Alexander T. Bakker − Department of Chemistry, Molecular
Sciences Research Hub, Imperial College London, London
W12 0BZ, U.K.
Maria Maneiro − Department of Chemistry, Molecular
Sciences Research Hub, Imperial College London, London
W12 0BZ, U.K.; orcid.org/0000-0001-6229-9157
Reiko J. Tanaka − Department of Bioengineering, Imperial
College London, London SW7 2AZ, U.K.
Charlotte L. Bevan − Department of Surgery and Cancer,
Imperial Centre for Translational and Experimental
Medicine, Imperial College London, Hammersmith Hospital,
London W12 0NN, U.K.; orcid.org/0000-0002-7533-
0552
Judith A. Clements − Australian Prostate Cancer Research
Centre-Queensland (APCRC-Q), Institute of Health &
Biomedical Innovation and School of Biomedical Sciences,
Faculty of Health, Queensland University of Technology,
Translational Research Institute, Woolloongabba, QLD
4102, Australia
(8) Fuhrman-Luck, R. A.; Loessner, D.; Clements, J. A. Kallikrein-
Related Peptidases in Prostate Cancer: From Molecular Function to
Clinical Application. EJIFCC 2014, 25, 269−81.
(9) Lilja, H.; Ulmert, D.; Vickers, A. J. Prostate-specific antigen and
prostate cancer: prediction, detection and monitoring. Nat. Rev.
Cancer 2008, 8, 268−278.
(10) Darst, B. F.; et al. The Four-Kallikrein Panel Is Effective in
Identifying Aggressive Prostate Cancer in a Multiethnic Population.
Cancer Epidemiol., Biomarkers Prev. 2020, 29, 1381−1388.
(11) Mize, G. J.; Wang, W.; Takayama, T. K. Prostate-specific
kallikreins-2 and −4 enhance the proliferation of DU-145 prostate
cancer cells through protease-activated receptors-1 and -2. Mol.
Cancer Res. 2008, 6, 1043−1051.
(12) Yonou, H.; et al. Prostate-specific antigen stimulates
osteoprotegerin production and inhibits receptor activator of nuclear
factor-KappaB ligand expression by human osteoblasts. Prostate 2007,
67, 840−848.
(13) Nadiminty, N.; et al. Prostate-Specific Antigen Modulates
Genes Involved in Bone Remodeling and Induces Osteoblast
Differentiation of Human Osteosarcoma Cell Line SaOS-2. Clin.
Cancer Res. 2006, 12, 1420−1430.
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.1c03950
(14) Lose, F.; et al. The kallikrein 14 gene is down-regulated by
androgen receptor signalling and harbours genetic variation that is
associated with prostate tumour aggressiveness. Biol. Chem. 2012,
393, 403−412.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
(15) Kryza, T.; et al. The molecular function of kallikrein-related
peptidase 14 demonstrates a key modulatory role in advanced
prostate cancer. Mol. Oncol. 2020, 14, 105−128.
■
The authors thank L. Haigh (Department of Chemistry Mass
Spectrometry Facility, Imperial College London) for assistance
in acquiring high-resolution mass spectrometry (HRMS) data.
S.L. was supported by the EPSRC Centre for Doctoral
Training in Physical Sciences Innovation in Chemical Biology
for Bioindustry and Healthcare (Grant EP/LO15498/1) and
an EPSRC Doctoral Prize fellowship. M.M. thanks the Xunta
de Galicia for a postdoctoral fellowship. E.D.V. is supported by
a H2020 MSCA-IF fellowship (Grant 890900). L.Z. was
supported by Cancer Research UK (Grant C24523/A25192).
E.W.T. thanks Worldwide Cancer Research for support (Grant
19-0059). A.N. was supported by a postdoctoral mobility
fellowship from the Swiss National Science Foundation. C.L.B.
acknowledges infrastructure support from the CRUK Imperial
Centre (Grant CS24523/A25147). The table of contents and
abstract graphic was created with BioRender.com.
(16) Yoon, H.; et al. Activation Profiles and Regulatory Cascades of
the Human Kallikrein-related Peptidases. J. Biol. Chem. 2007, 282,
31852−31864.
(17) Yoon, H.; et al. A completed KLK activome profile:
investigation of activation profiles of KLK9, 10, and 15. Biol. Chem.
2009, 390, 373−7.
(18) Kasparek, P.; et al. KLK5 and KLK7 Ablation Fully Rescues
Lethality of Netherton Syndrome-Like Phenotype. PLoS Genet. 2017,
13, e1006566.
(19) Goettig, P.; Magdolen, V.; Brandstetter, H. Natural and
synthetic inhibitors of kallikrein-related peptidases (KLKs). Biochimie
2010, 92, 1546.
(20) Rawlings, N. D.; et al. The MEROPS database of proteolytic
enzymes, their substrates and inhibitors in 2017 and a comparison
with peptidases in the PANTHER database. Nucleic Acids Res. 2018,
46, D624−D632.
́
(21) Maslanka, M.; Mucha, A. Recent Developments in Peptidyl
REFERENCES
Diaryl Phoshonates as Inhibitors and Activity-Based Probes for Serine
Proteases. Pharmaceuticals 2019, 12, 86.
■
(1) Pascale, M.; Azinwi, C. N.; Marongiu, B.; et al. The outcome of
prostate cancer patients treated with curative intent strongly depends
on survival after metastatic progression. BMC Cancer 2017, 17, 651.
(2) Ryan, C. J.; et al. Abiraterone acetate plus prednisone versus
placebo plus prednisone in chemotherapy-naive men with metastatic
castration-resistant prostate cancer (COU-AA-302): final overall
survival analysis of a randomised, double-blind, placebo-controlled
phase 3 study. Lancet Oncol. 2015, 16, 152−160.
(22) Kojtari, A.; Shah, V.; Babinec, J. S.; Yang, C.; Ji, H.-F. Structure-
Based Drug Design of Diphenyl α-Aminoalkylphosphonates as
Prostate-Specific Antigen Antagonists. J. Chem. Inf. Model. 2014, 54,
2967−2979.
(23) LeBeau, A. M.; Banerjee, S. R.; Pomper, M. G.; Mease, R. C.;
Denmeade, S. R. Optimization of peptide-based inhibitors of prostate-
specific antigen (PSA) as targeted imaging agents for prostate cancer.
Bioorg. Med. Chem. 2009, 17, 4888−4893.
(24) Kasperkiewicz, P.; Altman, Y.; D’Angelo, M.; Salvesen, G. S.;
Drag, M. Toolbox of Fluorescent Probes for Parallel Imaging Reveals
Uneven Location of Serine Proteases in Neutrophils. J. Am. Chem. Soc.
2017, 139, 10115−10125.
(25) Wright, M. H.; Sieber, S. A. Chemical proteomics approaches
for identifying the cellular targets of natural products. Nat. Prod. Rep.
2016, 33, 681−708.
(26) de Bruin, G.; et al. A Set of Activity-Based Probes to Visualize
Human (Immuno)proteasome Activities. Angew. Chem., Int. Ed. 2016,
55, 4199−203.
(3) Rice, M. A.; Malhotra, S. V.; Stoyanova, T. Second-generation
antiandrogens: From discovery to standard of care in castration
resistant prostate cancer. Front. Oncol. 2019, 10, 801.
(4) Body, J.-J.; Casimiro, S.; Costa, L. Targeting bone metastases in
prostate cancer: improving clinical outcome. Nat. Rev. Urol. 2015, 12,
340−356.
(5) Watson, P. A.; Arora, V. K.; Sawyers, C. L. Emerging
mechanisms of resistance to androgen receptor inhibitors in prostate
cancer. Nat. Rev. Cancer 2015, 15, 701−711.
(6) Morova, T.; et al. Androgen receptor-binding sites are highly
mutated in prostate cancer. Nat. Commun. 2020, 11, 832.
8922
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(27) Malm, J.; Hellman, J.; Hogg, P.; Lilja, H. Enzymatic action of
prostate-specific antigen (PSA or hK3): substrate specificity and
regulation by Zn(2+), a tight-binding inhibitor. Prostate 2000, 45,
132−9.
(28) LeBeau, A. M.; Singh, P.; Isaacs, J. T.; Denmeade, S. R. Potent
and selective peptidyl boronic acid inhibitors of the serine protease
prostate-specific antigen. Chem. Biol. 2008, 15, 665−74.
(29) Poreba, M.; Salvesen, G. S.; Drag, M. Synthesis of a HyCoSuL
peptide substrate library to dissect protease substrate specificity. Nat.
Protoc. 2017, 12, 2189−2214.
(30) Kasperkiewicz, P.; et al. Design of ultrasensitive probes for
human neutrophil elastase through hybrid combinatorial substrate
library profiling. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 2518−2523.
(31) van de Plassche, M. A. T.; et al. Use of Non-Natural Amino
Acids for the Design and Synthesis of a Selective, Cell-Permeable
MALT1 Activity-Based Probe. J. Med. Chem. 2020, 63, 3996−4004.
(32) Poreba, M.; et al. Caspase selective reagents for diagnosing
apoptotic mechanisms. Cell Death Differ. 2019, 26, 229−244.
(33) van Soom, J.; et al. The first potent diphenyl phosphonate
KLK4 inhibitors with unexpected binding kinetics. MedChemComm
2015, 6, 1954−1958.
(34) Skala, W.; et al. Structure-function analyses of human kallikrein-
related peptidase 2 establish the 99-loop as master regulator of
activity. J. Biol. Chem. 2014, 289, 34267−34283.
(35) Ménez, R.; et al. Crystal Structure of a Ternary Complex
between Human Prostate-specific Antigen, Its Substrate Acyl
Intermediate and an Activating Antibody. J. Mol. Biol. 2008, 376,
1021−1033.
(36) Molecular Operating Environment (MOE); Chemical Computing
Group ULC (1010 Sherbooke St. West, Suite No. 910, Montreal, QC,
Canada), 2020.
(37) Singh, P.; LeBeau, A. M.; Lilja, H.; Denmeade, S. R.; Isaacs, J.
T. Molecular insights into substrate specificity of prostate specific
antigen through structural modeling. Proteins: Struct., Funct., Genet.
2009, 77, 984−993.
(38) de Veer, S. J.; Swedberg, J. E.; Parker, E. A.; Harris, J. M. Non-
combinatorial library screening reveals subsite cooperativity and
identifies new high-efficiency substrates for kallikrein-related
peptidase 14. Biol. Chem. 2012, 393, 331−341.
(39) de Veer, S. J.; et al. Selective Substrates and Inhibitors for
Kallikrein-Related Peptidase 7 (KLK7) Shed Light on KLK
Proteolytic Activity in the Stratum Corneum. J. Invest. Dermatol.
2017, 137, 430−439.
(40) Nguyen, H. M.; et al. LuCaP Prostate Cancer Patient-Derived
Xenografts Reflect the Molecular Heterogeneity of Advanced Disease
and Serve as Models for Evaluating Cancer Therapeutics. Prostate
2017, 77, 654−671.
(41) Nevedomskaya, E.; Baumgart, S. J.; Haendler, B. Recent
advances in prostate cancer treatment and drug discovery. Int. J. Mol.
Sci. 2018, 19, 1359.
(42) Lu, Y.; et al. Osteoblasts induce prostate cancer proliferation
and PSA expression through interleukin-6-mediated activation of the
androgen receptor. Clin. Exp. Metastasis 2004, 21, 399−408.
(43) Sung, S.-Y.; et al. Coevolution of prostate cancer and bone
stroma in three-dimensional coculture: implications for cancer growth
and metastasis. Cancer Res. 2008, 68, 9996−10003.
(44) Nguyen, D. P.; Li, J.; Tewari, A. K. Inflammation and prostate
cancer: The role of interleukin 6 (IL-6). BJU International 2014, 113,
986−992.
(45) Bock, N.; et al. Engineering osteoblastic metastases to delineate
the adaptive response of androgen-deprived prostate cancer in the
bone metastatic microenvironment. Bone Res. 2019, 7, 13.
(46) Shokoohmand, A.; et al. Microenvironment engineering of
osteoblastic bone metastases reveals osteomimicry of patient-derived
prostate cancer xenografts. Biomaterials 2019, 220, 119402.
(47) Killian, C. S.; Corral, D. A.; Kawinski, E.; Constantine, R. I.
Mitogenic Response of Osteoblast Cells to Prostate-Specific Antigen
Suggests an Activation of Latent TGF-β and a Proteolytic Modulation
of Cell Adhesion Receptors. Biochem. Biophys. Res. Commun. 1993,
192, 940−947.
(48) Dallas, S. L.; et al. Preferential production of latent
transforming growth factor ?-2 by primary prostatic epithelial cells
and its activation by prostate-specific antigen. J. Cell. Physiol. 2005,
202, 361−370.
(49) Yonou, H.; et al. Prostate-Specific Antigen Induces Osteoplastic
Changes by an Autonomous Mechanism. Biochem. Biophys. Res.
Commun. 2001, 289, 1082−1087.
́
́
(50) Cumming, A. P.; Hopmans, S. N.; Vukmirovic-Popovic, S.;
Duivenvoorden, W. C. PSA affects prostate cancer cell invasion in
vitro and induces an osteoblastic phenotype in bone in vivo. Prostate
Cancer Prostatic Dis. 2011, 14, 286−294.
(51) Guise, T. A.; et al. Basic Mechanisms Responsible for
Osteolytic and Osteoblastic Bone Metastases. Clin. Cancer Res.
2006, 12, 6213s−6216s.
(52) Iwamura, M.; Hellman, J.; Cockett, A. T. K.; Lilja, H.;
Gershagen, S. Alteration of the hormonal bioactivity of parathyroid
hormone-related protein (PTHrP) as a result of limited proteolysis by
prostate-specific antigen. Urology 1996, 48, 317−325.
(53) Williams, S. A.; Xu, Y.; De Marzo, A. M.; Isaacs, J. T.;
Denmeade, S. R. Prostate-specific antigen (PSA) is activated by KLK2
in prostate cancer ex vivo models and in prostate-targeted PSA/KLK2
double transgenic mice. Prostate 2010, 70, 788−96.
(54) Mundy, G. R. Metastasis to bone: Causes, consequences and
therapeutic opportunities. Nat. Rev. Cancer 2002, 2, 584−593.
(55) Sotiropoulou, G.; Pampalakis, G. Targeting the kallikrein-
related peptidases for drug development. Trends Pharmacol. Sci. 2012,
33, 623−34.
(56) Beaufort, N.; et al. Interdependence of kallikrein-related
peptidases in proteolytic networks. Biol. Chem. 2010, 391, 581−7.
(57) Ohler, A.; Debela, M.; Wagner, S.; Magdolen, V.; Becker-Pauly,
C. Analyzing the protease web in skin: meprin metalloproteases are
activated specifically by KLK4, 5 and 8 vice versa leading to
processing of proKLK7 thereby triggering its activation. Biol. Chem.
2010, 391, 455−60.
(58) Tanaka, R. J.; Ono, M.; Harrington, H. A. Skin Barrier
Homeostasis in Atopic Dermatitis: Feedback Regulation of Kallikrein
Activity. PLoS One 2011, 6, e19895.
(59) Li, N.; et al. Relative quantification of proteasome activity by
activity-based protein profiling and LC-MS/MS. Nat. Protoc. 2013, 8,
1155−1168.
(60) Kasperkiewicz, P.; Poreba, M.; Groborz, K.; Drag, M. Emerging
challenges in the design of selective substrates, inhibitors and activity-
based probes for indistinguishable proteases. FEBS J. 2017, 284,
1518−1539.
(61) Shaw, J. L.; Diamandis, E. P. Distribution of 15 Human
Kallikreins in Tissues and Biological Fluids. Clin. Chem. 2007, 53,
1423−1432.
(62) Furio, L.; et al. Transgenic kallikrein 5 mice reproduce major
cutaneous and systemic hallmarks of Netherton syndrome. J. Exp.
Med. 2014, 211, 499−513.
(63) Shah, R. B.; et al. Androgen-independent prostate cancer is a
heterogeneous group of diseases: lessons from a rapid autopsy
program. Cancer Res. 2004, 64, 9209−16.
(64) Arora, V. K.; et al. Glucocorticoid receptor confers resistance to
antiandrogens by bypassing androgen receptor blockade. Cell 2013,
155, 1309−22.
(65) Guo, Z.; et al. A Novel Androgen Receptor Splice Variant Is
Up-regulated during Prostate Cancer Progression and Promotes
Androgen Depletion-Resistant Growth. Cancer Res. 2009, 69, 2305−
2313.
(66) Emami, N.; Diamandis, E. P. Potential role of multiple
members of the kallikrein-related peptidase family of serine proteases
in activating latent TGF beta 1 in semen. Biol. Chem. 2010, 391, 85−
95.
(67) Costello, L. C.; Feng, P.; Milon, B.; Tan, M.; Franklin, R. B.
Role of zinc in the pathogenesis and treatment of prostate cancer:
critical issues to resolve. Prostate Cancer Prostatic Dis. 2004, 7, 111−7.
8923
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(68) Obiezu, C. V.; et al. Human Kallikrein 4: Quantitative Study in
Tissues and Evidence for Its Secretion Into Biological Fluids. Clin.
Chem. 2005, 51, 1432−1442.
(69) Theodorescu, D.; et al. Discovery and validation of urinary
biomarkers for prostate cancer. Proteomics: Clin. Appl. 2008, 2, 556−
570.
(70) Webb, M.; et al. Methodology for the at-home collection of
urine samples for prostate cancer detection. BioTechniques 2020, 68,
65−71.
(71) Eitelhuber, A. C.; et al. Activity-based probes for detection of
active MALT1 paracaspase in immune cells and lymphomas. Chem.
Biol. 2015, 22, 129−38.
(72) Nguyen, Q. T.; Tsien, R. Y. Fluorescence-guided surgery with
live molecular navigation  a new cutting edge. Nat. Rev. Cancer
2013, 13, 653−662.
(73) Oresic Bender, K.; et al. Design of a Highly Selective Quenched
Activity-Based Probe and Its Application in Dual Color Imaging
Studies of Cathepsin S Activity Localization. J. Am. Chem. Soc. 2015,
137, 4771−4777.
(74) Serim, S.; Baer, P.; Verhelst, S. H. L. Mixed alkyl aryl
phosphonate esters as quenched fluorescent activity-based probes for
serine proteases. Org. Biomol. Chem. 2015, 13, 2293−2299.
8924
https://doi.org/10.1021/jacs.1c03950
J. Am. Chem. Soc. 2021, 143, 8911−8924