Simultaneous fluorescent monitoring of proteasomal subunit catalysis

Article abstract of DOI:10.1021/ja907226n

The proteasome, a multicatalytic protease, displays distinct chymotrypsin-like, caspase-like, and trypsin-like activities at three different subunits of the multimeric complex. Fluorescent substrates for each of these active sites have been described. However, since the fluorescent properties of these substrates are very similar, it is not possible to simultaneously monitor catalysis of two or more activities. We have developed a long wavelength ( _λex = 600 nm, _λem = 700 nm) fluorescent substrate for the chymotrypsinlike active site via a combinatorial library strategy. This peptide-based substrate is a highly selective proteasomal chymotrypsin-like sensor, as assessed by a series of proteasomal active site mutants in yeast cell lysates. A corresponding caged analog of the sensor has been prepared, which is resistant to proteolysis until activated by 349 nm light. The latter affords the opportunity to assess proteasomal activity with a high degree of temporal control. The distinct photophysical properties of the sensor allow the chymotrypsin-like activity to be simultaneously monitored during caspase-like or trypsin-like catalysis. We have found that chymotrypsin-like activity is enhanced in the presence of the trypsin-like substrate but reduced in the presence of caspase-like substrate. Furthermore, the chymotrypsin-like sensor hinders the activity of both the caspase- and trypsin-like active sites. Coincident monitoring of two catalytic active sites furnishes twothirds coverage of total proteasomal activity, which should provide the means to address if and how the distinct active sites of the proteasome influence one another during catalysis.

Full text of DOI:10.1021/ja907226n

Published on Web 01/15/2010
Simultaneous Fluorescent Monitoring of Proteasomal Subunit
Catalysis
†
‡
†
§
Aya Wakata, Hsien-Ming Lee, Philipp Rommel, Alexei Toutchkine,
,
†
,‡
Marion Schmidt,* and David S. Lawrence*
Department of Biochemistry, Albert Einstein College of Medicine, Bronx, New York 10461,
Sigma-Aldrich Co., 3 Strathmore Rd., Natick, Massachusetts 01760, and Department of
Chemistry, DiVision of Medicinal Chemistry and Natural Products, Department of
Pharmacology, UniVersity of North Carolina, Chapel Hill, North Carolina 27599
Received August 26, 2009; E-mail: mschmidt@aecom.yu.edu; lawrencd@email.unc.edu
Abstract: The proteasome, a multicatalytic protease, displays distinct chymotrypsin-like, caspase-like, and
trypsin-like activities at three different subunits of the multimeric complex. Fluorescent substrates for each
of these active sites have been described. However, since the fluorescent properties of these substrates
are very similar, it is not possible to simultaneously monitor catalysis of two or more activities. We have
developed a long wavelength (λex ) 600 nm, λem ) 700 nm) fluorescent substrate for the chymotrypsin-
like active site via a combinatorial library strategy. This peptide-based substrate is a highly selective
proteasomal chymotrypsin-like sensor, as assessed by a series of proteasomal active site mutants in yeast
cell lysates. A corresponding caged analog of the sensor has been prepared, which is resistant to proteolysis
until activated by 349 nm light. The latter affords the opportunity to assess proteasomal activity with a high
degree of temporal control. The distinct photophysical properties of the sensor allow the chymotrypsin-like
activity to be simultaneously monitored during caspase-like or trypsin-like catalysis. We have found that
chymotrypsin-like activity is enhanced in the presence of the trypsin-like substrate but reduced in the
presence of caspase-like substrate. Furthermore, the chymotrypsin-like sensor hinders the activity of both
the caspase- and trypsin-like active sites. Coincident monitoring of two catalytic active sites furnishes two-
thirds coverage of total proteasomal activity, which should provide the means to address if and how the
distinct active sites of the proteasome influence one another during catalysis.
Aberrant pathways responsible for cancerous cell growth are
remarkably plastic, endowing transformed cells with an ex-
traordinary resilience against inhibitory (anticancer) agents. As
a consequence, there is emerging evidence that the future of
cancer chemotherapy lies in targeting multiple members of
three catalytically distinct protease active sites: chymotrypsin-
like (Ch-L), caspase-like (Ca-L), and trypsin-like (T-L).
2
Previous studies suggest that some of these active sites can
4
allosterically regulate one another, although others have
5
questioned this interpretation. Fluorophore-labeled substrates
1
abnormal biochemical networks. In an analogous vein, the
have been designed for each of the individual active sites and,
upon hydrolysis, a fluorescent response is observed. However,
the fluorophores employed in these substrates possess similar
photophysical properties and consequently it is not possible to
simultaneously monitor the activity of multiple catalytic activi-
ability to simultaneously measure the catalytic activity at several
sites in a network offers the means to detect aberrant biochemi-
cal behavior, screen for new inhibitory agents on a more global
biochemical scale, and potentially predict the efficacy of various
chemotherapeutic cocktails for individual patients. Conspiring
biochemical activities may reside at distinct intracellular sites,
within the same organelle, or as part of a single protein complex.
An especially beautiful, yet challenging, example of the latter
is the proteasome, which serves as the primary protein digestive
6
ties. We describe herein the acquisition and characterization
of a long wavelength Ch-L sensor, the synthesis of the
corresponding caged analog, and the simultaneous observation
of activities from proteolytically distinct proteasomal active sites.
2
Results and Discussion
apparatus in the cytoplasm and nuclei of eukaryotic cells.
Inhibitors of the proteasome (bortezomib, salinosporamide A,
carfilzomib) have received considerable attention as anticancer
agents. The highly oligomeric proteasomal complex houses
Some observations regarding substrate specificity of the Ch-L
subunit have been reported, such as the preference for hydro-
3
(
4) Kisselev, A. F.; Akopian, T. N.; Castillo, V.; Goldberg, A. L. Mol.
Cell 1999, 4, 395–402.
†
Albert Einstein College of Medicine.
University of North Carolina.
Sigma-Aldrich Co.
‡
(5) Schmidtke, G.; Ench, S.; Groettrup, M.; Holzh u¨ tter, H.-G. J. Biol.
Chem. 2000, 275, 22056–63.
§
(
(
1) Grant, S. Best Pract. Res. Clin. Haematol. 2008, 21, 629–37.
2) Marques, A. J.; Palanimurugan, R.; Matias, A. C.; Ramos, P. C.;
Dohmen, R. J. Chem. ReV. 2009, 109, 1509–36.
(6) (a) Orlowski, M.; Cardozo, C.; Hidalgo, M. C.; Michaud, C.
Biochemistry 1991, 30, 5999–6005. (b) Kessler, B. M.; Tortorella,
D.; Altun, M.; Kisselev, A. F.; Fiebiger, E.; Hekking, B. G.; Ploegh,
H. L.; Overkleeft, H. S. Chem. Biol. 2001, 8, 913–29. (c) Kisselev,
A. F.; Goldberg, A. Methods Enzymol. 2005, 398, 364–78.
(
3) Yang, H.; Zonder, J. A.; Dou, Q. P. Expert. Opin. InVestig. Drugs
2
009, 18, 957–71.
1
578 9 J. AM. CHEM. SOC. 2010, 132, 1578–1582
10.1021/ja907226n  2010 American Chemical Society

Fluorescent Monitoring of Proteaosomal Catalysis
A R T I C L E S
a
Table 1. Four Lead Peptides from the Ch-L Sensor Library
V
max
k
(M
cat/K
s )
m
peptide
Flc increase
m
K (µM)
-1 -1
(
nmol/min·mg)
3, Ac-HWSL-Dap(Fl) 22.4-fold 56 ( 11
4, Ac-HWSL-Dab(Fl) 27.1-fold 80 ( 13
5, Ac-HWSL-Lys(Fl) 24.9-fold 81 ( 8
6, Ac-WHSL-Lys(Fl) 23.5-fold 73 ( 15
1.4 ( 0.2
1.0 ( 0.1
2.7 ( 0.2
1.7 ( 0.3
300 ( 70
150 ( 30
390 ( 50
280 ( 70
a
Flc increase
) fold increase in fluorescence intensity upon
incubation with proteasome. The excitation and emission spectra of
sensor 5 is furnished in Figure S-2 (Supporting Information).
phobic residues (Leu > Tyr > Phe > Trp > Ile) on the N-terminal
Figure 1. Yeast cell lysate assessment of the Ch-L selectivity of peptide
5 and time-dependent photolytic liberation of 10 [fold increase in
fluorescence intensity upon proteasome treatment (REL. FLC) as a function
of time]. (A) Peptide 5 (9.25 µM) with cell lysates containing wild-type
7
side of the scissile bond (P1). However, it is also clear that
the P1 amino acid is not the only factor determining the location
8
of the cleavage site. On the basis of the screening of a two-
(
red), defective T-L/Ca-L (blue), defective Ch-L (brown), defective T-L/
position library, Harris et al. identified Pro-Glu-Gly-Phe and
His-His-Ser-Leu as two peptide sequences in which Ch-L site
substrate specificity is optimized, where the bold typeface
Ca-L proteasome with the proteasome inhibitor MG132 (5 µM, green); no
lysate (black). (B) Peptide 10 (5.65 µM) treated with wild-type cell lysate
(black), and preirradiated for 2 min (red) or 10 min (blue).
8
b
represents the site of proteolysis. In order to identify a
proteasome sensor for Ch-L activity, we prepared a small library
of peptides based on two considerations: (1) previously dem-
thereby providing the better “fit” between the substrate and the
catalytic residues of the enzyme.
8
b
onstrated Ch-L substrates and (2) the ability of Trp to quench
The selectivity of peptide 5 as a Ch-L proteasomal substrate
was assessed using a series of yeast cell lysates. Cell lysate
autofluorescence does not interfere with the observed fluores-
cence change owing to the characteristic long wavelength
excitation (λex ) 663 nm) and emission (λem ) 678 nm) of the
9
the fluorescence of oxazine fluorophores. We sequentially
replaced each residue in the two standard peptide sequences
His-His-Ser-Leu-P1′ and Pro-Glu-Gly-Phe-P1′ with tryptophan.
Furthermore, the side chain length at the P1′ site, which contains
1
0
-1
the oxazine fluorophore 1, was varied [i.e., L-2,3-diamino-
propionic acid (Dap), L-2,4-diaminobutanoic acid (Dab), Orn,
and Lys]. A structural outline of the Ch-L sensor library is
depicted in 2, where n ) 1, 2, 3, and 4 on the P1′ side chain.
oxazine probe, as well as its extinction coefficient (119 000 M
-
1
10
cm ) and quantum yield (0.33). Even submicromolar con-
centrations of sensor 5 are sufficient to observe hydrolytic
activity (Figure S-1, Supporting Information), reducing the
11
likelihood of interfering with endogenous biochemical activity.
By comparison, previously described coumarin and ꢀ-naphthy-
lamide-based substrates are employed at concentrations of 100
6
µM or greater. The three distinct proteasome active sites possess
N-terminal Thr residues that are essential for catalytic activity.
Cell lysates of a double mutant strain were prepared, in which
the catalytic Thr was converted to an Ala in both the Ca-L and
1
2
T-L subunits. In this strain the only functional active site of
the proteasome is the Ch-L catalytic subunit and, as expected,
lysates of this mutant strain and that of the wild-type strain
catalyze hydrolysis of peptide 5 at equal rates (Figure 1A, cf.
blue and red curves). In addition, the potent and proteasome-
specific inhibitor MG132 completely abolishes hydrolytic activ-
ity (Figure 1A, green curve), consistent with the notion that
proteolysis is due to the proteasome and not a consequence of
other proteases present in the lysate. Finally, we prepared lysates
of the defective Ch-L subunit (Thr-to-Ala mutant). Only a barely
perceptible fluorescence change was observed (Figure 1A,
brown), thereby establishing peptide 5 as a selective sensor of
the proteasome that can sample Ch-L activity even in crude
lysates. Liquid chromatography-mass spectrometry (LC-MS)
revealed the formation of oxazine-labeled Lys (Figure S-3,
Supporting Information), the expected product generated via
proteolysis at the Leu-Lys amide linkage. We have established
a linear correlation between the proteasome-induced fluores-
cence change and formation of the oxazine-labeled Lys (as
assessed by HPLC; Figure S-4, Supporting Information). An
analogous correlation exists between fluorescence change and
sensor 5 consumption (as assessed by HPLC, Figure S-4,
Supporting Information). Ac-HWSL, the peptide cleavage
The proteasome-induced fluorescence changes for the four
purified leads acquired from library 2 are provided in Table 1
and for the entire crude library in Table S-2 (Supporting
Information). Isolated yeast 20S proteasome was employed for
screening and kinetic characterization. Trp, positioned at P3 and
P4, facilitates efficient quenching and generated, upon protea-
some-mediated proteolysis, fluorescent increases of greater than
2
0-fold. The K
m
values of 3-6 are comparable. Peptide 4
furnishes the largest fluorescence increase but displays a
significantly reduced Vmax relative to peptide 5. The longer Lys
side chain of 5 (relative to the Dap and Dab side chains) may
place the bulky fluorophore away from the active site pocket,
(
(
(
7) Nussbaum, A. K.; Dick, T. P.; Keilholz, W.; Schirle, M.; Stevanovic,
S.; Dietz, K.; Heinemeyer, W.; Groll, M.; Wolf, D. H.; Huber, R.;
Rammensee, H. G.; Schild, H. Proc. Natl. Acad. Sci. U.S.A. 1998,
9
5, 12504–9.
8) (a) Emmerich, N. P.; Nussbaum, A. K.; Stevanovic, S.; Priemer, M.;
Toes, R. E.; Rammensee, H. G.; Schild, H. J. Biol. Chem. 2000, 275,
2
1140–8. (b) Harris, J. L.; Alper, P. B.; Li, J.; Rechsteiner, M.; Backes,
B. J. Chem. Biol. 2001, 8, 1131–41.
9) Doose, S.; Neuweiler, H.; Sauer, M. ChemPhysChem 2005, 6, 2277–
(11) Sharma, V.; Lawrence, D. S. Angew. Chem. Intl. Ed. Engl. 2009, 48,
7290–2.
8
5.
(
10) Toutchkine, A. PCT. Int. Appl., PCT/US2009/46238.
(12) Arendt, C. S.; Hochstrasser, M. EMBO J. 1999, 18, 3575–85.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 5, 2010 1579

A R T I C L E S
Wakata et al.
Scheme 1. Solid-Phase Peptide Synthesis (SPPS) of the Caged Ch-L Sensor 10
product that results from this proteasome-catalyzed hydrolysis,
undergoes subsequent hydrolysis to Ac-HWS (Figure S-3,
Supporting Information).
until activated by light (Figure 1B, black curve). Moreover, the
amount of active substrate released can be controlled by the
time of irradiation [2 min (red) versus 10 min (blue)]. HPLC
and mass spectrometry confirmed that irradiation of the sample
generates sensor 5 and that, in the presence of proteasome,
compound 5 is converted to the hydrolyzed product Lys(ox-
azine) (Figure S-5, Supporting Information). The quantum yield
(0.08) for the conversion of 10 to 5 was determined using a
caged fluorescein derivative as a reference (Figure S-6, Sup-
porting Information).
Although sensor 5 serves as a selective Ch-L substrate, its
ultimate utility as a fluorescent probe of intracellular proteasomal
activity in living cells would be greatly facilitated by the ability
to control when proteasomal activity is observed (e.g., as a
function of cell cycle, following exposure to environmental
stress, etc.). We have established that light-activatable
1
3
(
“caged”) sensors of protein kinases can be used to probe
catalytic activity in living cells in an investigator-controlled
Finally, we employed sensor 5 to simultaneously monitor
Ch-L activity in the presence of previously described fluorescent
substrates for the Ca-L (Z-LLE-ꢀna, 11) and T-L (Boc-LRR-
amc, 12) active sites. The presence of either of these substrates
does not affect the fluorescence of the oxazine fluorophore
14
fashion. However, previously described caged peptides contain
side chain functionality required for catalytic activity (e.g., a
Ser hydroxyl moiety that serves as a phosphoryl acceptor in
protein kinase-catalyzed reactions). By contrast, sensor 5 lacks
a side chain moiety that, upon modification, would render the
peptide impervious to hydrolysis. Recently, we demonstrated
that caged peptides are easily prepared during solid-phase
peptide synthesis via incorporation of a photocleavable moiety
(
Figure S-7, Supporting Information). Ch-L activity increases
in the presence of the T-L substrate but decreases upon exposure
to the Ca-L substrate (Figure 2A). Conversely, the Ch-L
substrate reduces the rate at which both the Ca-L and T-L active
sites process their respective substrates (Figure 2B). Previous
studies have described the effect of noncognate inhibitors and
substrates on catalytic activity. In particular, the catalytic activity
at the Ch-L site is repressed in the presence of a Ca-L substrate
1
5
at a key peptide backbone amide bond. Consequently, we
reasoned that incorporation of an o-nitrobenzyl moiety at the
demonstrated scissile bond in 5 would render the sensor resistant
to proteasome-mediated hydrolysis, until activated by photolysis.
The caged sensor 10 was prepared on-resin using a reductive
alkylation step to generate an N-o-nitrobenzyl intermediate,
followed by acylation of the secondary amine with the acid
chloride of Fmoc-Leu (Scheme 1). All other steps employ
standard Fmoc chemistry. The activity assay with purified yeast
4
and vice versa. This observation has lead, in part, to a model
depicting allosteric crosstalk between the Ca-L and Ch-L active
4
sites. However, this interpretation has been called into ques-
5
tion. Sensor 5, which is both catalytically and spectrally
orthogonal to existing sensors, allows two proteasomal activities
2
0S proteasome revealed that 10 is impervious to hydrolysis
(Ch-L and Ca-L; Ch-L and T-L) to be simultaneously observed.
Consequently, we have been able to not only confirm the
previous report of a reciprocal negative effect of Ca-L and Ch-L
substrates but we have also observed the heretofore unreported
potentiation of Ch-L activity by a T-L substrate. The coincident
monitoring of two catalytic active sites furnishes two-thirds
coverage of total proteasomal activity. The latter, in conjunction
with appropriate inhibitors and various active site mutants, offers
(
(
13) Lee, H. M.; Larson, D. R.; Lawrence, D. S. ACS Chem. Biol. 2009, 4,
4
09–27.
4
14) (a) Dai, Z.; Dulyaninova, N. G.; Kumar, S.; Bresnick, A. R.; Lawrence,
D. S. Chem. Biol. 2007, 14, 1254–60. (b) Wang, Q.; Dai, Z.; Cahill,
S. M.; Blumenstein, M.; Lawrence, D. S. J. Am. Chem. Soc. 2006,
1
28, 14016–7.
15) Nandy, S. K.; Agnes, R. S.; Lawrence, D. S. Org. Lett. 2007, 9, 2249–
2.
(
5
1
580 J. AM. CHEM. SOC. 9 VOL. 132, NO. 5, 2010

Fluorescent Monitoring of Proteaosomal Catalysis
A R T I C L E S
a diode-pumped, Q-switched UV laser that produces 5 ns pulses at
3
49 nm with a peak repetition rate of 1 kHz and 100 mJ pulse
energy (Explorer, Spectra-Physics, Mountain View, CA). Mass
spectra by ESI and MALDI were acquired at the Mass Spectrometry
Laboratory of the Albert Einstein College of Medicine and/or at
Sigma Aldrich Co. HPLC analysis was performed using a Waters
6
00 solvent delivery system and Waters Delta 600 controller with
a 996 photodiode array detector. Analyses were carried out either
on analytical (Altech Apollo C18, 5 µm, 4.6 × 250 mm) or
semipreparative (GraceVydac, Everest C18 HPLC Column, 300 Å,
1
0 µm, 250 × 10 mm) scales.
Peptide Synthesis. All peptides were manually synthesized using
a standard Fmoc solid-phase peptide synthesis protocol using
CLEAR Rink amide resin. The side chains of Glu and Ser were
protected with O-tert-Bu. The side chain of Lys, L-2,3-diamino-
propionic acid (Dap), L-2,3-diaminobutyric acid (Dab), and ornithine
Figure 2. Simultaneous monitoring of proteasomal subunit catalysis [fold
increase in fluorescence intensity upon proteasome treatment (REL. FLC) as a
function of time]. (a) Ch-L activity as assessed with peptide 5 (30 µM, λex
)
6
00 nm, λem ) 700 nm) in the presence of a T-L substrate (100 µM Boc-
LRR-amc, blue), alone (black), or in the presence of a Ca-L substrate (100
(
Orn) were protected with a 1-(4,4-dimethyl-2,6-dioxocyclohexy-
µM Z-LLE-ꢀna, red). (b) Ca-L activity as assessed with Z-LLE-ꢀna (100 µM,
lidene)-3-methylbutyl (ivDde) group. The side chain of His and
Trp were protected with butyloxycarbonyl (Boc) group. Each amino
acid was attached via a standard addition/deprotection stepwise
protocol followed by covalent labeling with the fluorophore and
final cleavage (steps a and b, vide infra). For caged peptides, the
reductive alkylation procedure (c) and coupling of the subsequent
Fmoc-residue (d) were performed under the specified conditions
before the dye modification (b) was carried out.
λ
ex ) 342 nm, λem ) 425 nm) in the absence (red) or presence (orange) of the
Ch-L substrate (30 µM peptide 5) and T-L activity as assessed with Boc-LRR-
amc (100 µM, λex ) 380 nm, λem ) 460 nm) in the absence (blue) or presence
(green) of the Ch-L substrate (30 µM peptide 5).
the means to address if and how the distinct active sites of the
proteasome influence one another during catalysis. These studies
will be reported in due course.
Step a. The amino acid coupling conditions were as follows:
amino acid (3 equiv), HCTU (2.9 equiv), 6-Cl-HOBt (2.9 equiv),
and N,N-diisopropylethylamine (DIPEA) (5.8 equiv) in DMF for
0
.5-2 h at room temperature. Between each coupling the resins
were washed sequentially with DMF (3×), 2-propanol (3×), and
CH Cl
2
2
(3×). Deprotection of the Fmoc group on the growing
peptide chain was effected with 30% piperidine in DMF (20 min)
followed by sequential washes with DMF (3×), 2-propanol (3×),
and CH
2
Cl
2
(3×).
Step b. Following peptide synthesis, each individual peptide-resin
61.25 µmol, 1 equiv) was treated with 1% hydrazine in DMF to
(
In summary, we have prepared a highly selective long
wavelength sensor of proteasomal Ch-L activity. The sensor
has been used, in combination with previously described
fluorescent substrates, to concurrently measure the catalytic
activity of two distinct proteasome active site regions. In
addition, the caged analog of the sensor provides the means to
intracellularly load the reagent in the absence of confounding
background hydrolysis as well as subsequently establish a well-
resolved “in-cell” enzymatic start point.
selectively deprotect the side chain amine of Dap, Dab, Orn, or
Lys (3 × 3 min, 2 × 5 min), followed by washing with DMF (8 ×
2
min), 2-propanol (3×), and CH
2
Cl
2
(3×). The peptidyl resins
were shaken for 5 min with DMF containing 5% DIPEA. The free
amine in each peptide was covalently labeled with the fluorophore
[
∼55 mg, 1.4 equiv; HATU (35 mg, 1.5 equiv), DIPEA (32 µL, 3
equiv) in DMF (∼1.5 mL) on the shaker, overnight]. The peptides
were then cleaved and deprotected via exposure to TFA:H O:
2
triisopropylsilane (TIS) in a ratio of 95:2.5:2.5 and purified by
HPLC. The peptides were isolated via filtration of the resin,
precipitation with ice-cold diethyl ether, and centrifugation. The
precipitates were air-dried and purified by reverse-phase HPLC
using a linear gradient (3%-40% acetonitrile in water with 0.1%
TFA over 40 min). The peak corresponding to the desired peptide
was collected, frozen, and lyophilized. The resulting white, floc-
culent peptides were characterized by MS-ESI.
Step c. For reductive alkylation of Lys(ivDde)-resin with 4,5-
dimethoxy-6-nitrobenzaldehyde (DMNB), the free primary amine
of the side chain protected Lys(ivDde)-resin (12.25 µmol) was
washed with a few milliliters of DMF:MeOH:AcOH (9:9:2) and
then mixed with DMNB (6.25 mg, ∼2.4 equiv) in 600 µL of DMF:
MeOH (1:1) for 50 min. The solvent was then removed and the
imine-forming reaction repeated (2.4-3 equiv up to 2 h). Following
solvent removal, the resin was washed with 3 mL of dry DMF
Materials and Methods
Materials. Materials and chemicals were obtained from Fisher
and Aldrich, except for 1H-benzotriazolium 1-[bis(dimethylami-
no)methylene]-5-chlorohexafluorophosphate (1-),3-oxide (HCTU),
N-{(dimethylamino)-1H-1,2,3-triazolo[4,5-b]pyridino-1-ylmethyl-
ene}-N-methylmethanaminium hexafluorophosphate N-oxide (HATU),
and 2-(7-aza-1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate) 1-hydroxy-6-chlorobenzotriazole (6-Cl-HOBt),
protected amino acids, CLEAR Rink amide resin (100-200 mesh,
0
.43 mmol/g), and NovaSyn TGR resin (0.20 mmol/g), which were
obtained from Advanced ChemTech, Novabiochem, or Peptides
International. 20S particles (wild-type and R3R7N mutant) from
yeast Saccharomyces cereVisiae (strains sDL135 and yMS160,
1
6
respectively) were purified as described previously. See Table
S-1 (Supporting Information) for a list of yeast strains employed
in this study. Fluorogenic proteasomal substrates (except those
synthesized in this study) were purchased from Bachem. Fluores-
cence assays were performed using a SpectraMAX MS fluorescent
plate reader (Molecular Device), and irradiation experiments utilized
(5×), and then 5 equiv of NaBH CN (3.84 mg) in 500 µL of DMF:
3
MeOH:AcOH (9:9:2) was added to the resin and mixed at room
temperature for 20 min. The solvent was removed and resin
successively washed with DMF, DMF/H O, H O, MeOH/CH Cl ,
2
2
2
2
CH Cl , and DMF. The reaction progress was monitored by HPLC
2
2
and ESI-MS analyses of cleaved product from a few milligrams of
resin (95% aqueous TFA). The reaction and the subsequent peptide
synthesis were performed in reaction vessels wrapped in aluminum
foil.
(
16) Schmidt, M.; Haas, W.; Crosas, B.; Santamaria, P. G.; Gygi, S. P.;
Walz, T.; Finley, D. Nat. Struct. Mol. Biol. 2005, 12, 294–303.
J. AM. CHEM. SOC. 9 VOL. 132, NO. 5, 2010 1581

A R T I C L E S
Wakata et al.
Step d. Amino acid coupling to the resin N-DMNB peptide was
fluorescence measured after 1.5 h of incubation (on the shaker)
with the enzyme at 30 °C.
15,17
performed as previously described.
was dried over phosphorus pentoxide in a desiccator overnight. A
50 mL round-bottom flask was charged with Fmoc-Leu-OH (2.26
g, 6.4 mmol, 1 equiv) and purged with Ar. Five milliliters of SOCl
ca. 64 mmol, 10 equiv) was added via a cannula along with 46
µL of anhydrous DMF (0.64 mmol, 0.1 equiv). The clear mixture
was stirred for 1 h at room temperature. Excess SOCl was
evaporated. The acid chloride of Fmoc-Leu was precipitated with
mL of cold CH Cl followed by 70 mL of hexane and dried under
Briefly, the Fmoc-Leu-OH
Activity Assay of Purified 20S Proteasome with Peptide 5.
The enzymatic activity of the proteasome toward 5 was assessed
in a 96-well plate format: 200 µL (per well) containing buffer A
2
2
(
(
2
50 mM Tris-HCl (pH 7.5, 5 mM MgCl , 0.5 mM EDTA) and
peptide 5 (concentrations as indicated in Figure S-1, Supporting
Information). The reaction was initiated by the addition of 2-3
µL enzyme (latent CP, 7.2 µg/µL) and the fluorescence change
monitored (λex ) 550 nm, λem ) 700 nm) at 30 °C. The relative
fluorescence was obtained by simultaneously monitoring the
fluorescence of the blank sample (no enzyme) at the corresponding
substrate concentrations. The correlation between the rate of product
formation and that of fluorescence change was made by HPLC and
MS-ESI analysis of the peptide cleavage products.
2
7
2
2
vacuum. After washing thoroughly with THF, 20 equiv of freshly
prepared Fmoc-Leu-Cl in 125 µL of THF (5 mL/g resin; previously
swelled in THF) was added and the mixture was shaken for 30
min. Forty equivalents of DIPEA was subsequently added and the
mixture allowed to react for an additional 2 h. The reaction was
confirmed by HPLC and ESI-MS analyses of cleaved product from
a few milligrams of resin (95% aqueous TFA).
Cell Lysate Assay. Cells were harvested in log phase, resus-
pended in lysis buffer A (50 mM Tris-HCl, pH 7.4, 1 mM EDTA,
Compound 3, Ac-His-Trp-Ser-Leu-Dap(Fl)-amide (m/z calculated
271, found 1274); compound 4, Ac-His-Trp-Ser-Leu-Dab(Fl)-
5
2
mM MgCl ) and drop-frozen in liquid nitrogen. Frozen yeast cells
1
were lysed by cryolysis using an MM301 grinding mill (Retsch,
Germany) following the manufacturer protocol. Cell extracts were
cleared at 14 000 rpm for 20 min at 4 °C. Protein concentration
was assessed using a Bradford reagent (BioRad). In a 96-well plate,
equal concentrations of the different cell lysates (∼1 mg protein/
amide (m/z calculated 1285, found 1288); compound 5, Ac-His-
Trp-Ser-Leu-Lys(Fl)-amide (m/z calculated 1313, found 1316);
compound 6, Ac-Trp-His-Leu-Lys(Fl)-amide (m/z calculated 1313,
found 1316); compound 10, Ac-His-Trp-Ser-Leu-N(DMNB)-
Lys(Fl)-amide (m/z calculated 1508.7, found 1510.2).
Library Synthesis and Characterization. Synthesis of 32
peptidyl resins was carried out on a semiautomatic peptide
synthesizer (Syro, Sigma-Aldrich Co.) using NovaSyn TGR resin
at 50 µmol scale and a standard Fmoc solid-phase peptide synthesis
protocol (a). The sequences of the peptides are listed in Table S-2
mL in buffer A) were supplemented with 9.2 µM sensor 5 (at 2×
1
concentration in buffer A,
2
/ total assay volume), and the
fluorescence change was monitored via a fluorescent plate reader
(
λ
ex ) 600 nm, λem ) 700 nm).
Simultaneous Assay of Two Catalytic Activities. Each well
of a 96 well plate contained 150 µL of buffer A with various
combinations of substrates [100 µM Z-LLE-ꢀna (Ca-L substrate);
(
Supporting Information). The peptidyl resins (2.5 µmol, ∼13 mg)
were then transferred to a modified 96-well plate (with filter bottom)
for covalent fluorophore labeling in the following manner. First,
the peptidyl resins in the wells were washed with DMF on a shaker
1
00 µM Boc-LRR-amc (T-L substrate); 30 µM ChT-probe (5)].
The reaction was initiated by the addition of 4 µL/well purified
0S proteasome (0.29 mg/mL). The fluorescence change of respec-
tive probes was simultaneously monitored in the plate reader at λex
2
(
4×, 30 s), drained, and treated with 2% hydrazine in DMF to
selectively deprotect the side chain amine of Dap, Dab, Orn, or
Lys (3 × 3 min, and 2 × 5 min). The resins were then washed
with DMF (4 × 30 s). For the fluorophore labeling, the mixtures
of the fluorophore in the succinimidyl ester form (2.9 mol/well,
)
342 nm, λem ) 425 nm (Ca-L); λex ) 380 nm, λem ) 460 nm
(
T-L); λex ) 600 nm, λem ) 700 nm (Ch-L) at 30 °C.
Light-Driven Conversion of Caged Probe to Active Probe.
1
.1 equiv), hydroxybenzotriazole (HOBt) (1 equiv), diisopropyl-
The uncaging setup consisted of a UV-pulsed laser source il-
luminating the sample contained in a quartz cuvette. The laser was
a diode-pumped, Q-switched UV laser that produces 5 ns pulses at
349 nm with a peak repetition rate of 1 kHz and 100 mJ pulse
energy (Explorer, Spectra-Physics, Mountain View, CA). The beam
was expanded and collimated to ∼1 cm diameter and illuminated
the side face of a 160 µL quartz cuvette through a 2 × 8 mm
window (16.160F-Q-10, Starna Cells, Atascadero, CA). The un-
caging dose was controlled by varying the number of pulses
delivered to the sample and comparing to a reference standard
[R-carboxy-o-nitrobenzyl(CMNB)-fluorescein] (Figure S-3, Sup-
porting Information). Assay of peptide 10, both prior to and
following photolysis, was performed as follows: 5 µM peptide 10
(unexposed to light or photoactivated as indicated above) was added
to wild-type yeast cell lysates (0.9 mg protein/mL lysate) and the
fluorescence change monitored in a fluorescent plate reader.
Photolysis times were 2 and 10 min at 1 kHz.
carbodiimide (DIC) (1 equiv), and DIPEA (1 equiv) in DMF (125
µL/well) were manually added to the wells. After 2 h of shaking,
1
0 µL of DIC (10% in DMF) was added to each well, sealed with
a foil, and left on the shaker overnight. A few resin beads were
microcleaved (95:5 TFA:water) to monitor the reaction progress
by MS-MALDI. Ten microliters of DIC (10% in DMF) was added
again and the reaction was left for an additional 3 h. The reaction
mixture was then drained, and resins were washed with DMF (8×)
and acetonitrile (3×) and dried under vacuum for 15 min. The
peptides were cleaved and deprotected via eight cycles of 5 min
exposure to TFA:H
well; cycles 2-7, 100 µL/well) using N
2
O:TIS in a ratio of 93:5:2 (cycle 1, 200 µL/
2
pressure to drain and
collect the peptide solution into a new 96-well plate. TFA was
evaporated, and the peptides were isolated via precipitation with
ice-cold diethyl ether and resuspended in TFA for mass spectrom-
etry analysis (LC-MS and MS-MALDI). The buffer containing
2
5% acetonitrile, 75% water, and 0.1% TFA was added to the wells,
Acknowledgment. D.S.L. (CA79954 and GM067198) and M.S.
and peptides were frozen and lyophilized.
(
GM084228) thank the NIH for financial support, Dr. D. Larson
The peptide library was dissolved in 10 µL of DMSO and 90
and Prof. J. Condeelis for use of their laser, and Prof. M.
Hochstrasser for several of the mutant yeast strains.
µL of water, and 96-well plates containing the samples at various
dilutions in assay buffer (buffer A: 50 mM Tris-HCl, 5 mM MgCl ,
2
and 0.5 mM EDTA, pH 7.5) (1/100, 1/250, and 1/1000) were
Supporting Information Available: Details of the yeast
strains, the 32-member peptide library, fluorescent and enzy-
mological analysis of sensor 5, and photoconversion of caged
prepared to monitor the changes in fluorescence (λex ) 550 nm,
λ
em ) 700 nm) after addition of 2 µL of purified 20S proteasome
(
4 mg/mL). Fluorescence change was obtained from the fluores-
cence measured before the addition of the enzyme and the
1
0 to its uncaged counterpart. This material is available free of
charge via the Internet at http://pubs.acs.org.
(
17) Carpino, L. A.; Cohen, B. J.; Stephens, K. E., Jr.; Sadat-Aalaee, Y.;
Tien, J.-H.; Langridge, D. C. J. Org. Chem. 1986, 51, 3732–34.
JA907226N
1
582 J. AM. CHEM. SOC. 9 VOL. 132, NO. 5, 2010