Proteasome Subunit Selective Activity-Based Probes Report on Proteasome Core Particle Composition in a Native Polyacrylamide Gel Electrophoresis Fluorescence-Resonance Energy Transfer Assay

Article abstract of DOI:10.1021/jacs.6b04207

Most mammalian tissues contain a single proteasome species: constitutive proteasomes. Tissues able to express, next to the constitutive proteasome catalytic activities (β1c, β2c, β5c), the three homologous activities, β1i, β2i and β5i, may contain numerous distinct proteasome particles: immunoproteasomes (composed of β1i, β2i and β5i) and mixed proteasomes containing a mix of these activities. This work describes the development of new subunit-selective activity-based probes and their use in an activity-based protein profiling assay that allows the detection of various proteasome particles. Tissue extracts are treated with subunit-specific probes bearing distinct fluorophores and subunit-specific inhibitors. The samples are resolved by native polyacrylamide gel electrophoresis, after which fluorescence-resonance energy transfer (FRET) reports on the nature of proteasomes present.

Full text of DOI:10.1021/jacs.6b04207

Subscriber access provided by RMIT University Library
Article
Proteasome subunit selective activity-based probes report on
proteasome core particle composition in a native-PAGE FRET assay
Gerjan de Bruin, Bo-Tao Xin, Bogdan I. Florea, and Herman S. Overkleeft
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b04207 • Publication Date (Web): 18 Jul 2016
Downloaded from http://pubs.acs.org on July 18, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Proteasome subunit selective activity-based probes report on pro-
teasome core particle composition in a native-PAGE FRET assay
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Gerjan de Bruin, BoꢀTao Xin, Bogdan I. Florea, Herman S. Overkleeft*
Gorlaeus Laboratories, Leiden Institute of Chemistry, Einsteinweg 55, 2333 CC Leiden, The Netherlands
ABSTRACT: Most mammalian tissues contain a single proteasome species: constitutive proteasomes. Tissues able to express, next
to the constitutive proteasome catalytic activities (β1c, β2c, β5c), the three homologous activities, β1i, β2i and β5i, may contain
numerous distinct proteasome particles: immunoproteasomes (composed of β1i, β2i and β5i) and mixed proteasomes containing a
mix of these activities. This work describes the development of new subunitꢀselective activityꢀbased probes and their use in an
activityꢀbased protein profiling assay that allows the detection of various proteasome particles. Tissue extracts are treated with
subunitꢀspecific probes bearing distinct fluorophores and subunitꢀspecific inhibitors. The samples are resolved by native polyꢀ
acrylamide gel electrophoresis after which fluorescenceꢀresonance energy transfer (FRET) reports on the nature of proteasomes
present.
INTRODUCTION
Mathematically, 33 different CP particles are possible (26/2
+1), however, β1i and β2i can only be incorporated together
with β5i.6 Taking this restriction in account, five different βꢀ
rings are possible, and thus 15 different proteasome types may
exist simultaneously in cells expressing all cCP and iCP subuꢀ
nits (Table 1).
26S proteasomes are responsible for the degradation of the
majority of cytoplasmic and nuclear proteins in eukaryotic
cells.1 Proteins destined for degradation are tagged with polyꢀ
ubiquitin chains for recognition by 19S (PA700) caps. Subseꢀ
quently and in an ATPꢀdependent process, proteasome subꢀ
strates are unfolded and funneled through the αꢀrings to the
inner side of 20S proteasome core particles (CP), where they
are degraded.
CPs are 28ꢀmer multiꢀprotein complexes consisting of four
heptameric rings: two outer αꢀrings onto which 19S caps can
dock and two inner βꢀrings in which the catalytic subunits
reside. Each βꢀring of the constitutive proteasome (cCP, Figꢀ
ure 1), constitutively expressed in all eukaryotic cells, contains
three different active subunits: β1c (caspaseꢀlike, cleaving
preferentially Cꢀterminal of acidic residues), β2c (trypsinꢀlike,
cleaving preferentially after basic residues) and β5c (chymoꢀ
trypsinꢀlike, cleaving preferentially after hydrophobic resiꢀ
dues).
Proteasomes produce oligopeptides varying in length between
3ꢀ12 amino acid residues.2 These are further processed by
aminopeptidases and in part escape to the ER lumen, where
they bind to major histocompatibility complex class I (MHCꢀI)
heterodimers for antigen presentation. Another type of proꢀ
teasomes, immunoproteasomes (iCP, Figure 1), are constituꢀ
tively expressed in bone marrow derived cells and can be
induced in other tissues by the inflammatory cytokines, interꢀ
feronꢀγ (IFNꢀγ) and tumor necrosis factorꢀα (TNFꢀα).3 iCPs
generate peptide pools containing a comparatively (with reꢀ
spect to cCPꢀproduced oligopeptide pools) higher number of
peptides prone to bind to MHCꢀI complexes.
In tissue expressing both active βꢀsubunit sets (β1c/β2c/β5c
and β1i/β2i/β5i)s iCPs and cCPs are not formed exclusively
but also mixed proteasomes (mCPs) containing cCP and iCP
catalytic subunits can be formed (Figure 1). mCPs can either
be symmetric (msCP, identical βꢀrings) or asymmetric (maCP,
different βꢀrings).4,5
Figure 1. 20S proteasome subtypes and examples of ABPꢀbased
FRET. Only βꢀrings are shown.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 8
es has not been fully exploited.12 The distances between all
Table 1. Theoretical possible proteasome subtypes.*
active site threonine residues fall well within the FRET range
(<100 Å).13 NativeꢀPAGE separation of proteasomes has proꢀ
vided important insights in proteasomal composition, assemꢀ
bly and binding characteristics.14 On native gel, proteasome
complexes separate in three bands, corresponding to doubly
capped 30S proteasomes, singly capped 26Sꢀproteasomes and
20S proteasome CPs. These complexes are revealed by either
Western blotting or inꢀgel fluorogenic substrate assays.14 We
reasoned that, in analogy to SDSꢀPAGE, it should be possible
to visualize intact proteasome complexes on nativeꢀPAGE
using ABPs. Indeed (Figure S1, lane 1ꢀ3), clear labeling of
both 26S proteasomes and 20S proteasomes was observed in
crude cell lysate using either Cy5ꢀNC001 (β1ꢀselective),
BODIPY(FL)ꢀLU112 (β2ꢀselective) or BODIPY(TMR)ꢀ
NC005 (β5ꢀselective) (See Figure S1 for excitation/emission
wavelengths and Figure 2 and S1 for structures).15 In the first
instance we investigated whether FRET signals emerge from
proteasomes exposed to combinations of these probes and next
resolved by nativeꢀPAGE. For this purpose, lysates were treatꢀ
ed with each of the three combinations of two probes simultaꢀ
neously. Clear FRET signals were observed for each combinaꢀ
tion (Figure S1, lane 4ꢀ6, Cy2ꢀCy3, Cy3ꢀCy5 and Cy2ꢀCy5
channels). However, due to the spectral overlap with both Cy2
excitation and Cy5 emission, the use of BODIPY(TMR) as
either FRET donor or acceptor proved suboptimal (Figure S1).
Hardly any background signal was observed in the samples
treated with BODIPY(FL)ꢀ and Cy5ꢀmodified probes (Figure
S1, lane 1 and 2, Cy2ꢀCy5). As well, FRET efficiency beꢀ
tween these fluorophores appeared close to 100%, indicating
near complete quenching of BODIPY(FL) fluorescence (Figꢀ
ure S1a, Cy2).
1
2
3
4
5
6
7
8
9
Subtype
cCP
β1
β2
β5
β1
β2
β5
iCP
msCP1
msCP2
msCP3
maCP1
maCP2
maCP3
maCP4
maCP5
maCP6
maCP7
maCP8
maCP9
maCP10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
*Given that β5i is required for incorporation of β1i and β2i, 15
subtypes are theoretically allowed. Grey: constitutive proteasome
subunit. White: immunoproteasome subunit. msCP: mixed symꢀ
metric core particle; maCP: mixed asymmetric core particle.
mCPs containing either β5iꢀβ1cꢀβ2c or β5iꢀβ1iꢀβ2c βꢀrings are
encountered most often and have been identified in human
liver, colon, small intestine and kidney tissues.7 mCPs produce
peptide pools distinct from both those produced by cCPs and
iCPs, thus adding to the diversity of MHCꢀI ligands and thereꢀ
by to a broad CD8+ Tꢀcell repertoire. Tumorꢀspecific antigenic
peptides7,8 as well as virally encoded antigenic peptides9 have
been identified that appear to be produced uniquely by mCPs.
A rapid and accurate assay to detect mixed proteasomes and
that would report on the nature of their composition would be
of considerable use to get insight in the contribution of these
in protein turnover and MHCꢀI antigenic peptide pool producꢀ
tion. Here, we describe a nativeꢀPAGE Fluorescence Resoꢀ
nance Energy Transfer (FRET) assay that reports on proꢀ
teasome CP composition of crude cell lysates. For this purpose
proteasomeꢀsubunit selective irreversible inhibitors were
equipped with suitable fluorophores to yield a panel of activiꢀ
tyꢀbased probes (ABPs) for FRET mediated detection of proꢀ
teasome compositions.
Given these results, we decided to develop BODIPY(FL) and
Cy5 ABPs for each subunitꢀpair (β1c/β1i, β2c/β2i and
β5c/β5i). Structures of the ABPs used in this study are shown
in Figure 2. In keeping with the tradition of naming our comꢀ
pounds, the last digit indicates which subunits/subunit pairs
are targeted (β1, β2 or β5) and ‘c’ or ‘i’ indicates respectively
cCP or iCP selectivity. BODIPY(FL)ꢀNC00116
2 and
BODIPY(FL)ꢀLU11217 4 have been described previously,
whereas Cy5ꢀLU112 3 was readily synthesized following
established procedures (see Supporting information). Cy5ꢀ
LU015 5 and BODIPY(FL)ꢀLU015 6 were used to selectively
label β5c/β5i (see Supporting information for their synthesis).
Furthermore, in order to study maCPs, ABPs selective for a
single catalytic subunit, namely BODIPY(FL)ꢀLU001c 7 (β1cꢀ
selective), Cy5ꢀLU001i
8 (β1iꢀselective), BODIPY(FL)ꢀ
LU015c 9 (β5cꢀselective) and Cy5ꢀLU035i 10 (β5iꢀselective),
were developed (see Supporting Information). The single
subunit selective ABPs are based on our previously reported
subunit selective inhibitors.15,18 The selectivity window and
concentrations required for complete labeling of the respective
subunits by ABPs 1ꢀ10 was assessed in Rajiꢀ and HEK cell
lysates (ABP 1ꢀ6) (Figure S2, Table S1). β2ꢀselective probes 3
and 4 as well as β1ꢀselective probe 2 are partially crossꢀ
reactive towards β5c and β5i at concentrations required for full
labeling. To avoid this to happen, the β5 subunits are to be
blocked previous to treatment with 2, 3 or 4 by either a β5ꢀ
selective inhibitor (NC005 13, Table S1) or by β5 probes 5 or
6 (neither of which are crossꢀreactive). β5c selective probe
BODIPY(FL)ꢀLU015c 9 partially labels both β2 subunits
(Figure S2m), which however can be prevented by preꢀ
treatment with the β2ꢀselective inhibitor, LU102 12.
RESULTS
Development of FRET donor and acceptor ABPs. FRET is
a physical process in which energy is transferred from a donor
fluorophore to an acceptor fluorophore via dipoleꢀdipole couꢀ
pling. This nonꢀradiative energy transfer depends on whether
the fluorophores are in close proximity (>100 Å); whether
there is substantial overlap between the donor emission and
acceptor excitation spectra and whether the fluorophores are
properly oriented (the dipoles of the fluorophores should be
approximately parallel).10 FRET has been widely used to study
proteinꢀprotein interactions and conformational changes,11 but
its potential to determine the composition of protein complexꢀ
ACS Paragon Plus Environment

Page 3 of 8
Journal of the American Chemical Society
entirely (Figure 3b), with concomitant return of fluorescence
1
2
3
4
5
6
7
8
of the donor ABPs. This result confirms the occurrence of
intraꢀproteasomal FRET and the suitability of ABPs 1ꢀ6 for
nativeꢀPAGE FRET analysis of proteasome compositions.
Remarkably, mutual differences in fluorescence intensity on
nativeꢀPAGE between samples treated with a single ABP were
observed, while on SDSꢀPAGE the intensities are similar
(Figure 3a,c; compare lanes 1ꢀ6 in both gels). For instance,
BODIPY(FL)ꢀNC001 2 shows the highest fluorescent signal
with the intensities for BODIPY(FL)ꢀLU112
4
and
9
BODIPY(FL)ꢀLU015 6 being respectively 1.5 and 4.0 times
lower (Figure 3a; compare lane 2, 4 and 6). However, after gel
fixation the fluorescence intensity for the three probes became
almost equal.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
These differences may be caused by either selfꢀquenching19 or
homoꢀFRET,20 processes that can take place when fluoroꢀ
phores have sufficient overlap in their excitation and emission
spectra. Since proteasomes encompass two copies of each
subunit (pair), two fluorophores are brought in close proximiꢀ
ty, which may allow selfꢀquenching or homoꢀFRET processes
to occur. Due to differences in the mutual orientation ofꢀ and
distances between the two fluorophores, the efficiency of selfꢀ
quenching of the ABPs may vary, resulting in different fluoꢀ
rescence intensities.
NativeꢀPAGE FRET allows the detection of mixed proꢀ
teasomes. In order to detect mixed proteasomes by nativeꢀ
PAGE FRET, each donor or acceptor ABP should bind to a
single cCP or iCP subunit. The presence of mixed proteasome
core particles in a given sample is revealed by FRET when, for
instance, a donor ABP is bound to a cCP subunit and an acꢀ
ceptor ABP to an iCP subunit assembled in the same CP.
Selective binding of ABPs 1ꢀ6 to a single subunit can be atꢀ
tained by making use of our recently published panel of subuꢀ
nitꢀselective inhibitors (Table S1).15
Figure 2. Structures of FRET acceptor (Cy5) and (BODIPY(FL))
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
activityꢀbased probes used in this study.
Evaluation of ABPs 1ꢀ6 as nativeꢀPAGE FRET proteasome
probes. From the pool of ABPs 1ꢀ6, six FRET donor/acceptor
pairs can be assembled. We evaluated all these pairs in Raji
and HEKꢀ293 lysates on their behavior as FRET couples in a
nativeꢀPAGE fluorescence readout setting. In the first step,
both β5 subunits were either inhibited with NC005 13 (in case
FRET signals emerging from ABP labeling of β1 and β2 were
sought for) or labeled with ABPs 6 or 7 (β1ꢀβ5 or β2ꢀβ5 labelꢀ
ing) for 1 hour. Subsequently, β1 and/or β2 targeting probes 1,
2, 3 and/or 4 were added and the samples were again incubatꢀ
ed for 1 hour. One half of each sample was resolved by nativeꢀ
PAGE (Figure 3a) and the other half by SDSꢀPAGE (Figure
3c). Clear FRET signals and near complete quenching of
FRET donor ABPs were observed for each FRET pair. Folꢀ
lowing quantification of the fluorescence bands of the acceptor
ABPs (Cy2 channel), the FRET efficiencies (E) were calculatꢀ
ed, and high FRET efficiencies for each FRET pair (E> 0.8,
Table S2) were revealed. Swapping the FRET donor
(BODIPYꢀFL) and acceptor (Cy5) on the subunitꢀselective
ABPs did not result in significant differences in FRET effiꢀ
ciency. In order to verify whether true intraꢀproteasomal
FRET signals are observed, the nativeꢀPAGE slab was transꢀ
ferred to a fixing solution (5:4:1 H2O/MeOH/AcOH) and
heated in a microwave oven. This process results in denaturaꢀ
tion of the proteins, and separation of the fluorophores. Inꢀ
deed, after fixation, FRET signals have disappeared almost
Figure 3. Evaluation of six FRET donor/acceptor pairs in HEKꢀ
293 and Raji lysates. A. NativeꢀPAGE analysis. B. NativeꢀPAGE
analysis after gel fixation. C. SDSꢀPAGE analysis to verify corꢀ
rect labelling. Gels were imaged using Cy2, Cy5 or Cy2ꢀCy5
settings, see Figure S1. See Table S2 for calculated FRET effiꢀ
ciencies.
ACS Paragon Plus Environment

Journal of the American Chemical Society
With the exception of LUꢀ002c (targeting β2c) the selectivity
Page 4 of 8
1
2
3
4
5
6
7
8
windows for all inhibitors are sufficiently large to allow selecꢀ
tive and complete blocking of their target subunits. With the
panel of five inhibitors selective for β1c, β1i, β2i, β5c, or β5i
(Table S1), eight combinations of two inhibitors can be made.
With these and together with ABPs 1ꢀ6 eight different proꢀ
teasome subunit combinations can in theory be detected. Each
possible inhibitor combination was assessed in Raji cell lyꢀ
sates using two FRET ABP pairs (Figure 4). Since both interꢀ
and intraꢀβꢀring FRET can take place, the observed FRET
signal is a sum of several possible FRET pathways that
emerge from, either, two, three or four ABPs present in a
proteasome particle. Interestingly, however, clear FRET sigꢀ
nals were observed for each combination, which implies that,
next to cCP, iCP also mCPs are present. The various ABP
couples yield FRET signals of similar intensities, but subtle
differences are observed (see also Table 2).
The high β2c over β2i ratio (see Table 2) is reflected in the
relatively high FRET intensities emerging from ABPs bound
to β2cꢀβ1c/β5c compared to β2cꢀβ1i/β5i, indicating that β2c is
preferentially incorporated together with β1c and β5c. The low
FRET signal between β2cꢀβ1i reflects the preferential incorpoꢀ
ration of β2i together with β1i. The FRET intensities for β1cꢀ
β5c and β1iꢀβ5i are higher than those for β1iꢀβ5c and β1cꢀβ5i,
indicating preferential formation of β1cꢀβ5c and β1iꢀβ5i conꢀ
taining βꢀrings.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Asymmetric mixed proteasomes (maCPs) can be detected
using nativeꢀPAGE FRET. When applied at appropriate
concentrations, ABPs 7, 8, 9 and 10 selectively and completeꢀ
ly block a single proteasome subunit, namely β1c, β1i, β5c and
β5i respectively. Inclusion of these compounds in the nativeꢀ
PAGE FRET experiments allows labeling of for instance β1c
with a FRET donor and β1i with a FRET acceptor ABP. FRET
signals emerging from samples treated in this way and reꢀ
solved on nativeꢀPAGE, can only be caused by maCPs. Clear
FRET signals were observed for both FRET pairs 7/8 and
9/10, confirming the presence of one or more maCPs (see
Figure 5). Although relative amounts of maCPs could not be
established due to lack of appropriate reference samples, this
method demonstrates the presence of maCPs in a given sample
in an unambiguous fashion.
Figure 4. Detection of mixed proteasomes in Raji cell lysates.
Samples were preꢀincubated with indicated inhibitors, followed
by labelling of residual proteasome activity by indicated ABPs.
The samples were analyzed by nativeꢀPAGE and SDSꢀPAGE.
As well, substantial β5iꢀβ5c and β1iꢀβ1c FRET signals were
observed (Figure 5), indicating the presence of maCPs. Most
likely, the majority of these maCPs contain, a single β5i
(maCP1), or both one β5i and one β1i in the same βꢀring
(maCP2), as witnessed by the observed FRET signal between
β5c and β1i. After exposure to IFNꢀγ, a strong increase in β5iꢀ
β1i FRET signal is observed, while β5iꢀβ1c, β5cꢀβ1i and β5cꢀ
β1c combinations are slightly decreased. This decrease does
not necessarily reflect a decrease of the absolute amounts of
these subunitꢀpairs, but is probably caused by the increase in
expression and incorporation into proteasome particles of both
β5iꢀsubunits and β1iꢀsubunits, with an increased total FRET
intensity as the result. Given the long halfꢀlife of cCPs (up to 5
days),21 IFNꢀγ induced expression of iCP subunits presumably
leads to a net increase of the total proteasome amount. When
for convenience it is assumed that limited proteasome degraꢀ
dation takes place during the 24 h exposure with IFNꢀγ, the
slight decrease in β5iꢀβ1c and β5cꢀβ1iꢀderived FRET signals
indicates that no new proteasomes containing these subunitꢀ
pairs are formed, that newly expressed β5i is mainly incorpoꢀ
rated together with β1i and that newly formed CPs are symꢀ
metric with respect to their β1/β5 subunit composition. Interꢀ
estingly, the β2cꢀβ5i FRET signals significantly increase (Figꢀ
ure 6) and their intensities become closer to the β2cꢀβ5c FRET
signals. The same applies to β1iꢀβ2c, indicating that β5i is to
some extend incorporated with β2c, which is also reflected by
the lower net increased amount of β2i (+37%) compared to β1i
and β5i (both +44%).
Assessment of proteasome composition after induction of
iCPs by IFNꢀγ. The expression of iCP subunits can be inꢀ
duced by exposure to the inflammatory cytokine interferonꢀγ
(IFNꢀγ). Dahlmann and coꢀworkers reported that HeLa cells
exposed to IFNꢀγ express both msCPs and maCPs.5 These
observations were reꢀevaluated using the aboveꢀdescribed
nativeꢀPAGE FRET assay. For this, HeLa cells were either
exposed to IFNꢀγ for 24 h or left untreated. Next, both samples
were subjected to various inhibitor/ABP combinations and
evaluated by nativeꢀPAGE FRET as described above. When
not exposed to IFNꢀγ, HeLa cells express only small amounts
of β1i and β5i while β2i could not be detected. Following
exposure to IFNꢀγ, a substantial increase of the amount of all
iCP subunits was found (Figure 6, Table 2). As expected, nonꢀ
exposed HeLa cells show high interꢀcCP subunit FRET sigꢀ
nals. Since the FRET signals emerging from β5iꢀβ1c are highꢀ
er than those observed for β5iꢀβ1i, it is likely that the majority
of the β5i subunits in these nonꢀexposed cells are present in
proteasomes that also contain at least one β1c subunit.
ACS Paragon Plus Environment

Page 5 of 8
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Figure 5. Asymmetric mixed proteasomes in Raji lysates and in
HeLa cell lysates (before and after exposure to IFNꢀγ for 24 h).
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Samples were incubated with a β2ꢀselective inhibitor (LU102, in
case of ABPs 9, 10) or a β5ꢀselective inhibitor (NCꢀ005, in case
of ABPs 7, 8) followed by treatment with ABPs. *Ctrl: control.
Control samples contained twice the amount of protein compared
to normal samples were incubated with LUꢀ102 or NCꢀ005 as for
normal samples, followed by incubation with one of the ABPs.
Next, remaining proteasome activities were blocked by a mixture
of NC001, LU102 and NC005 and the samples treated with ABP
7/8 and 9/10.
Figure 6. Mixed proteasomes in HeLa cell lysates, with or
without exposure to IFNꢀγ for 24 h. Samples were preꢀincubated
with indicated inhibitors, followed by labelling of residual proꢀ
teasome activity by indicated ABPs. SDSꢀPAGE analysis shows
the relative amounts of iCP and cCP subunit before and after
exposure to IFNꢀγ. To label β5c or β5i specifically, either β5i was
blocked by LUꢀ035i or β5c by LUꢀ005c.
Altogether it can be concluded that, following exposure to
IFNꢀγ, HeLa cells predominantly produce two distinct proꢀ
teasome types: mCPs featuring βꢀrings composed of either
β1iꢀβ2iꢀβ5i or β1iꢀβ2cꢀβ5i. This observation is further conꢀ
firmed by the lower relative amount of maCPs found after
exposure to IFNꢀγ (lower signal/noise ratio compared to no
IFNꢀγ, Figure 5), indicating that the newly formed proteasome
are symmetric with respect to their β1 and β5 subunit compoꢀ
sition.
As β5i is required for incorporation of β1i, β5c cannot be
incorporated in the same βꢀring together with β1i, and thereꢀ
fore this result shows that either maCP2 or maCP4, or both, are
present. Moreover, using selective β1cꢀβ1i and β5cꢀβ5i targetꢀ
ing FRET donorꢀacceptor pairs, maCPs were visualized that
are asymmetric in their β1 and β5 subunit composition
(maCP1ꢀ4: asymmetric in β5 composition; maCP2,4,5,7,8,10
asymmetric in β1 composition) (see Figure 5).
:
DISCUSSION
Altogether, these results reveal a complex mixture of proꢀ
teasome subtypes to exist in Raji cells. In Raji cells, iCPꢀiCP
and cCPꢀcCP subunitꢀpairs show higher relative FRET intensiꢀ
ties compared to iCPꢀcCP subunitꢀpairs for β1ꢀβ5, indicating
preferential formation of β1iꢀβ5i and β1cꢀβ5c containing βꢀ
rings. Compared to Raji cells, IFNꢀγ exposed HeLa cells do
show much lower relative FRET intensities of β1ꢀβ5 iCPꢀcCP
subunitꢀpairs, and the relative FRET intensities indicate prefꢀ
erential formation of proteasomes containing βꢀrings comꢀ
posed of β5iꢀβ1iꢀβ2i and β5iꢀβ1iꢀβ2c. Interestingly, both Rajiꢀ
and IFNꢀγ exposed HeLa cells express similar amounts of all
subunits, indicating that more mCPs are formed when all
subunits are constitutively expressed compared to induction of
iCP subunits in otherwise low iCP expressing cells.
Dahlmann and coꢀworkers identified CPs asymmetric in their
β1 subunit composition in IFNꢀγ exposed HeLa cells, while
the nonꢀexposed HeLa in their hands did not express detectaꢀ
ble amount of immunoproteasome subꢀunits.5 However, in this
study it was found that after IFNꢀγ exposure HeLa cells preꢀ
dominantly express proteasomeꢀcontaining βꢀrings symmetric
in their β1 subunit composition.
We have reported here an inꢀdepth analysis on the use of
ABPs to determine the composition of large protein complexes
using a nativeꢀPAGE FRET assay. Proteasome subunitꢀpair
selective ABPs equipped with suitable FRET donor and accepꢀ
tor fluorophores were selected, which target β1c/β1i, β2c/β2i
or β5c/β5i. Crude cell extracts were treated with combinations
of these FRET donor/acceptor ABPs and resolved on nativeꢀ
PAGE, after which FRET signals can be measured by fluoresꢀ
cence imaging of the gel. In HEKꢀ293 cells, expressing excluꢀ
sively cCPs, all FRET pairs gave clear FRET signals with high
FRET efficiencies (E> 0.8), confirming that β1c; β2c and β5c
are present in stoichiometric amounts.
Though our assay does not enable the detection of each specifꢀ
ic (mixed) proteasome particle, many of the possible particles
in fact show up in our Native PAGE FRET assays. In lysates
of Raji cells, which express all six proteasome subunits, FRET
signals were observed for iCPꢀiCP and cCPꢀcCP subunitꢀpairs.
Importantly, also FRET signals of all possible iCPꢀcCP subuꢀ
nitꢀpairs were detected, demonstrating the presence of mCPs.
Remarkably, a substantial FRET signal was observed for β1iꢀ
β5c (Figure 4, Table 2).
ACS Paragon Plus Environment

Journal of the American Chemical Society
Table 2. Relative FRET intensities in Raji and HeLa lysates.
Page 6 of 8
1
2
3
4
5
Subunits
Relative FRET intensity (% of FRET no inhibitor)*
HeLa Raji HeLa
Ratio subunits**
Raji
ꢀIFNꢀγ
+IFNꢀγ
ꢀIFNꢀγ
+IFNꢀγ
Raji
6
7
8
9
Cy5ꢀNC001 – BDPꢀLU112 (1ꢀ4)
BDPꢀNC001 – Cy5ꢀLU112 (2ꢀ3)
β1c/β1i= 1.24 ± 0.03 (55/45)
β2c/β2i= 2.28 ± 0.07 (70/30)
β5c/β5i= 1.04 ± 0.03 (51/49)
Hela ꢀIFNꢀγ
ABPs
β2c
53
18
89
22
57
20
50
17
102
20
58
19
β1c
β1i
β2c
ABPs
β5i
Cy5ꢀNC001 – BDPꢀLU015 (1ꢀ6)
BDPꢀNC001 – Cy5ꢀLU015 (2ꢀ5)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
26
28
31
23
29
14
54
17
21
41
33
7
23
35
44
20
32
13
67
15
22
40
40
10
β1c/β1i= 5.52 ± 0.04 (85/15)
β2c/β2i= 1.00 ± 0.00 (100/0)
β5c/β5i= 2.47 ± 0.12 (72/28)
Hela +IFNꢀγ
β1c
β1i
β1c
β1i
β5i
β5c
β5c
ABPs
β5c
Cy5ꢀLU112 – BDPꢀLU015 (3ꢀ6)
BDPꢀLU112 – Cy5ꢀLU015 (3ꢀ5)
β1c/β1i= 1.44 ± 0.07 (85/59)
β2c/β2i= 2.72 ± 0.01 (100/37)
β5c/β5i= 0.98 ± 0.04 (72/72)
39
27
72
30
36
30
50
29
74
19
47
31
β2c
β2c
β5i
*Determined by quantification of FRET signals from Figure 4 (Raji) or Figure 5 (HeLa). ** Determined by quantification of SDSꢀ
PAGE
Dahlmann and coꢀworkers used β1cꢀZZ transfected HeLa cells
to allow specific precipitation of β1ꢀZZ by binding to IgG and
subsequent analysis of subunit composition of precipitated
proteasomes. This may result in higher β1cꢀZZ than normal
β1c expression, causing higher incorporation of β1cꢀZZ in
newly formed proteasomes resulting in asymmetric proꢀ
teasome formation.
cells and no selective inhibitors were applied to inhibit β5i and
β1c.
In conclusion, the nativeꢀPAGE FRET assay we report here
adds to existing methods for the assessment of proteasome
core particle composition. The method provides semiꢀ
quantitative insights in the abundance of ten different proꢀ
teasome subunitꢀpairs and can do so in any crude cell extract.
Current methods to identify proteasome CP composition are
based on either chromatographic separation of proteasome
subtypes,5,22 isoelectric focusing electroꢀphoresis,23 or antiꢀ
body mediated depletion of a βꢀsubunit,7 followed by determiꢀ
nation of CP composition by either immunostaining, substrate
hydrolysis assays or mass spectrometry analysis of purified
proteasomes.24 Compared to these methods, the method deꢀ
scribed here has several advantages. For instance, van den
Eynde and coꢀworkers used subunit depletion and subsequent
immunoblotting to determine proteasome composition.7 Alterꢀ
natively, they calculated the quantity of mCPs based on the
assumption that β5i can be incorporated as the only iCP subuꢀ
nit or together with β1i, but that β2i is always incorporated
together with β1i and β5i. However, in these approaches
asymmetric proteasomes are not taken into account and thereꢀ
fore several proteasome subtypes are possibly overlooked. Our
FRETꢀbased approach, besides being more sensitive, is also
much faster, straightforward and less time consuming than the
methods relying on chromatographic separation of proꢀ
teasomes.
The measurement of FRET signals inꢀnative PAGE also repreꢀ
sents, in our opinion, a major improvement to the methodoloꢀ
gy developed by Kim and coꢀworkers.12 In this method, FRET
signals were measured in a plate reader, which required reꢀ
moval of unbound probes by filtration. Moreover, whereas
β5cꢀβ1i FRET signals were detected, the detection of other
subunitꢀpairs was hampered by the lack of truly subunit selecꢀ
tive probes and inhibitors. We feel that their method to prove
β5cꢀβ1i containing proteasomes may be compromised, since a
β5i selective probe (LKSCy5)25 was previously claimed to
selectively label β5c. As well, a β1c/β1i targeting probe
(UKPCy3) was used to selectively label β1i in RPMIꢀ8226
ASSOCIATED CONTENT
Supporting Information.
Figures, Tables, biochemical procedures, synthetic procedures,
1Hꢀ and 13C NMR spectra and LCꢀMS data. This material is availꢀ
able free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*h.s.overkleeft@chem.leidenuniv.nl
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
We thank the Netherlands Organization for Scientific Research
(NWOꢀCW, TOPꢀPUNT Grant to HSO), the China Scholarship
Council (CSC, PhD fellowship to B.ꢀT. X) and the European
Research Council (ErCꢀ2011ꢀAdGꢀ290836, to HSO) for financial
support.
REFERENCES
(1) Hershko, A.; Ciechanover, A. Annu. Rev. Biochem. 1998, 67,
425.
(2) Kisselev, A. F.; Akopian, T. N.; Woo, K. M.; Goldberg, A. L.
J. Biol. Chem. 1999, 274, 3363.
(3) Ferrington, D. A.; Gregerson, D. S. Prog. Mol. Biol. Transl.
Sci. 2012, 109, 75.
(4) Dahlmann, B.; Ruppert, T.; Kuehn, L.; Merforth, S.; Kloetzel,
P.ꢀM. J. Mol. Biol. 2000, 303, 643.
(5) Klare, N.; Seeger, M.; Janek, K.; Jungblut, P. R.; Dahlmann, B.
J. Mol. Biol. 2007, 373, 1.
ACS Paragon Plus Environment

Page 7 of 8
Journal of the American Chemical Society
(6) Griffin, T. A.; Nandi, D.; Cruz, M.; Fehling, H. J.; Kaer, L. V.;
Monaco, J. J.; Colbert, R. A. J. Exp. Med. 1998, 187, 97.
(7) Guillaume, B.; Chapiro, J.; Stroobant, V.; Colau, D.; Van
Holle, B.; Parvizi, G.; BousquetꢀDubouch, M.ꢀP.; Théate, I.;
Parmentier, N.; Van den Eynde, B. J. Proc. Natl. Acad. Sci. 2010,
107, 18599.
(8) Guillaume, B.; Stroobant, V.; BousquetꢀDubouch, M.ꢀP.;
Colau, D.; Chapiro, J.; Parmentier, N.; Dalet, A.; Van den Eynde, B.
J. J. Immunol. 2012, 189, 3538.
(9) Zanker, D.; Waithman, J.; Yewdell, J. W.; Chen, W. J.
Immunol. 2013, 191, 52.
(10)ChirioꢀLebrun, M.ꢀC.; Prats, M. Biochem. Educ. 1998, 26,
320.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(11)Piston, D. W.; Kremers, G.ꢀJ. Trends Biochem. Sci., 32, 407.
(12)Park, J. E.; Wu, Y.; Carmony, K. C.; Miller, Z.; Sharma, L.
K.; Lee, D.ꢀM.; Kim, D.ꢀY.; Lee, W.; Kim, K.ꢀB. Mol. BioSyst. 2014,
10, 196.
(13)Borissenko, L.; Groll, M. Chemical Reviews 2007, 107, 687.
(14)Elsasser, S.; Schmidt, M.; Finley, D. In Methods Enzymol.;
Academic Press: 2005; Vol. Volume 398, p 353.
(15)de Bruin, G.; Xin, B. T.; Kraus, M.; van der Stelt, M.; van der
Marel, G. A.; Kisselev, A. F.; Driessen, C.; Florea, B. I.; Overkleeft,
H. S. Angew. Chem. Int. Ed. 2016, 55, 4199.
(16)Verdoes, M.; Willems, L. I.; van der Linden, W. A.;
Duivenvoorden, B. A.; van der Marel, G. A.; Florea, B. I.; Kisselev,
A. F.; Overkleeft, H. S. Org. Biomol. Chem. 2010, 8, 2719.
(17)Geurink, P. P.; van der Linden, W. A.; Mirabella, A. C.;
Gallastegui, N.; de Bruin, G.; Blom, A. E.; Voges, M. J.; Mock, E. D.;
Florea, B. I.; van der Marel, G. A.; Driessen, C.; van der Stelt, M.;
Groll, M.; Overkleeft, H. S.; Kisselev, A. F. J. Med Chem. 2013, 56,
1262.
(18)de Bruin, G.; Huber, E. M.; Xin, B.ꢀT.; van Rooden, E. J.; Alꢀ
Ayed, K.; Kim, K.ꢀB.; Kisselev, A. F.; Driessen, C.; van der Stelt, M.;
van der Marel, G. A.; Groll, M.; Overkleeft, H. S. J. Med. Chem
2014, 57, 6197.
(19)Meer, B. W. v. d.; III, G. C.; Chen, S.ꢀY. S. VCH Publishers,
New York 1994, 5.
(20)Bader, A. N.; Hoetzl, S.; Hofman, E. G.; Voortman, J.; van
Bergen en Henegouwen, P. M. P.; van Meer, G.; Gerritsen, H. C.
ChemPhysChem 2011, 12, 475.
(21)Heink, S.; Ludwig, D.; Kloetzel, P.ꢀM.; Krüger, E. Proc. Natl.
Acad. Sci. 2005, 102, 9241.
(22)Dahlmann, B.; Ruppert, T.; Kloetzel, P. M.; Kuehn, L.
Biochimie 2001, 83, 295.
(23)Drews, O.; Wildgruber, R.; Zong, C.; Sukop, U.; Nissum, M.;
Weber, G.; Gomes, A. V.; Ping, P. Mol. Cell. Proteomics 2007, 6,
2021.
(24)Pelletier, S.; Schuurman, K. G.; Berkers, C. R.; Ovaa, H.;
Heck, A. J. R.; Raijmakers, R. Molecular BioSystems 2010, 6, 1450.
(25)Sharma, L. K.; Lee, N.ꢀR.; Jang, E. R.; Lei, B.; Zhan, C.ꢀG.;
Lee, W.; Kim, K.ꢀB. ChemBioChem 2012, 13, 1899.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 8
TOC file
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
8
ACS Paragon Plus Environment