A Photoactivatable Probe for Super-Resolution Imaging of Enzymatic Activity in Live Cells

Article abstract of DOI:10.1021/jacs.7b07748

A dual-Activatable, fluorogenic probe was developed to sense esterase activity with single-molecule resolution. Without enzymatic pre-Activation, the diazoindanone-based probe has an electron-poor core and, upon irradiation, undergoes Wolff rearrangement to give a ring-expanded xanthene core that is nonemissive. If the probe is pre-Activated by carboxylesterases, the tricyclic core becomes electron-rich, and the photoinduced Wolff rearrangement produces a highly emissive rhodol dye. Live-cell and solution studies confirmed the selectivity of the probe and revealed that the photoactivated dye does not diffuse away from the original location of activation because the intermediate ketene forms a covalent bond with surrounding macromolecules. Single-molecule localization microscopy was used to reconstruct a super-resolved image of esterase activity. These single-molecule images of enzymatic activity changed significantly upon treatment of the cells with inhibitors of human carboxylesterase I and II, both in terms of total number of signals and intracellular distribution. This proof-of-principle study introduces a sensing mechanism for single-molecule detection of enzymatic activity that could be applied to many other biologically relevant targets.

Full text of DOI:10.1021/jacs.7b07748

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES
Article
A Photoactivatable Probe for Super-Resolution
Imaging of Enzymatic Activity in Live Cells
Elias A. Halabi, Zacharias Thiel, Nils Trapp, Dorothea Pinotsi, and Pablo Rivera-Fuentes
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b07748 • Publication Date (Web): 18 Aug 2017
Downloaded from http://pubs.acs.org on August 18, 2017
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 17
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
A Photoactivatable Probe for Super-Resolution Imaging of
Enzymatic Activity in Live Cells
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
Elias A. Halabi,†,‡ Zacharias Thiel,†,‡ Nils Trapp,† Dorothea Pinotsi,§ and Pablo RiveraꢀFuentes*,†
†Laboratory of Organic Chemistry, VladimirꢀPrelogꢀWeg 3, and §Scientific Center for Optical and Electron Microscopy,
OttoꢀSternꢀWeg 3, ETH Zurich, CHꢀ8093, Zurich, Switzerland.
ABSTRACT: A dualꢀactivatable, fluorogenic probe was developed to sense esterase activity with singleꢀmolecule resolution.
Without enzymatic preꢀactivation, the diazoindanoneꢀbased probe has an electron poor core and, upon irradiation, undergoes Wolff
rearrangement to give a ringꢀexpanded xanthene core that is nonꢀemissive. If the probe is preꢀactivated by carboxylesterases, the
tricyclic core becomes electronꢀrich and the photoinduced Wolff rearrangement produces a highly emissive rhodol dye. Liveꢀcell
and solution studies confirmed the selectivity of the probe and revealed that the photoactivated dye does not diffuse away from the
original location of activation because the intermediate ketene forms a covalent bond with surrounding macromolecules. Singleꢀ
molecule localization microscopy was used to reconstruct a superꢀresolved image of esterase activity. These singleꢀmolecule imagꢀ
es of enzymatic activity changed significantly upon treatment of the cells with inhibitors of human carboxylesterase I and II, both in
terms of total number of signals and intracellular distribution. This proofꢀofꢀprinciple study introduces a sensing mechanism for
singleꢀmolecule detection of enzymatic activity that could be applied to many other biologically relevant targets.
INTRODUCTION
chemical conversion. Herein, we address this challenge by
developing a dual enzymeꢀ and photoꢀactivatable probe that
provides superꢀresolved subcellular imaging of the activity of
esterases, which are important enzymes involved in detoxificaꢀ
tion of xenobiotics and proꢀdrug activation.16 We further
demonstrate that besides producing superꢀresolved images,
singleꢀmolecule localization data collected with this probe can
be used to obtain dynamic information and quantify the subꢀ
cellular localization of microdomains of enzymatic activity.
Light microscopy has revolutionized the study of living sysꢀ
tems but its classical resolution is limited by diffraction of
light. This limit is proportional to the wavelength of light diꢀ
vided by twice the numerical aperture of the objective (reꢀ
solved distance >200 nm).1 Several strategies have been deꢀ
veloped to break this barrier and their implementation
achieves superꢀresolution imaging of nanometric cellular
structures.2 Stochastic optical reconstruction microscopy
(STORM)3 and photoactivated localization microscopy
(PALM)4 are closely related methods that achieve singleꢀ
molecule resolution by sequential activation, readout and deꢀ
activation of photochemically active fluorophores. Many phoꢀ
toactive fluorescent proteins5 and smallꢀmolecule dyes6 have
been developed for STORM and PALM. These emitters have
facilitated the localization and tracking of macromolecules in
live specimens,7,8 enhancing the resolution of optical microsꢀ
copy well beyond the limit of diffraction. Despite these adꢀ
vances, sensing and quantification of biochemical activity with
nanometer resolution is a challenge that remains largely unmet
because current photoactivatable fluorophores are not responꢀ
sive to specific enzymes.
RESULTS AND DISCUSSION
Design and synthesis. Belov, Wurm, Hell, and coꢀworkers
reported that xanthene dyes can be locked in a dark isomer
upon conversion of the lactone into diazoindanones.17 These
molecules undergo Wolff rearrangement upon photolysis,
which restores the conjugation in the xanthene core (Scheme
1A). The diazoindanone photoremovable group is a small,
nonꢀpolar modification well suited for liveꢀcell superꢀ
resolution imaging.18 More recently, Lavis and coꢀworkers
observed that diazoindanone derivatives of silicon rhodamines
undergo Wolff rearrangement to give primarily a ring expanꢀ
sion product that is only very weakly fluorescent.19 Silicon
xanthenes are more electrophilic than oxygenꢀsubstituted xanꢀ
thenes,20 making them more prone to attack by the putative
carbene intermediate of the Wolff rearrangement to give a
dark, sevenꢀmembered ring photoproduct (Scheme 1A). We
envisioned that this differential reactivity could be harnessed
as a sensing mechanism. As proofꢀofꢀprinciple, we designed
the electronꢀpoor probe 1, which should photoisomerize to a
dark, sevenꢀmembered ring photoproduct (Scheme 1B). Upon
removal of the acetyl group by carboxylesterases, compound 2
is obtained, which is electronꢀrich and should photoisomerize
to give a bright rhodol emitter (Scheme 1B). Such probe
would effectively be a fluorescent probe for endogenous esterꢀ
Fluorescent sensors are molecules that respond to a stimulus
by exhibiting a measurable change in their photophysical
properties.9 A variety of proteinꢀbased and smallꢀmolecule
sensors have been developed to detect metal ions,10,11 small
signaling molecules,12 pH,13 and redox states,14 among many
other stimuli. These probes have been applied in diffractionꢀ
limited microscopy with great success, but sensors that are
specifically designed for application in superꢀresolution miꢀ
croscopy have only started to appear very recently.15 To the
best of our knowledge, there are no smallꢀmolecule sensors
that can be directly applied to STORM/PALM imaging, beꢀ
cause these techniques require photoactivatable fluorophores
and the sensing mechanism must be coupled to the photoꢀ
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 17
ase activity and could be employed in singleꢀmolecule localiꢀ tions. Whereas a significant increase in fluorescence was deꢀ
1
2
3
4
5
6
zation imaging, as we show below.
tected in the photolysis of 2, only a small change was measꢀ
ured for compound 1, indicating that acetylation of the phenolꢀ
ic group suppresses the formation of emissive species
Scheme 1. A) Transformations of oxygen and silicon rho-
damines protected with a diazoindanone photoremovable
group. B) Proposed mechanism of enzyme sensing and
photoactivation using a substituted, rhodol-based di-
azoindanone.
7
8
A
R
N
R
N
R
N
R
N
O
Si
O
O
9
R
R
R
R
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
h
ν
N+ N–
R = alkyl
COO–
O
R
N
R
N
R
N
R
N
R
Si
R
R
R
h
ν
N+ N–
R = alkyl
O
O
B
N
O
O
N
OH
N+ N–
esterase
O
N+ N–
O
O
1
2
h
ν
hν
dark photoproduct
fluorescent photoproduct
Compound 1 was synthesized in two steps (Scheme 2) starting
from rhodol 3.21 Treatment of this compound with acetic anꢀ
hydride and pyridine gave the acetylated rhodol 4. The diꢀ
azoindanone group was formed by conversion of the lactone
into the acid chloride and treatment with (trimethylsiꢀ
lyl)diazomethane. Deprotection of 1 was achieved with NaOH
to yield deacetylated product 2.
Scheme 2.
Synthesis of compounds
DIPEA = diisopropylethylamine; TMS = trimethylsilyl.
1
and 2.
Figure 1. A) Fluorescence increase upon irradiation (~300 nm) of
solutions of 1 and 2 (0.25 ꢀM, phosphateꢀbuffered saline, pH =
7.4, in quartz cuvettes). Three independent measurements are
depicted. B) Main photoreactions of compounds 1 and 2, and
structure of compound 8. C) Xꢀray structure of photoproduct 5
(CCDC: 1564214). D) Xꢀray structure of photoproduct 6 (CCDC:
1564215).
Photochemistry and photophysics. The photoinduced Wolff
rearrangements of 1 and 2 were investigated in buffered soluꢀ
tions (0.25 ꢀM, phosphateꢀbuffered saline, pH = 7.4). The
samples were irradiated with UV light (~300 nm, for details
see Materials and Methods) and fluorescence emission was
recorded at defined time intervals (Figure 1A). The increase in
fluorescence signal plateaued after 25 min for both photoreacꢀ
. Isolation of the major photoproducts (see the SI) revealed
that under these conditions, photolysis of compound 1 yielded
predominantly product 5, whereas compound 2 gave mostly
compound 6 (Figure 1B), in which the product of the Wolff
ACS Paragon Plus Environment

Page 3 of 17
Journal of the American Chemical Society
rearrangement underwent further decarboxylation. The identiꢀ Colocalization analysis with various organelle markers reꢀ
1
2
3
4
fication of these compounds was established by spectroscopic
methods and singleꢀcrystal Xꢀray diffraction (Figures 1C and
vealed that compound 1 is activated most efficiently in the
endoplasmic reticulum (ER), with some fluorescence being
detected also in mitochondria and the Golgi apparatus (for
details see Material and Methods, Figure S18 and Table S2).
1D). Product 5 displays an absorption maximum (λabs) at 565
nm with an extinction coefficient (ε) of 3.8 × 103 M–1 cm–1,
and exhibits essentially no fluorescence (Figure S7). In conꢀ
trast, compound 6 displays λabs = 541 nm, ε = 2.9 × 104 M–1
cm–1 and an emission quantum yield (φem) of 36% (Figure S8).
A few other minor photoproducts were isolated from both
reactions and some of them characterized (see the SI), but the
only considerably emissive photoproducts were isolated from
the reaction of compound 2. This distribution of photoproducts
explains the large increase in fluorescence intensity upon irraꢀ
diation of 2 but not 1.Quantum yields of photoactivation were
estimated for the disappearance of starting materials 1 and 2
and for the appearance of some photoproducts (Figure S1).
Upon irradiation with a 405 nm LED both compound 1 and 2
were photolyzed with small quantum yields (φPA = 0.3% and
φPA = 0.2%, respectively). Under these conditions, the main
photoproduct of 1 was compound 5 (φPA = 0.05%) and the
main product of 2 was 7 (φPA = 0.16%). Decarboxylation of
compound 7 to give 6 did not occur at 405 nm (Figure 1B).
Preparative amounts of 7 could not be obtained and therefore a
stable, methyl ester derivative thereof (compound 8, Figure
1B) was prepared independently (see Scheme S3) to estimate
the photophysical properties of 7. The quantum yields and
photophysical parameters of all main photoproducts are disꢀ
played in Table S1. These experiments confirm that the acetyl
group on compound 1 shuts down the formation of emissive
rhodol products upon photoinduced Wolff rearrangement. A
mechanistic proposal that explains the divergent photochemisꢀ
try of compounds 1 and 2 is provided in Scheme S4.
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
Figure 2. A) HeLa cells imaged at 561 nm laser excitation after 10
min of incubation with compound 1 (10 ꢀM), before photoactivaꢀ
tion. B) Same cells as in panel A after photoactivation (405 nm,
120 mW, 20%, 1 s). C) *Normalized fluorescence (fluorescence
of compound 1 before photoactivation = 1) with 561 nm laser
excitation after photoactivation (405 nm, 120 mW, 20%, 1 s) of
HeLa cells incubated with the indicated compound (10 ꢀM of
each). 13 to 15 values of independent measurements, mean values
and standard deviations are shown. D) Structures of inhibitors 9
and 10. Scale bars = 10 ꢀm.
Enzymatic and Live-Cell Activation. Compound 1 could be
converted into compound 2 by recombinant human carboxyꢀ
lesterase II (hCEII) in solution. Irradiation of samples of 1 in
the presence of hCEII gave a strong fluorescent signal,
whereas a combination of compound 1, hCEII and βꢀ
lapachone (9, Figure 2D), which is a known inhibitor of carꢀ
boxylesterases,22 gave nearly no change in fluorescence (Figꢀ
ure S19).
Intracellular Diffusion Dynamics. Determining the precise
localization of an enzymatic event with smallꢀmolecule fluoꢀ
rescent sensors can be challenging if the activated probe difꢀ
fuses away of the site of enzymatic reaction before it can be
imaged. The Wolff rearrangement of compound 2 (enzymatiꢀ
cally activated form of compound 1) leads to a ketene interꢀ
mediate that can be trapped by nucleophiles (see synthesis of
compound 8 in the SI). We envisioned that if photoactivation
occurred immediately after compound 1 was converted into
compound 2 by an enzyme, the ketene intermediate could be
trapped by the esterase itself or at least within its vicinity. To
test this hypothesis, we first irradiated a region of interest
(ROI, Figure 3A) within a cell that had been preꢀtreated with
compound 1 and recorded the fluorescence of the whole cell
over a period of 15 min (Figure 3 and Movie S1). Immediately
after photoactivation, there is very fast diffusion of fluorescent
molecules to nonꢀactivated regions of the cell. This fluoresꢀ
cence, however, is significantly weaker than the fluorescence
within the ROI (Figure 3D). A line profile of the fluorescence
intensity over the 15 min period reveals that there is no broadꢀ
ening of the ROI (Figure 3D), indicating that most of the phoꢀ
toactivated molecules stay within the region in which they
were converted.
The ability of endogenous human esterases to convert comꢀ
pound 1 into 2 within live cells was evaluated by confocal
optical microscopy. Cells were incubated with compound 1 at
a concentration of 10 ꢀM, which is lower than the IC50 values
determined for this compound, either in the dark (20 ꢀM) or
upon photoactivation (17 ꢀM, Figure S22). After only 10 min
of incubation, photoactivation of 1 with light of 405 nm (120
mW, 20%, 1 s) led to a substantial increase in fluorescence
(Figure 2). Preꢀtreatment of cells with inhibitor 9 (10 ꢀM),
however, did not decrease significantly the fluorescence after
photoactivation (Figure 2C). In contrast, preꢀtreatment of cells
with compound 10 (10 ꢀM, Figure 2D), which is an inhibitor
of human carboxylesterase I (hCEI),23 decreased substantially
the fluorescence intensity after photoactivation of compound 1
(Figure 2C). This experiment suggests that either compound 1
is a better substrate of hCEI than hCEII or compound 9 is not
an effective inhibitor of hCEII in live cells. In either case, it is
clear that the photoactivation of compound 1 depends to a
large extent on the activity of at least hCEI and therefore can
be used as a fluorescent probe of esterase activity.
Super-Resolution Imaging. As we showed before, compound
1 can only be photoconverted into a fluorophore after preꢀ
activation by esterases and it does not diffuse from the site of
photoactivation. Based on these observations, we concluded
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 17
that a superꢀresolved image of intracellular esterase activity
confirming that singleꢀmolecule signals correspond to esterase
activity.
1
2
3
4
5
6
7
8
could be generated by addition of compound 1 to live cells
that were already under constant cycles of irradiation at 405
nm (photoactivation) and 561 nm (readout) using highly inꢀ
clined laminated optical sheet (HILO) ilumination.24 We enviꢀ
sioned that in such an experiment, molecules of compound 1
would diffuse into the cell and as soon as they react with esꢀ
terases, they would be rapidly photoconverted, trapped and
detected within the vicinity of the enzyme. Singleꢀmolecule
localization microscopy of those emitters would reveal the
intracellular locations of esterase activity.
Singleꢀmolecule data can be used to extract dynamic inforꢀ
mation about the sites of activation of compound 1 as it difꢀ
fuses into the cell. The 30,000 PALM frames (5 min) were
divided in five time intervals of 1 min each and plotted using
different colors for each time interval (Figure 4E). Inspection
of a magnified section of this timeꢀlapse, colorꢀcoded image
(Figure 4F) reveals that early localizations are very few and
sparsely distributed (1 min, blue). As time progresses, activity
can be detected in vesicleꢀlike structures (2, 3, and 4 min;
green, yellow, and red), and at later time points, the tubular
structure of what seems to be the ER starts to become populatꢀ
ed (5 min, white).
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
Figure 3. A) HeLa cell imaged with 561 nm laser excitation after
10 min of incubation with compound 1 (10 ꢀM), before photoacꢀ
tivation. Region of interest (ROI, fuchsia circle), outline of the
cell (cyan line), line for intensity profile (dashed green line) are
shown. B) Same cell as in panel A, 561 nm laser excitation, but
immediately after photoactivation (405 nm, 120 mW, 20%, 1 s)
within the ROI. C) Same cell as in panel A, 15 min after photoacꢀ
tivation within the ROI and imaged with 561 nm laser excitation.
D) Line profile (dashed green line in panel A) of the intracellular
fluorescence intensity immediately after (black squares) and 15
min after photoactivation (red triangles). Scale bars = 5 ꢀm
We performed PALM imaging in HILO mode employing live
HeLa cells. An activation pulse of 405 nm (100 ꢀs duration)
was followed by 100 readout camera frames at 561 nm excitaꢀ
tion (10 ms exposure time). This sequence was repeated every
second. Compound 1 was added to the cells approximately 10
s after the beginning of acquisition and the first singleꢀ
molecule signals started to appear only a few seconds after
addition. A total of 30,000 frames and 360,771 localizations
were obtained over a period of 300 s. Using these localizaꢀ
tions, a superꢀresolved image of esterase activity was reconꢀ
structed (Figure 4A), with an average localization precision of
single molecules of 36±11 nm (Figure S23). This image has a
much higher resolution than its diffractionꢀlimited (>200 nm),
deconvolved counterpart (Figure 4B). Moreover, acquisition
of images under the same conditions, but employing cells that
had been preꢀtreated with inhibitor 10 (10 ꢀM), gave signifiꢀ
cantly fewer signals than untreated cells (Figure 4C and 4D),
Figure 4. A) PALM superꢀresolved image of esterase activity in a
live HeLa cell treated with compound 1 (1 ꢀM, 5 min acquisition
time). B) Diffractionꢀlimited, deconvolved image of the cell
shown in panel A. C) PALM superꢀresolved image of esterase
activity in a live HeLa cell treated with compound 1 (1 ꢀM, 5 min
acquisition time) and inhibitor 10 (10 ꢀM, 30 min incubation
time). D) Diffractionꢀlimited, deconvolved image of the cell
shown in panel C. E) Timeꢀlapse, colorꢀcoded image of the cell
shown in panel A (see color bar in panel F). F) Magnification of
the area inside the magenta square in panel E. Scale bars = 3 ꢀm
(panels AꢀE); 300 nm (panel F).
In the experiments described before, it was shown that the
fluorescent molecules do not diffuse away from the site of
activation and hence represent the distribution of esterase acꢀ
tivity. This finding can be further corroborated by singleꢀ
particle tracking analysis of PALM data (sptPALM).25 In
ACS Paragon Plus Environment

Page 5 of 17
Journal of the American Chemical Society
sptPALM, singleꢀmolecule data can be used to determine
Moreover, reconstructed trajectories can be used to calculate
the diffusion coefficient (D) of single molecules, allowing us
1
2
3
4
which localizations correspond to the same molecule over
multiple frames, and such information can be used to reconꢀ
struct trajectories of diffusing particles after photoactivation.
to identify any rapidly diffusing smallꢀmolecules (D > 20 ꢀ
m2
s–1)26 that were not trapped in the vicinity of the activating
esterases.
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
Figure 5. A) Single particle tracks of photoactivated and trapped compound 1, overlaid on a brightfield image of an untreated live HeLa
cell (tracks are colored randomly). B) Distribution of track “births” for the cell in panel A. C) Voronoï tessellation of panel B. Polygons are
colored according to their areas, see color bar (units: nm2). D) Objects (blue) and clusters (green, for definition of objects and clusters see
Materials and Methods) identified by Voronoï tessellation in panel C. E) Single particle tracks of photoactivated and trapped compound 1,
overlaid on a brightfield image of a HeLa cell that was preꢀtreated with inhibitor 10 (10 ꢀM). F) Distribution of track “births” for the cell
in panel E. G) Voronoï tessellation of panel F. Polygons are colored according to their areas, see color bar (units: nm2). H) Objects (green)
and clusters (blue) identified by Voronoï tessellation in panel G. Scale bars = 5 ꢀm.
To generate single tracks, 20,000 consecutive frames (200 s)
were analyzed using DiaTrack27 and singleꢀmolecule tracks
are displayed in Figure 5A. Trajectories with a lifetime
mentation of an image of cells treated only with compound 1
identified 22 clusters (Figure 5D, green areas) within regions
(objects) of high density of localizations (Figure 5D, blue
areas, for definition of clusters and objects, see Materials and
Methods). In comparison, segmented images for compound
1 in cells preꢀtreated with inhibitor 10 yielded only one clusꢀ
ter (Figure 5H, green area, results using inhibitor 9 can be
found in Figure S25). Regardless of the total number of loꢀ
calizations, this analysis reveals that clusters of active esterꢀ
ases are lost upon treatment with an inhibitor. This inforꢀ
mation would have been lost in diffractionꢀlimited microsꢀ
copy, and we envision that superꢀresolved sensing could be
used to evaluate the ability of inhibitors to reach intracellular
microdomains of enzymatic activity, with potential applicaꢀ
tions in drug discovery. These results demonstrate that quanꢀ
titative information about the extent and subcellular organiꢀ
zation of esterase activity can be extracted from singleꢀ
molecule localization data employing compound 1 as a suꢀ
perꢀresolution probe.
smaller than 3 s and with diffusion coefficients D > 20
s–1 were excluded from further analysis. A mean diffusion
coefficient of 0.013
m2 s–1 was estimated for the calculated
ꢀ
m2
ꢀ
tracks (Figure S24), which confirms that fluorophores are
attached to slowlyꢀdiffusing, possibly membraneꢀbound
macromolecules.28 sptPALM connects multiple signals to
obtain the trajectory of a single molecule and, in the present
case, intrinsically contains quantitative information about
how many molecules of 1 were converted by esterases and
photoactivated. Using DiaTrack, we identified the first signal
of each track (‘’birth’’) and generated an image of singleꢀ
molecule activations (Figure 5B). An equivalent image was
obtained from an experiment in which the cells were treated
with inhibitor 10 (Figure 5F, for results with inhibitor 9, see
Figure S25). These images show significantly different sinꢀ
gleꢀmolecule activation events and also very different intraꢀ
cellular distributions. Whereas untreated cells display clusꢀ
ters, cells treated with the inhibitor seem to contain mostly
sparsely distributed signals. We quantified the formation of
clusters in both images by Voronoï tessellation using SRꢀ
Tessler software.29 Voronoï tessellation is a segmentation
procedure that has been used, for instance, to study the nanoꢀ
architecture of synaptic release of neurotransmitters.30 Segꢀ
SUMMARY AND CONCLUSIONS
We developed a dualꢀactivatable, smallꢀmolecule probe for
superꢀresolved sensing of esterase activity. The sensing
mechanism relies on diverging pathways of the photochemiꢀ
cal Wolff rearrangement depending on the electronic properꢀ
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 17
bovine serum (FBS, 10%) and penicillinꢀstreptomycin (0.1%) at 37
ties of substituted diazoindanones. We were able to control
the photochemical outcome of the reaction by appending an
esteraseꢀcleavable substituent that controls the electronic
properties of the probe. A combination of solution and liveꢀ
cell experiments confirmed that the fluorescence response
depends on esterase activity and that fluorescent molecules
do not diffuse away from their sites of photoactivation. Theꢀ
se features allowed us to perform PALM imaging to obtain
superꢀresolved images of esterase activity in live cells. From
singleꢀmolecule localization data, we identified tracks of
single molecules, which led us to confirm that these smallꢀ
molecule probes do not diffuse substantially from their actiꢀ
vation sites. Moreover, the formation of clusters of enzymatꢀ
ic activity could be quantified.
°C in a 95% humidity atmosphere under 5% CO2 environment.
Cells were grown to 90% confluence and seeded onto Nunc Labꢀ
Tek II chambered cover glass plates 48 h prior to imaging experiꢀ
ments. Before imaging, the growth medium was removed and the
cells were washed with PBS (0.5 mL). Cells were subsequently
incubated for the indicated time with 0.5 mL of imaging medium
(Fluorobrite, Thermo Fisher Scientific) containing the fluorescent
probe (10 ꢀM) and imaged in the same medium.
1
2
3
4
5
6
7
8
Cell Viability Assay. The viability of HeLa cells was determined
using methylthiazoluldiphenylꢀtetrazolium bromide (MTT). Two
96ꢀwell plates were prepared with each well containing 9000 cells
previously grown to 90% confluency and attached during a 24 h
incubation in growth medium. The cells were treated with comꢀ
pound 1 at different concentrations (Figure S22). Solutions of comꢀ
pound 1 (< 110 ꢀM) were prepared in DMSO with final concentraꢀ
tion of 0.25% DMSO in growth medium. From the resulting soluꢀ
tion, 5 ꢀL were added to each well containing the aforementioned
cells in growth medium. Positive controls were prepared with
DMSO (0.25%, 100% viability). After addition of compound 1, one
plate was irradiated in a transꢀilluminator (410 nm) for 30 min,
while the second plate was kept in the dark. After 72 h incubation at
37 oC and CO2 (5%), the cells were treated with 10% MTT solution
(5 g L–1) in imaging medium (Fluorobrite) and incubated for 3 h.
The supernatant was carefully removed and replaced with 2ꢀ
propanol (200 ꢀL). The plates were shaken at 450 rpm in a microꢀ
plate shaker (VWR) and the absorbance of each well was read out
with a plate reader (SPARK 10M, TECAN). Duplicates of tripliꢀ
cates were sampled for every concentration. Absolute IC50 values
were determined with Prism 6 (GraphPad) and error values were
based on standard deviations.
Confocal Fluorescence Microscopy. Fluorescent images of cells
were taken using a Nikon Eclipse Ti Light Microscope equipped
with a Yokogawa spinningꢀdisk confocal scanner unit CSUꢀW1ꢀT2,
two sCMOS cameras (Orca Flash 4.0 V2) and a LUDLPrecision2
stage with a piezo focus. Diodeꢀpumped solidꢀstate lasers (DPSS)
were used as light sources: 405 nm (120 mW, 20% laser power),
445 (100 mW, 40%) and 561 nm (200 mW, 20%). The exposure
time, acquisition time and laser powers were kept constant unless
stated otherwise. All images were collected using a 100x CFI Apo
TIRF (NA = 1.49) objective with oilꢀimmersion. Emitted light was
filtered using the following filters: BP 450/50, BP 470/24, and BP
605/52. The microscope was operated using VisiVIEW (Metaꢀ
morph) software. Quantification of fluorescence intensity was carꢀ
ried out using Fiji (ImageJ 1.5d, NIH) or ICY (GPLv3, Institute
Pasteur). The cell body was selected as the region of interest (ROI)
and the integrated intensity within the ROI was measured. An addiꢀ
tional ROI of the same size and shape was used to obtain the inteꢀ
grated intensity of the background (region with no cell). The backꢀ
ground intensity was subtracted from that of the cellꢀcontaining
ROI. Significant outliers were identified by the ROUT method (Q =
1%)31 and excluded from further analysis. All data sets were anaꢀ
lyzed by unpaired, twoꢀtailed, Student’s tꢀtest. The statistical analꢀ
yses were carried out using Prism 6 (GraphPad). Error bars repreꢀ
sent standard deviation, P values are given and results are depicted
as mean.
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
This proofꢀofꢀprinciple study introduced a mechanism for
singleꢀmolecule localization sensing. The present findings
open the path for the development of smallꢀmolecule probes
to detect, at the nanometer scale, other hydrolytic enzymes,
reactive signaling molecules and metal ions.
MATERIALS AND METHODS
General Methods. All reagents were purchased from commerꢀ
cial sources and used without further purification. Solvents were
obtained from SigmaꢀAldrich and used as received. NMR spectra
1
were acquired on a Bruker AV300 or Bruker 400 instruments. H
NMR chemical shifts are reported in ppm relative to SiMe4 (
and were referenced internally with respect to residual protons in
the solvent ( = 7.26 for CD3Cl and = 3.31 for CD3OD). Coupling
constants are reported in Hz. 13C NMR chemical shifts are reported
in ppm relative to SiMe4 ( = 0) and were referenced internally with
respect to solvent signal ( = 77.16 for CD3Cl and 49.00 for
δ = 0)
δ
δ
δ
δ
CD3OD). Peak assignments are based on calculated chemical shifts,
multiplicity and 2D experiments. IUPAC names of all compounds
are provided and were determined using CS ChemBioDrawUltra 15.
Highꢀresolution mass spectra (HRMS) were recorded by staff at the
Molecular and Bioanalytical (MoBiAS) center (ETH Zurich) emꢀ
ploying a Bruker maXis–ESIꢀQqꢀTOFꢀMS (ESI).
Optical Spectroscopy. Stock solutions in dimethyl sulfoxide
(DMSO) were prepared at concentrations of 1 mM, stored at –20 °C
and thawed immediately before each experiment. Spectroscopic
measurements were conducted in phosphateꢀbuffered saline (PBS)
pH = 7.4. UVꢀVisible spectra were obtained employing a Cary 500
Scan spectrometer using quartz cuvettes from Thorlabs (10 mm path
length). Fluorescence spectra were acquired using a Fluorolog 3
fluorimeter (Horiba JobinꢀYvon). All measurements were conductꢀ
ed at room temperature and under red light ambient illumination to
avoid photoactivation of the compounds. Extinction coefficients
were determined by linear fit of 5 different concentrations of comꢀ
pound in PBS/CH3CN (1:1). Absolute fluorescence quantum yields
were determined in PBS/CH3CN (1:1) by means of an integrating
sphere (Horiba JobinꢀYvon). All spectroscopic experiments were
carried out in triplicate.
Photoactivation of Compounds 1 and 2 in Solution. Photoactiꢀ
vation of compounds 1 and 2 was performed in an RMRꢀ600 photoꢀ
reactor (Rayonet) equipped with eight RMRꢀ3000 Å ultraviolet light
bulbs. For fluorescence analysis, solutions of 1 and 2 were prepared
in PBS/CH3CN (1:1, 0.25 ꢀM) and transferred to a quartz fluoresꢀ
cence cuvette (Thorlabs) to a final volume of 2.5 mL. The solutions
were irradiated (32 W, 25 °C) for a total of 25 min. Fluorescence
intensity was measured (λex = 540 nm and λem = 550–650 nm) beꢀ
fore activation (t = 0 min) and at fixed time intervals (t = 1, 2, 3, 4,
5, 6, 7, 8, 9, 10, 15, 20 and 25 min). The data were plotted as the
integrated fluorescence intensity normalized to the initial intensity (I
= 1). For the procedure to determine quantum yields of photoactivaꢀ
tion, see the SI.
Inhibition of Human Carboxylesterases hCEI and hCEII.
Carboxylesterase inhibitors 9 and 10 were synthesized according to
the reported procedures.22,23 DMSO stock solutions (1 mM) were
prepared. Inhibition experiments in solution were carried out in
black 96ꢀwell plates with transparent bottom (VITARIS AG) for
compound 1 (10 ꢀM). Different samples containing 1, hCEII (Sigꢀ
ma, final concentration 10 ꢀg mL–1), and inhibitor 9 (10 ꢀM) were
prepared (Figure S19). The plate was irradiated with 410 nm light
(LED transꢀilluminator) for 30 min. Dark (nonꢀirradiated) solutions
containing compound 1 were added to the plate after irradiation.
Fluorescence intensities were measured with a plate reader coupled
to a fluorimeter (Horiba JobinꢀYvon). Triplicates of each sample
were measured and the average values were reported with their
respective standard deviation (Figure S19). For experiments in live
Cell Culture. HeLa cells (ATCC CCL2) were grown in Dulbecꢀ
co’s Modified Eagle Medium (DMEM) supplemented with fetal
ACS Paragon Plus Environment

Page 7 of 17
Journal of the American Chemical Society
cells, HeLa cells were incubated for 30 min (10 ꢀM) in growth
medium followed by washing with PBS. Probes 1 and 2 were incuꢀ
Crystallographic information of compounds 1, 5, 6, S3, and S4
(CIF)
1
2
3
4
5
6
bated for 10 min (100 nM) in fresh imaging medium prior to imagꢀ
ing. Pictures were taken at 561 nm excitation with a 500 ms expoꢀ
sure time prior to photoactivation. Photoactivation of the dye was
carried out with the 405 nm laser with 1 s irradiation time. Pictures
after photoactivation were taken at 561 nm excitation with a 500 ms
exposure time.
Supporting Information (PDF)
Movie S1 of intracellular diffusion after photoactivation (AVI)
AUTHOR INFORMATION
Corresponding Author
Intracellular Diffusion Experiment. Cells were incubated with
compound 1 (10 ꢀM) for 10 min. Without any further washing step,
an ROI was selected for irradiation using the FRAP (fluorescence
recovery after photobleaching) module available in VisiVIEW
software. A time lapse imaging experiment was recorded in which
the FRAP pulse (405 nm, 120 mW, 20%, 1 s) was applied.
7
8
9
*
pablo.riveraꢀfuentes@org.chem.ethz.ch
Author Contributions
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
‡ E. A. H. and Z. T. contributed equally.
Super-Resolution Microscopy. HeLa cells were plated in pheꢀ
nol redꢀfree growth medium (Gibco). Cells were washed with PBS
(0.5 mL) before imaging and refilled with imaging medium (0.25
mL). Compound 1 (1 ꢀM in imaging medium) was added to the
cells within the first 10 s of acquisition (for details see below). For
inhibition experiments, cells were incubated with compound 9 or 10
(10 ꢁM in imaging medium) 30 min prior to imaging and the same
imaging procedure was followed. Cells were imaged using a Nikon
NꢀSTORM microscope (Nikon, UK Ltd.) equipped with an SR
Apochromat TIRF 100x 1.49 N.A. oil immersion objective lens.
Activation and excitation lasers with wavelengths of λ = 405 nm
and λ = 561 nm illuminated the sample in HILO mode. Fluoresꢀ
cence was detected with an iXon DU897 (Andor) EMꢀCCD camera
(16 × 16 ꢁm2 pixel size) with an exposure time of 10 ms for the
laser excitation. An inꢀbuilt focusꢀlock system was used to prevent
axial drift of the sample during data acquisition. The emission was
collected and passed through a laser QUAD filter set for TIRF apꢀ
plications (Nikon CꢀNSTORM QUAD 405/488/561/647) comprisꢀ
ing laser cleanꢀup, dichroic and emission filters.
The laser excitation was at 561 nm, with a peak power density of
1.2 kW cm–2. Photoconversion was achieved with 100 ꢀsꢀlong actiꢀ
vation pulses of 405 nm laser of 46 W cm–2 peak power density,
which were directed to the sample every 1 s. The pulses were synꢀ
chronized with the excitation laser using a DAQ Board and a PXI
1033 controller (National Instruments). Typically, 30,000 camera
frames were recorded. From each image stack, a reconstructed suꢀ
perꢀresolved image was generated as described below.
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
This work was supported by ETH Zurich (grant ETHꢀ02 16ꢀ1 to
P. R.ꢀF.). All microscopy work was carried out at the Scientific
Center for Optical and Electron Microscopy (ScopeM) at ETH
Zurich. We thank Dr. Giovanni Bassolino (ETH Zurich) for
assistance measuring quantum yields of photoactivation, Dr.
Pascal Vallotton (ETH Zurich) for advise regarding particle
tracking and Dr. Szymon Stoma (ETH Zurich) for help with
colocalization analysis.
REFERENCES
(1)
(2)
Abbe, E. Arch. Mikrosk. Anat. 1873, 9, 413.
Hell, S. W.; Sahl, S. J.; Bates, M.; Zhuang, X.; Heintzmann,
R.; Booth, M. J.; Bewersdorf, J.; Shtengel, G.; Hess, H.;
Tinnefeld, P.; Honigmann, A.; Jakobs, S.; Testa, I.; Cognet,
L.; Lounis, B.; Ewers, H.; Davis, S. J.; Eggeling, C.;
Klenerman, D.; Willig, K. I.; Vicidomini, G.; Castello, M.;
Diaspro, A.; Cordes, T. J. Phys. D Appl. Phys. 2015, 48,
443001.
(3)
(4)
Rust, M. J.; Bates, M.; Zhuang, X. W. Nat. Methods 2006, 3,
793.
Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.;
Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincottꢀ
Schwartz, J.; Hess, H. F. Science 2006, 313, 1642.
LippincottꢀSchwartz, J.; Patterson, G. H. Trends Cell Biol.
2009, 19, 555.
Chozinski, T. J.; Gagnon, L. A.; Vaughan, J. C. FEBS Lett.
2014, 588, 3603.
Sauer, M.; Heilemann, M. Chem. Rev. 2017, 117, 7478.
von Diezmann, A.; Shechtman, Y.; Moerner, W. E. Chem.
Rev. 2017, 117, 7244.
Specht, E. A.; Braselmann, E.; Palmer, A. E. Annu. Rev.
Physiol. 2017, 79, 93.
Carter, K. P.; Young, A. M.; Palmer, A. E. Chem. Rev. 2014,
114, 4564.
Aron, A. T.; RamosꢀTorres, K. M.; Cotruvo Jr., J. A.; Chang,
C. J. Acc. Chem. Res. 2015, 48, 2434.
Chan, J.; Dodani, S. C.; Chang, C. J. Nat. Chem. 2012, 4, 973.
Hou, J.ꢀT.; Ren, W. X.; Li, K.; Seo, J.; Sharma, A.; Yu, X.ꢀQ.;
Kim, J. S. Chem. Soc. Rev. 2017, 46, 2076.
Single-Molecule Localization and Image Analysis. Molecules
were localized using the ThunderSTORM plugin of Fiji (ImageJ).32
Signals were detected searching the local intensity maxima in each
frame, which were fit using an integrated Gaussian point spread
function. Localizations with a localization precision above 40 nm
and a sigma above 250 nm were filtered out. The first 15 s (1500
frames) were omitted from the analyses. Particle tracking and calcuꢀ
lation of diffusion coefficients were performed using DiaTrack.27
The following parameters were used for particle tracking: maximum
particle jump of 6 pixels (0.48 ꢁm) between frames and a minimum
track length of 5 frames . A “birth” map of the tracks with diffusion
coefficients (D < 20 ꢁm2 sꢀ1) was exported as a text file and cluster
(segmentation) analysis was performed using SRꢀTesseler softꢀ
ware.29 In brief, Voronoï tessellation was applied to the image.
Objects were defined as regions that enclose Voronoï polygons in
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
which the local density δ1 (particles per unit of area) is larger than
i
(14)
(15)
Kaur, A.; Kolanowski, J. L.; New, E. J. Angew. Chem. Int. Ed.
2016, 55, 1602.
twice the average density of the whole image (2δ). Clusters were
defined as regions within the objects in which the density of moleꢀ
cules was higher than the average density of the object.
Mo, G. C. H.; Ross, B.; Hertel, F.; Manna, P.; Yang, X.;
Greenwald, E.; Booth, C.; Plummer, A. M.; Tenner, B.; Chen,
Z.; Wang, Y.; Kennedy, E. J.; Cole, P. A.; Fleming, K. G.;
Palmer, A.; Jimenez, R.; Xiao, J.; Dedecker, P.; Zhang, J. Nat.
Methods 2017, 14, 427.
ASSOCIATED CONTENT
Supporting Information
(16)
(17)
(18)
Ross, M. K.; Crow, J. A. J. Biochem. Mol. Toxicol. 2007, 21,
187.
Supporting figures and schemes, synthetic procedures, spectroꢀ
scopic characterization of all compounds, singleꢀcrystal xꢀray
diffraction methods and data, and NMR spectra. The Supporting
Information is available free of charge on the ACS Publications
website.
Belov, V. N.; Wurm, C. A.; Boyarskiy, V. P.; Jakobs, S.; Hell,
S. W. Angew. Chem. Int. Ed. 2010, 49, 3520.
Belov, V. N.; Mitronova, G. Y.; Bossi, M. L.; Boyarskiy, V.
P.; Hebisch, E.; Geisler, C.; Kolmakov, K.; Wurm, C. A.;
Willig, K. I.; Hell, S. W. Chem. Eur. J. 2014, 20, 13162.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 17
(19)
(20)
Grimm, J. B.; English, B. P.; Choi, H.; Muthusamy, A. K.;
5, 155.
Mehl, B. P.; Dong, P.; Brown, T. A.; LippincottꢀSchwartz, J.;
Liu, Z.; Lionnet, T.; Lavis, L. D. Nat. Methods 2016, 13, 985.
Hirabayashi, K.; Hanaoka, K.; Takayanagi, T.; Toki, Y.;
Egawa, T.; Kamiya, M.; Komatsu, T.; Ueno, T.; Terai, T.;
Yoshida, K.; Uchiyama, M.; Nagano, T.; Urano, Y. Anal.
Chem. 2015, 87, 9061.
Dickinson, B. C.; Huynh, C.; Chang, C. J. J. Am. Chem. Soc.
2010, 132, 5906.
Inagaki, R.; Ninomiya, M.; Tanaka, K.; Koketsu, M.
ChemMedChem 2015, 10, 1413.
Hatfield, M. J.; Chen, J.; Fratt, E. M.; Chi, L.; Bollinger, J. C.;
Binder, R. J.; Bowling, J.; Hyatt, J. L.; Scarborough, J.;
Jeffries, C.; Potter, P. M. J. Med. Chem. 2017, 60, 1568.
Tokunaga, M.; Imamoto, N.; SakataꢀSogawa, K. Nat. Methods
2008, 5, 159.
(26)
Mastro, A. M.; Babich, M. A.; Taylor, W. D.; Keith, A. D.
Proc. Natl. Acad. Sci. U. S. A. 1984, 81, 3414.
1
2
3
4
5
6
7
8
(27)
(28)
Vallotton, P.; Olivier, S. Microsc. Microanal. 2013, 19, 451.
Knight, J. D.; Lerner, M. G.; MarcanoꢀVelázquez, J. G.;
Pastor, R. W.; Falke, J. J. Biophys. J. 2010, 99, 2879.
Levet, F.; Hosy, E.; Kechkar, A.; Butler, C.; Beghin, A.;
Choquet, D.; Sibarita, J.ꢀB. Nat. Methods 2015, 12, 1065.
Tang, A.ꢀH.; Chen, H.; Li, T. P.; Metzbower, S. R.;
MacGillavry, H. D.; Blanpied, T. A. Nature 2016, 536, 210.
Motulsky, H. J.; Brown, R. E. BMC Bioinformatics 2006, 7,
123.
(29)
(30)
(31)
(32)
(21)
(22)
(23)
9
Ovesný, M.; Křížek, P.; Borkovec, J.; Švindrych, Z.; Hagen,
G. M. Bioinformatics 2014, 30, 2389.
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
(24)
(25)
Manley, S.; Gillette, J. M.; Patterson, G. H.; Shroff, H.; Hess,
H. F.; Betzig, E.; LippincottꢀSchwartz, J. Nat. Methods 2008,
ACS Paragon Plus Environment

Page 9 of 17
Journal of the American Chemical Society
Figure for the Table of Contents
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
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 17
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
95x198mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 11 of 17
Journal of the American Chemical Society
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
100x89mm (299 x 299 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 17
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
100x100mm (299 x 299 DPI)
ACS Paragon Plus Environment

Page 13 of 17
Journal of the American Chemical Society
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
100x151mm (299 x 299 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 17
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
203x103mm (299 x 299 DPI)
ACS Paragon Plus Environment

Page 15 of 17
Journal of the American Chemical Society
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
80x44mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 17
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
88x96mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 17 of 17
Journal of the American Chemical Society
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
53x34mm (300 x 300 DPI)
ACS Paragon Plus Environment