An Integrated Chemical Cytometry Method: Shining a Light on Akt Activity in Single Cells

Article abstract of DOI:10.1002/anie.201606914

Tools to evaluate oncogenic kinase activity in small clinical samples have the power to guide precision medicine in oncology. Existing platforms have demonstrated impressive insights into the activity of protein kinases, but these technologies are unsuitable for the study of kinase behavior in large numbers of primary human cells. To address these limitations, we developed an integrated analysis system that utilizes a light-programmable, cell-permeable reporter deliverable simultaneously to many cells. The reporter's ability to act as a substrate for Akt, a key oncogenic kinase, was masked by a 2-4,5-dimethoxy 2-nitrobenzyl (DMNB) moiety. Upon exposure to ultraviolet light and release of the masking moiety, the substrate sequence enabled programmable reaction times within the cell cytoplasm. When coupled to automated single-cell capillary electrophoresis, statistically significant numbers of primary human cells were readily evaluated for Akt activity.

Full text of DOI:10.1002/anie.201606914

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201606914
German Edition: DOI: 10.1002/ange.201606914
Biosensors
An Integrated Chemical Cytometry Method: Shining a Light on Akt
Activity in Single Cells
Emilie R. Mainz, Qunzhao Wang, David S. Lawrence, and Nancy L. Allbritton*
Abstract: Tools to evaluate oncogenic kinase activity in small
clinical samples have the power to guide precision medicine in
oncology. Existing platforms have demonstrated impressive
insights into the activity of protein kinases, but these technol-
ogies are unsuitable for the study of kinase behavior in large
numbers of primary human cells. To address these limitations,
we developed an integrated analysis system that utilizes a light-
programmable, cell-permeable reporter deliverable simultane-
ously to many cells. The reporterꢀs ability to act as a substrate
for Akt, a key oncogenic kinase, was masked by a 2-4,5-
dimethoxy 2-nitrobenzyl (DMNB) moiety. Upon exposure to
ultraviolet light and release of the masking moiety, the substrate
sequence enabled programmable reaction times within the cell
cytoplasm. When coupled to automated single-cell capillary
electrophoresis, statistically significant numbers of primary
human cells were readily evaluated for Akt activity.
the very small sample size of most clinical specimens and
limited throughput of analysis has precluded the routine
measurement of kinase activity in single human primary
cells.[2] The requirements for these measurements are two-
fold: 1) a sensor that is selective for the kinase of interest,
deliverable simultaneously to single cells, and resistant to
phosphorylation until the desired assay start time, and 2)
a platform that can analyze statistically relevant numbers of
single cells.
Fluorescent peptide sequences are utilized as reporters to
directly assay kinase activity and stand to bridge the gap
between small molecules, which are poor substrates of limited
specificity, and genetically engineered proteins, which are
difficult to transfect uniformly into cells.[3] Peptide reporters
phosphorylated by kinases within cells are highly sensitive,
quantitative, and possess great multiplexing potential when
coupled to capillary electrophoresis (CE). To date, the
delivery of these reporters into the cell has relied on low-
throughput methods, such as microinjection, or on cell-
penetrating peptides (CPPs), which largely deliver cargo to
endosomal compartments.[4] Additionally, reporters delivered
by CPPs may be acted upon by the kinase of interest during
the delivery process, thus the reaction time in the cell is
unknown. Delivering a reporter into the cytosol of large
numbers of cells while retaining control of assay initiation
would dramatically improve overall throughput when cou-
pled to an automated single-cell assay platform. The latter
would enable sampling from sufficient numbers of cells for
statistical analyses and identification of cellular subpopula-
tions.
T
he appropriate prescription of targeted therapeutics for
cancer treatment relies on the accurate identification of
aberrant signaling pathways, most frequently those involving
protein kinases. However, the selection of patients for
pathway inhibition is mostly empirical or reliant upon
sequencing to identify mutations within the genome. A
targeted selection process that reveals those patients whose
cells possess increased kinase activity would lessen costs as
well as forego drug toxicity in individuals not likely to be
responsive to pathway inhibition. Thus, kinase-activity sen-
sors and their attendant analytical platforms to identify
aberrant kinase activities in primary cells independent of their
genomic mutation status would be of high value.[1] However,
We addressed these opportunities by integrating chemical
and instrumental innovations. Specifically, we have developed
a cell-permeable, photoactivatable reporter that is chemically
masked (Scheme 1) and integrated into an automated anal-
ysis platform. The active reporter acts as a substrate for the
protein kinase Akt, a kinase which plays critical roles in
tumor formation and progression.[5] The native substrate (NS)
Akt peptide reporter 6FAM-GRP-(R)-AFTF-(A)-Amide (R
and A represent N-methylated versions of amino acids) has
favorable phosphorylation kinetics, specificity, and a demon-
strated resistance to intracellular degradation.[6]
NS possesses a single hydroxy group that, when coupled to
a photoremovable DMNB group, should block phosphoryla-
tion by Akt and thus silence the reporter as a substrate
(Figure 1b). Figure 1a and the Supporting Information, Fig-
ure S1 demonstrate the photochemical conversion of inactive
to active reporter. Importantly, the active reporter 2 is
phosphorylated at a rate indistinguishable from that of NS,
both in vitro and in cell lysates (Figure 1b). The inactive Akt
reporter 1 is not phosphorylated nor spontaneously converted
[*] Prof. N. L. Allbritton
Department of Chemistry and Pharmacology
University of North Carolina
Chapel Hill, NC 27599 (USA)
and
Joint Department of Biomedical Engineering, University of North
Carolina and North Carolina State University
Chapel Hill and Raleigh, NC (USA)
E-mail: nlallbri@unc.edu
Dr. E. R. Mainz
Department of Chemistry, University of North Carolina
Chapel Hill, NC 27599 (USA)
Prof. Q. Wang, Prof. D. S. Lawrence
Department of Chemistry, Division of Chemical Biology and
Medicinal Chemistry and Department of Pharmacology
University of North Carolina
Chapel Hill, NC 27599 (USA)
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
http://dx.doi.org/10.1002/anie.201606914.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
Figure 1. In vitro characterization of caged AKT reporter. a) Photolysis
of 1 (10 mm) in extracellular buffer (ECB). b) Phosphorylation assays
with recombinant Akt. In vitro assays in buffer were performed with
peptides NS (&), 1 (*), and 2 (~). Experiments in PANC-1 lysates
were performed with peptides NS (&), 1 (*), and 2 (~) (Supporting
Information, Section S3.5).
Scheme 1. Native substrate (NS) was chemically caged with DMNB to
generate 1, which is delivered intracellularly. Photolysis releases the
active sensor 2 (identical to NS), which is phosphorylated (3) and
measured in single cells by automated single-cell capillary electro-
phoresis.
gesting that the addition of DMNB may be sufficient to
render peptides of similar hydrophobicity cell-permeant.
We next evaluated the ability of this reporter to detect
Akt activation after photolysis within intact PANC-1 cells.
Akt is upstream of DNA repair mechanisms and is known to
be a stress-activated kinase. Thus, we first assessed whether
the UV exposure needed for photoactivation might cause
DNA damage and/or activate Akt. CE-analysis revealed that
UV illumination (0.4 JcmÀ2) of cells transforms 96 Æ 2.2% of
1 into active 2. UV exposure under these conditions does not
induce cytotoxicity or DNA pyrimidine lesions (Supporting
Information, Figures S8 and S9). Using these conditions, we
assessed Akt activation and inhibition in single PANC-1 cells
(Figure 2d,f). After photolysis, the contents of the single cells
were assayed by CE and the phosphorylated sensor was
quantified. Without endogenous stimuli, minimal Akt activity
to its active counterpart in the dark during the assays
(Supporting Information, Figure S6).
Having demonstrated that 1 is photoactivated and sub-
sequently phosphorylated by Akt, we explored the delivery
options for this reporter. Since NS is excluded from PANC-
1 cells, prior work required microinjection of the reporter into
cells, which is a protracted process. Many CPPs have been
reported, but most are taken up through endosomal pathways,
resulting in significant peptide degradation and little to no
delivery to the cytosol, which is the primary location of Akt.[7]
These challenges were manifested in our initial attempts to
render NS cell-permeable, with multiple
CPPs[8] appended to the reporter resulting
in a punctate distribution of the reporter
after uptake into the cell, which is indicative
of a non-cytoplasmic location (Supporting
Information, Figure S4). However, NS,
when modified with DMNB (1), is suffi-
ciently hydrophobic to enter into the cytosol
when incubated with PANC-1 cells (Fig-
ure 2a and the Supporting Information,
Figure S5). The diffuse (rather than punc-
tate) appearance of the fluorescence after
loading peptide into the cells (Supporting
Information, Figure S5) combined with the
observed phosphorylation of the reporter
(post photoactivation) (Figure 2d) suggests
delivery into the cytosol rather than a cyto-
solic subcompartment. Quantification of the
amount of intracellular peptide by CE
revealed a correlation (R2 = 0.79) between
the total moles of fluorescent reporter
Figure 2. Intracellular characterization of peptide 1. Images of cells incubated at 48C with
a) 1, b) 2, and c) without peptide. d) Single-cell Akt activity measured by chemical cytometry
with the electrophoretic peaks labeled as 1, 2, and 3 as denoted in Scheme 1. 6FAM-labeled
proteolytic products of 1 are labeled as i–viii (Supporting Information, Table S2). The STND
trace shows the migration time of a standard solution of the peptides. Traces below are
from individual PANC-1 cells that were deprived of serum (basal), stimulated with
100 ngmLÀ1 TNFa, or treated with 10 mm LY294002 + 100 ngmLÀ1 TNFa. e) The amount
(in attomoles) of reporter loaded as a function of cell diameter (R2 =0.79). f) Akt activity
for a population of single cells analyzed 20 min post-photolysis. Prior to loading with 1,
(ntotal = ncaged + nphosphorylated + ndegraded
)
and
the cell diameter (Figure 2e). NS modified
with a DMNB-Ser moiety (rather than
DMNB-Thr) is likewise cell-permeable
(Supporting Information, Figure S7), sug- photolysis, and analysis, the cells were treated as described in (d).
2
www.angewandte.org ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
was detected (Figure 2 f), which suggests
that the illumination protocol does not
activate Akt.[5] We also examined Akt
pathway stimulation and inhibition in
response to TNFa and TNFa + LY294002,
respectively. Significant heterogeneity in
reporter phosphorylation in response to
TNFa stimulation (0–100%) was observed,
consistent with reports of mosaic Akt
activation in individual cells of a cell pop-
ulation.[6,9] The PI3K inhibitor, LY294002,
blocked reporter phosphorylation in
response to TNFa, as expected since PI3K
is downstream of the TNFa receptor and
upstream of Akt. Having produced a cell-
permeable and photoactivatable Akt
reporter, we sought to harness these
unique properties to enhance the through-
put of single-cell Akt activity measure-
ments. Previous work describing the micro-
injection of peptide-based protein kinase
sensors offered a very limited throughput of
0.5 cellhÀ1, which is not biomedically
useful.[6] A recently reported automated
single-cell CE system quantified kinase
activity as rapidly as 3.5 cellsminÀ1, rivaling
microfluidic single-cell electrophoresis sys-
tems and offering excellent (10À20 mol)
detection limits.[10] A limitation of the
automated system, however, is that report-
ers are active substrates during the loading
process into the cell, so that the reaction
time in the cell is unknown. These chal-
Figure 3. Automated single-cell analysis. a) The microfabricated integrated system used for
automated single-cell CE analysis. Cell traps (45 mm deep, inset) were loaded with PANC-
1 cells and the peptide within the cell was photoactivated. Cells were then assayed.
b) Separation of intracellular contents from consecutively analyzed cells. Trace shows peaks
for 2, 3, and proteolytic products of 1 (i–viii) for four single cells (E–H). Inset: Trace
showing the separated contents of 11 cells (A–K). c) Reporter metabolism in PANC-1 cells.
Each colored data point represents a single cell with the color denoting the amount of
reporter (in moles) loaded into the cell. d) Reporter metabolism at various time points after
photoactivation. Each overlay (color) represents the metabolism from a single cell.
lenges are uniquely addressed by reporter 1, which is
membrane-permeant with a programmable light-triggered
reaction initiation.
paired with single-cell CE, the amount of the different
peptide species are readily quantified (Figure 3c).
Rather than simply increasing as a function of reporter
incubation time, the rate of Akt reporter phosphorylation in
single cells was highly variable (0.0–0.024 zmolpgÀ1 sÀ1) at all
times measured (Figure 3c the Supporting Information, Fig-
ure S9). While we and others have demonstrated heteroge-
neous Akt activity in PANC-1 and other cell lines at
individual endpoints,[6,9,12] reporter 1 enabled time-resolved
studies of this heterogeneity, suggesting that peptide phos-
phorylation may reach an equilibrium determined by the
balance of kinase and opposing phosphatase activity, with
a unique equilibrium point for each cell (Supporting Infor-
mation, Figure S11). As Akt is a major promoter of cell
survival,[13] this type of heterogeneity has been suggested to
promote survival of a subset of cells under stressful conditions
and/or confer resistance to chemotherapeutics.[14] The inte-
grated analysis system also delivers insight into how peptides
are modified over time within single cells (Figure 3d; frag-
ment naming in the Supporting Information, Table S2), and
the formation of key fragments can be monitored (Supporting
Information, Section S4).
We conducted a pilot study in which the unique properties
of compound 1 were combined with the automated single-cell
CE system. All cells in a population were simultaneously
photo-activated to assess Akt activity as a function of time.
Cells were loaded with compound 1, illuminated, and serially
analyzed (n = 109) (Figure 3). The photoactivation efficiency
is 97 Æ 2.8% within intact cells, with 42.4 Æ 42.7 amol of the
reporter loaded per cell. The amount of reporter (amol)
detected correlated with the cell volume (Figure 2e) and
appeared to be independent of the dwell time (R2 = 0.018;
Supporting Information, Figure S10), which suggests that the
reporter is not exported by the cell over the analysis period.
An oft-overlooked commonality among peptides, even when
engineered for stability, is their eventual degradation within
the intracellular environment. Phosphorylation, degradation,
and other intracellular alterations or environments can have
similar effects on sensor fluorescence properties, creating
a challenge for many sensors relying solely on fluorescence
properties as a proxy for sensor phosphorylation.[11] In
contrast, the single-cell CE system measures intact and
degraded reporter as well as phosphorylated reporter to
accurately quantify the phosphorylation rate (Figure 3b).
When peptide standards of known concentration are also
The cell-permeable, photoactivated reporter, combined
with an integrated platform, enabled the analysis
(7.2 cellshÀ1) of kinase activity in a population of cells at
a substantially higher throughput than previous studies. The
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
[6] A. Proctor, G. Herrera-Loeza, Q. Wang, D. S. Lawrence, J. J.
Yeh, N. L. Allbritton, Anal. Chem. 2014, 86, 4573 – 4780.
[7] H. Rꢀꢀgel, P. Sꢀꢀlik, M. Pooga, Biochim. Biophys. Acta
Biomembranes 2010, 1798, 2240 – 2248.
[8] a) K. Montrose, Y. Yang, X. Sun, S. Wiles, G. W. Krissansen, Sci.
Rep. 2013, 3, 1661; b) J. Cruz, M. Mihailescu, G. Wiedman, K.
Herman, P. C. Searson, W. C. Wimley, K. Hristova, Biophys. J.
2013, 104, 2419 – 2428.
protein kinase reporter is rapidly loaded into the cytoplasm of
cells and converted, on demand, into the active reporter. We
expect that outfitting peptide sensors with DMNB or related
moieties to properly tune the amphiphilic properties of
peptide-based protein kinase reporters, will prove useful in
revealing aberrant kinase activity in small samples of clinical
interest, especially when combined with automated analysis
systems.
[9] T. L. Yuan, G. Wulf, L. Burga, L. C. Cantley, Curr. Biol. 2011, 21,
173 – 183.
[10] A. J. Dickinson, P. M. Armistead, N. L. Allbritton, Anal. Chem.
2013, 85, 4797 – 4804.
[11] T. N. N. Van, M. C. Morris, Fluorescent Sensors of Protein
Kinases: From Basics to Biomedical Applications, Academic
Press, New York, 2013, p. 315.
[12] a) E. R. Mainz, D. S. Serafin, T. T. Nguyen, T. K. Tarrant, C. E.
Sims, N. L. Allbritton, Anal. Chem. 2016, 86, 7786 – 7792; b) R.
Meyer, L. A. DꢁAlessandro, S. Kar, B. Kramer, B. She, D.
Kaschek, B. Hahn, D. Wrangborg, J. Karlsson, M. Kvarnstrom,
M. Jirstrand, W. D. Lehmann, J. Timmer, T. Hofer, U. Kling-
muller, Front. Physiol. 2012, 3, 451.
Acknowledgements
We thank the NIH (CA203032 and CA177993) for financial
support.
Keywords: biosensors · chemical cytometry ·
kinase activity probes · photochemistry · single-cell analysis
[13] S. R. Datta, A. Brunet, M. E. Greenburg, Genes Dev. 1999, 13,
2905 – 2927.
[14] a) T. L. Yuan, L. C. Cantley, Oncogene 2008, 27, 5497 – 5510;
b) M. G. Schlieman, B. N. Fahy, R. Ramsamooj, L. Beckett, R. J.
Bold, Br. J. Cancer 2003, 89, 2110 – 2115; c) M. Mortenson, J. M.
Galante, M. Schlieman, R. J. Bold, Cancer Ther. 2004, 2, 227.
[1] S. S. Rubakhin, E. V. Romanova, P. Nemes, J. V. Sweedler, Nat.
Methods 2011, 8, S20 – S29.
[2] N. C. Turner, J. S. Reis-Filho, Lancet Oncol. 2012, 13, 178 – 185.
[3] N. P. Damayanti, L. L. Parker, J. M. K. Irudayaraj, Angew.
Chem. Int. Ed. 2013, 52, 3931 – 3934; Angew. Chem. 2013, 125,
4023 – 4026.
[4] W. F. Veldhuyzen, Q. Nguyen, G. McMaster, D. S. Lawrence, J.
Am. Chem. Soc. 2003, 125, 13358 – 13589.
Received: July 16, 2016
[5] C. Garcia-Echeverria, W. R. Sellers, Oncogene 2008, 27, 5511 –
5526.
Revised: August 20, 2016
Published online: && &&, &&&&
4
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Biosensors
E. R. Mainz, Q. Wang, D. S. Lawrence,
N. L. Allbritton*
&&&& — &&&&
An Integrated Chemical Cytometry
Method: Shining a Light on Akt Activity in
Single Cells
See the light: An integrated analysis
system that utilizes a light-programma-
ble, cell-permeable reporter that can be
delivered simultaneously into many cells
was developed. The reporter is a sub-
strate for Akt, a key oncogenic kinase, and
was used to evaluate the Akt activity of
statistically significant numbers of pri-
mary human cells through automated
single-cell capillary electrophoresis.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!