Fluorescent Kinase Probes Enabling Identification ... - ACS Publications

Nov 18, 2016 - Department of Chemistry, University of Miami, Coral Gables, Florida ... of Biochemistry and Molecular Biology, University of Miami, Mia...
0 downloads 0 Views 3MB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Letter

Fluorescent Kinase Probes Enable Identification and Dynamic Imaging of HER2(+) Cells Heajin Lee, Wenjun Liu, Adrienne S Brown, Ralf Landgraf, and James N Wilson Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b03836 • Publication Date (Web): 18 Nov 2016 Downloaded from http://pubs.acs.org on November 21, 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.

Analytical Chemistry 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 5

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

Analytical Chemistry

Fluorescent Kinase Probes Enable Identification and Dynamic Imaging of HER2(+) Cells. Heajin Lee,† Wenjun Liu,‡ Adrienne S. Brown,† Ralf Landgraf, ‡* James N. Wilson†* †

Department of Chemistry, University of Miami, Coral Gables, Florida 33146, United States



Department of Biochemistry and Molecular Biology, University of Miami, Miami, Florida 33101, United States

* Email [email protected], fax +1-305-284-4571; *Email [email protected], fax +1-305-243-3955

KEYWORDS: receptor tyrosine kinase, fluorescence, kinase inhibitor, Her2, ERBB2 ABSTRACT: The EGFR/ERBB family of receptor tyrosine kinases is central to many signaling pathways and a validated chemotherapy target in multiple cancers. While EGFR/ERBB-targeted therapies, including monoclonal antibodies, e.g. trastuzumab, and small molecule kinase inhibitors, such as lapatinib, have been developed, rapid identification and classification of cancer cells is key to identifying the best treatment regime. We report ERBB2 (also HER2) targeting kinase probes that exhibit a “turn-on” emission response upon binding. These live cell compatible probes differentiate ERBB2(+) cells from low-level, ERBB2(-) cells by targeting the intracellular ATP-binding pocket of ERBB2 with therapeutic inhibitor like specificity. Beyond kinase expression levels, probe signal is linked to the phosphotyrosine-correlated activation state of the ERBB2 population. Additionally, the rapid signaling capability of the probes can report changes in activation state in live cells providing a unique type of complimentary information to immunohistochemical assays of receptor kinase populations.

A cornerstone of advanced chemotherapies is the ability to selectively target signaling pathway that are uniquely dysregulated in cancer cells.1-5 Examples of this paradigm are monoclonal antibodies as well as small molecule kinase inhibitors. By contrast, modern diagnostic and analytical tools rely heavily on the specificity of antibodies, yet the analogous application of small molecule pharmacophores for the identification or analysis of cancer biomarkers has not been realized. This is despite the fact that less specific fluorescent molecular probes, such as nuclear or membrane stains, have high utility in live cell imaging.6-9 Combining the ease of application and live cell compatibility of these existing probes with the high selectivity of kinase-specific small molecule chemotherapeutics could yield molecular reporters capable of interrogating cancer-relevant signaling pathways in ways that are currently not accessible. Kinase signaling is typically assessed by end-point measurements via antibody based phosphorylation state assays, in most cases by western blot, however, this approach only provides data on cell populations. On a single cell level, immunohistochemistry approaches cannot provide continuous real-time reporting without a series of rapidly executed endpoint measurements on permeabilized and fixed samples.3 To address these limitations, we have developed fluorescent kinase probes targeted at the EGFR/ERBB receptor tyrosine kinases, which are validated chemotherapeutic targets in a many head, neck and breast cancers. We not only demonstrate that these live cell compatible probes compliment existing bioanalytical assays as they can identify ERBB2-overexpressing, i.e. HER2(+), cell, but also provide real time imaging of systematic perturbations of ERBB2 functional states and population heterogeneities.

The quinazoline core of EGFR/ERBB-targeted inhibitors4,10 is amenable to chemical modification and by extending the conjugation of the aromatic core via the 6-position, we have produced several fluorescent derivatives (Figure 1). The addition of a π-electron donating moiety, such as the 4dimethylaminophenyl group, generates a push-pull system with the p-electron withdrawing core. In an electronically excited state, the donor-acceptor interaction can lead to a polarized excited state with significant charge transfer (CT) character. In polar media, such as the aqueous intra- and extracellular environment, the CT state is stabilized and relaxes non-radiatively, likely via a twisted intramolecular CT (TICT) state.11 Binding to the ATP pocket of ERBB2 can minimize nonradiative processes, by limiting the extent of CT character in the relatively nonpolar and solvent excluding binding fold, or by eliminating the TICT state through conformational restriction. pharmacophore arm

A)

R1

R1 N

R2

π electron donor HN

R3

B) R2

R1 R1

N

HN

N planarized kinase-bound

N

ON

R3 N

π electron acceptor

free rotation in solution

N

OFF

Figure 1. Design of an OFF/ON fluorescent kinase inhibitor. A) Building on the N-phenyl-4-aminoquinazoline scaffold, an electrondonating arm is appended at the 6-position to create a donor-acceptor fluorophore. When bound to a geometrically confined and solventexcluding binding pocket, the π-system is should be planar and emissive, while in solution B) the polarized excited state is stabilized,

ACS Paragon Plus Environment

Analytical Chemistry

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

likely through twisted intramolecular charge transfer (TICT) state and emission is quenched.

A prototype probe, 1 (Figure 2), exhibited a “turn-on” emission response to ERBB2 kinase domain in solution, however, live cell imaging proved inaccessible due to the low binding affinity of this probe and other first generation probes (Ki ≈ 5-10 µM)12 as well as their propensity to aggregate at concentrations required to achieve even partial occupancy of the target binding site.13 Two modifications significantly improved the binding affinity of these probes and improved their imaging capability. First, the addition of an N-methyl piperazine moiety in place of the dimethylamino substituent serves to improve solubility in aqueous solutions. Second, one of two pharmacophore arms was utilized in place of the N-phenyl arm: to generate a Type I or active state inhibitor,4,10 a 3-chloro-4-fluorophenyl arm (3, 5) was employed instead; a Type II, or inactive state inhibitor,10 could be generated by the inclusion of a longer 3-chloro-4-[(3-fluorophenyl)methoxy]-phenyl arm (4, 6) arm. These two pharmacophore arms provide an opportunity to potentially assess the distribution and dynamic changes of the active and inactive conformations of the ERBB2 kinase domain. Each of these modifications was investigated for its impact on the Ki of the probes in receptor stimulation assays. The addition of the N-methyl piperazine arm effectively lowers the Ki from ~10 µM for 1 to 360 nM for 2. The addition of the Type I pharmacophore arm has a more modest effect, lowering the Ki to about 3 µM, for 3, while a Type II arm has a more pronounced effect, with a Ki of 500 nM for 4. The fully optimized Type I and Type II probes, compounds 5 and 6, respectively, have Ki values of 71 nM and 27 nM, comparable to both our experimentally measured and reported values (20 nM and 13 nM, respectively) for lapatinib.14 This allows partial target occupancy at low probe concentrations, thus minimizing limitations imposed by solubility and unspecific binding. The fluorophore component of the probes remained unchanged and the photophysical parameters of the probes varied only moderately from the parent compound, 1.12 Absorption maxima, obtained in CHCl3, ranged from 316 to 325 nm, with a significant shoulder extending beyond 400 nm, allowing for excitation with a 405 nm laser line. Emission maxima covered a somewhat broader range from 475 to 504 nm. The most significant difference was the addition of the Type II pharmacophore arm, which lowered the quantum yield of 4 and 6 relative to the Type I probes, likely through an excited state electron transfer process (Table S1). The fluorescence response of the optimized Type I and II probes, 5 and 6, respectively, was also evaluated in PBS solutions +/- purified, soluble HER2 kinase domain (Figure S7). Both probes exhibited a significant turn-on emission response (11fold and 4-fold for 5 and 6, respectively) in the presence of kinase domain and the emission maxima are moderately blue-shifted by approximately 40 nm, relative to CHCl3 solutions. Additional confirmation of the binding-induced emission response comes from the excitation spectra, which show pronounced peaks at 280 nm, which are absent in the absorption spectra and are likely due to energy transfer from Tyr and Trp residues surrounding the ATP binding pocket . With high affinity, fluorescent EGFR/ERBB-binding probes in hand, we next evaluated their ability to target these receptor tyrosine kinases in a live cell setting. The EGFR/ERBB family is implicated in a number of head and neck cancers, while ERBB2 is prognostic biomarker found in about 20% of breast cancers. BT474 cells (106 receptors/cell) are classified as HER2(+) and were

Page 2 of 5

utilized in dual labeling experiments to compare the fluorescent kinase probes against immunohistochemistry. Figure 4 depicts BT474 stained with ERBB2 antibody, then treated with fluorescent kinase probe 6 (see Figures 5 and S8 for the application of 5). The overlay of the two imaging channels shows significant overlap between the anti-ERBB2, shown in red, and 6, in green, with colocalization in yellow. This result demonstrates the probes' affinity of is maintained even in the complex intraceullar environment with a high degree of selectivity for ERBB2 preserved. Time course measurements (Figure S9) revealed that the binding of 5 is more rapid than that for 6 which is in good agreement with models of Type I and Type II binding kinetics.3

N N

N

HN

HN

N

N

N

improve solubility

N

1 improve binding

2

N

F HN

O N

Cl N

4

F

Cl N

Type I Arm

N

3

HN

Type II Arm

N

improve solubility F

N N

HN

Cl

O

N N

HN

N

5

F

Cl N

6

N

N

Figure 2. Structure key and optimization scheme for fluorescent EGFR/ERBB inhibitors.

[compound]/nM 0

39 156 625 2500 10K

1 2

I

3 5

Lapatinib 4

II

6

Figure 3. Western blot analysis of ERBB2 pTyr levels demonstrates that successive modifications of the parent compound, 1 (see Figure 2, above), improves binding affinities of Type I and Type II probes grouped in the top and bottom panels, respectively. Addition of the pharmacophore arms alone (3 and 5) has a modest effect, while appending the N-methyl piperazine moiety has a more pronounced effect on lowering the Ki values for 2, 5, 6. The fully optimized probes, 5 and 6, possess Ki values in the low nM range comparable to clinically relevant inhibitors such as lapatinib.

ACS Paragon Plus Environment

ERBB2

overlay

probe 6

20 µm

Figure 4. Costaining of BT474 cells with ERBB2-directed antibody (panel A) and fluorescent kinase probe, 6, (panel B) demonstrates a high degree of overlap. [6] = 2 µM, λex = 405 nm, λem = 450-550 nm.

We next evaluated the ability of a fluorescent kinase probe to differentiate between the ERBB2-overexpressing BT474 cells and another breast cancer cell line, MCF7, carrying ERBB2 at low levels (104 copies/cell), are classified clinically as HER2(-). As 5 exhibited more rapid uptake and higher brightness than 6, it was selected for these experiments. Figure 5 shows coseeded BT474 and MCF7 cells which were costained and imaged under similar conditions as the BT474 cells in Figure 4. Significant overlap is seen between the fluorophore-tagged antibody and kinase probe emission in the BT474 cells. The MCF7 cells can readily be discerned by eye, both by their morphology, as well as the near absence of anti-ERBB2 staining, compared to the BT474 cells. Relatively weak and disperse emission from 5 is visible in the MCF7 cells, which is not surprising as they do express ERBB2, albeit at much lower levels, however, the BT474 cells are readily distinguished by their bright emissive patches. This can be seen qualitatively in Figure 5B and 5C, and quantitatively in Figure 5D, demonstrating the ability of 5 to identify ERBB2(+) cells.

A) ERBB2

The higher level of ERBB2 expression in BT474 cells clearly allows for increased binding of 5, which is reflected in the higher emission intensity from BT474 cells compared to MCF7 cells. The emission histogram (Figure 5D) reveals considerable heterogeneity in the kinase probe signal within the BT474 population. This may simply reflect receptor levels or could report cell-to-cell variability in the number of receptors in an activated, i.e. ATP-binding state. We therefore carried out fluorescence-activated cell sorting (FACS) on BT474 cells treated with 5. Cells were sorted into two populations displaying high and low probe signal respectively (Figure 6A), followed by a 24 hour recovery prior to western blot analysis of ERBB2 receptor levels and phosphotyrosine state. Even after the 24 h recovery period, cells with high probe emission intensity showed a relative receptor phosphorylation that was 25% higher than the cells classified as low intensity, suggesting that, besides receptor levels, the basal activation state reflected a stable and cell specific feature. Thus, in addition to being able to discriminate ERBB2-overexpressing cells from cells expressing normal levels, fluorescent kinase probes can also report cell to cell variation in the receptor activation states within a population. This readout provides a unique and versatile complement to traditional, static immunohistochemical labeling. 240 220 200 A) 200

low

160 160 140 120 120

B) probe 5

100 80 80 60 40 40

100

C) emission intensity

low

BT474 cells high

60 50

+ NaOV/mock

40 30 20 10 0

0 25

15 10 5

100

0 0 emission intensity 100

Figure 5. Co-seeded MCF7 and BT474 cells were treated with antiERBB2 and labeled secondary antibody (red, panel A) and probe 5 (green, panel B). While HER2(+) cells are readily visible in both channels, HER2(-) MCF7 cells, encircled in panel C, are not. D) Quantitative analysis of the cell population from 10 experiments (250 cells) reveals that BT474 cells exhibit higher emission intensity than MCF7 cells and marked signal heterogeneity. The “high” signal BT474 cells, evaluated following FACS, show 25% more phosphotyrosine than the “low” population demonstrating that 5 is capable of reporting overall activation state in live cells (see Figure 6, below).

120

240

1200

360

time/sec

480

0 0

D) 100

0

MCF7 cells

0

200 300 400 400 500 600 600 700 800 800 9001000 1000 1100 200 cpd 5 intensity

probe + NaOV probe + 0.1% DMSO

70

D) 100

probe inhibitor + probe

20

20 0

C) overlay

B)

high

180

emission intensity

C)

B)

emission intensity

A)

% cell population

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

Analytical Chemistry

cells per bin

Page 3 of 5

600

30

60

90

time/sec

120

150

480

600

high response low response

80 60 40 20 0 0

120

240

360

time/sec

Figure 6. A) The FACS histogram of BT474 cells treated with 5 reveals a distribution of emission intensities that correlates well with fluorescence microscopy analysis (see Fig 4D above) and can be linked to ERBB2 activation state via phosphotyrosine analysis. B) Pretreatment with an ERBB2 inhibitor, CI-1033, significantly lowers the binding-induced signal, which saturates in untreated cells in 2-3 minutes with 2 µM solutions of 5. Error bars show S.D. C) Tyrosinephosphatases regulate ERBB2 activation and the addition of a phosphatase inhibitor will lead to an increase in the active state population. The dynamics of this activation can be monitored in real time following the addition of NaOV, a phosphatase inhibitor, which induces a rapid increase in emission intensity of BT474 cells treated with 5. D) The error bars, showing S.D., in panel C reflect the population heterogeneity, which can be clearly seen in single cell analysis of "high" and "low" responding cells.

ACS Paragon Plus Environment

Analytical Chemistry

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

The FACS data combined with western blot analysis indicate that differences in activation states can be detected on a subpopulation level as a function of time. We therefore evaluated the ability of 5 to track rapid perturbations of the activation states at the level of individual cells. First, we established that despite the heterogeneity in emission response, the fluorescence intensity in BT474 cells rapidly reaches a maximum within several minutes of probe addition (Figure 6B). This increase in fluorescence does not occur in cells pretreated with the ERBB2-directed, Type I kinase inhibitor, CI-1033 (Canertinib), validating probe specificity in a complex, cellular setting. Although BT474 cells carry a large number of auto-activated ERBB2 receptors, the fraction of activated receptors is kept low by competing actions of tyrosinephosphatases. Blocking of tyrosine phosphatases is known to rapidly shift the equilibrium towards a predominance of activated states, conventionally measured by endpoint measurements of tyrosine phosphorylation.15-17 When BT474 cells are pre-incubated with 5 for 5 minutes, the addition of sodium orthovanadate (NaOV), a tyrosine phosphatase inhibitor, results in a rapid increase in probe signal (Figure 6C). One key interpretation of this NaOV-induced activation is that the fluorescent kinase probe, although inhibitory at saturating concentrations, does not block the activation response. Both Figure 6B and 6C represent that average data from 20 cells in three separate imaging experiments. The variability of the signal is not a reflection of the signal to noise ratio of individual measurements, rather it reveals cell-to-cell variability in activation states, information that is typically lost in bulk analyses. Figure 6D showing traces of two individual cells that represent high and low responders that are consistent with Figure 6A. This data represents a novel route towards the analysis of cellular factors that control actual activation states of an oncogenic receptor on a single cell level. In summary, we have generated live cell compatible, high affinity kinase probes that identify ERBB2-overexpressing, i.e. HER2(+), cells through a binding-induced emission response. The combined observations utilizing the optimized active state probe, 5, demonstrate that kinase-targeted molecular reporters are capable of stratifying individual live cells by their dynamic response to activation. With the increasing emphasis on cancer heterogeneity and single cell signaling studies this adds a powerful tool to the available toolset. This capability can be merged with rapidly developing single cell expression and genetic analyses to probe dynamic responses in receptor kinase signaling and perturbation of this signaling by agonists and antagonists.

Page 4 of 5

This work was supported by the National Cancer Institute Innovative Molecular Analysis Technologies Program, CA182341-01 (J. N. W. and R. L.). J. N. W. also acknowledges seed funding from the American Cancer Society (98-277-07) R.L. also acknowledges the support of National Cancer Institute, CA98881-05 and the Braman Family Breast Cancer Institute.

REFERENCES (1) Tkaczuk, K. H. Clin. Ther. 2009, 31, 2273. (2) Traxler, P; Furet, P. Pharmacol. Ther. 1999, 82, 195. (3) Kleiman, L.B.; Maiwald, T.; Conzelmann, H.; Lauffenburger, D.A.; Sorger, P.K. Mol. Cell. 2011, 43, 723. (4) Rewcastle, G.W.; Denny, W.A.; Bridges, AJ; Zhou, H.; Cody, D.R.; McMichael, A.; Fry, D.W. J. Med. Chem. 1995, 38, 3482. (5) Kumar, A.; Petri, E.T.; Halmos, B.; Boggon, T.J. J. Clin. Oncol. 2008, 26, 1742. (6) Sinkeldam, R. W.; Greco, N. J.; Tor, Y. Chem. Rev. 2010, 110, 2579. (7) Ryan, T.A. Curr. Opin. Neurobiol., 2001, 11, 544-9 (8) Kapuscinski, J. Biotech. Histochem. 1995, 70, 220. (9) Gonçalves, M. S. T. Chem. Rev. 2009, 109, 190. (10) Liu, Y.; Gray, N. S. Nat. Chem. Biol. 2006, 2, 358-64. (11) Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W. Chem. Rev. 2003, 103, 3899. (12) Dhuguru, J.; Liu, W.; Gonalez, W. G.; Babinchak, W. M.; Miksovska, J.; Landgraf, R.; Wilson, J. N. J. Org. Chem. 2014, 79, 4940. (13) Wilson, J. N.; Liu, W.; Brown, A. S.; Landgraf. R. Org. Biomol. Chem. 2015, 13, 5006. (14) Wood, E.R.; Truesdale, A.T.; McDonald, O.B.; Yuan, D.; Hassell, A.; Dickerson, S.H.; Ellis, B.; Pennisi, C.; Horne, E.; Lackey, K.; Alligood, K.J.; Rusnak, D.W.; Gilmer, T.M.; Shewchuk, L. Cancer Res. 2004, 64, 6652. (15) Gerling, N.; Culmsee, C.; Klumpp, S.; Krieglstein, J. Neurochem. Int. 2004, 44, 505. (16) Kim, J.H.; Do, H.J.; Wang, W.H.; Macháty, Z.; Han, Y.M.; Day, B.N.; Prather, R.S. Biol. Reprod. 1999, 61, 900. (17) Reynolds AR, Tischer C, Verveer PJ, Rocks O, Bastiaens PI. Nat Cell Biol. 2003, 5, 447.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Synthesis, characterization data and spectra of 2-6, additional experimental details, including Table S1 and Figures S1-10.

AUTHOR INFORMATION Corresponding Author

*[email protected], [email protected] Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT

ACS Paragon Plus Environment

Page 5 of 5

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

Analytical Chemistry

For TOC only. F

N N

HN N

F N

N

HN

N

Cl N

N

unbound kinase probe

OFF

Cl

ERBB2 ERBB2-bound

ON

ACS Paragon Plus Environment