A Phenotypic Screening Approach Using Human ... - ACS Publications

Subscriber access provided by UNIV OF TEXAS DALLAS

Article

A phenotypic screening approach using human Treg cells identified regulators of forkhead box p3 expression Mei Ding, Johan Brengdahl, Madelene Lindqvist, Ulf Gehrmann, Elke Ericson, Stefan von Berg, Lena Ripa, and Rajneesh Malhotra ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.9b00075 • Publication Date (Web): 26 Feb 2019 Downloaded from http://pubs.acs.org on February 27, 2019

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 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 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.

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 25 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

ACS Chemical Biology

A phenotypic screening approach using human Treg cells identified regulators of forkhead box p3 expression

Mei Ding1 *, Johan Brengdahl1, Madelene Lindqvist2, Ulf Gehrmann3, Elke Ericson1, Stefan von Berg4, Lena Ripa4, Rajneesh Malhotra3 *

1

Discovery Sciences, IMED Biotech Unit, AstraZeneca, Gothenburg, 431 83 Mölndal,

Sweden, 2 Bioscience,

Respiratory, Inflammation and Autoimmunity, IMED Biotech Unit,

AstraZeneca; Gothenburg, 431 83 Mölndal, Sweden; 3 Target

& Translational Science, Respiratory, Inflammation and Autoimmunity, IMED

Biotech Unit, AstraZeneca, Gothenburg, 431 83 Mölndal, Sweden; 4 Medicinal

Chemistry, Respiratory, Inflammation and Autoimmunity, IMED Biotech Unit,

AstraZeneca, Gothenburg, 431 83 Mölndal, Sweden

*Corresponding authors: Mei Ding ([email protected]) Rajneesh Malhotra ([email protected])

Key words: phenotypic screen, human Treg, flow cytometry, FOXP3, CTLA4, target identification

1 ACS Paragon Plus Environment

ACS Chemical Biology 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

ABSTRACT Regulatory T (Treg) cells, expressing the transcription factor forkhead box p3 (FOXP3), are the key cells regulating peripheral autoreactive T lymphocytes by suppressing effector T cells. FOXP3+ Treg cells play essential roles controlling immune responses in autoimmune diseases and cancer. Several clinical approaches, e.g. polyclonal expansion of Treg cells with anti-CD3 and anti-CD28 coated beads in presence of drugs are under evaluation. However, expression of FOXP3, recognized as the master regulator of Treg cells, in induced Treg cells have been shown to be instable, and molecular targets involved in regulating FOXP3 expression and Treg cell function have not been well defined. Thus, new targets directly regulating FOXP3 expression and the expression of its downstream genes, e.g. cytotoxic T-lymphocyte-associated protein 4 (CTLA4), have the potential to stabilize the Treg cell phenotype and function. This report describes the development of an automated medium throughput 384-well plate flow cytometry phenotypic assay meauring the protein expression of FOXP3 and CTLA4 in human Treg cells. Screening a library of 4213 structurally diverse compounds allowed us to identify a variety of compounds regulating FOXP3 and CTLA4 expression. Further evaluation of these and related small molecules, followed by confirmation using siRNA-mediated gene knockdown, revealed three targets: euchromatic histone-lysine N-methyltransferase (EHMT2) and glycogen synthase kinase 3 alpha/beta (GSK3α/β) as potent positive regulators of FOXP3 expression, and bromodomain and extra-terminal domain (BET) inhibitors as negative regulators of FOXP3 and CTLA4 expression. These targets have potential implications for establishing novel therapies for autoimmune diseases and cancer.


Page 2 of 25

Page 3 of 25 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


INTRODUCTION Regulatory T (Treg) cells are a subset of T lymphocytes specialized in suppressing immune responses and thereby maintaining tolerance to self-antigens and homeostasis. Naturally occurring CD25+CD4+ Treg cells are antigen-primed and functionally mature in the thymus, constitutively express the transcription factor forkhead box p3 (FOXP3) and specialize in suppressing various effector lymphocytes, especially helper T (Th) cell subsets: Th1, Th2, Th17, and follicular Th (Tfh) cells 1, 2. Numerous studies have demonstrated the essential roles of FOXP3+ Treg cells in the control of a variety of physiological and pathological immune responses including inflammatory conditions, autoimmune diseases, and cancer

3-5.

Reduced

Treg cell number or deficiency in Treg function leading to down regulation of FOXP3 expression has been reported in a variety of human autoimmune diseases including systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA)6, 7. Furthermore, germline FOXP3 point mutations result in a fatal multiorgan autoimmune disease, termed scurfy disease in mice and immunodysregulation, polyendocrinopathy, enteropathy X-linked (IPEX) syndrome in humans8, 9. In addition, increased numbers of Treg cells within the tumor and tumor microenvironment has been described in many cancers including breast, lung, colon, and pancreatic cancer. The increased tumor Treg cells suppress tumor immunity and correlate with the progression of the disease and reduced survival or treatment response10. Therefore, modulation of Treg cell function could offer new approaches for treating autoimmune diseases and cancer. In addition to natural Treg cells, Treg cells can be differentiated from naïve conventional T (Tconv) cells under inflammatory conditions following stimulation with transforming growth factor beta (TGF-β) and interleukin-2 (IL-2)

11.

FOXP3-expressing Treg cells differentiated

from naïve Tconv exert suppressive activity and are refered to as induced Tregs. The transcription factor FOXP3 has long been considered as the ‘master regulator’ for controlling Treg cell development and function, and continuous expression of FOXP3 is required for maintenance of the Treg cell phenotype 4. Despite the recognized importance of FOXP3 expression for Treg cell function, molecular targets regulating FOXP3 expression and its downstream genes in human Treg cells are not well established. Recent studies have shown that Treg cells are unstable

12, 13.

Helios-deficient Tregs exert defective suppressive activity

accompanied by reduced FOXP3 expression in mice during inflammatory responses

12.

Deficiency of cytotoxic T-lymphocyte-associated protein 4 (CTLA4), a Treg cell marker that has been shown to be important for Treg suppressive activity 13, impairs suppressive function 3 ACS Paragon Plus Environment


Page 4 of 25

of mouse Treg in vivo and in vitro 13. In addition, several groups have reported that FOXP3 expression per se might not be sufficient for stably maintaining Treg cell phenotype and suppressive function. For example, neuropilin-1 (NRP1)-deficient Tregs were functional but fragile with reduced immuno-suppressive function but retained FOXP3 expression 14, TGF-βinduced Tregs expressed FOXP3 but showed poor suppressive ability and were unable to elicit the full Treg cell transcriptional signature 15, and transfecting human CD4+CD25- T cells with CTLA4 conferred suppression while cells transfected with FOXP3 were not suppressive

16.

These results suggest the involvement of multiple elements in regulating Treg phenotype stability and function. Several clinical approaches are under evaluation to modulate or expand ex vivo Treg cell function. Polyclonal expansion of Tregs with anti-CD3 and anti-CD28 coated beads can be achieved ex vivo in the presence of drugs (such as rapamycin and phosphatidylinositide 3-kinases (PI3K) inhibitors) 17. These strategies have provided protocols for expansion of Treg cells for clinical delivery. For example, low-dose IL-2 together with rapamycin to expand and stabilize Treg cells showed improved stability of the circulating Treg cells of diabetes mellitus type 1 patients, at least one-year post-treatment

18.

In this phase 1

clinical trial, despite a rise in circulating Tregs cells, a significant increase in effector T cells and in natural killer cells was observed, indicating loss of Treg phenotype. FOXP3 protein stability in human CD4+ T cells after T-cell activation is epigenetically regulated by DNA demethylation of conserved non-coding sequence intronic regions of the FOXP3 gene 19. Identifying other ways to regulate Treg cell FOXP3 and CTLA4 expression and hence Treg cell function have therapeutic potential for treating autoimmune diseases and cancer. For example, in patients with autoimmune disease, an increase in Treg cell FOXP3 or CTLA4 expression to suppress the immune system would likely benefit patients. On the contrary, a decrease in Treg cell FOXP3 and CTLA4 expression is desired in oncology to attenuate their suppressive activity and enhance the immune response to tumors. To identify novel regulators that stabilize (increase in FOXP3 or CTLA4 expression) or destabilize (decrease in FOXP3 and CTLA4 expression) the Treg cell phenotype, we developed a medium throughput 384-well plate phenotypic assay using primary human Treg cells in a high throughput flow cytometry system. A compound library consisting of 4213 small molecules including known epigenetic, kinase, and other target class modulators was screened for effects on the expression of FOXP3 and its downstream protein CTLA4. In this work, we describe the identification of positive regulators of FOXP3 expression (euchromatic histone-lysine N-methyltransferase 2 (EHMT2) inhibitor, glycogen synthase 4 ACS Paragon Plus Environment

Page 5 of 25 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


kinase 3 alpha/beta (GSK3α/β)) inhibitors) as well as negative regulators of FOXP3 expression (bromodomain and extra-terminal domain (BET) inhibiting compounds). This is the first report of the identification of FOXP3 expression positive and negative regulators in primary human Treg cells using a phenotypic screening approach measuring multiple readouts. The identified FOXP3 expression regulators have potential implications for establishing novel therapies for autoimmune diseases and cancer.

RESULTS and DISCUSSION Development of a medium-throughput flow cytometry Treg cell assay When considering how a proper Treg cell assay would be best set up, we reasoned that a phenotypic screening approach where physiologically relevant cells and biologically relevant in vitro assays are used to study phenotypic changes could be a fruitful approach for identifying novel targets potentially useful in treating human disease20. In our study, the most physiologically relevant cells to use would be human primary Treg cells. Since these cells are in suspension, we decided to use flow cytometry as the technology for our primary readout. Naïve Treg cells were isolated from fresh blood as described in the Methods in the supporting information (SI). Isolated naïve Treg cells (CD4+CD25highCD127lowCD45RO-) were expanded for 2 weeks. Expanded Treg cells were characterized by evaluating FOXP3 and CTLA4 protein expression as well as IL-2 production using flow cytometry-based methods. The high expression levels of FOXP3 and CTLA4 (Figure 1A) and the lack of IL-2 production (Figure 1B) in response to anti-CD3/Anti-CD28 stimulation confirmed a Treg cell phenotype. During the expansion phase, FOXP3 and CTLA4 expression in the Treg cells decreased. Continued reduction in FOXP3 and CTLA4 expression in Treg cells exposed to solvent only (0.1% DMSO) during day 14 to 17 was observed when comparing to cells at day 10 of expansion, indicating that the Treg cells were not stable over time in vitro. This was true for Treg cells from all the tested donors (SI, Figure S1). The observed decline in FOXP3 and CTLA4 expression provided a phenotypic change that could be exploited to screen for compounds stabilizing the Treg cell phenotype. Because the yield of naïve Treg cells was insufficient for screening large number of compounds using standard flow cytometry which requires using up to a few hundred thousand cells per test (we typically obtained 200,000 to 500,000 cells from 200 mL of human blood, and 2-10 million cells from one donor after the 2-week Treg cell expansion), we developed a Treg cell assay using a lower number of cells per test. The plate-based high-throughput flow cytometry 5 ACS Paragon Plus Environment


platform iQue Screener Plus (IntelliCyt, Albuquerque, NM) was used to set up a medium throughput flow cytometry Treg assay using a low number of cells per test. The Treg cell assay requires a cell staining procedure with repeated liquid addition and removal steps prior to sample acquisition on iQue Screener Plus. Because Treg cells are suspension cells, the cell staining procedure requires centrifugation to pellet cells prior to liquid removal which presents a high risk of a significant cell loss compared to assays using adherent cells. Utilization of an automated system for cell staining proved successful for minimizing cell loss. To better understand the effect of compound treatment on Treg cell stability, we se up a multiparameter assay measuring the effect on FOXP3 expression, CTLA4 expression and cell viability simultaneously. Our Treg cell assay has several advantages over standard flow cytometry assays: it uses as few as 4000 cells/well in the screen and Treg cells from a single donor can be used for screening 1000-1500 compounds. The assay was optimized with regard to Treg cell expansion time, compound treatment time, cell seeding density, cell staining parameters and iQue sampling parameters. In the optimized assay condition, we could generate a reproducible assay with the coefficient of variation (CV) in vehicle (DMSO)-treated cells ranging from 3% to 7% for CTLA4 MFI, and from 10% to 13% for FOXP3 MFI (across different experiment occassions using Tregs from different donors). This level of reproducibility was considered

sufficient to proceed to the compound screen, and the assay allowed us to identify small molecules regulating FOXP3 or/and CTLA4 expression (Figure 1C-D). The optimized assay protocol is described in detail in the Methods (SI) and involves a 2-week expansion of the isolated Treg cells prior to plating cells in 384-well assay plates (Figure 1E). Treg assay validation Another potential challenge using primary cells in a screen could be donor to donor variation. To explore this, we tested a subset of 290 compounds at 5µM using Treg cells from four donors. This compound set contained small molecules identified from the literature, including marketed drugs, clinical candidates and chemical probes for novel targets. A satisfactory correlation of compound effect on FOXP3 and CTLA4 expression in cells isolated from different donors was observed, with pairwise correlations of R2= 0.66-0.75, 0.75-0.86 for FOXP3 and CTLA4 respectively (Figure 2A-D). Besides being sufficiently reproducible across donors, the assay validation using this compound subset showed that small molecules that can increase or decrease FOXP3 or CTLA4 expression could be identified using the robust Z-score (Figure 2E) (The method for calculating robust Z-score 21 is described in the SI Methods under “Data and


Page 6 of 25

Page 7 of 25 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


statistical analysis”). The assay was robust with reproducible FOXP3 and CTLA4 MFI data in DMSO-treated cells (Figure 2F-G). Previously published data from Beier and coworkers demonstrated that both deletion of Sirturin 1 (SIRT1) and inhibition of SIRT1 resulted in increased FOXP3 expression in Treg cells, enhanced their suppressive function and prolonged allograft’s overall survival.22 To validate the assay, we knocked down the SIRT1 gene using siRNA in Treg cells. Knockdown of SIRT1 resulted in a 70% knockdown of SIRT1 mRNA (Figure 2H). As expected, we observed an increase in FOXP3 and CTLA4 protein expression in Treg cells isolated from two independent donors (Figure 2I-J). Taken together, we conclude that the assay developed here can be used as a reliable tool to identify FOXP3 and CTLA4-regulators in human Treg cells. Treg cell phenotypic screen of compounds and evaluation of actives and related compounds The Treg cell assay was used to screen a small molecule library of 4213 structurally diverse compounds (including the subset of 290 compounds used for Treg assay validation) which have an annotated mechanism of action (Figure 3) (the selection of this compound library is described in the SI Methods under “Compounds selected for Treg cell phenotypic screen”) at a concentration of 5 µM using 0.1% DMSO as solvent. The same concentration of DMSO (0.1%) was included on each plate as a control. The Treg cell iQue flow cytometry assay allowed for simultaneous measurement of multiple parameters, including cell viability and density, expression of FOXP3 and CTLA4, the percentage of FOXP3+ and CTLA4+ cells. The screen results were analysed using GeneData Screener software (GeneData, Basel, Switzerland). The effect of the compound was normalized to percent activity as compared to the MFI of live cells for the on-plate DMSO controls. Potential compound-mediated effects on cell viability were assessed using a fixable live-dead dye allowing counting the number of live cells per well. FOXP3 and CTLA4 median fluorescence intensity (MFI) of live cells was used for assessing FOXP3 and CTLA4 expression respectively. From the initial primary screen, we identified 310 actives defined as compounds causing an increase in FOXP3 expression based on ≥ 4x Robust Z-score of FOXP3 expression without negatively affecting cell viability (≥ 85% cell viability required). We also identified 352 compounds reducing FOXP3 expression based on < -3x Robust Z-score for FOXP3 expression and ≥ 85% cell viability. These initial actives were further evaluated in seven-point concentration response (CR) in the Treg screen using cells from two to three donors, as well as in a Jurkat cell compound fluorescence counter assay (an assay used for identifying fluorescent compounds that would yield a false positive in the Treg screen 7 ACS Paragon Plus Environment


Page 8 of 25

where no effect on FOXP3 or CTLA4 expression was expected. The Jurkat cell assay is described in the Methods under “Fluorescent compounds counter screen assay”). The CR screen confirmed actives as compounds that either increased or reduced FOXP3 expression in a concentration-dependent manner without showing any effect in the Jurkat cell counter screen. We separated the confirmed compounds into chemical structure clusters and as can be expected for such a small number of compounds, only a few small clusters and many singletons were identified. We then looked in our AstraZeneca compound collection for structurally related analogues to the identified clusters and singletons to expand the structural scope. These analogues, selected based on multiple fingerprint methods and substructure searches (structurally similar and structurally diverse compounds with similar biological activity), were subsequently tested in the Treg cell screen and Jurkat cell counter screen in an iterative manner. Confirmation of the suspected targets of the compounds using siRNA-mediated target knockdown enabled us to identify GSK3α/β and EHMT2 as positive regulators of FOXP3 expression, and BET as negative regulator of FOXP3 and CTLA4 expression. GSK3α and/or GSK3β: Compound-A

23

and LY317615

24

were confirmed as actives up-

regulating the expression of FOXP3 (Figure 4). Both compounds are known kinase inhibitors with activity on several kinases (data not shown), emphasizing the need to deconvolute what target is responsible for the observed increase in FOXP3 expression. Therefore, in the followup screen, we tested not only structurally similar compounds like SB216763 25, but also other structurally diverse kinase inhibitors with dissimilar kinase selectivity profiles (SI, Figure S2). In this way we hoped to be able to pinpoint the kinase responsible for the displayed phenotype. Among these structurally diverse known kinase inhibitors, Compound-B (Example 1 in the patent application WO 2011059388A1) 26, Compound-C 27, Compound-D 27 and Compound-E 28

all induced a significant increase in FOXP3 expression at concentrations that did not affect

cell viability as compared to the vehicle (Figure 4). Out of those, Compound-C and especially Compound-D have a reported specific kinase selectivity profile (SI, Figure S2), in principle being active only on GSK3α and GSK3β in comparison with ~400 other kinases. GSK3α activity is only reported for a small fraction of the compounds in our collection. Where the activity of a compound for both isoforms is known, they are in general in the same potency range. All of the compounds mentioned above show high binding affinity towards GSK3β with reported IC50 or Ki values of

A Phenotypic Screening Approach Using Human ... - ACS Publications

Recommend Documents