Areas of interest/types of work being carried out at the host institution
GSK researches, develops and manufactures innovative pharmaceutical medicines, vaccines and consumer healthcare products. Within the pharmaceuticals and vaccines businesses, they are focused on the delivery of medicines in six core areas: HIV and infectious diseases, oncology, immuno-inflammation, vaccines, respiratory and rare diseases.
The Screening, Profiling and Mechanistic Biology (SPMB) department to which I belonged is a subset of the Platform Technology and Science (PTS) group within pharmaceuticals research and development. PTS provides an end-to-end scientific and technical platform upon which GSK therapy area units (TAU) discover, test and develop new medicines for patients.
The therapy area units they work with are infectious diseases, respiratory, oncology and neurosciences. The infectious diseases TAU focus their discovery efforts primarily on four areas: HIV, hepatitis B (HBV), antibacterials and modulation of host defense responses to promote pathogen clearance and limit disease pathology. The respiratory TAU focuses on asthma, COPD and lung fibrosis and acute lung injury. In the oncology TAU they focus on answering questions pertaining to how epigenetic changes drive cancer development and progression, how agents targeting epigenetic pathways can be used to treat cancer, how to harness the body’s own immune system to attack cancer and finding drug combinations that may have the potential to reduce treatment resistance and help to provide durable activity. In neurosciences they work towards discovering and developing disease modifying medicines which slow or halt progression of neurodegenerative diseases, specifically Alzheimer’s disease and Parkinson’s disease.
PTScontributes to the work of the therapy areas to choose and validate targets by generating genetic evidence and greater understanding of cellular biology to recognise the links between targets and disease, and increasing proportion of targets with strong biological validity. PTS also contributes to the selection of the modality and the candidate by identifying and recommending optimal modalities for the disease state, and influence candidate selection based on understanding of efficacy, safety and developability. PTS provides an understanding of how the medicine interacts with biology to improve nonclinical to clinical translation of efficacy and safety, and predicting, measuring and minimising off-target interactions.
Within SPMB the main focus areas are assay development and compound profiling to meet the needs of project teams. They build assays to support hit identification (high throughput screens, focused set screens), lead optimisation (primary structure-activity relationship, selectivity, ortholog), developability and mechanistic profiling. They work with both cellular and biochemical assays. Cellular assays are typically with recombinant mammalian cells (mostly cryo-preserved), but native cells (e.g. astrocytes, blood cells) and yeast (proliferation) are also used routinely. They use a wide breadth of detection technologies including reporter genes (e.g. luciferase, β-lactamase), intracellular Ca2+ measurements (FLIPR, Lumilux aequorin) and homogeneous antibody-based systems (e.g. AlphaLisa). Biochemical assays make use of IMAP, FP, TR-FRET, FLINT, FRET, SPA, MSD and AlphaLisa detection methods for enzymes, kinases and integrins. Assays are optimised for biological relevance, robustness (reproducibility and precision), ease of use, throughput and cost. Once assays have been developed, they can be used for compound profiling. Compound Profiling is the measurement of compound effects, by means of in vitro assays, to develop structure-activity relationships for target potency, to measure efficacy, or to minimise selectivity issues. At any one time over 300 separate target assays are being run to support program teams. These targets cover all therapeutic areas and are derived from all organs. Where relevant, primary targets may also have assays available for the key orthologues (usually mouse or rat), typically to support in vivo compound testing in disease or toxicology models.
Optimisation of cell-based assays targeting the JAK signalling family implicated in inflammatory diseases
Inflammatory diseases are caused by a combination of factors including disturbances in the inflammatory response of the immune system. One of the pathways involved in the inflammatory response is the JAK X pathway required for the cytokine signalling of IL12 and IL23 in T lymphocytes and natural killer cells. Impairment of this pathway results in over-activation of the adaptive immune system leading to excessive inflammation, a characteristic of inflammatory diseases. This makes the JAK X pathway a potential therapeutic target for these diseases. JAK X is 1 of 4 highly homologous kinases whichselectively phosphorylate STAT proteins. The aim of this project was to optimise assays that would be used to test small molecules and find inhibitors of the JAK X pathway, without implicating other related pathways. MSD technology was used in all cases, the principle of which is similar to that of an ELISA. The use of LanthaScreen TR-FRET for detection of phosphorylated STAT proteins was also assessed with no success. Efforts were made to optimise two cellular assays and one phenotypic assay. All successful assays had 0≤Z’≤1 and at least 9-fold signal window hence, were put into production.
1.1 The immune response, inflammation and inflammatory diseases.
Inflammation and its characteristics
Inflammation is the general term used to describe the local accumulation of fluid, plasma proteins and leucocytes in a tissue. It is initiated by physical injury, infection or a local immune response. When pathogens, such as bacteria, enter a tissue, they are recognised by resident macrophages that activate the rest of the innate immune response system leading to a cascade of reactions.1 Macrophages express receptors for many bacterial constituents. The binding of bacteria to macrophage receptors initiates the release of pro-inflammatory cytokines (interleukin (IL)-1, IL6 and tumor necrosis factor alpha (TNF-α)) and chemokines. The macrophages also take in the infectious agent into a phagosome which then fuses with lysosomes to form a phagolysosome that kills the pathogen (phagocytosis).2,3 The chemokines released act as signals for other inflammatory cells which migrate to the infection area by chemotaxis. TNF-α acts on vascular endothelial cells of the surrounding capillaries to increase vasodilation as well as vascular permeability. Inflammatory cells extravasate into the infected tissue, perform phagocytosis and release more cytokines. The adaptive immune response is initiated when dendritic cells ingest antigens and migrate to a local lymph node to activate resting T cells.3Phagocytosis continues until the pathogens are eradicated. In normal cases, this leads to the resolution of inflammation by the suppression of pro-inflammatory gene expression, leukocyte migration and activation, followed by inflammatory-cell clearance by apoptosis and phagocytosis.4
Inflammation is characterised by swelling, erythema (reddening of the skin), heat and pain. These lead to a general feeling of sickness, exhaustion and fever.5As vascular permeability of the infected site increases, blood, cells and proteins leak from the surrounding blood vessels causing the tissue to swell1,6. The increased blood volume heats the tissue and causes erythema. TNF-α, IL1 and IL6 also induce fever – elevated body temperature inhibits replication of certain pathogens. The pain experienced is due to tissue expansion that causes mechanical pressure on nerve cells, as well as the presence of pain mediators released by the inflammatory cells.1-3
At times, the body fails to recognise signals to cease the inflammatory response and/or the immune system mistakes the body’s cells as foreign. In these cases, the prolonged inflammation can lead to the pathogenesis of various inflammatory diseases. These include rheumatoid arthritis (RA), pelvic inflammatory disease (PID), psoriasis, gout and Crohn’s disease5 to name a few. Inappropriate inflammatory response can also lead to a loss of tissue or organ failure as seen in chronic bronchitis, asthma and emphysema4. Inflammatory diseases can last for several years or even a lifetime, occurring in varying degrees of severity and activity5, and all have different prevalence. For example, in 2014 4.4% of sexually experienced women of reproductive age (18–44 years) were living with PID in USA7. In 2011, in UK, 417 per 200,000 females within the same age range were diagnosed with PID8. Approximately 0.8% of the UK population aged over 16 years have RA9 and 2.49% have gout10.
Several factors contribute to the malfunction of the inflammatory response that leads to inflammatory diseases. These include microbiota dysbiosis, alterations and disturbances in the immune response pathways, genetic variations, lifestyle and environmental factors. Many of the inflammatory diseases that have increased in incidence in recent years have been associated with microbiota dysbiosis, mainly intestinal dysbiosis.11 Intestinal homeostasis depends on complex interactions between the microbiota, the intestinal epithelium and the host immune system. Diverse regulatory mechanisms cooperate to maintain this homeostasis. The general hypothesis is that inflammatory diseases develop as a result of persistent, inappropriate perturbation of this complex interaction, resulting in dysbiosis 1213and mucosal inflammation. The altered microbiota can be sensed by the host and lead to an inappropriate activation of the immune system11. It has been shown through genetic knockout studies that Janus Kinases (JAKs) and signal transducers and activators of transcription (STAT) proteins have various specific roles in the regulation of the immune inflammatory response14. Impairment of this system results in over-activation of the adaptive immune system leading to excessive pro-inflammatory cytokine production derived from CD4+ T cells over and above the response normally associated with tolerance and immunoregulation15,16.
A large-scale genetic association study carried out by, Cargill et al showed that IL12B and IL23R are risk genes for psoriasis17. Casanova et al have also reviewed the inborn errors of JAKs and STATs in humans18 that might lead to inflammatory diseases. Lifestyle factors such as excessive alcohol consumption, poor diet and obesity have been marked as major risk factors for gout10. In one particular study, Fagundes et al investigated whether depressive symptoms interacted with acute stress to induce pro-inflammatory cytokine production. They found that individuals with more depressive symptoms had larger stress induced increases in IL6 to a standardised laboratory speech and mental arithmetic stressor. It was concluded that people with more depressive symptoms show enhanced inflammation in response to stress compared with those with fewer depressive symptoms. Accordingly, these depressive symptoms are more likely to enhance the stress response system in ways that promote excessive inflammation.19
Pharmaceutical treatment of inflammatory diseases currently includes five major categories – anti-inflammatory drugs, immunosuppressants, biological agents, antibiotics and drugs for symptomatic relief20.Some of these drugs although useful, come with side effects. For example, the pro-inflammatory cytokine TNF-α has been identified as playing a pivotal role in the inflammatory cascade that causes chronic inflammatory diseases 21.One of the available treatments isTNF-α antagonist22.These drugs amongst others have been found to increase the susceptibility to certain cancers, such as colorectal and skin cancers 23,24.The most popular anti-inflammatory agents currently used are glucocorticosteroids25. Non-steroidal anti-inflammatory drugs are also a common therapy option26.
1.2 The JAK/STAT pathway
The JAK family which consists of 4 non-receptor tyrosine kinases whichselectively phosphorylate STAT proteins. JAKs have 7 regions of high homology 27and the function of some of these have been identified (fig. 1A). JH1 encodes the kinase. JH2 is a pseudokinase domain, which is required for JH1 catalytic activity.28 JH3–JH7 are thought to be involved in receptor association.29,30JAKs are associated with membrane receptors on the intracellular side of the cell.
Figure 1. Structure of Janus Kinases and Signal Transducers and Activators of Transcription. JAKs share seven regions of high homology, JH1–JH7. JH1 encodes the kinase. JH2 represents a pseudokinase domain. JH3–JH7 are involved in receptor association. STATs also share several conserved domains, including an amino-terminal domain, a coiled-coil domain, the DNA binding domain, a linker domain, an SH2 domain, and a tyrosine activation domain (TAD, circled P). The carboxy-terminal transcriptional activation domain is conserved in function but not in sequence.30
There are 7 STAT proteins (STAT1-6, including STAT5a and STAT5b), all of which share some functionally and structurally conserved domains. This includes the amino-terminal domain, the coiled-coiled domain, the DNA binding domain, the linker domain and the SRC homology 2 (SH2; tyrosine activation) domain (fig. 1B). The carboxy-terminal represents the transcriptional activation domain and differs amongst the proteins, contributing to their specificity.30,31 Specific cytokines activate specific kinases. Different kinase combinations activate different STAT proteins.
When a cytokine binds to its receptor, it induces a conformational change which in turn leads to the associated JAKs to come into close proximity with each other. This leads to kinase activation by phosphorylation, which in turn leads to phosphorylation of specific tyrosine residues on the receptors. STATs and other molecules that recognise these phosphorylated residues are recruited to the site and are activated by phosphorylation. Activated STATs dissociate, dimerise and translocate to the nucleus31,32, directed by a nuclear localisation signal 30. Once in the nucleus the phosphorylated STATs alter the expression of their target genes, thereby regulating the innate and adaptive host immune response.33 The JAK X pathway is one of the several pathways activated by numerous cytokines in the body. It is required for the cytokine signalling of IL12 and IL23 in T lymphocytes and natural killer (NK) cells. These are implicated in the pathology of some inflammatory diseases such as psoriasis. JAK X is therefore a potential therapeutic target for these diseases.31,34
1.3 Drug discovery process and project aims
Figure 2. Drug discovery flow diagram. The drug discovery process can be split into phases. 1)Target ID & validation – research leads to identification of a potential targetand studies conducted in cells, tissues and animal models to determine whether the target can be influenced by a medicine. 2)Hit identification – identifycompounds that have the desired activity in a compound screen and whose activity is confirmed upon retesting. 3) Lead generation and optimisation – refinement of hit series in order to improve compound potency and selectivity. 4) Pre-clinical phase – testing in animals to determine if the drug is suitable for human testing. 5) Clinical trials – testing in human volunteers done in three phases. 6) Regulatory approval – appropriate regulatory body reviews complete data sets and grants approval for drug marketing.35
The drug discovery process takes about 10-15 years from basic research to a new drug that can be administered to patients, and can be split into six stages (fig. 2). Basic research leads the identification of a possible drug target, which can be a gene or protein implicated in a specific disease or condition. Target identification involves searching through large amounts of data to select and prioritise possible drug targets. Approaches used include the use of proteomics and genomics technologies, as well as bioinformatics approaches. The role of the target is then clearly defined. This can be achieved through in vitro (e.g. using antisense oligonucleotides) or in vivo testing (e.g. using transgenic animals).36 The next stage (hit identification) sees the development of screening assays used to look at the effect of compounds at the target of interest. This leads to high throughput screening (HTS) of approximately two million compounds in a biochemical format. Identified hits are optimised through medicinal chemistry (structure-activity relationship – the modification of a molecule’s structure to try and improve its activity). During lead identification, the aim is to find a chemical structure or series of structures that demonstrate activity and selectivity in a pharmacological or biochemically relevant screen. This forms the basis for a focused medicinal chemistry effort for lead optimisation. About 250 compounds make it to the next stage where toxicology and safety are examined in animal models. The clinical trials that follow are split into three stages of drug testing in humans. Approximately five compounds make it to this stage. In phase 1, drug safety is assessed in healthy volunteers (20-80 people). Phase 2 assesses drug efficacy in a small number of patients (100-300 patients). In phase 3 a larger cohort of patients (500-3000 patients) is used to further assess efficacy and adverse events. A request is then sent to the appropriate regulating body for approval to market the new drug (Medicines and Healthcare Products Regulatory Agency for marketing in UK, and Food and Drug Administration for USA).35
The work discussed in this report falls within the lead generation and optimisation stage of drug discovery. The aim of this project was to optimise cell-based assays that would be used to test small molecules and find inhibitor compounds of the JAK X pathway without implicating other related pathways. These assays will be used to validate JAK X hits from the HTS run prior to this. This report discusses efforts to optimise two cellular assays – JAK X/JAK Y pathway stimulation by IL12, producing phosphorylated STAT4 (pSTAT4); JAK Y pathway stimulation by erythropoietin (EPO), producing phosphorylated STAT5 (pSTAT5); and one phenotypic assay – JAK Y/JAK Z pathway stimulation by IL12 and IL18, producing IFN-γ. Any compounds active in the pSTAT4 assay but not the pSTAT5 assay are taken to be JAK X specific. The phenotypic assay differs from the cellular assays in that it uses a disease relevant cell line. This helps in further understanding the target.
IL12 induced pSTAT4
IL12 is a heterodimeric cytokine produced mostly by phagocytic cells in response to bacteria, bacterial products, intracellular parasites and to some degree, by B lymphocytes. IL12 elicits its response through the activation of STAT430,37via JAK X and JAK Y 30,33. IL12 acts primarily at 3 stages during the innate and adaptive immune response to infection: 1) early on in the infection, it is produced and induces IFN-γ production from NK and T cells, which contributes to phagocytic cell activation and inflammation; 2) IL12 and IL12-induced IFN-γ favour TH1 cell differentiation by priming naive CD4+ T cells for high IFN-γ production;38,39 and 3) It contributes to optimal IFN-γ production and to proliferation of differentiated TH1 cells in response to antigens 39-41. An existing 96w cellular Meso Scale Discovery assay (MSD) will be miniaturised and validated to determine if it is fit for purpose. A large number of compounds were anticipated to be screened through this assay. The purpose of miniaturising the assay was therefore, to increase the number of compounds tested per run (without increasing the number of plates), thereby reducing costs and time taken to run the assay.
EPO induced pSTAT5
EPO is a glycoprotein hormone that has also been recognised as a member of the cytokine type 1 superfamily 30. It is synthesised predominantly in the kidney and secreted by renal cortical interstitial cells. It promotes the survival, proliferation and differentiation of erythroid progenitor cells. EPO participates in a classic feedback control system, as its production is regulated by impaired oxygen delivery to the kidney. Hypoxia causes an increase in EPO gene transcription, enhancing erythropoiesis.42,43 In erythroid progenitor cells, the response to EPO is initiated upon binding of picomolar concentrations of this cytokine to the receptor homodimers. This enables the activation of STAT5 via JAK Y pathway.30,44An existing 96w MSD assay will be miniaturised to a 384w LanthaScreen TR-FRET (Time-Resolved Fluorescence Resonance Energy Transfer) assay and validated to determine if it is fit for purpose. As with the pSTAT4, a large number of compounds ( ̴2000) were anticipated to be screened through this assay hence, the need for miniaturisation.
IL12/IL18 induced IFN-γ production
IL18 is secreted from activated macrophages and other cells39as an inactive precursor molecule that becomes functional following cleavage by caspase-1.45,46 This molecule, originally identified as IFN-γ inducing factor, is produced by splenocytes, liver lymphocytes and TH1 cells. IL18 enhances NK cell activity and proliferation of activated T cells.39,47 IL12 and IL18 are thought to work in a synergistic manner whereIL12 up-regulates IL18 receptor expression on T cells, TH1 cells and B cells.48It is speculated that naive T helper cells, when primed with IL12, express the IL18 receptor and then develop into fully differentiated TH1 cells in response to both cytokines.47 T cells stimulated with IL12 and IL18 produce IFN-γ without the need of T cell receptor stimulation.48,49 Upon cytokine ligation with receptor, JAK Y and JAK Z get phosphorylated, resulting in STAT1 activation and IFN-γ production.48The assay discussed in this report used peripheral blood mononuclear cells (PBMC), which contain a portion of IFN-γ-producing CD4+ T cells. An existing 96w MSD assay will be validated to determine if it is fit for purpose.
1.4 Technology employed
Most of the assays described in this paper employ Meso Scale Discovery technology. The principle of an MSD assay is similar to that of an enzyme-linked immunosorbent assay (ELISA). Figure 3 shows how the assay works and the different formats available. In the assays discussed in this report SULFO-TAG directly conjugated to the detection antibody was used, as shown in figure 3A. MSD provides Multi-Spot plates precoated with capture antibody on each spot of the well. Each well can have up to 10 spots each coated with a different capture antibody. In the assays discussed here, only one analyte was being assessed, so only one spot had the antibody and the rest were pre-blocked with bovine serum albumin (BSA). The sample lysate and a solution containing detection antibody conjugated with an electrochemiluminescence (ECL) label (MSD SULFO-TAGTM) are added to the capture antibody over the course of two incubation periods. Analytes in the sample bind to the capture antibody immobilised on the electrode surface. The detection antibodies are then recruited by the bound analytes to complete the sandwich (fig.3A). An MSD read buffer is added to provide the appropriate chemical environment for ECL. Once inside the MSD plate reader (Sector imager) a voltage is applied to the plate electrode, causing the SULFO-TAG, which is close to the surface of the well, to emit light through a series of reduction and oxidation reactions (fig. 3D). The imager measures the intensity of emitted light to provide a quantitative measure of analytes in the sample. 50
Figure 3. MSD immunoassay formats using capture antibodies on MULTI-ARRAY and MULTI-SPOT plates, and sample detection. A: MSD SULFO-TAG is directly conjugated to the detection antibody. B: Biotinylated detection antibody binds to SULFO-TAG Streptavidin. C: Detection antibody binds to SULFO-TAG-conjugated anti-species antibody. D: A top-view image of a well in a MSD MULTI-ARRAY plate depicting the electrochemiluminescence reaction. 50
LanthaScreen TR-FRET technology was assessed for use in one of the assays. In these cellular assays specific target proteins are expressed as fusions with green fluorescent protein (GFP) (or fluorescein), a suitable TR-FRET acceptor for the excited fluorophore. Invitrogen’s LanthaScreen™ TR-FRET kinase assay platform uses a long lifetime terbium (Tb) chelate fluorophore as a donor species. After the cells are lysed, the terbium and GFP (or fluorescein) labelled molecules (antibodies and stimulus-induced phosphorylated proteins, respectively) are brought into close proximity, energy transfer takes place causing an increase in acceptor (GFP) fluorescence and a decrease in donor (Tb) fluorescence (fig. 4). These fluorescent signals can be read in a time-resolved manner to reduce assay interference and increase data quality. The resulting TR-FRET value is a dimensionless number that is calculated as the ratio of the acceptor (GFP) signal to the donor (Tb) signal. The amount of antibody that is bound to the tracer is directly proportional to the amount of phosphorylated substrate present. In this manner, kinase activity can be detected and measured by an increase in the TR-FRET value.51
Figure 4. Schematic of the LanthaScreen™ TR-FRET kinase activity assay platform. A fluorescein (or green fluorescent protein)-labelled substrate peptide is incubated with kinase and ATP. Terbium-labelled antibody is then added and phosphorylation detected by an increase in the TR-FRET ratio (acceptor:donor signal).51Similar principles are used in the cellular assays.
2.0 Materials and Methods
Where reagents have been produced by GSK’s Protein, Cellular and Structural Sciences (PCSS) department, a GSK biological catalogue (BioCat) and/or global reagent inventory and tracking system (GRITS) number has been quoted.
IL12 induced pSTAT4
A 1ml vial of cryopreserved human natural killer lymphoma (NK92) cells (BioCat, # 114669; GRITS, # N38492-7-1 and N38492-9-1) was resuspended in 20ml assay medium. The assay medium contained 10ng/ml IL2 (Thermofisher, #PHC0026) in alpha minimum essential medium (MEM) without nucleosides (Invitrogen, # 12561-056). Cells were centrifugated twice for 5 minutes at 1300rpm. The supernatant was removed and the cells resuspended in 20ml medium after each centrifugation. A cell count was performed on a Vicell cell counter (Beckman Coulter) and cells adjusted to desired densities in assay medium.
Cell concentration and stimulant optimisation
50µl of cell suspension was added to each well of a 384well clear V bottom plate (compound plate; Greiner, # 781280) at various cell densities. 8 point dose-response curves (DRCs; relationship between the size of dose and the response to it) were generated for IL12 (R&D Systems, # 219 IL-025) ranging from 100ng/ml to 0ng/ml, diluted in 100% dimethyl sulfoxide (DMSO). 10µl/w of the dose-response curves was transferred to the compound plate to stimulate the cells. Cells were stimulated at various time points between 0 and 60 minutes, and incubated at 37oC, 5% CO2.
EPO induced pSTAT5
Cell culture and starvation
MSD: Erythroleukemic cells from human bone marrow (TF-1 cells; BioCat, #138889) were passaged twice a week (Monday and Thursday) and seeded at 9×106 cells in 100ml culture medium. The assay medium contained 10% Australian fetal bovine serum (FBS) (Invitrogen, # 10099-141), 1mM sodium pyruvate (Invitrogen, # 11360-039), 2mM L-GlutaMAX (Invitrogen, # 35050-038), 100U/ml penicillin and 100μg/ml streptomycin (Invitrogen, # 15140-122), and 2ng/ml GM-CSF (Invitrogen, # PHC2015). This was added to Roswell Park Memorial Institute (RPMI)-1640 culture medium (Invitrogen, # 31870-025). GM-CSF was added on the day of passage. Cells were incubated in T175 tissue culture flasks (Fisher Scientific, Nunc, # TKT-130-220Q) at 37oC, 5% CO2.
LanthaScreen: GFP-tagged TF-1 cells were passaged in the same way as mentioned above. The assay medium contained 10% dialysed FBS (Invitrogen, # 26400-044), 1mM sodium pyruvate, 0.1mM non-essential amino acids (NEAA; Invitrogen, # 11140-050), 100U/ml penicillin, 100μg/ml streptomycin, 2ng/ml GM-CSF and 5µg/ml blasticidin (Invitrogen, # R210-01). This was added to RPMI-1640 culture medium with glutaMAX (Invitrogen, # 72400-047). GM-CSF and blasticidin were added on the day of passage. Cells were incubated in T175 tissue culture flasks at 37oC, 5% CO2.
Cells were transferred to falcon tubes and centrifugated at 300g for 5 minutes. The supernatant was removed and cells resuspended in assay medium to a concentration of ̴0.5‐1×106 cells/ml. The assay medium used in the MSD assay was the same as the culture medium with the absence of FBS and GM-CSF. The assay medium used in the LanthaScreen assay was Opti-MEM®I reduced serum without phenol red (Invitrogen, # 11058-021) supplemented with the same reagents as the culture medium with the exception of FBS, GM-CSF and blasticidin. Cells were then transferred to T175 tissue culture flasks and incubated overnight at 37oC, 5% CO2.
Cell concentration and stimulant optimisation
MSD: On day 2, the serum starved cells were spun at 300g for 5 minutes and resuspended in assay medium. A cell count was performed and the concentration was adjusted as required. 80µl of cell suspension was added to each well of a 96well clear V bottom plate (compound plate; Greiner, # 651201) at different cell densities. 8 point dose-response curves were generated for EPO (Cell Signaling Technology, # 6980LF) ranging from 2.7µg/ml to 0µg/ml, diluted in 100% DMSO. 20µl/w of the dose-response curves was transferred to the compound plate to stimulate the cells. Cells were stimulated at various time points between 0 and 50 minutes, and incubated at 37oC, 5% CO2.
LanthaScreen: On day 2, the serum starved cells were spun at 300g for 5 minutes and resuspended in assay medium to a concentration of 3.125×106c/ml. 16μl of cell suspension was added to each well of a 384well white polystyrene MaxiSorp plate (compound plate; Nunc, # 6980LF) (5×104c/w). 8 point dose-response curves were generated for EPO, ranging from 1µg/ml to 0 µg/ml, diluted 100% DMSO. 4µl/w of the dose-response curves was transferred to the compound plate to stimulate the cells. Cells were stimulated at various time points between 0 and 60 minutes, and incubated at 37oC, 5% CO2. 1% protease inhibitor mix (SIGMA-Aldrich, # P8340), 1% phosphatase inhibitor mix (SIGMA-Aldrich, # P2850), 10nM anti-mouse antibody (Invitrogen, # PV3766) and 5nM anti-pSTAT5 antibody (Invitrogen, # PV5262) were added to LanthaScreen™ Tcellular assay lysis buffer (Thermo Fisher, # PV5598) to make the complete lysis buffer. 15µl/w of the buffer was added to the compound plate. The plate was incubated at room temperature (RT) for 2 hours before being read on a BMG Labtech PHERASTAR plate reader.
IL12/IL18 induced IFN-γ production
Cell plating and stimulation
Human peripheral blood was drawn from healthy volunteers by GSK Stevenage blood donation units. Blood was collected in sodium heparin (1ml/100ml of blood). Peripheral blood mononuclear cells (PBMCs) were isolated and cryopreserved in 90% serum, 10% DMSO by the PCSS department. A 1ml vial of cryopreserved PBMCs (BioCat and GRITS # various) was resuspended in 10ml assay medium. The assay medium contained 2mM L-glutamine (Invitrogen, # 25030-024), 0.1mM NEAA, 1mM sodium pyruvate, 0.05mM 2-mercaptoethanol (Invitrogen, # 31350-010), 100U/ml penicillin and 100μg/ml streptomycin in Iscove’s Modified Dulbecco’s Medium (IMDM) containing 25mM HEPES (Invitrogen, # 21980-032). Cells from two donors were pooled and centrifugated twice for 10 minutes at 1000rpm as described previously. A cell count was performed and cells were adjusted to 1×106 c/ml in assay medium. 100μl of cell suspension was added to each well of a 96well clear V bottom plate (compound plate) (1×105c/w). Cells were stimulated with 10ng/ml IL12 and 1ng/ml IL18 (Medical & Biological Laboratories, # B001-5) added as a 20μl/w solution. The stimulated cells were incubated for 48 hours at 37oC, 5% CO2.
In all assays cells were incubated for 1 hour at 37oC, 5% CO2 after plating. Cytokine stock solutions were made in phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA). All media was pre-warmed to 37oC before use. Compound plates for optimising conditions typically contained 0.5µl/w 100% DMSO initially. Pharmacology plates were pre-stamped with 0.5µl 10mM compounds diluted 1 in 3 in 100% DMSO across 10 points (96w plate) or 11 points (384w plate). 0.5µl 100% DMSO was used as the negative control and 0.5µl 10mM (3.3mM for pSTAT5) known inhibitor as the positive control. Pre-stamped plates are plates supplied by GSK’s Discovery Supply department, containing reagents added automatically, as desired by the scientist.
2.2 Titer detection using MSD kits
Unless otherwise stated, the named reagents were supplied by MSD as part of the assay kit.
IL12 induced pSTAT4 (MSD, # L21CA-1)
After cell stimulation, the compound plate was centrifugated for 2 minutes at 2000rpm and the supernatant aspirated from the cells using a Biomek FX machine (Beckman Coulter). 1X complete lysis buffer was made up using 100X phosphatase inhibitor 1, 100X phosphatase inhibitor II and 100X protease inhibitor solution added to 1X Tris Lysis buffer. 25µl of this solution was added to each well of the compound plate. Plates were then incubated for 30 minutes at 4oC. 10X Tris wash buffer was diluted to 1X in MilliQ water (MQH2O) and used to make 3% blocker A solution from blocker A powder. 35μl of the solution was added to each well of a 384well MSD plate and incubated for an hour. 10µl of cell lysate was transferred from the compound plate to the pre-blocked MSD plate using a Biomek FX machine. The MSD plate was then incubated for 2 hours. 50X SULFO-TAG anti- total pSTAT4 (detection) antibody was diluted to 1X in a solution made up of 3/8 blocker A solution and 5/8 1X Tris wash buffer. 10µl/w of the 1X detection antibody was added to the MSD plate and incubated for a further 2 hours. 4X read buffer was diluted to 1X in MQH2O and 35µl/w added to the MSD plate.
EPO induced pSTAT5(MSD, # K150IGD-3)
After cell stimulation, 30µl/w of 4X complete lysis buffer (made up as previously described) was added and the plate incubated for 30 minutes at 4oC. 150μl of 3% blocker A solution was added to each well of a 96well MSD plate and incubated for an hour. 25µl of cell lysate was transferred from the compound plate to the pre-blocked MSD plate using a Biomek FX machine, and incubated for 1 hour. 50X SULFO-TAG anti-total pSTAT5 antibody was diluted to 1X in a solution made up of 1/3 blocker A solution and 2/3 1X Tris wash buffer. 25µl/w of the 1X detection antibody was added to the MSD plate and incubated for another hour, followed by addition of 150µl/w 1X read buffer.
IL12/IL18 induced IFN-γ production (MSD, # K151AEB-2)
1% blocker B solution was made up in PBS without Ca2+ or Mg2+ (Sigma-Aldrich, # D8537). 150μl of the solution was added to each well of a 96well MSD plate and incubated for an hour. 25µl of cell lysate was transferred from the compound plate to the pre-blocked MSD plate using a Biomek FX machine, and incubated for 2 hours. Diluent 100 was used to make 1X detection antibody, which was added to the MSD plate as 25µl/w. This was incubated for another 2 hours followed by the addition of 150µl/w 2X read buffer.
All MSD plate incubations were at RT, with shaking at 300-1000rpmon an orbital shaker. After each incubation period, MSD plates were washed three times with 1X Tris wash buffer (70µl/w pSTAT4, 150µl/w pSTAT5). IFN-γ plates were washed with 150µl/w PBS containing 0.05% v/v Tween 20 (Sigma-Aldrich, # T2700). Plates were read on the Sector Imager 6000 MSD plate reader after the addition of the read buffer. The methods above were consistent with the manufacturer’s protocol, with the exception of the lysis buffer concentration used in the pSTAT5 assay. Therefore, only the biological aspects of the assays required optimisation. In the pSTAT4 and pSTAT5 assays, the MSD detection step could be carried out straight after lysis, or lysates could be frozen at -80oC. Frozen plates were thawed at 4oC on the day of detection (1.5hrs for pSTAT4 and 1hr for pSTAT5). Human biological samples were sourced ethically; their research use was in accordance with the terms of the informed consents, and were handled in accordance with the Human Tissue Act.
Cell Titer Glo
The cell viability assay was run in parallel with the IFN-γ MSD step, using the stimulated PBMC supernatant. 40μl/w of the supernatant was transferred from the compound plate to a 96well white polypropylene plate (Greiner, # 655207) using a Biomek FX machine. The CellTiter-Glo® reagent was made by reconstituting 1 vial of lyophilized CellTiter-Glo® substrate in 100ml of CellTiter-Glo® buffer (Promega, # G7573). 40μl/w of the CellTiter-Glo® reagent was added and the plate incubated with shaking at RT for 10 minutes. The plate was then read on a Perkin Elmer Wallac EnVision 2104 Multilabel plate reader.
2.3 Data analysis
Results from the initial experiments to identify the optimal assay conditions were analysed manually in Microsoft Excel using Excel tools and GSK’s proprietary curve fitting (XC50) module. The XC50 module uses a four parameter logistical fit model to plot curves that best fit the data (equation 1). Full curve compound data was analysed automatically in ActivityBase (ABASE) XE (which also uses the XC50 module) using equation 1. This equation was used to calculate IC50, which is the concentration of the compound that gives 50% inhibition. In ABASE, each individual raw data point was normalised to the control data using equation 2. The 0% response/negative control was calculated from the DMSO data (column 6 of a 384w plate or column 11 of a 96w plate). The 100% response/positive control was calculated from the known inhibitor data (column 18 of a 384w plate or column 12 of a 96w plate). When fitting curves, points were removed where scientifically justified. Data points were left unfit or connected by straight lines where curves were unable to be drawn. The Excel templates and ABASE XE were used to determine plate quality by generating a Z prime statistic (Z’) using equation 3. The calculation depends on a normal distribution of the negative and positive controls being within three standard deviations of their respective means. A high Z’ value is obtained when there is a large distinction between the mean of the two controls. This suggests that there is a significant difference between the positive and negative controls signifying that the data quality is high.52 Plates with a 0˃Z’˃1 did not meet the required quality criteria and the data was rejected from further analysis. Both Microsoft Excel templates and ABASE XE also used equation 4 to calculate signal to background (S:B/signal window). S:B is used as another measure of assay quality, only taking into account the mean values of the controls. The higher the value generated, the larger the difference between the two controls, indicating high assay quality. Where there was only one set of control data, equation 5 was used to calculate the coefficient of variation (%CV) to determine the plate quality. %CV is a precision measurement used to assess the variation of data across a plate. A low %CV indicates high precision; an acceptable %CV is generally below 10%. A correlation coefficient was calculated in Microsoft Excel using equation 6. The calculated value indicates how well two sets of data correlate. This value between 0 and 1 can be thought of as the proportion of variability explained by the model. The correlation of two data sets is considered excellent if R2>0.81, good if 0.64≤R2≤0.81 and poor if R2<0.64.53
Equation 1. Full four parameter logistical fit – where A is the minimum y signal, B is the maximum y signal, C is the slope factor, x is the log10 molar compound concentration and y is the chemiluminescence or fluorescence signal.44
Equation 2. Response data normalisation – Where N is the 0% response control and P is the 100% inhibition control.44
Equation 4. Signal to background calculation – Where µP is mean of the 100% inhibition control and µN is the mean of the 0% response control.43
Equation 3. Z’ calculation – Where σ is the standard deviation, N is the 0% response control and P is the 100% inhibition control.44
Equation 6. Linear regression fitting – Where R2 is the coefficient of determination, SStotal is the sum of squares of residuals in the initial predicted curve and SSresidual is the sum of squares of residuals in the final predicted curve.44
Equation 5. Coefficient of variation calculation – Where σ is the standard deviation and µ is the maximum signal control.44
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this dissertation and no longer wish to have your work published on the UKDiss.com website then please: