Advertisement

A multi-parametric high throughput assay for detecting beta-cell proliferation in dispersed primary islets

Open AccessPublished:January 13, 2023DOI:https://doi.org/10.1016/j.slasd.2023.01.002

      Abstract

      Identification of novel compounds to selectively induce pancreatic beta-cell proliferation has the potential to restore functional beta-cell mass and insulin secretory demand in type 2 diabetes. The rarity of islet cell clusters (comprising of only 1% of the total pancreas mass) makes such a discovery a challenge. To address this obstacle a highthroughput, 384 well, plate-based multi-parametric imaging assay was developed to capture ex vivo primary islet proliferation, allowing positive identification of compounds that can selectively enhance islet beta-cell proliferation. The use of microscopy-based, high-content imaging technology enables acquisition of additional multi-parametric information such as proliferating populations in the islet beta and non beta-cells, insulin intensity, and cell counts, improving understanding of on and off target effects in primary tissue.
      The protocol requires access to a high-throughput microscopy platform for automated image acquisition of treated islet cells in assay plates. High content image analysis software is required to extract multiparametric cellular features and aid identification of therapeutically relevant small molecules and perturbants. Several putative beta-cell proliferative compounds have validated in this high throughput assay format, including the pleiotropic hormone prolactin [[1]] and the small molecule DYRK1A inhibitor harmine [[2]]. It is recommended to include one, or both, as positive controls to provide a reference for image analysis, give confidence in assay performance and capture potential assay variability during experimental runs.
      The protocol outlined specifically focuses on the multiparametric assessment of betacell proliferation in mouse and rat ex vivo islets and provides the methodology required for the collection of high quality cellular material. The high throughput, plate based assay can additionally be adapted to evaluate and quantify other disease relevant endpoints by high content microscopy and be applied to other downstream measurements. One of the caveats of a high-throughput, 384 microplate beta-cell proliferative assay is its limitations to facilitate human beta-cell proliferation detection, especially for weak activators. Adult human beta-cell proliferation is an extremely rare biological event and assessment experimentally can be donor dependent. In addition lower human islet beta-cell subpopulations require large numbers of cells for accurate rare event measurement.

      Keywords

      Introduction

      Identification of novel compounds to selectively induce pancreatic beta-cell proliferation has the potential to restore functional beta-cell mass and insulin secretory demand in type 2 diabetes. The rarity of islet cell clusters (comprising of only 1% of the total pancreas mass) makes such a discovery a challenge. To address this obstacle a high-throughput, 384 well, plate-based multi-parametric imaging assay was developed to capture ex vivo primary islet proliferation, allowing positive identification of compounds that can selectively enhance islet beta-cell proliferation. The use of microscopy-based, high-content imaging technology outlined in this protocol enables acquisition of additional multi-parametric information such as proliferating populations in the islet beta and non beta-cells, insulin intensity, and cell counts, improving understanding of on and off target effects in primary tissue. While other methodologies for quantifying beta-cell proliferation have been described in the literature, there have been none which describe the use of high content image analysis to capture a multi-parametric endpoint assessment of beta-cell and non-beta cell proliferation in addition to valuable relevant information on compounds under investigation in the assay, such as quantification of insulin intensity.
      The protocol requires access to a high-throughput microscopy platform for automated image acquisition of treated islet cells in assay plates. High content image analysis software is required to extract multiparametric cellular features and aid identification of therapeutically relevant small molecules and perturbants. To date several putative compounds have been reported to promote beta-cell proliferation capacity including the pleiotropic hormone prolactin [[
      • Brelje TC
      • Parsons JA
      • Sorenson RL.
      Regulation of islet beta-cell proliferation by prolactin in rat islets.
      ]] and the small molecule DYRK1A inhibitor harmine [[
      • Wang P
      • Alvarez-Perez JC
      • Felsenfeld DP
      • Liu H
      • Sivendran S
      • Bender A
      • Kumar A
      • Sanchez R
      • Scott DK
      • Garcia-Ocaña A
      • Stewart AF.
      A high-throughput chemical screen reveals that harmine-mediated inhibition of DYRK1A increases human pancreatic beta cell replication.
      ]] [[
      • Shen W
      • Taylor B
      • Jin Q
      • Nguyen-Tran V
      • Meeusen S
      • Zhang YQ
      • Kamireddy A
      • Swafford A
      • Powers AF
      • Walker J
      • Lamb J
      • Bursalaya B
      • DiDonato M
      • Harb G
      • Qiu M
      • Filippi CM
      • Deaton L
      • Turk CN
      • Suarez-Pinzon WL
      • Liu Y
      • Hao X
      • Mo T
      • Yan S
      • Li J
      • Herman AE
      • Hering BJ
      • Wu T
      • Martin Seidel H
      • McNamara P
      • Glynne R
      • Laffitte B
      Inhibition of DYRK1A and GSK3B induces human β-cell proliferation.
      ]], the latter of which was identified through screening for BrdU incorporation in beta-cell models and validated in dispersed primary islets [[
      • Shen W
      • Tremblay MS
      • Deshmukh VA
      • Wang W
      • Filippi CM
      • Harb G
      • Zhang YQ
      • Kamireddy A
      • Baaten JE
      • Jin Q
      • Wu T
      • Swoboda JG
      • Cho CY
      • Li J
      • Laffitte BA
      • McNamara P
      • Glynne R
      • Wu X
      • Herman AE
      • Schultz PG
      Small-molecule inducer of β cell proliferation identified by high-throughput screening.
      ]] [[
      • Robitaille K
      • Rourke JL
      • McBane JE
      • Fu A
      • Baird S
      • Du Q
      • Kin T
      • Shapiro AM
      • Screaton RA.
      High-throughput Functional Genomics Identifies Regulators of Primary Human Beta Cell Proliferation.
      ]]. Both prolactin and harmine have validated in our multi-parametric high throughput imaging assay protocol. It is recommended to include one, or both, as positive controls to provide a reference for image analysis, give confidence in assay performance and capture potential assay variability during experimental runs.
      The protocol outlined specifically focuses on the multiparametric assessment of beta-cell proliferation in mouse and rat ex vivo islets and provides the methodology required for the collection of high quality cellular material. The high throughput, plate based assay can additionally be adapted to evaluate and quantify other disease relevant endpoints by high content microscopy and be applied to other downstream measurements. One of the caveats of a high-throughput, 384 microplate beta-cell proliferative assay is its limitations to facilitate human beta-cell proliferation detection, especially for weak activators. Adult human beta-cell proliferation is an extremely rare biological event and assessment experimentally can be donor dependent. In addition lower human islet beta-cell subpopulations require large numbers of cells for accurate rare event measurement. An additional caveat of this protocol is that the cellular model utilized for this assay is a 2D system rather than the 3D islet structure. The use of a 2D system confers the advantage of acquiring the high-resolution images required for high content image analysis with the disadvantage that this may not fully replicate the native in vivo conditions.

      Materials

      Reagents

      Tabled 1
      Cell culture reagents
      RPMI 1640Sigma-AldrichR0883
      Penicillin/StreptomycinSigma-AldrichP0781
      L-GlutamineSigma-AldrichG7513
      Fetal Bovine Serum (non-USA origin)Sigma-AldrichF7524
      Corning® Ultra-Low Attachment 75cm² U-FlaskCorning®3814
      Poly-D lysine-coated PhenoPlate™ 384-wellPerkinElmer6057500
      Digestion reagents
      Egg white trypsin inhibitor (To 250 mg pot of egg white trypsin inhibitor add 1.25 ml distilled water (200 mg/ml))Sigma-AldrichT9253
      Soybean trypsin inhibitor (To 250 mg pot of soybean trypsin inhibitor add 1.25 ml distilled water (200 mg/ml))Sigma-AldrichT6522
      Liberase™ TL Research Grade (8 mg/ml solution prepared in water on day of use e.g. 1 pot in 625 ul)Roche05401020001
      Histopaque 1077Sigma-Aldrich10771
      TrypLE™ Express EnzymeGibco12605010
      Proliferation assay reagents
      Click-iT™ EdU Alexa Fluor™ 488 HCS AssayThermo Fisher ScientificC10351
      Insulin Rabbit mAb (use at 1:200)Cell Signaling TechnologyC27C9
      Goat anti-Rabbit IgG (H+L) Secondary Antibody, Alexa Fluor™ 594 (use at 1:500)Thermo Fisher ScientificA11012
      Hoechst (use at 1:2000)Thermo Fisher ScientificH3570
      Krebs Ringer Phosphate Hepes 0.1% BSA (KRH 0.1% BSA)
      Tabled 1
      ReagentFinal concentrationg per 1 L
      NaCl129 mmol/l7.5 g
      KCl4.8 mmol/l0.36 g
      KH2PO41.2 mmol/l0.16 g
      NaHCO35.0 mmol/l0.42 g
      MgSO41.2 mmol/l0.14 g
      HEPES (pH 7.4)10 mmol/l2.38 g
      Prepare using sterile H20, pH to 7.4, store @ 4⁰C for up to 1 month
      Additions on day
      Tabled 1
      CaCl2 (1 M)2.5 mmol/l2.5 ml
      BSA Fraction V0.1%1 g
      Glucose5.6 mmol/l1 g
      Store @ 4⁰C/on ice and use on day of experiment
      Culture media
      Tabled 1
      ReagentFinal concentrationAmount
      RPMI 1640500 ml
      FCS (non-USA origin)10%50 ml
      Pen/Strep1%5 ml
      L-Glutamine1%5 ml
      Total volume560 ml
      Prepare under sterile conditions, store @ 4⁰C for up to 1 month
      Modified Blocking Buffer
      Tabled 1
      ReagentFinal concentrationAmount
      PBS500 ml
      Triton-X 1000.2%1 ml
      BSA1.1%5.5 g
      Store @ 4⁰C for up to 1 month
      Neutralization Buffer
      Tabled 1
      ReagentFinal concentrationAmount
      PBS50 ml
      FCS5%2.5 ml
      Total volume52.5 ml
      Prepare under sterile conditions, store @ 4⁰C for up to 1 month
      Digest 1: (Two rat pancreases) Liberase (final concentration) = 1.1 mg/ml
      8 mg/ml liberase solution prepared in water on day of use e.g. 1 × 5 mg pot in 625 ul
      Tabled 1
      ReagentFinal concentrationAmount
      Egg white trypsin inhibitor20 mg100 ul
      Soybean trypsin inhibitor20 mg100 ul
      Liberase (8 mg/ml)1 mg/ml1.5 ml
      KRH 0.1% BSA9.2 ml
      Total volume10.9 ml
      Make up fresh, store @ 4⁰C/on ice and use on day of experiment
      Digest 2: (Two rat pancreases) Liberase (final concentration) = 0.6 mg/ml
      Tabled 1
      ReagentFinal concentrationAmount
      Egg white trypsin inhibitor20 mg100 ul
      Soybean trypsin inhibitor20 mg100 ul
      Liberase (8 mg/ml)0.6 mg/ml930 ul
      KRH 0.1% BSA11.7 ml
      Total volume12.83 ml
      Make up fresh, Store @ 4⁰C/on ice and use on day of experiment

      CRITICAL: Use of Paraformaldehyde

      Paraformaldehyde solutions (8%) can emit toxic formaldehyde gas above permissible limits which poses significant health hazards. All handling of paraformaldehyde should be done inside chemical fume hoods with adequate ventilation using appropriate chemically protective gloves such as nitrile.

      CRITICAL: Use of Liberase

      Liberase is a blend of highly purified enzymes and may cause respiratory irritation when aerosols are generated. It is recommended to prepare the liberase solution in a biological safety cabinet or fume hood for respiratory protection.

      Equipment

      A Yokogawa CellVoyager CV8000 High Content Screening System with high speed confocal scanning was used for the development and application of this protocol. Any alternative high content high resolution imaging instruments may be used which are deemed suitable for quantitative high throughput 384 well image-based screening.

      Alternatives: Substitution of Liberase for Collagenase XI

      The success of the protocol relies on the acquisition of high quality viable pancreatic islets. The liberase digestion method has been optimized to acquire such cellular material. The protocol may also be performed with a substitution of liberase using Collagenase XI from Clostridium histolyticum. With a collagenase digestion methodology only a single digestion step should be performed with an approximate 8-minute digestion time and a digest volume of 5 ml for a rat pancreas and 3 ml for a mouse pancreas. With this enzymatic digest solution there is no requirement for inclusion of trypsin inhibitors in the digest mix.
      Collagenase Digest: (Two rat pancreases – 5 ml each, three mouse pancreases – 3 ml each)
      Collagenase XI (final concentration) = 1 mg/ml
      Tabled 1
      ReagentFinal concentrationAmount
      Collagenase XI1 mg/ml10 mg
      KRH 0.1% BSA10 ml
      Make up fresh, store @ 4⁰C/on ice and use on day of experiment

      Alternatives: Substitution of Poly-D-Lysine coated microplates for plates coated with Matrigel

      Assay ready Poly-D-Lysine (PDL) coated microplates are included in this protocol for enhanced islet cell attachment and ease of use. Dispersed islet cells are low adherent and will not bind well to tissue culture treated plates. PDL is a synthetic molecule which promotes the adhesion of cells to a solid substrate by enhancing the frequency of positively charged cell binding sites. An alternative to PDL coated microplates is to use Matrigel (1 ml/100 ml media), an extracellular matrix protein mixture which will promote cell anchoring and optimal attachment. Plates may be pre-coated with the addition of 5 ul Matrigel (1 ml/100 ml media) per well, followed by plate centrifugation at 300 g for 1 min at room temperature. Matrigel is then removed by plate inversion and tapping dry on absorbent paper to remove all liquid. Culture media is then added to the plates for use in the protocol.
      Matrigel is a reconstituted basement membrane extract and its composition may contain biologically active elements which themselves may stimulate physiological growth and proliferation [[
      • Kleinman HK
      • Martin GR.
      Matrigel: basement membrane matrix with biological activity.
      ]] [[
      • Aisenbrey EA
      • Murphy WL.
      Synthetic alternatives to Matrigel.
      ]]. The effects of Matrigel on islet cell growth was not assessed in the development of this protocol.

      Software

      The protocol uses Columbus™ (PerkinElmer), a commercial imaging analysis software. The analysis steps outlined in the protocol were performed on a Columbus™ version 2.9.1 using a Microsoft® Windows™ operating system. This Columbus™ software allows a sequential image analysis pipeline through the use of building blocks. The user can employ any high content image analysis software including the free open-source software CellProfiler.

      Procedure

      The protocol describes a procedure to isolate high quality islets, followed by dispersal and plating of islets for downstream high throughput assays. A method for proliferation assessment by EdU incorporation and immunofluorescence staining is outlined. A high content analysis pipeline to quantify multi-parametric endpoints is described.

      Islet Isolation Procedure

      Timing: 6 hours

      The primary goal of the islet isolation procedure is to yield high quantities of purified islets with preservation of islet morphology and functional integrity. The liberase digestion and purification method has been optimized to cleanly separate pancreatic islets from surrounding exocrine tissue, prevent over digestion and fragmentation of islets, and improve the viability and functionality of isolated islets. All steps, unless otherwise mentioned, are performed at room temperature.
      • 1
        Digestion procedure
        • a
          Terminally anaesthetize rat/mouse using rising CO2 and then use cervical dislocation. Rapidly remove the pancreas, with the spleen and any adipose tissue attached. Place in cold KRH 0.1% BSA in a 50 ml centrifuge tube and keep tube on ice.
        • b
          Place tissue in a petri dish and trim pancreas (darker tissue) of fat, spleen and any other tissues.
        • c
          Inflate the pancreas with 5 ml for rat, or 3 ml for mouse, cold Digest 1 cocktail in a new petri dish using a syringe and needle (suck back up and reinject leaked cocktail if needed). Repeat until pancreas has been injected and inflated in all parts. Transfer inflated pancreas and any remaining digest cocktail (with syringe) into a 50 ml centrifuge tube.
        • d
          Incubate for 3 min in a 37°C water bath, swirling continuously by hand.
        • e
          Remove tube from water bath and shake manually for 1 min. The shaking style should be slightly vigorous to aid with the digestion of the pancreatic tissue.
        • f
          Return to the 37°C water bath and incubate for a further 3 min swirling continuously by hand.
        • g
          Shake manually for 1 min.
        • h
          Make the volume up to 30 ml with KRH 0.1% BSA and centrifuge at 300 g for 1 min at room temperature.
        • i
          Pour off the supernatant and resuspend the pellet in the second digest solution (6.4 ml).
        • j
          Incubate for 2 min in the 37°C water bath with gentle swirling.
        • k
          Shake by hand for 1 min then check the digest using a black petri dish previously cleaned with 70% Ethanol. To aid visualization 1 ml of digest can be added to 3 ml KRH 0.1% BSA in the black petri dish. The islets should be separated from the acinar tissue. If islets have not fully detached it is recommended to return all digest to the 50 ml centrifuge tube, return to the 37°C water bath and incubate for 1 min with swirling. This is followed by tube removal and shaking for 1 min before checking the digest again. Repeat, if required, until islets are separated from exocrine tissue.
        • l
          Make the volume up to 30 ml with cold KRH 0.1% BSA and centrifuge at 300 g for 1 min at room temperature. Pour off supernatant gently, make volume up to 30 ml again and repeat this 2 times. This is necessary to remove all traces of digestion enzymes and prevent over digestion of the pancreatic islets.
      • 2
        Purification of islets
        • a
          Add 7.5 ml of Histopaque 1077 to 2 × 15 ml centrifuge tubes.
        • b
          Resuspend pellet in 10 ml of KRH 0.1% BSA for rat and 5 ml for mouse.
        • c
          Carefully add 5 ml of sample to the Histopaque by pipetting down the sides of the tube extremely slowly so as not to mix with the Histopaque (use autopipettor on slow setting). There should be two distinct layers (see Figure 2).
        • d
          Centrifuge in a refrigerated centrifuge (4°C) at 1200g for 10 mins with no brake and turning speed up slowly by hand to give slower acceleration (or keeping acceleration/deceleration rate at the lowest setting).
        • e
          Remove tubes from centrifuge. At this point it is normal to have white exocrine contamination. Remove the band occurring at the gradient interface, along with some of the surrounding supernatant (totaling about 7 ml) to a new tube containing 20 ml of KRH 0.1% BSA using an autopipettor. Be careful to not disturb the pellet which contains the exocrine component (see Figure 2).
        • f
          Place purified islet solution into a black petri dish to observe. The black petri dish should be previously cleaned with 70% Ethanol. To aid visualization 1 ml of digest can be added to 3 ml KRH 0.1% BSA in the black petri dish.
        • g
          Add 5 ml culture medium per well to a 6 well ultra-low attachment plate
        • h
          Using an illuminated microscope hand pick 30-50 islets per well to further purify islets from exocrine tissue (see Figure 3). It is recommended to use a 20x/40x stereomicroscope (also known as a dissection microscope) which will allow low magnification observation of the digest and pipette tip. As islets can vary considerably in size and shape it is best to use a p200 pipette with a 200 μL pipette tip to avoid islet shearing. Islets can be distinguished from exocrine tissue by their spherical shape and fawn color. If exocrine tissue is too abundant for visualization and picking of individual islets it is recommended to further dilute the digest with KRH 0.1% BSA, for example adding 1 ml of digest to 5 ml KRH 0.1% BSA in the black petri dish. Any small pieces of exocrine acinar tissue which is picked to culture dishes should dissipate overnight. With the liberase digestion method expected yields are 300-400 islets per mouse pancreas and 700-900 islets per rat pancreas.
        • i
          Incubate 6 well plates containing islets at 37°C, 5% CO2, 90% humidity overnight.
      Note: All experiments must be performed with appropriate ethical approval and in accordance with the local ethical laws.
      CRITICAL: An overnight incubation period is necessary for recovery from the digestion process and purification procedure (see Figure 4). It is recommended to use islets for functional assessment assays the day following isolation.
      Note: Culturing islets in ultra-low attachment surface plates will prevent islet attachment to plastic and greatly facilitate collection and pooling of islets in subsequent steps.
      Pause point: Following purification of islets from exocrine tissue by Histopaque density gradient, and removal of interface containing islets to KRH 0.1% BSA, the protocol can be paused for 30 minutes before continuation with manual purification. It is recommended to not extend longer than 30 minutes.

      Dissociation and seeding of islets

      Timing: 6 hours

      Dispersed islet cell 2-D monolayer cultures are a valuable model for the study of single cell phenotypic changes and functional responses in islet cell sub-populations. Dissociation of islets is performed by TrypLE Express, a gentle enzymatic solution which increases dissociation specificity and reduces damage to individual cells. All steps, unless otherwise mentioned, are performed at room temperature.
      • 1
        Dissociation of islets to a single cell solution
        • a
          Following overnight culture hand pick islets to a 15 ml centrifuge tube using an illuminated dissection microscope. This step can be performed outside a laminar flow cabinet.
        • b
          Allow islets to sediment or centrifuge at 300 g for 2 min at room temperature and gently remove the supernatant taking care not to disturb the cell pellet. Transfer 15 ml centrifuge tube to a laminar flow cabinet for subsequent steps.
        • c
          Add 1 ml TrypLE to the centrifuge tube and place in a 37°C incubator for 7 min.
        • d
          Using a p1000 pipette set to 1000 µl gently pipette up and down for 2 min to physically disperse islets into single cells. Check the cell suspension under a microscope by removing contents of tube to one well of a 6 well ultra-low attachment plate. If whole islets or aggregates remain return plate to a 37°C incubator for 2 min followed by an additional 1 min of pipetting.
        • e
          Collect well contents of dispersed cells to a 15 ml centrifuge tube and add 10 ml neutralization buffer to cells to neutralize the TrypLE (use 2-3 ml of this to rinse 6 well plate to collect all cells.
        • f
          Centrifuge for 4 minutes at 400 g to pellet the dispersed islet cells.
        • g
          Resuspend in 15 ml culture medium and take a small aliquot (10-20 µl) to count the cells. Cell viability can be assessed on counting (e.g. trypan blue exclusion, propidium iodide) and it is recommended that cell viability should be at least 80% in order to achieve a good recovery after seeding
      • 2
        Seeding of dispersed islet cells
        • a
          Following counting adjust volume to a final concentration of 8000 cells/40 µl culture media (2 × 10^5 cells/ml)
        • b
          Seed cells manually in Poly-D-Lysine (PDL) coated clear-bottom black 384 microplates with a multi-channel pipette set to 40 µl followed by gentle agitation of the plate to ensure even distribution of cells in the well. A low volume automated liquid handling system, such as a Dragonfly Discovery liquid dispenser (SPT Labtech), can be employed for this step which will enable seeding in 384 well plates with minimum dead volume of cell suspension.
        • c
          Following seeding allow the plate to rest at room temperature in the tissue culture hood for 60 minutes before moving to a 37°C incubator with a humidified atmosphere of 5% CO2 in air. This will aid attachment and consistent cell distribution in wells for high content imaging endpoints [[
          • Shen W
          • Taylor B
          • Jin Q
          • Nguyen-Tran V
          • Meeusen S
          • Zhang YQ
          • Kamireddy A
          • Swafford A
          • Powers AF
          • Walker J
          • Lamb J
          • Bursalaya B
          • DiDonato M
          • Harb G
          • Qiu M
          • Filippi CM
          • Deaton L
          • Turk CN
          • Suarez-Pinzon WL
          • Liu Y
          • Hao X
          • Mo T
          • Yan S
          • Li J
          • Herman AE
          • Hering BJ
          • Wu T
          • Martin Seidel H
          • McNamara P
          • Glynne R
          • Laffitte B
          Inhibition of DYRK1A and GSK3B induces human β-cell proliferation.
          ]].
      Note: To eliminate edge effects (caused by evaporation in the outer wells which may occur in 384 multiwell plates) it is recommended to seed cells in the interior wells and fill the outer rows and columns with media. Any wells not containing cells should also be filled with media. Alternatively, a microclimate plate lid can be used with will preserve the concentrations and volumes of solutions in the outside wells and minimize edge effects.
      Note: The protocol routinely obtains 700-900 islets per rat pancreas. Each islet contains 1000-2000 cells with the majority of cells being beta cells (65–85%) and alpha cells 10–25%), and the remaining consisting of delta and other cell types [[
      • Carter JD
      • Dula SB
      • Corbin KL
      • Wu R
      • Nunemaker CS.
      A practical guide to rodent islet isolation and assessment.
      ]]. With a seeding density of 8000 cells well, the source tissue of two rat pancreases should provide cells for one 384 well plate. However, this can vary depending on the age and strain of the rodents.

      Proliferation assay

      Timing: 96 hours

      Detection of beta-cell proliferation is performed with a high-content imaging-based assay. The image-based assay allows quantitative assessment of the proliferative capacity of beta-cells in response to pharmacological agents and the high content readout allows incorporation of additional multiparametric endpoints such as proliferation in non-beta islet cells. Proliferation is measured through EdU staining. EdU (5-Ethynyl-2′-deoxyuridine) is a thymidine analog which is incorporated into newly synthesized DNA in proliferating cells. Detection of EdU is accomplished through click reaction with Click-iT Plus technology and labelling with an Alexa Fluor dye. Identification of the beta-cell islet population is performed by immunofluorescence staining of insulin. All steps, unless otherwise mentioned, are performed at room temperature.
      • 1
        Addition of compounds
        • a
          Addition of compounds and pharmacologic agents should be performed >3 hour after cell seeding or following overnight culture. Dispense 10 µl of 5X compound or control in media to the appropriate wells of the assay plate. Compounds may also be added with automated liquid handling or by incorporating automated solutions for compound dispensing.
        • b
          Incubate at 37°C, 5% CO2 for 48 hours.
      • 2
        Addition of EdU
        • a
          Prepare 6X EdU in culture media (60 µM EdU from 10 mM EdU stock). This should be adjusted based on the volume in the assay plates following compound dispensing.
        • b
          Add 10 μl/well with a multi-channel pipette.
        • c
          Return plate to the 37°C incubator for 24 hours.
      • 3
        Fixation
        • a
          Cell proliferation is assessed 72 hours of compound incubation and 24 hours following EdU incorporation.
        • b
          Fix cells by addition of 60 μl 8% Paraformaldehyde in PBS onto the assay plates (4% final PFA concentration). Handling of paraformaldehyde should be done inside a chemical fume hood.
        • c
          Incubate cells for 20 minutes at room temperature in the dark for covalent cross-links to form.
        • d
          After fixation aspirate formaldehyde.
        • e
          Wash plates 2X with 80 µl PBS using slow and careful manual dispensing followed by aspiration. Dispense the PBS while lightly touching the pipette tip against the side of the receiving well to prevent disturbing the islet cell monolayer.
      • 4
        Permeabilization
        • a
          Permeabilize cells with addition of 100 μl modified blocking buffer to each well using slow and careful manual dispensing with a multi-channel pipette.
        • b
          Incubate plate at room temperature for 30 minutes.
        • c
          Wash cells 2X with PBS using gentle manual dispensing followed by aspiration.
      • 5
        Click-iT EdU reaction staining with Click-iT® EdU Alexa Fluor® 488 Imaging Kit
        • a
          Prepare 1X Click-iT® EdU reaction buffer by diluting the 10X solution 1:10 in deionized water.
        • b
          Prepare 1X Click-iT® EdU buffer additive by diluting the 10X solution created 1:10 in deionized water.
        • c
          Prepare Click-iT® reaction cocktail according to the table below. Add ingredients in the order listed in the table.
        • d
          Add 25 μl/well with a multi-channel pipette with careful manual dispensing.
        • e
          Incubate for 30 minutes at room temperature, protected from light.
        • f
          Remove the reaction cocktail by aspiration.
        • g
          Add 100 μl of Click-iT® reaction rinse buffer by careful manual dispensing.
        • h
          Remove the reaction rinse buffer by aspiration.

      Click-iT reaction cocktail

      Tabled 1
      Reaction componentsNumber of 384 well plates
      0.250.51
      1X Click-iT® EdU reaction buffer2550 μl5100 μl10300 μl
      CuSO4 (Component D)120 μl240 μl480 μl
      Alexa Fluor® azide (Component B)7.5 μl15 μl30 μl
      1X Click-iT® EdU buffer additive300 μl600 μl1200 μl
      Total2,977.5 μl5,955 μl12,010 μl
      Make up fresh, store at room temperature and use on day of experiment
      • 1
        Antibody staining
        • a
          When reaction rinse buffer is removed add 100 μl neutralization buffer by careful manual dispensing.
        • b
          Incubate for 30 minutes at room temperature, protected from light.
        • c
          Prepare 1:100 dilution of rabbit anti-insulin antibody (C27C9) in modified blocking buffer.
        • d
          Remove modified blocking buffer from the wells by aspiration.
        • e
          Add 50 μl of antibody/modified blocking buffer to each well by gentle manual dispensing.
        • f
          Incubate for one hour at room temperature, protected from light.
        • g
          Wash 3X in modified blocking buffer adding 100 μl modified blocking buffer by careful manual dispensing against the side of the assay wells, incubating in modified blocking buffer for three minutes, and aspirating.
        • h
          Prepare 1:500 dilution of secondary Alexa Fluor® 594 (A11012) plus 1:2000 HCS Nuclear Mask (1:2000 Hoechst may be used as an alternative) in modified blocking buffer.
        • i
          Remove modified blocking buffer by aspiration.
        • j
          Add 100 μl of secondary/block to each well by gentle manual dispensing with a multi-channel pipette.
        • k
          Incubate for 45 minutes at room temperature protected from light.
        • l
          Remove secondary/block by aspiration.
        • m
          Wash 2X in PBS w/o Mg2+/ Ca2+ with gentle manual dispensing against the side of the assay wells and aspiration.
        • n
          Add 80 μl PBS w/o Mg2+/ Ca2+ to each well.
        • o
          Cover plate with a black plate seal.
        • p
          Store plate at 4°C prior to imaging.
      CRITICAL: It is recommended to perform aspiration steps using an automated liquid handling system with a slow aspiration protocol designed for 384 plates with 10 μl residual volume. If no automated liquid handling systems are available, fluid can be discarded by plate inversion over an appropriate container with remaining liquid removed by gently tapping on absorbent paper.
      Optional: The following reference compounds may be included to ensure assay to assay accuracy and to calculate assay robustness - Negative control: 0.1% DMSO (or appropriate compound vehicle solution); Stimulatory control: 5 μM GNF4877 [[
      • Shen W
      • Taylor B
      • Jin Q
      • Nguyen-Tran V
      • Meeusen S
      • Zhang YQ
      • Kamireddy A
      • Swafford A
      • Powers AF
      • Walker J
      • Lamb J
      • Bursalaya B
      • DiDonato M
      • Harb G
      • Qiu M
      • Filippi CM
      • Deaton L
      • Turk CN
      • Suarez-Pinzon WL
      • Liu Y
      • Hao X
      • Mo T
      • Yan S
      • Li J
      • Herman AE
      • Hering BJ
      • Wu T
      • Martin Seidel H
      • McNamara P
      • Glynne R
      • Laffitte B
      Inhibition of DYRK1A and GSK3B induces human β-cell proliferation.
      ]]. DMSO tolerance above 0.1% has not been tested in the development of this protocol. Screening quality can be measured by calculating the Z-factor of the assay (separation between negative and stimulatory controls) according to the method described by Zhang and colleagues [[
      • Zhang JH
      • Chung TD
      • Oldenburg KR.
      A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays.
      ]].
      Pause point: Following fixation and PBS washing steps plates may be stored at 4°C containing 80 μl PBS/well. Plates can be stored for up to 1 week without substantially reducing image quality. count

      Image acquisition and analysis

      Timing: 8 hours

      A imaging system is utilized for automated capture of a large numbers of images from the assay plate. High-content image analysis software enables acquisition of quantifiable data from biological images. This protocol describes an image analysis sequence using Perkin Elmer Columbus which utilizes building blocks for cell segmentation and analysis.
      • 1
        Image acquisition
      • Acquire images with a 20X objective capturing multiple adjacent fields-of-view to cover as much of the well as possible. Development of the protocol was conducted with a 20X long working distance objective lens (0.45 NA), capturing 16 field of view per well of a 384 well plate. A z-stack with a 1 μm step and maximum projection processing is recommended as some islet cells may not sit on the same focal plane. In protocol development a total of 10 slices per stack with a step size of 1 µm proved optimal. It is recommended to use confocal scanning if possible, which will provide the highest signal to noise ratio and high resolution imaging. The protocol has not been tested on a widefield fluorescence microscope.
      Tabled 1
      HCS NuclearMaskTM Blue StainAlexa Fluor® 488Alexa Fluor® 594
      Excitation/Emission (in nm)350/461495/519590/615
      Standard filter setDAPIFITCTRITC
      • 1
        Image analysis
        • a
          The table below describe the steps for high content image analysis and capture of multi-parametric endpoints (see Figure 5). The image analysis pipeline outlined uses the commercial Columbus™ (PerkinElmer) analysis software.
      Tabled 1
      Analysis sequenceDescription
      Find NucleiUsing DNA stain (Nuclear Mask or Hoechst) ensure correct thresholds to capture all nuclei, and split factor to split adjacent nuclei in separate objects without spurious splitting of whole single nuclei (see Figure 6).
      Optional: Remove border objects with a ‘select population’ building block – these incomplete cells can interfere with the subsequent analysis
      Find CytoplasmThe ‘background’ staining in the Hoechst/Nuclear Mask channel can be used to define the cytoplasm. It is best to reduce the ‘Individual Threshold’ down to 0.05 if using this type of background staining (see Figure 6).
      Alternatively use a ‘Calculate Image’ building block to add up staining from all available channels and use this to define the cytoplasm.
      Optional: A whole cytoplasmic stain such as Cell Mask can be added to the remaining fluorescence channel, and this channel can be used to segment the cytoplasm. This should be considered if using an alternative image software which may not accurately identify cell boundaries from the faint Hoechst background staining.
      Calculate Intensity PropertiesThe insulin immunofluorescence measurement can be quantified with the ‘Calculate Intensity Properties’ which will be used to identify the insulin expressing beta-cells.
      Select population (to select beta-cell population)Within the ‘Select Population’ building block, the insulin intensity measurement can be used as a filter to select the beta-cell population. The threshold can be determined by observing clusters in the scatter plot and histogram functions.
      Select population (to select the non beta-cell population)Another ‘Select Population’ building block can be used to select the non beta-cell population. Here the insulin intensity measurement can be used as a filter and the operator tool can be changed to a ‘less than’ (<) function.
      Calculate Intensity PropertiesThe EdU fluorescence measurement can be quantified with the ‘Calculate Intensity Properties’ which will be used to identify the proliferative populations.
      Select population (to select proliferative beta-cell population)Using a ‘Select Population’ building block, the pre-defined beta-cell population can be selected. The EdU intensity measurement may then be used as a filter to select the proliferative beta-cell population. The threshold for EdU can be determined by observing clusters in the scatter plot and histogram functions (see Figure 6).
      Select population (to select the proliferative non beta-cell population)Another ‘Select Population’ building block can use the pre-defined non beta-cells to quantify the proliferative non beta-cell population. It is recommended to use the same EdU intensity threshold as that for the proliferative beta-cell population (see Figure 6).
      Define resultsAny properties calculated by the preceding building blocks can be included in the ‘Define results’ building block. In addition a formula output can be selected which allows entry of a formula with defined variables. This feature can be used to quantify endpoints like percentage beta-cell proliferation.
      Note: EdU incorporation can cause a reduction in the intensity of DNA stains [[
      • Lundholt BK
      • Scudder KM
      • Pagliaro L.
      A simple technique for reducing edge effect in cell-based assays.
      ]] [[
      • Kubbies M
      • Rabinovitch PS.
      Flow cytometric analysis of factors which influence the BrdUrd-Hoechst quenching effect in cultivated human fibroblasts and lymphocytes.
      ]]. This may confound any data extracted from high content analysis of nuclear stains, for example if analyzing cell cycle profiles, the reduction in DNA stain intensity and shift in apparent DNA content will specifically cause perturbation of the detected cell cycle phase in EdU positive populations.
      Note: Capturing beta-cell proliferation frequency is key to ensuring robust and reproducible data. Rat and mouse proliferation is a rare event, and human beta-cell proliferation is even less frequent. More dimensions may be added to the image acquisition and analysis method to ensure optimal rare event detection. During the imaging stage it is recommended to capture as much of the well as possible through the acquisition of multiple fields of view. This will enable analysis of more cell objects and increase the frequency of rare event detection. In addition to fields of view, increasing the sample replicates in multiple wells may greatly increase the capture of total cell events, though this will reduce the number of compounds that can be screened in the assay. Ensuring that image analysis protocol is refined will minimize the generation of nonspecific false positive rare events. It is important to select the right thresholds for EdU staining and the correct parameters for nuclei capture. In addition to the image analysis workflow outlined, a machine learning approach can be employed to automate detection and classification of rare phenotypes and thus can utilized to detect the presence or absence of a proliferation event. However, a deep learning method will limit the multi-parametric approach for the quantitative characterization of multiple cellular features in the assay.

      Troubleshooting

      Problem 1

      Islet cell detachment occurs in the 384 well assay plates during the course of the protocol leaving little remaining cells for image acquisition.

      Potential solution

      Use Poly-D-Lysine (PDL) coated microplates to facilitate adherence to plastic. Reduce incubator temperature fluctuations which can prevent successful attachment of cells post seeding. Do not subject cells to harsh washing forces. Leave a small remaining volume in assay plates after aspirating to avoid disturbing the cell monolayer.

      Problem 2

      Inadequate pancreas tissue digestion.

      Potential solution

      Keep reconstituted liberase enzyme on ice and use on day of experiment to avoid loss in activity. Increase digestion phase with 1 minute waterbath and 1 minute shaking cycles. Use microscope observation to determine separation of islets from exocrine tissue.

      Problem 3

      Some residual acinar tissue remains attached to islets following density gradient separation.

      Potential solution

      Try gentle mechanical resuspension with a pipette during hand-picking phase. Islets with small acinar tissue evident should still be hand-picked to media containing plates as this may separate away during overnight culture period.

      Problem 4

      Poor islet yields

      Potential solution

      This is likely due to over-digestion of islets during the isolation protocol. Avoid over forceful shaking of digest mix during incubation with digestion cocktail. Use microscope observation to determine separation of islets from exocrine tissue.

      Problem 5

      No EdU staining visible in microscope images.

      Potential solution

      Refer to manufacturer's instructions on Click-iT® EdU Alexa Fluor® 488 Imaging Kit.

      Problem 6

      Inconsistency in nuclei/cell segmentation during the image analysis step

      Potential solution

      Accurate nuclei/cell segmentation is vital for downstream analysis, robust data and plate to plate consistency. Segmentation can occasionally be erroneous due to artifacts or noise in the image. If this occurs it is possible to apply a morphological/intensity filtering step in the analysis to eliminate background clutter and only select objects that conform to true nuclei features (roundness, size etc.).

      Anticipated Results

      The protocol has been optimized with markers chosen to give a high fluorescence signal. Fluorescence staining of EdU, insulin and nuclear mask should be uniform and bright making the markers readily distinguishable for high content quantification.
      A recommended positive control (DYRK1A inhibitor - GNF4877 5 μM) [[
      • Wang P
      • Alvarez-Perez JC
      • Felsenfeld DP
      • Liu H
      • Sivendran S
      • Bender A
      • Kumar A
      • Sanchez R
      • Scott DK
      • Garcia-Ocaña A
      • Stewart AF.
      A high-throughput chemical screen reveals that harmine-mediated inhibition of DYRK1A increases human pancreatic beta cell replication.
      ]] should exhibit a large increase in proliferation which should be visually observable in the imaging output.

      Limitations

      The protocol utilizes tissue-derived primary cells, and although every effort has been made to purify the islet cell population it is possible that remains of some other non-islet tissue (such as capillaries) will be carried and plated in the assay. On occasion this may result in a false positive and thus it is recommended that manual inspection of images should be performed as a quality control measure. This visual inspection, especially of images from any hit compounds, may require researcher time and thus is a limitation of this assay.
      Another limitation is the use of an islet cell monolayer rather than the 3D islet structures that resemble the in vivo physiologically relevant condition. While this assay format increases the throughput, and thus the quantities of compounds that can be screened, it is accepted that a monolayer does not fully reflect the in vivo characteristics of islets.

      Author contributions

      N.M. contributed substantially to the conception and design of the protocol and drafted the manuscript.
      D.M.S. contributed substantially to the conception and design of the protocol and contributed to manuscript revision.
      D.G. directed and supervised the research and contributed to manuscript revision.
      All authors read and approved the final manuscript.

      Uncited Link

      Figure 1, Figure 7
      Fig 1:
      Fig 1A simplified schematic workflow of a high throughput assay in dispersed primary islets.
      Fig 2:
      Fig 2Islets are separated from exocrine tissue by density gradient centrifugation. Histopaque is utilized to create a gradient with pancreas digest added carefully. Samples are centrifuged using the slowest possible acceleration/deceleration speed setting. Purified islets are carefully collected from the Histopaque/buffer interface using a pipette.
      Fig 3:
      Fig 3Islets are purified by handpicking. This step is performed outside a laminar flow cabinet. A dissecting microscope at 20x is used to identify the islets (A). Handpicking of islets is greatly assisted with a gooseneck light fixture and black petri dish which provide optimal visualization conditions. Islets are handpicked with a 200 μL pipette tip.
      Fig 4:
      Fig 4Representative images of islets post Histopaque purification (A), following handpicking to separate islets from exocrine tissue (B), and following overnight recovery in culture media (C).
      Fig 5:
      Fig 5The use of microscopy-based high-content imaging technology enables acquisition of multi-parametric information such as proliferating populations in the islet non beta-cells (insulin negative) improving understanding of on and off target effects in primary ex vivo tissue.
      Fig 6:
      Fig 6A high-content image analysis approach is used to identify objects of interest and extract metrics for these objects. Information on islet cell proliferation in response to compound perturbation is readily obtainable. In addition to proliferation endpoints, extraction of additional features such as intensity and morphology can provide valuable information on islet cell responses to compounds or external perturbants.
      Fig 7:
      Fig 7Representative images of dispersed islet proliferation in cells treated with positive control DYRK1A inhibitor - GNF4877 (10 μM)

      Uncited References

      [
      • Rabinovitch PS.
      Regulation of human fibroblast growth rate by both noncycling cell fraction transition probability is shown by growth in 5-bromodeoxyuridine followed by Hoechst 33258 flow cytometry.
      ]

      Declaration of Competing Interests

      The authors N.M., D.M.S., and D.G. declare the following competing interest: they are employees of, and shareholders in, AstraZeneca.
      The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
      Niamh Mullooly reports financial support was provided by AstraZeneca. Niamh Mullooly reports a relationship with AstraZeneca that includes: equity or stocks. Davide Gianni reports a relationship with AstraZeneca that includes: equity or stocks. David M. Smith reports a relationship with AstraZeneca that includes: equity or stocks. The authors received no financial support for the research, authorship, and/or publication of this article. All authors are employed by AstraZeneca, and their research and authorship of this article was completed within the scope of their employment with AstraZeneca. The authors N.M., D.M.S., and D.G. declare the following competing interest: they are employees of, and shareholders in, AstraZeneca.

      Acknowledgments

      The authors received no financial support for the research, authorship, and/or publication of this article. All authors are employed by AstraZeneca, and their research and authorship of this article was completed within the scope of their employment with AstraZeneca.

      Appendix. Supplementary materials

      References

        • Brelje TC
        • Parsons JA
        • Sorenson RL.
        Regulation of islet beta-cell proliferation by prolactin in rat islets.
        Diabetes. 1994 Feb; 43 (PMID: 7904577): 263-273https://doi.org/10.2337/diab.43.2.263
        • Wang P
        • Alvarez-Perez JC
        • Felsenfeld DP
        • Liu H
        • Sivendran S
        • Bender A
        • Kumar A
        • Sanchez R
        • Scott DK
        • Garcia-Ocaña A
        • Stewart AF.
        A high-throughput chemical screen reveals that harmine-mediated inhibition of DYRK1A increases human pancreatic beta cell replication.
        Nat Med. 2015 Apr; 21 (Epub 2015 Mar 9. PMID: 25751815; PMCID: PMC4690535): 383-388https://doi.org/10.1038/nm.3820
        • Shen W
        • Taylor B
        • Jin Q
        • Nguyen-Tran V
        • Meeusen S
        • Zhang YQ
        • Kamireddy A
        • Swafford A
        • Powers AF
        • Walker J
        • Lamb J
        • Bursalaya B
        • DiDonato M
        • Harb G
        • Qiu M
        • Filippi CM
        • Deaton L
        • Turk CN
        • Suarez-Pinzon WL
        • Liu Y
        • Hao X
        • Mo T
        • Yan S
        • Li J
        • Herman AE
        • Hering BJ
        • Wu T
        • Martin Seidel H
        • McNamara P
        • Glynne R
        • Laffitte B
        Inhibition of DYRK1A and GSK3B induces human β-cell proliferation.
        Nat Commun. 2015 Oct 26; 6 (PMID: 26496802; PMCID: PMC4639830): 8372https://doi.org/10.1038/ncomms9372
        • Shen W
        • Tremblay MS
        • Deshmukh VA
        • Wang W
        • Filippi CM
        • Harb G
        • Zhang YQ
        • Kamireddy A
        • Baaten JE
        • Jin Q
        • Wu T
        • Swoboda JG
        • Cho CY
        • Li J
        • Laffitte BA
        • McNamara P
        • Glynne R
        • Wu X
        • Herman AE
        • Schultz PG
        Small-molecule inducer of β cell proliferation identified by high-throughput screening.
        J Am Chem Soc. 2013 Feb 6; 135 (Epub 2013 Jan 25. Erratum in: J Am Chem Soc. 2013 Mar 20;135(11):4573. PMID: 23330637): 1669-1672https://doi.org/10.1021/ja309304m
        • Robitaille K
        • Rourke JL
        • McBane JE
        • Fu A
        • Baird S
        • Du Q
        • Kin T
        • Shapiro AM
        • Screaton RA.
        High-throughput Functional Genomics Identifies Regulators of Primary Human Beta Cell Proliferation.
        J Biol Chem. 2016 Feb 26; 291 (Epub 2016 Jan 6. PMID: 26740620; PMCID: PMC4813485): 4614-4625https://doi.org/10.1074/jbc.M115.683912
        • Kleinman HK
        • Martin GR.
        Matrigel: basement membrane matrix with biological activity.
        Semin Cancer Biol. 2005 Oct; 15 (PMID: 15975825): 378-386https://doi.org/10.1016/j.semcancer.2005.05.004
        • Aisenbrey EA
        • Murphy WL.
        Synthetic alternatives to Matrigel.
        Nat Rev Mater. 2020 Jul; 5 (Epub 2020 May 27. PMID: 32953138; PMCID: PMC7500703): 539-551https://doi.org/10.1038/s41578-020-0199-8
        • Carter JD
        • Dula SB
        • Corbin KL
        • Wu R
        • Nunemaker CS.
        A practical guide to rodent islet isolation and assessment.
        Biol Proced Online. 2009 Dec 3; 11 (PMID: 19957062; PMCID: PMC3056052): 3-31https://doi.org/10.1007/s12575-009-9021-0
        • Zhang JH
        • Chung TD
        • Oldenburg KR.
        A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays.
        J Biomol Screen. 1999; 4 (PMID: 10838414): 67-73https://doi.org/10.1177/108705719900400206
        • Lundholt BK
        • Scudder KM
        • Pagliaro L.
        A simple technique for reducing edge effect in cell-based assays.
        J Biomol Screen. 2003 Oct; 8 (PMID: 14567784): 566-570https://doi.org/10.1177/1087057103256465
        • Kubbies M
        • Rabinovitch PS.
        Flow cytometric analysis of factors which influence the BrdUrd-Hoechst quenching effect in cultivated human fibroblasts and lymphocytes.
        Cytometry. 1983 Jan; 3 (PMID: 6185287): 276-281https://doi.org/10.1002/cyto.990030408
        • Rabinovitch PS.
        Regulation of human fibroblast growth rate by both noncycling cell fraction transition probability is shown by growth in 5-bromodeoxyuridine followed by Hoechst 33258 flow cytometry.
        Proc Natl Acad Sci U S A. 1983 May; 80 (PMID: 6190165; PMCID: PMC393951): 2951-2955https://doi.org/10.1073/pnas.80.10.2951