Protocol for screening and validating antibodies specific to protein phosphorylation sites using a set of yeast biopanning approaches

Summary

Developing antibodies with high specificity against post-translationally modified epitopes remains a challenge. Yeast biopanning is well suited in screening for high-specificity binders. Here, we present a protocol for screening and validating antibodies specific to protein phosphorylation sites using a set of yeast biopanning approaches. We describe steps for screening a yeast surface display library for antibodies and other binders. We then detail procedures for validating the antibodies found by analyzing their specificity through whole-well image analysis in 96-well plates.

For complete details on the use and execution of this protocol, please refer to Arbaciauskaite et al.¹

Graphical abstract

Highlights

• •A yeast biopanning approach for screening post-translational modification-specific binders
• •Rigorous quantitative validation of post-translational modification-specific binders
• •A whole-well image analysis pipeline for yeast biopanning

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Developing antibodies with high specificity against post-translationally modified epitopes remains a challenge. Yeast biopanning is well suited in screening for high-specificity binders. Here, we present a protocol for screening and validating antibodies specific to protein phosphorylation sites using a set of yeast biopanning approaches. We describe steps for screening a yeast surface display library for antibodies and other binders. We then detail procedures for validating the antibodies found by analyzing their specificity through whole-well image analysis in 96-well plates.

Before you begin

Yeast surface display protocols are widely used for screening antibody fragments² or other binder proteins.^3,4 Yeast biopanning is based on yeast surface display and has been primarily used for screening antibody libraries against cell surface proteins.^5,6,7 Since yeast cells have low non-specific binding to other cell surfaces,^5,8 yeast biopanning enables the screening of large binder libraries to find rare clones. Since high-specificity antibody clones that bind to protein phosphorylation or other post-translational modification sites are hard to find,^9,10,11 a method that allows streamlined antibody screening and validation is desired.

The two part protocol described below aims to address this need. In the first part, a protocol for screening a large yeast surface display binder library using 6-well plates is described. Peptides containing the target phosphorylation site are immobilized on a layer of human embryonic kidney (HEK) 293FT cell surface using biotinylation followed by streptavidin incubation. The peptide antigen sequence can be determined following the guidelines used for animal immunization using phospho-peptides.^12,13,14,15 Many commercial sources are available that synthesize custom sequence peptides with desired modifications. A peptide with the same amino acid sequence but without phosphorylation is also needed to perform a negative selection to remove clones that bind to non-phosphorylated protein. We note that yeast biopanning on peptides immobilized directly on a cell culture plate without an adherent cell layer was not successful.¹ We have shown that the degree of cell-surface biotinylation is a key parameter for leaving binding sites and successful immobilization of biotinylated peptides.¹

Note: This first part of the protocol describes the screening process for finding post-translational modification (PTM) site-specific antibodies or other binders from a library. Previously established yeast surface display libraries such as the human single-chain variable region fragment (scFv) or nanobody libraries (see key resources table) can be used for the screen. The maintenance, growth, and induction of these libraries have been previously described in detail.^2,4 Buffers and cell culture plates listed in the materials and equipment section should be prepared before beginning this part.

Prepare yeast cells

Timing: 3–4 days

Note: Yeast cell growth and induction should follow the protocol for each library.^2,4 Yeast cell growth prior to induction may take 2–3 days. The induction of the yeast cells to display the scFv or nanobodies typically takes 36–48 h.^2,4 These steps should be conducted before day 1.

• 1.Grow and induce yeast cells expressing binders such as a scFv library.
• 2.Measure optical density at 600 nm (OD₆₀₀) using a spectrophotometer to determine cell density.
• 3.
  Collect yeast cells at 10-fold excess of library size (or the number of clones recovered from the previous round).

  • a.
    For example, if the estimated library size is 5 × 10⁸, collect 5 × 10⁹ yeast.
• 4.Pellet and wash yeast cells twice in phosphate-buffered saline containing calcium, magnesium, and bovine serum albumin (PBSCM-BSA).
• 5.
  Resuspend yeast cells in PBSCM-BSA.

  • a.
    Achieve a volume of 1 mL per 3 × 10⁸ cells.
  • b.
    Keep cells on ice until needed for the next step.

Note: The induced yeast cells can be stored in PBSCM-BSA for 1–2 days at 4°C.

Prepare HEK293FT cells

Timing: 2 days

Note: The HEK293FT cells should be thawed and passaged according to the supplier’s instructions. Freshly thawed cells should be passaged at least twice before use for this biopanning step. For the biopanning experiment, we cultured the HEK293FT cells in DMEM supplemented with 10% (v/v) FBS (D10). These steps are to be conducted on day 1.

• 6.Dilute Matrigel 50-fold in cold DMEM. Diluted Matrigel can be stored at 4°C for two weeks.
• 7.
  Add 1 mL of diluted Matrigel per well to 6-well plates needed for the experiment and incubate at 37°C for at least 1 h.

  • a.
    Should prepare one well per 3 × 10⁸ yeast cells.
  • b.
    Meanwhile, warm up the necessary amounts of trypsin and D10 (5 mL per well).
• 8.Aspirate the diluted Matrigel from wells and passage HEK293FT cells into the Matrigel-coated wells. To obtain confluent cells, passage approximately 6 × 10⁵ cells per well in a 6-well plate.
• 9.Incubate the plate at 37°C overnight.

Note: Although we used the HEK293FT cells which have a faster growth rate, HEK293 or HEK293T cells may be used. The diluted Matrigel helps the adhesion of HEK293FT cells to the plate surface and is needed to prevent the loss of HEK293FT cells during washing. On the day of the biopanning experiment, it is recommended to check that the HEK29FT cells have grown to 100% confluency, creating a monolayer in the wells. If there are areas in the well that do not contain any cells, the plate can be left to incubate longer and checked later in the day. Additionally, an important thing to look for is any areas where the cells seem overgrown. Yeast cells used during biopanning may not bind well to these areas.

Prepare biopanning reagents

Timing: 30 min

• 10.
  Prepare streptavidin reagent.

  • a.
    Reconstitute streptavidin to 1 mg/mL following the manufacturer’s protocol.

    • i.
      For example, streptavidin can be resuspended in 1 mL H₂O.
  • b.
    Dilute the streptavidin stock to 6.25 μM in PBSCM-BSA to achieve a total volume of 1 mL per well in a 6-well plate.
• 11.
  Prepare biotinylation reagent.

  • a.
    Add 170 μL dimethyl sulfoxide (DMSO) to 2 mg NHS-PEG4-Biotin to make a 20 mM stock solution.

    Note: The biotinylation stock solution can be kept in −20°C and used for up to 14 days.

  • b.
    Dilute the stock to 125 μM in PBSCM-BSA to achieve a total volume of 1 mL per well in a 6-well plate.
• 12.
  Prepare biotinylated peptide.

  • a.
    Dilute biotinylated peptide to 0.1 μM in PBSCM-BSA to achieve a total volume of 1 mL per well in a 6-well plate.

Note: It is extremely important that all reagents are well mixed. All reagents should be kept on ice during the experiment.

Note: The second part of the protocol describes a 96-well plate experiment to validate the specificity of PTM site-specific antibodies or other binders that result from the yeast biopanning screening. Here, a 96-well plate is used to increase the throughput and enable rapid image-based validation of the binding specificity. Although we used this protocol to validate anti-tau PTM-specific scFvs, it is possible to validate PTM-site-specific scFvs or binders against other proteins.

Prepare plasmids

Timing: 1 week

Note: The plasmids below allow robust detection of yeast cells displaying a binder by intracellularly expressing a fluorescent protein. We provide two bi-directional yeast surface display plasmids (Addgene #218827 and #218828), each expressing a distinct fluorescent protein. Using these plasmids allows distinguishing the binder clone interaction from that of a negative control binder within the same well. Instead of cloning the isolated binder clone to these plasmids, the isolated binder yeast cell can be directly used in 96-well biopanning. For this, follow the preparation in the “prepare yeast cells (when validating clones from a library)” section below.

• 13.
  Use restriction enzyme cloning to insert scFv of interest into pBEVY-GFP backbone.

  • a.
    Restriction enzymes to use: NheI, BsrGI.
• 14.Verify plasmid sequence using Sanger or whole plasmid sequencing methods.
• 15.Transform plasmid construct into S. cerevisiae strain EBY100 using the Frozen-EZ Yeast Transformation II Kit following the manufacturer’s instructions.
• 16.Transform plasmid encoding pBEVY-4420-mCherry into S. cerevisiae strain EBY100.

Note: 4420 refers to the anti-fluorescein scFv 4-4-20¹⁶ used as a control scFv. If interested in switching out this scFv, use restriction sites NheI and BsrGI.

• 17.Grow on SD-CAA agar plates for 2–3 days.

Prepare yeast cells (when using bi-directional plasmids)

Timing: 3 days

• 18.Pick a single yeast colony and grow in 3 mL of SD-CAA medium at 30°C with shaking at 250 rpm overnight.
• 19.Measure optical density at 600 nm (OD₆₀₀). OD₆₀₀ = 1 roughly corresponds to 10⁷ yeast cells/mL.
• 20.Centrifuge cells at 4,000 rpm (3,000 × g) for 10 min and remove the spent medium by decanting.
• 21.Resuspend 10⁷ cells in 3 mL SG-CAA medium and incubate at 20°C with shaking at 250 rpm for 36–48 h. The induction time should follow that of the library used.

Note: Steps 22–25 should be carried out on the morning of the biopanning experiments (day 2).

• 22.Measure optical density at 600 nm (OD₆₀₀).

Note: It is recommended to check for the expression of the plasmids. This can be done by looking at the yeast cells under a fluorescence microscope. In the case of yeast cells expressing pBEVY-4420-mCherry, expression can be confirmed visually as the yeast cells may appear to be pink.

• 23.
  Separate number of yeast cells necessary for the biopanning experiment.

  • a.
    The number is determined by (10⁶ cells) × (# of wells in the 96-well plate included in the experiment).
• 24.Wash cells once in 500 μL PBSCM-BSA.
• 25.
  Resuspend yeast cells in PBSCM-BSA.

  • a.
    Achieve a volume of 50 μL per 10⁶ cells.
  • b.
    Mix equal volumes of the yeast cells displaying the binder clone to validate (e.g., cells transformed with pBEVY-pT231 scFv-GFP) and the control yeast cells (e.g., cells transformed with pBEVY-4420-mCherry).
  • c.
    Keep cells on ice until needed for the experiment.

Prepare yeast cells (when validating clones from a library)

Timing: 1 day

For this we still recommend using the control yeast transformed with the plasmid pBEVY-4420-mCherry (Addgene #218828) as described above. Since the binder yeast clones to be validated do not express a fluorescent protein, they are labeled using an anti-epitope tag antibody (see key resources table) expressed with the binder construct. Each library may use a different epitope tag to detect the expression of the binder. Commonly used epitopes include c-myc, hemagglutinin (HA), and FLAG tags.

• 26.Pick a colony of control yeast colony and inoculate in 3 mL of SD-CAA medium.
• 27.Pick yeast colonies and inoculate each in 3 mL of SD-CAA medium.
• 28.Grow cells at 30°C with shaking at 250 rpm overnight.
• 29.Measure optical density at 600 nm (OD₆₀₀).

Note: OD₆₀₀ = 1 roughly corresponds to 10⁷ yeast cells/mL.

• 30.Centrifuge cells at 4,000 rpm (3,000 × g) for 10 min and remove the spent medium by decanting.
• 31.For the control yeast, resuspend 5 × 10⁷ cells in 15 mL SG-CAA medium. This will be enough for validating 100 library colonies.
• 32.For each library clone, resuspend 10⁷ cells in 3 mL SG-CAA medium.
• 33.Incubate at 20°C with shaking at 250 rpm for 36–48 h. The induction time should follow that of the library used.

Note: The PBSCM-BSA buffer used below should be kept on ice at all times.

• 34.Measure OD₆₀₀.
• 35.For the control yeast, transfer to a conical tube and centrifuge at 4,000 rpm (3,000 × g) for 10 min and remove the spent medium. Resuspend the control yeast in PBSCM-BSA at a density of 2 × 10⁷ cells/mL (10⁶ cells/50 μL). Set aside the control yeast on ice.

Note: The following steps are for fluorescent staining of the library clones. The control yeast already expresses mCherry so staining is unnecessary.

• 36.For each library clone, transfer 2 × 10⁶ cells for each clone to a 1.5 mL microcentrifuge tube.
• 37.Centrifuge cells at 14,000 rpm for 1 min using a microcentrifuge.
• 38.Remove the spent medium by aspiration and resuspend cells in 500 μL PBSCM-BSA.
• 39.In PBSCM-BSA, dilute an antibody that binds to the epitope tag (e.g., mouse anti-c-Myc tag antibody).

Note: Typically, purified antibodies at 1 mg/mL concentration are diluted 100-fold. This may vary depending on the library used.^2,4,17 For each 2 × 10⁶ yeast cell sample, prepare 100 μL of diluted peptide/antibody mixture.

• 40.Centrifuge cells at 14,000 rpm for 1 min and remove the supernatant by aspiration.
• 41.Resuspend cells using the diluted antibody.
• 42.Incubate the cells for 1 h on ice.
• 43.Dilute secondary detection reagent in PBSCM-BSA.

Note: Anti-mouse IgG conjugated with Alexa 488 (aM488) is used for detecting the anti-epitope tag antibody. aM488 is diluted 100-fold in PBSCM-BSA. For each 2 × 10⁶ yeast cell sample, prepare 100 μL of diluted detection reagents.

• 44.Centrifuge cells at 14,000 rpm for 1 min and remove the supernatant by aspiration.
• 45.Wash cells by resuspending in 500 μL PBSCM-BSA.
• 46.Centrifuge cells at 14,000 rpm for 1 min and remove the supernatant by aspiration.
• 47.Repeat washing steps 44–46.
• 48.Resuspend cells using the diluted secondary detection reagent.
• 49.Incubate the cells for 30 min on ice.
• 50.Wash cells as in steps 44–47.
• 51.Resuspend cells in 100 μL PBSCM-BSA.
• 52.Mix equal volumes of the library clone to validate and the control yeast cells.
• 53.Keep cells on ice until needed for the experiment.

Prepare HEK293FT cells (required for all 96-well biopanning)

Timing: 2 days

Note: The note on HEK293FT cultivation in the 6-well biopanning described above applies here.

• 54.Dilute Matrigel 50-fold in cold DMEM. Diluted Matrigel can be stored at 4°C for two weeks.
• 55.
  Add 50 μL of diluted Matrigel to wells in 96-well tissue culture plates needed and incubate at 37°C for at least 1 h.

  • a.
    Meanwhile, warm up the necessary amounts of trypsin and D10.
• 56.Passage HEK293FT cells and add cells to wells containing the diluted Matrigel.
• 57.Incubate plate at 37°C overnight.

Note: On the day of the biopanning experiment, it is recommended to check that the HEK293FT cells have grown to 100% confluency, creating a monolayer in the wells. If there are areas in the well that do not contain any cells, the plate can be left to incubate longer and checked later in the day. Additionally, an important thing to look for is any areas where the cells seem overgrown (multiple layers). Yeast cells used during biopanning may not bind well to these areas.

Prepare biopanning reagents (required for all 96-well biopanning)

Timing: 30 min

• 58.
  Prepare streptavidin reagent.

  • a.
    Prepare streptavidin stock solution in accordance with the manufacturer’s protocol.

    • i.
      For example, streptavidin (Cat. No. 85878-1MG, Sigma) can be resuspended in 1 mL H₂O.
  • b.
    Dilute stock to the necessary concentration in PBSCM-BSA to achieve a total volume of 50 μL per well used.
  • c.
    Final streptavidin concentrations depend on the affinity of the scFv and the following should be used:

    • i.
      High affinity scFv (K_d sub-nM): 155 nM.
    • ii.
      Moderate affinity scFv (K_d low nM): 1.25 μM.
    • iii.
      Low affinity scFv (K_d mid to high nM): 6.25 μM.

Note: If the binding affinity is unknown, use the low affinity concentration.

• 59.
  Prepare biotinylation reagent.

  • a.
    Add 170 μL DMSO to 2 mg NHS-PEG4-Biotin to make a 20 mM stock solution.

    Note: Stock solution can be kept in −20°C and used for up to 14 days.

  • b.
    Dilute stock to 125 μM in PBSCM-BSA to achieve a total volume of 50 μL per well used.
• 60.
  Prepare biotinylated peptide.

  • a.
    Dilute biotinylated peptide to 0.1 μM in PBSCM-BSA to achieve a total volume of 50 μL per well used.

Note: It is extremely important that all reagents are well mixed. All reagents should be kept on ice during the experiment.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-c-myc mouse monoclonal antibody, clone 9E10 (1:100 dilution)	Thermo Fisher Scientific	Cat#13-2500
Anti-FLAG mouse monoclonal antibody, clone M2 (1:500 dilution)	MilliporeSigma	Cat#F1804
Anti-HA mouse monoclonal antibody, clone 12CA5 (1:100 dilution)	Thermo Fisher Scientific	Cat#50-139-3662
Anti-mouse IgG goat secondary antibody, Alexa 488 conjugate (1:100 dilution)	Thermo Fisher Scientific	Cat#A-21236
Chemicals, peptides, and recombinant proteins
Streptavidin	MilliporeSigma	Cat#85878-1MG
Streptavidin, R-phycoerythrin conjugate	Thermo Fisher Scientific	Cat#S866
Biotinylation reagent, EZ-link NHS-PEG4-Biotin	Thermo Fisher Scientific	Cat# 21330
Matrigel	Corning	Cat#354234
Dimethyl sulfoxide (DMSO)	MilliporeSigma	Cat#D2650; CAS: 67-68-5
Tetracycline hydrochloride	MilliporeSigma	Cat#T7660; CAS: 64-75-5
Carbenicillin disodium salt	MilliporeSigma	Cat#C1389; CAS: 4800-94-6
Streptomycin sulfate salt	MilliporeSigma	Cat#S6501; CAS: 3810-74-0
Glycerol	Thermo Fisher Scientific	Cat#AC158920010; CAS: 56-81-5
pT231 phospho-tau peptide (KKVAVVR(pT)PPKSPSSAK-biotin)	Peptide 2.0	N/A
Non-phospho tau peptide (KKVAVVRTPPKSPSSAK-biotin)	Peptide 2.0	N/A
Dulbecco’s modified Eagle’s medium (DMEM)	Thermo Fisher Scientific	Cat#11965092
Fetal bovine serum	Cytiva	Cat#SH3008803
0.05% Trypsin/0.53 mM EDTA in HBSS without calcium and magnesium	Thermo Fisher Scientific	Cat#MT-25-051-CI
Frozen-EZ Yeast Transformation II Kit	Zymo Research	Cat#T2001
Experimental models: Cell lines
Human embryonic kidney (HEK) 293FT	Thermo Fisher Scientific	Cat# R70007; RRID: CVCL_6911
Recombinant DNA
pBEVY-pT231 scFv-GFP	Addgene	#218827
pBEVY-4420-mCherry	Addgene	#218828
Others
Yeast surface display human scFv library	Wittrup Lab; https://kdw-lab.mit.edu/reagents/	N/A
Yeast surface display nanobody library	Kerafast	EF0014-FP
Benchtop centrifuge	Eppendorf	5810
Microcentrifuge	Thermo Fisher Scientific	Sorvall Legend Micro 21
Incubator with a shaking platform	New Brunswick	G-25
Incubator	Thermo Fisher Scientific	S42809
NanoDrop 1000 spectrophotometer	Thermo Fisher Scientific	ND-1000
Keyence all-in-one fluorescence microscope	Keyence	BZ-X810
Flow cytometer	BD Biosciences	LSRFortessa X-20
Freezing container	Corning	CoolCell FTS30

Materials and equipment

Washing and dilution buffer (PBSCM-BSA)

Reagent	Final concentration	Amount
NaCl	137 mM	8 g
KCl	2.7 mM	0.2 g
Na2HPO4	10 mM	1.44 g
KH2PO4	1.8 mM	0.24 g
Bovine Serum Albumin (BSA)	0.1% (w/v)	1 g
CaCl2	1 mM	–
MgCl2	0.5 mM	–
ddH2O	–	1 L
Total	–	1 L

Store at 4°C for up to 6 months.

Quenching buffer (PBSCM + glycine)

Reagent	Final concentration	Amount
NaCl	137 mM	8 g
KCl	2.7 mM	0.2 g
Na2HPO4	10 mM	1.44 g
KH2PO4	1.8 mM	0.24 g
Glycine	0.1 M	–
CaCl2	1 mM	–
MgCl2	0.5 mM	–
ddH2O	–	1 L
Total	–	1 L

Store at 4°C for up to 6 months.

SD-CAA medium

Reagent	Final concentration	Amount
Dextrose	20 g/L	20 g
Yeast N2 Base	6.7 g/L	6.7 g
Cas Amino Acids	5 g/L	5 g
ddH2O	–	1 L
Total	–	1 L

Store at 4°C for up to 12 months.

SD-CAA low glucose

Reagent	Final concentration	Amount
Dextrose	5 g/L	20 g
Yeast N2 base	6.7 g/L	6.7 g
Cas amino acids	5 g/L	5 g
ddH2O	–	1 L
Total	–	1 L

Store at 4°C for up to 12 months.

SD-CAA plates

Reagent	Final concentration	Amount
Dextrose	20 g/L	20 g
Yeast N2 base	6.7 g/L	6.7 g
Cas amino acids	5 g/L	5 g
Agar	14 g/L	14 g
ddH2O	–	1 L
Total	–	1 L

Store at 4°C for up to 3 months.

SG-CAA medium

Reagent	Final concentration	Amount
Galactose	20 g/L	20 g
Yeast N2 base	6.7 g/L	6.7 g
Cas amino acids	5 g/L	5 g
ddH2O	–	1 L
Total	–	1 L

Store at 4°C for up to 12 months.

Yeast freezing medium

Reagent	Final concentration	Amount
Glycerol	2% (v/v)	2 mL
Yeast N2 base	6 g/L	0.6 g
ddH2O	–	100 mL
Total	–	100 mL

Store at 4°C for up to 12 months.

Step-by-step method details

Screening: 6-well plate yeast cell biopanning

Timing: 2 days (per round)

The steps described here are the “peptide immobilization” and “yeast cell biopanning” that occur on day 2. Yeast cells displaying an scFv library or other binders are screened via biopanning against immobilized peptide antigens.

• 1.Wash wells containing HEK293FT cells twice with 2 mL of PBSCM-BSA each time.

Note: PBSCM-BSA buffer should be kept on ice throughout the experiment to ensure that it is ice cold when used for washing. The biotinylation reagent and quenching buffer should be at room temperature.

• 2.
  Add 1 mL of the biotinylation reagent to each well.

  • a.
    Incubate for 30 min at room temperature.
  • b.
    Wash wells 2 times with 2 mL PBSCM-BSA.
• 3.
  To quench the reaction, add 1 mL of quenching buffer to each well.

  • a.
    Incubate for 10 min at room temperature.
  • b.
    Wash wells 2 times with 2 mL PBSCM-BSA.
• 4.
  Add 1 mL of the streptavidin reagent to each well.

  • a.
    Incubate for 30 min at room temperature.
  • b.
    Wash wells 2 times with 2 mL PBSCM-BSA.
• 5.
  Add 1 mL of biotinylated peptide to each well.

  • a.
    Incubate for 30 min at room temperature.
  • b.
    Wash wells 2 times with 2 mL PBSCM-BSA.
• 6.
  Add 1 mL of yeast cells (prepared at 3 × 10⁸ cells/mL) to each well.

  • a.
    Cover the plate lid with aluminum foil and incubate for 30 min at room temperature.

Note: For the following washing steps (steps 7–9), refer to supplementary Methods video S1.

• 7.
  Carry out the first round of washing.

  • a.
    Remove non-binding yeast by aspiration.
  • b.
    Dispense 1 mL of PBSCM-BSA dropwise to each well.
  • c.
    Gently rock the plate 25 times.
  • d.
    Gently rotate the plate 5 times.
  • e.
    Remove supernatant by aspiration.
• 8.
  Carry out the second round of washing.

  • a.
    Repeat step 7 with fresh 1 mL of PBSCM-BSA.
• 9.
  Carry out the third round of washing.

  • a.
    Dispense 1 mL of PBSCM-BSA dropwise to each well.
  • b.
    Gently rotate the plate 10 times.
  • c.
    Remove supernatant by aspiration.
• 10.Add 1 mL of PBSCM-BSA to each well.
• 11.Scrape bound cells off the wells, collect cells, transfer them to a 15 mL conical tube, and centrifuge at 4,000 rpm (3,000 × g) for 10 min using a benchtop centrifuge.
• 12.Resuspend the cells in 10 mL SD-CAA supplemented with tetracycline (10 μg/mL), carbenicillin (100 μg/mL), and streptomycin (500 μg/mL).
• 13.
  Calculate library size by doing serial dilution and plating on SD-CAA plates.

  • a.
    Let yeast grow for 2 h.

    Note: Letting the yeast cells grow is to reduce the risk of losing rare clones during initial rounds of screening. Since yeast cells may grow during this time, the estimated number of recovered clones during the initial rounds would be off by 2–4 fold. However, even a rough estimation is helpful in determining the number of cells to be screened in subsequent rounds.

  • b.
    Collect 10 μL, add 990 μL media, and mix well.
  • c.
    Collect 100 μL of first dilution (b), add 900 μL media, and mix well.
  • d.
    Collect 100 μL of second dilution (c), add 900 μL media, and mix well.
  • e.
    Plate 100 μL of (b), (c), and (d) on separate SD-CAA plates and grow for 3–4 days.

    • i.
      The dilution factor for plate resulting from (b) is 10⁴.
    • ii.
      The dilution factor for plate resulting from (c) is 10⁵.
    • iii.
      The dilution factor for plate resulting from (d) is 10⁶.
  • f.
    Calculate library size by multiplying the number of colonies present by the dilution factor.

    Note: If needed, further dilutions can be carried out. This marks the end of screening round 1 (day 2). For enriching the binder population, typically 3–4 rounds of screening is needed.

    Pause point: Yeast cells can be stored at 4°C for up to 1 week. Additionally, calculating the library size (number of collected yeast cells after biopanning) requires growing yeast cells on selective plates and could take 3–4 days. The next steps describe how to propagate the collected yeast cells for the next round and freeze them for long-term storage.

• 14.Incubate the yeast cells at 30°C overnight with shaking at 250 rpm to propagate the cells.
• 15.
  Passage cells.

  Note: For the next round, passaging the yeast cells is necessary to remove HEK293FT cell debris carried over during the scraping. Cells may be passaged 2–3 times to completely remove the debris. You may store the overnight culture at 4°C to wait for the plated colonies to grow and estimate the library size. If it is necessary to passage cells before calculating true library size, make sure to passage many cells to ensure that library diversity is preserved. We generally passage 10-fold excess of the library size.

  • a.
    Mix 1 mL of the overnight culture with 10 mL fresh SD-CAA, which generally ensures oversampling of the enriched pool.
  • b.
    Incubate the yeast cells at 30°C overnight with shaking at 250 rpm.
• 16.Measure OD₆₀₀ and calculate the volume of culture needed to collect at least 10-fold excess of the library size.
• 17.Centrifuge cells at 4,000 rpm (3,000 × g) for 10 min using a benchtop centrifuge.
• 18.Induce yeast cells overnight in SG-CAA at 20°C with shaking.
• 19.Repeat steps 1–18 for subsequent rounds of biopanning.

Note: After each round of biopanning, we recommend freezing the collected pool of yeast cells to check the enrichment of binders later. Also, the frozen pool can be used for screening in case the cells are contaminated during further rounds of biopanning.

• 20.To freeze the collected yeast cells, passage 10-fold excess of the library size in SD-CAA and grow at 30°C for 2 days.
• 21.Measure OD₆₀₀ and passage at least 10-fold excess of the library size in SD-CAA low glucose medium and grow at 30°C for 40–45 h.
• 22.Measure OD₆₀₀ and centrifuge all cells at 4,000 rpm (3,000 × g) for 10 min.
• 23.Resuspend cells in yeast freezing medium at a density of 10 times the library size per mL.
• 24.Transfer 1 mL into each cryovial and place the vials in a delayed freezer container such as the Corning CoolCell or Thermo Fisher Mr. Frosty container. The next day the vials can be transferred to a regular freezer box.

Note: After initial rounds of screening (1–2 rounds), negative (subtractive) screening should be performed by following the modifications outlined in step 25.

• 25.
  To carry out negative screening, use biotinylated non-phospho peptide (before you begin, step 12).

  • a.
    Collect cells from all washing steps (steps 7e, 8e, and 9c) and centrifuge at 4,000 rpm (3,000 × g) for 10 min.
  • b.
    Carry out steps 12–18 as previously described.

Note: We recommend checking the enrichment of binders after 3–4 rounds of biopanning. The enrichment can be assessed using flow cytometry or 96-well biopanning described below. Note that yeast biopanning allows the identification of low-affinity binders that may not show detectable binding in flow cytometry. Therefore, we recommend validating individual binder clones using the 96-well validation approach described below. Flow cytometry is suitable for determining the fraction of cells that display binders against the phospho-peptide or the non-phospho-peptide. The same biotinylated phospho- and non-phospho-peptides can be used for detecting peptide-binder interaction using flow cytometry. The following steps describe how to check the enrichment using flow cytometry.

• 26.To include frozen vials of cells from previous rounds, thaw the cells in a 30°C water bath (or water pre-heated in an incubator).
• 27.Grow the cells overnight in SD-CAA at 30°C with shaking at 250 rpm.
• 28.Dilute 1 mL of the overnight culture to 10 mL of SD-CAA and grow them overnight at 30°C with shaking. Repeat this twice to dilute dead cells.
• 29.Centrifuge cells at 4,000 rpm (3,000 × g) for 10 min and remove the spent medium by decanting.
• 30.Induce yeast cells by resuspending cells in SG-CAA at a density of 10⁷ cells/3 mL and incubating at 20°C with shaking at 250 rpm for 36–48 h. The induction time should follow that of the library used.

Note: After induction, the cells may be stored at 4°C for 2–3 days before flow cytometry analysis. Each library may use a different epitope tag to detect the expression of the binder. The PBSCM-BSA buffer used below should be kept on ice at all times.

• 31.Measure OD₆₀₀ and transfer 2 × 10⁶ cells from each pool in a 1.5 mL microcentrifuge tube.
• 32.Centrifuge cells at 14,000 rpm for 1 min using a microcentrifuge.
• 33.Remove the spent medium by aspiration and resuspend cells in 500 μL PBSCM-BSA.
• 34.In PBSCM-BSA, dilute the biotinylated peptide and an antibody that binds to the epitope tag for checking binder expression (e.g., mouse anti-c-Myc tag antibody).

Note: The epitope tag antibody and the biotinylated peptide are mixed to label the yeast cells. Typically, purified antibodies at 1 mg/mL concentration are diluted 100-fold. This may vary depending on the library used.^2,4,17 For the biotinylated peptide, dilute phospho- or non-phospho-peptides at 1 μM. Binding to both peptides should be tested to determine the fraction of non-specific binders. For each 2 × 10⁶ yeast cell sample, prepare 100 μL of diluted peptide/antibody mixture.

• 35.Centrifuge cells at 14,000 rpm for 1 min and remove the supernatant by aspiration.
• 36.Resuspend cells using the diluted peptide/antibody mixture.
• 37.Incubate the cells for 1 h on ice.
• 38.Dilute secondary detection reagents in PBSCM-BSA.

Note: Streptavidin conjugated with phycoerythrin (SA-PE) and anti-mouse IgG conjugated with Alexa 488 (aM488) are used for detecting the biotinylated peptide and the anti-epitope tag antibody, respectively. The fluorophore conjugates may be changed depending on the optical detection scheme of the flow cytometer used. SA-PE and aM488 are both diluted 100-fold in PBSCM-BSA. For each 2 × 10⁶ yeast cell sample, prepare 100 μL of diluted detection reagents.

• 39.Centrifuge cells at 14,000 rpm for 1 min and remove the supernatant by aspiration.
• 40.Wash cells by resuspending in 500 μL PBSCM-BSA.
• 41.Repeat washing steps once more.
• 42.Centrifuge cells at 14,000 rpm for 1 min and remove the supernatant by aspiration.
• 43.Resuspend cells using the diluted secondary detection reagents.
• 44.Incubate the cells for 30 min on ice.
• 45.Wash cells as in steps 39 to 41.
• 46.Resuspend cells in 500 μL PBSCM-BSA. The cells are ready for flow cytometry.

Note: The labeled cells can be stored on ice for a few hours. Flow cytometers such as the BD Biosciences LSRFortessa X-20 or Miltenyi Biotec MACS Quant VYB can be used. Details on performing flow cytometry of yeast cells have been published previously.¹⁸ Once binders have been enriched, the pool can be plated on SD-CAA plates to isolate single colonies.

Validation: 96-well plate yeast cell biopanning

Timing: 3–4 h

Although the negative screening step reduces the fraction of binders that cross-react with the non-phosphorylated peptide, it is essential to check the specificity of each clone obtained from the screen. The 96-well biopanning step allows streamlined specificity assessment of clones quickly. The following steps allow validating single clones isolated from the 6-well screening. The steps described here are the “peptide immobilization” and “yeast cell biopanning” that occur on day 2. In addition, the whole-well imaging step also should be completed on day 2, taking up all day. Therefore, we recommend starting the “peptide immobilization” early in the morning.

• 47.Wash wells containing HEK293FT cells 2 times with 100 μL of PBSCM-BSA.

Note: PBSCM-BSA buffer should be kept on ice throughout the experiment to ensure that it is ice cold when used for washing.

• 48.
  Add 50 μL of the biotinylation reagent to each well.

  • a.
    Incubate for 30 min at room temperature.
  • b.
    Wash wells 2 times, each with 100 μL PBSCM-BSA.
• 49.
  To quench the reaction, add 50 μL quenching buffer to each well.

  • a.
    Incubate for 10 min at room temperature.
  • b.
    Wash wells 2 times, each with 100 μL PBSCM-BSA.
• 50.
  Add 50 μL of the streptavidin reagent to each well.

  • a.
    Incubate for 30 min at room temperature.
  • b.
    Wash wells 2 times, each with 100 μL PBSCM-BSA.
• 51.
  Add 50 μL of biotinylated peptide to each well.

  • a.
    Incubate for 30 min at room temperature.
  • b.
    Wash wells 2 times, each with 100 μL PBSCM-BSA.
• 52.
  Add 50 μL of yeast cells to each well.

  • a.
    Cover the plate lid with aluminum foil and incubate for 30 min at room temperature.
  • b.
    Wash wells once with 100 μL PBSCM-BSA.

Note: For the following round of washing, the 96-well plate should be held by the experimenter and tilted in accordance with the sides of the well that are being targeted. Reference supplementary Methods video S2 for guidance.

• 53.
  Carry out the first round of washing.

  • a.
    Dispense 100 μL of PBSCM-BSA to one side of a well wall (for example, the east side) and aspirate liquid from the opposite side (for example, the west side).

    • i.
      Dispense and aspirate buffer from a different combination of well sides (for example, dispense to the north side and aspirate from the south side).
    • ii.
      Repeat once more using a different combination of well walls (for example, dispense to the northeast side and aspirate from the southwest side).
    • iii.
      Dispense buffer into a waste beaker.
• 54.
  Carry out the second round of washing.

  • a.
    Repeat step 53 with fresh 100 μL of PBSCM-BSA.

Note: Make sure to dispense buffer to well sides that were aspirated from previously. For example, if the first round of washing included dispensing to the north side of the well and aspirating from the south side of the well, then dispensing to the south side and aspirating from the north side of the well needs to be accounted for. Reference supplementary Methods video S2 for guidance.

Note: Make sure to include combinations of well sides that were not included previously.

• 55.
  Carry out the third and final round of washing.

  • a.
    Repeat step 53 with fresh 100 μL of PBSCM-BSA.

Note: At this point, each well should have been washed a total of four times—once from step 52 and three times from steps 53–55.

CRITICAL: It is extremely important that the rounds of washing described above are carried out with attention to dispensing to/aspirating from all possible combinations of the sides of the well (should equal 8 total combinations). Doing so ensures that yeast cells are washed from all possible angles. See supplementary Methods video S2 for reference and guidance on the washing steps.

• 56.Add 100 μL of PBSCM-BSA to the wells.

Full-well imaging

Timing: 1–6 h

We used the Keyence BZ-X810 microscope and the BZ-X800 viewer software to automatically capture fluorescence images from an entire well. Other automated fluorescence microscopes have similar functionalities.

Image 96 well plate using a scanning microscope that can image GFP and mCherry fluorescence.

• 57.
  Load 96 well plate into a scanning microscope.

  • a.
    Select 96 well microscope setting on microscope software under sample holder.
• 58.
  Move the plate so that the center of a well containing a sample is centered.

  • a.
    Use 10x objective lens and select the fluorescence channels to image (GFP or mCherry, Figure 1A).
  • b.
    Focus the microscope on the sample and adjust the exposure level as appropriate (Figure 1A).
• 59.
  Set capture area settings for each well using the “Set Edge Points” setting (setting area and focus points, Figures 1B and 1C).

  • a.
    Navigate the microscope to the edge of the well and focus the sample.

    • i.
      Save this setting to focus point 1 (Figure 1C). This will save the x, y, and z positions. This will allow imaging the well without autofocusing at each point.
    • ii.
      Switch to the mCherry channel and re-focus. Save the setting to focus point 1.

      Note: The BZ-X810 microscope has a navigation function, which will automatically take a crude image of the whole well. The crude image helps to quickly move to different areas of the well when setting the focus points. To use the navigation function, move to the center of the sample well and select navigation.

  • b.
    Continue setting different focus points within the same well.

    • i.
      If using the microscope’s autofocus function (Figure 1B) to capture the images, it is enough to set a total of 5 focus points: middle, north edge, west edge, south edge, and east edge of the well.
    • ii.
      If not using the microscope’s autofocus function to capture the images, it is recommended to set additional focus points between the middle of the well and the edge points.
  • c.
    To focus following wells, make sure to change the registering area.

    • i.
      Sample 1 would be area 1, sample 2 would be area 2, etc.
• 60.Start capture and wait for microscope to take images of the samples.

Note: Using the 10x objective, each well generally requires around 50 images to cover the entire area.

Pause point: Depending on the number of samples, the microscope may take a significant amount of time to finish imaging. In a typical experiment, it may take around 4–5 h to image 25 wells.

• 61.
  Stitch images that have multiple fluorescence channels (for example, mCherry and GFP).

  • a.
    Using the microscope’s image analyzer software, find the image stitching function.
  • b.
    Load in the group file of a sample’s image and stitch the image.
  • c.
    Save the stitched image that includes merged channels (Figure 2).

Note: The following image analysis approach was used in our published results for quantitative assessment of binder specificity.¹

Figure 1.

Example microscope setup

A setup is shown for automated whole-well imaging using the Keyence BZ-X810 microscope.

(A) Illumination and camera setup.

(B) Navigating to different well positions and marking the x, y positions.

(C) Setting the capture area using the “Set Edge Points” setting.

Figure 2.

An example of a stitched whole-well image

A stitched whole-well image of an overlay of GFP and mCherry channels obtained from a 96-well biopanning experiment conducted using yeast cells transformed with the pBEVY-pT231 scFv-GFP plasmid or the pBEVY-4420-mCherry plasmid.

(A) Stitched whole-well image from a well immobilized with the pT231 phospho-tau peptide. Green and red indicate GFP and mCherry fluorescence, respectively. A zoomed-in image of regions 1 (B) and 2 (C) boxed in yellow in panel (A). The scale bar in (A) indicates 500 μm, and the scale bars in (B) and (C) indicate 100 μm.

Image analysis

Timing: 1 h

Analyze whole-well images using CellProfiler¹⁹ and MATLAB.

• 62.Load in the stitched image with merged channels of the sample to analyze to CellProfiler.
• 63.
  Run the pipeline attached.

  • a.
    CellProfiler will first convert the loaded stitched image to two separate grayscale image files reading the red colored channel as cells expressing mCherry and the green colored channel as cells expressing GFP.
  • b.
    The Otsu thresholding method is then used to identify individual cells from the grayscale images (Figure 3).
  • c.
    The pipeline finally outputs three Excel spreadsheets containing the following information: total cell counts from each stitched image loaded into the pipeline, x, y coordinate data for the identified cells expressing GFP, and x, y coordinate data for the identified cells expressing mCherry.
• 64.Load the Excel files to MATLAB.
• 65.
  Run the attached MATLAB code.

  • a.
    The code will first rescale the location of each cell identified so that the radii of the wells are uniform with a unit of 1.
  • b.
    The code removes any imaging artifacts that appear outside of the threshold of the well.
  • c.
    Heat maps are then created using the rescaled location data of all the identified cells.
  • d.
    For a specified number of radii (in our case, we used n = 40), the fraction of total cells (either GFP-specific scFv or mCherry-control scFv) contained at or within each radius is calculated.
  • e.
    The fraction of cells encompassed is then averaged over the data set and plotted with error bars.
  • f.
    The MATLAB code will output figures showing the individual heat maps for the loaded Excel files and the average radial distribution plots (Figure 4).

Note: A clear difference should appear when looking at yeast cells interacting specifically with the target antigen vs. non-specifically with control antigens or a control scFv (Figure 4).

Figure 3.

An example image analysis result from the CellProfiler pipeline

The stitched whole-well image shown in Figure 2 was processed using the CellProfiler pipeline. Fluorescent objects identified from the GFP (A) and the mCherry (B) channels are shown. Also shown are the number of objects identified for the GFP (C) and mCherry (D) channels. The pipeline also generates Excel files containing the x and y coordinates (E) of the identified objects in each channel.

Figure 4.

Exemplar output from the MATLAB code

Heat maps that depict the normalized density of yeast cells in each well. (A) and (B) are heat maps generated using the objects identified using CellProfiler. (A) and (B) correspond to heat maps of Figures 4A and 4B, respectively.

(C) The average radial distributions of wells (A) (blue) and (B) (red).

Expected outcomes

Representative flow cytometry results after 4 rounds of 6-well plate biopanning are shown in Figure 5. This mock biopanning screen was executed on wells immobilized with the pT231 phospho-tau peptide. Yeast cells transformed with either pBEVY-pT231 scFv-GFP plasmid or the pBEVY-4420-mCherry plasmid mixed at a ratio of 1:1,000 (GFP:mCherry) and subjected to biopanning. The pBEVY plasmids take advantage of the bi-directional transcription activity of the GAL1-10 promoter for surface display of scFvs and intracellular expression of a fluorescent protein reporter.¹ The phospho-specific pT231 scFv used here has an affinity (K_D) of approximately 60 nM. In this example, the scFv displaying cells contained a fluorescent reporter which will not be the case when screening scFv or other binder libraries. Therefore, the enrichment of binders should be assessed using a peptide containing the target PTM modification. In most binder libraries, the binder protein contains an epitope tag to check for its surface display, which ensures the measurement of antigen binding to yeast cells displaying the binder protein.² After a significant binder population is enriched as in Figure 5, fluorescence-activated cell sorting can be used to further enrich the binder population.

Figure 5.

Result from a 6-well biopanning screen

A mock 6-well biopanning result starting from a mixed yeast population of cells carrying the pBEVY-pT231 scFv-GFP plasmid (0.1%) or the pBEVY-4420-mCherry plasmid (99.9%). After each round of biopanning, the fraction of yeast cells expressing GFP or mCherry was determined using flow cytometry.

Once binders are isolated, they can be validated using the 96-well biopanning. The 96-well biopanning result is quantified through the whole-well imaging and image analysis steps. A representative stitched whole-well image is shown in Figure 2. The image was obtained using yeast cells transformed with the pBEVY-pT231 scFv-GFP plasmid and those transformed with the pBEVY-4420-mCherry plasmid biopanned in a well immobilized with the pT231 phospho-tau peptide. The stitched GFP and mCherry images were overlayed. As can be seen from zoomed-in images of regions 1 and 2, non-specific binding as indicated by the mCherry positive cells is more prevalent towards the edge of the well. The whole-well image analysis was developed to quantify this pattern.

As seen in the heat maps (Figures 4A and 4B), yeast cells that express the pT231 scFv show a relatively even distribution within the well (Figure 4A), while the control yeast expressing the 4420 scFv are enriched toward the well edge (Figure 4B). The average radial distribution (Figure 4C) allows a quantitative comparison of this trend. In addition to the number of identified objects, the distribution of cells provides evidence of yeast binding due to specific scFv-peptide interaction. In cases where specific peptide-scFv interactions occur, bound cells can be found in all regions of the well. Cells that are non-specifically bound tend to be enriched towards the edge of the well.

Limitations

As seen in Figure 4, yeast biopanning is not free from incomplete washing of non-specifically bound yeast cells, particularly near the edge of the wells. Although we haven’t conducted a similar imaging study in 6-well plates, we anticipate similar patterns of incomplete washing. This may delay the enrichment of the binder population and limit the purity of binders after several rounds of biopanning. This problem can be overcome by conducting a fluorescence-activated cell sorting (FACS) screen of the final population of cells. Since we used biotinylated peptides, the same reagents can be used for detection using flow cytometry.

Troubleshooting

Problem 1

Enriched yeast cell population not detected after 3–4 rounds of 6-well biopanning (after step 46).

Potential solution

• •For enriching rare clones, more than 4 rounds of biopanning may be needed.
• •If you screened one library, try screening a different library.

Problem 2

Fraction of binder population (determined after step 46) does not increase after several rounds of 6-well biopanning.

Potential solution

• •This may be due to many unwashed cells toward the edge of the plate. You may increase the number of washing steps to increase the stringency.
• •If a positive binding population is detected using flow cytometry, use FACS to further increase the fraction of binder cells.

Problem 3

Yeast cells seem to not be binding to parts of the well as expected (assessed after step 61).

Potential solution

• •View the wells under a microscope and check HEK293FT confluency. Note if there are areas that do not contain HEK293FT cells or areas where cells seem to be overgrown, both could result in yeast cells not binding. Change HEK293FT incubation time or initial seeding volume accordingly for future experiments.

Problem 4

When imaging wells (steps 59–60), there are artifacts appearing to be fluorescent that are not yeast cells.

Potential solution

• •Clean the bottom of the plate with a wipe to remove any dust.

Problem 5

There is a very large amount of yeast cells collecting along the well edge (assessed after step 61).

Potential solution

• •Ensure the washing steps have all been carried out.
• •When dispensing and aspirating during washing, make sure the tip of the pipette is touching the well wall.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Yongku Cho (cho@uconn.edu).

Technical contact

Technical questions on executing this protocol should be directed to and will be answered by the technical contact, Monika Arbaciauskaite (monika.crowl@uconn.edu).

Materials availability

This protocol uses plasmids generated for the manuscript by Arbaciauskaite et al.¹ The plasmids are available through Addgene (#218827 and #218828).

Data and code availability

The MATLAB code generated for this protocol is available on Zenodo: https://doi.org/10.5281/zenodo.12667020.