Phage & yeast display
A library of fragments fused to a filamentous phage coat protein is incubated with immobilised target, the well is washed, bound phage are eluted with trypsin or acid, and the eluate is amplified in fresh bacteria to give the input for the next round. Three to four rounds moves a clone from one in a billion to the majority of the pool, provided the plate and the library are properly blocked and a no-target well is panned alongside.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
Phage display biopanning
A library of fragments fused to a filamentous phage coat protein is incubated with immobilised target, the well is washed, bound phage are eluted with trypsin or acid, and the eluate is amplified in fresh bacteria to give the input for the next round. Three to four rounds moves a clone from one in a billion to the majority of the pool, provided the plate and the library are properly blocked and a no-target well is panned alongside.
How it differs from the standard method
- Stringency is raised by washing longer, coating less target, or competing with soluble target rather than by adding rounds.
- Monovalent phagemid display reports off-rates; polyvalent display reports avidity and selects weak binders.
- The eluate must be amplified between rounds or the rare clone is physically lost.
- Selection can be done on beads, on cells, or in vivo, which changes which background binders dominate.
1A protein tied to its own gene
Display technologies solve one problem: a selection can only act on a protein, but the thing you need back is its gene. Tying the two together physically (a peptide or antibody fragment fused to a phage coat protein, with the phage carrying the DNA inside it, or a fragment fused to a yeast cell-wall protein, with the DNA in the nucleus behind it) means that separating proteins by binding also separates genes by binding. Everything else is plumbing, and the whole method lives or dies on how cleanly that separation is done.
Fusion to pIII or pVIII
Filamentous phage (M13, fd) tolerate a foreign protein fused to the pIII tip. A phagemid with wild-type pIII supplied by a helper phage gives mostly one copy per particle; a true phage vector gives up to five. Libraries of 10⁹–10¹¹, selections done in a microtitre well or on beads.
Fusion to Aga2
Displayed on the cell wall at about 10⁴ copies per cell, which is large enough to measure by flow cytometry. Libraries are capped near 10⁷–10⁸ by yeast transformation, but the readout is quantitative and a eukaryotic secretory pathway helps disulfide-containing fragments fold.
Cell-free alternatives
Libraries of 10¹²–10¹³ with no transformation step, and PCR-based diversification between rounds, at the cost of a fiddlier selection and no quality control from a secretory pathway.
Naive, immune and synthetic
Naive libraries from donor B cells give micromolar to nanomolar hits against almost anything; immune libraries from an immunised or infected donor start much tighter; synthetic libraries with randomised CDRs on one scaffold give a defined, reproducible repertoire.
2Panning selects on off-rate; sorting selects on occupancy
Biopanning is not an equilibrium measurement. A clone is captured according to its occupancy on the immobilised target, and then washed, and what survives washing is set by how slowly the complex falls apart.
3The arithmetic of rounds
Enrichment multiplies. If a round recovers a specific clone a thousand times more efficiently than the background, three rounds move a clone from one in a billion to the majority of the pool, and that is why three or four rounds is the standard answer.
4How it goes wrong
Target-unrelated binders (to the plastic, the blocking protein, the streptavidin, the biotin, the capture antibody) are the commonest failure, and the only reliable defence is a no-target well panned and titred alongside the real one every round, plus monoclonal rather than polyclonal screening of the output. Over-panning is the second: once the wanted clone dominates, further rounds select only for whatever grows and packages fastest, which is usually a deletion or a clone with no insert at all.
Key terms
Practise these as flashcards →One round of capture on immobilised target, washing, elution and amplification.
Copies of the displayed fragment per particle; monovalent display reports single-interaction off-rates, polyvalent display reports avidity.
The apparent tightening that comes from several weak interactions holding one particle; the particle only releases when every copy does.
A plasmid carrying the fusion plus a phage supplying wild-type coat proteins, which is how monovalent display is achieved.
Infectious phage recovered from a round; meaningful only beside the titre from the no-target well.
A clone selected for binding the plastic, the blocking agent, the capture reagent or the tag rather than the target.
The yeast display format: the fragment fused to Aga2, which disulfide-bonds to Aga1 in the cell wall.
Dividing the target signal by an epitope-tag signal, so the sort ranks occupancy rather than expression.
The share of events sorted; in an unenriched population this sets the per-round enrichment, at most about the reciprocal of the gate.
Loading the library, then washing in the presence of excess soluble target so that only slowly-dissociating clones remain bound.
When to use it
- Finding binders (antibody fragments, peptides, alternative scaffolds) against a purified target, from a library far too large to screen clone by clone.
- Affinity maturation of an existing binder, where a mutated library is selected monovalently under rising stringency.
- Mapping an epitope or finding a peptide ligand, by panning a random-peptide library and sequencing the consensus.
- Ranking a panel of clones by affinity quickly, which yeast display does well because the sort readout is quantitative.
Compared with alternatives
The same loop applied to catalysis or stability rather than binding; display is how its selection step is usually built.
The immunisation route to a monoclonal antibody: no library, affinity maturation done in vivo, and a host that will not tolerate a toxic or self antigen.
Where the affinity of a selected clone is actually measured: on-rate, off-rate and KD on purified protein.
The instrument behind a yeast display sort, and the source of its gating logic and its controls.
The other yeast-based interaction selection: a transcriptional readout inside the cell rather than a stain on its surface.
How the output of a panning campaign is screened clone by clone, and why an ELISA-positive clone is not necessarily a tight one.
Sub-techniques
Choosing one here switches the whole page, including the protocol and the lab, to it.
A library of fragments fused to a filamentous phage coat protein is incubated with immobilised target, the well is washed, bound phage are eluted with trypsin or acid, and the eluate is amplified in fresh bacteria to give the input for the next round. Three to four rounds moves a clone from one in a billion to the majority of the pool, provided the plate and the library are properly blocked and a no-target well is panned alongside.
- Stringency is raised by washing longer, coating less target, or competing with soluble target rather than by adding rounds.
- Monovalent phagemid display reports off-rates; polyvalent display reports avidity and selects weak binders.
- The eluate must be amplified between rounds or the rare clone is physically lost.
- Selection can be done on beads, on cells, or in vivo, which changes which background binders dominate.