Fig 2: Quantitative analysis of scFvs isolated from the CDRH1his library using yeast surface display.(a) Fraction TfR bound to M mutants after 10 minute incubation at pH 5.5. (n = 5 for M mutants and n = 13 for H7, and ***, p <0.001) (b) Fraction TfR bound to N and 5P mutants after 10 minute incubation at pH 5.5. (n = 8 for N and 5P mutants and n = 13 for H7 mutant as in panel a) and ***, p <0.001 *, p <0.05). (c) Apparent equilibrium binding affinity of select clones on the surface of yeast at pH 7.4. Mean data from five independent experiments are plotted along with the fitted equilibrium binding isotherms. The legend shows numeric values for the best-fit equilibrium binding affinity (Kd) and associated 95% CI.

Fig 3: Flow-cytometric screening of the CDRH1his library for scFvs having affected dissociation from TfR at pH 5.5.Dot plots depict the behavior of the various scFv populations after saturation with rhTfR followed by a 10 minute dissociation step in pH 5.5 or 7.4 buffers. TfR-binding is indicated on the y-axis while scFv surface expression level is indicated on the x-axis. Sample gates are drawn only for illustrative purposes for the reader to follow the sorting enrichment procedure. (a) Antigen binding and expression of wild-type H7 at pH 7.4. (b and d) Pool F was derived from the CDRH1his library via two rounds of sorting at pH 7.4 and comprises neutral pH TfR-binders. (c and also S1 Fig) A population comprising pH-insensitive binders exists in CDRH1his and can be visualized in gate F after the library is subject to pH 5.5 dissociation. (e) Antigen binding and expression of wild-type H7 at pH 5.5. (f) Both pools M and N were obtained by selecting for different populations within pool F, post dissociation at pH 5.5. (f and g) Gate N was placed near the maximum in TfR-binding signal to select for pH-insensitive scFvs and ultimately yielded pool N. (f and h) Alternatively, gate M was placed just above the no-antigen control, but below gate N to select for the pH-sensitive binders ultimately found in pool M (see also S1 Fig). (g and h) Flow-cytometric analysis of the M and N pools. Individual clones arising from these screening strategies are listed in Table 1 and quantitative clonal dissociation behavior is shown in Fig 3. Further detailed explanation of the screening strategy and pools depicted in the various panels can be found in the text.

Fig 4: Creation and screening methodology for a histidine-saturated CDRH1 library based on the wild-type anti-TfR scFv, H7.(a) The pESO-H7sdm vector contained a mutant scFv H7, that was specially designed to harbor a unique restriction site (SpeI) followed by two stop codons in CDRH1. SpeI digestion of pESO-H7sdm produced a linearized backbone that would be undetectable by flow cytometry if introduced as a reclosed vector into yeast (no c-myc epitope tag expression as a result of the double stop codon). A double-stranded DNA cassette containing the histidine-saturated CDRH1 was built using two cycles of PCR, from primers and a degenerate oligonucleotide (IP90T). Although depicted as a single entity for simplicity, IP90T was in fact a mixture of millions of unique ssDNA oligos representing all possible combinations of histidine in CDRH1. The CDRH1his cassette and linearized pESO-H7sdm were used to create the CDRH1his library by homologous recombination in yeast. (b) Screening and assessment of yeast-displayed scFvs was accomplished by saturating with recombinant human transferrin receptor (rhTfR) followed by a 10 minute incubation in acidic buffer (pH 4.0–6.5) and assay for TfR dissociation. A pH 5.5 buffer was used to simulate endosomal conditions relevant to transferrin-TfR dissociation [19,20]. A dissociation control was included by using a pH 7.4 buffer as opposed to acidic buffer at this step. The dissociation process was halted by addition of an excess volume of ice cold pH 7.4 buffer. Surface constructs were subsequently immunolabeled and analyzed by flow cytometry. The histogram depicts the response of yeast-displayed wild-type scFv H7 to 10 minute incubation with buffers having acidic pH.