Synthesis of libraries and multi-site mutagenesis using a PCR-derived, dU-containing template

Abstract

Directed DNA libraries are useful because they focus genetic diversity in the most important regions within a sequence. Ideally, all sequences in such libraries should appear with the same frequency and there should be no significant background from the starting sequence. These properties maximize the number of different sequences that can be screened. Described herein is a method termed SLUPT (Synthesis of Libraries via a dU-containing PCR-derived Template) for generating highly targeted DNA libraries and/or multi-site mutations wherein the altered bases may be widely distributed within a target sequence. This method is highly efficient and modular. Moreover, multiple distinct sites, each with one or more base changes, can be altered in a single reaction. There is very low background from the starting sequence, and SLUPT libraries have similar representation of each base at the positions selected for variation. The SLUPT method utilizes a single-stranded dU-containing DNA template that is made by polymerase chain reaction (PCR). Synthesis of the template in this way is significantly easier than has been described earlier. A series of oligonucleotide primers that are homologous to the template and encode the desired genetic diversity are extended and ligated in a single reaction to form the mutated product sequence or library. After selective inactivation of the template, only the product library is amplified. There are no restrictions on the spacing of the mutagenic primers except that they cannot overlap.

Keywords: DNA library synthesis; antibody design; directed evolution; multisite mutagenesis; protein engineering.

Figure 1.

Schematic overview of the SLUPT strategy. Step 1: The gene of interest is amplified with a 5′ phosphorylated top strand primer and dNTP’s containing dU (blue). The primer for the bottom strand is not phosphorylated. Optional, nonhomologous regions (e.g. to introduce restriction enzyme sites) are shown in green. Step 2: The phosphorylated strand is selectively degraded by lambda exonuclease to create the uracil-containing single stranded template. Step 3: An end-primer complementary to the 3′ terminus and 5′ phosphorylated internal primers containing altered bases are annealed to the uracil containing single strand template. Altered bases depicted as X’s in red box. Gap filling and ligation are performed by Phusion-U and Taq ligase to create a mutated, complementary strand. Step 4: The Uracil-containing single stranded template is digested by UDG. Step 5: The single-stranded product is made double stranded and amplified by PCR.

Figure 2.

Structure-based library design of a Cre-based recombinase. (Left) A structural model of a Cre variant (cyan) bound to its target DNA (surface representation in gray) with only one monomer of the Cre tetramer shown. The location of the desired mutations is indicated by red spheres located at the C-alpha coordinates. (Right) Close-up of the helix B (with amino acid 43) and helix D (with amino acids 89, 90, 93, and 94).

Figure 3.

Sample sequencing results of the amplified library wherein two mutated regions are separated by ∼130 bp. Sanger sequencing chromatograms of the library in the mutated helix B region (A) and the mutated helix D region (B) of the recombinase. The 5′ phosphorylated donor primer is shown in a purple box above the starting sequence (lowercase bold). The location of the mutations within the primer are indicated by a *. An expanded view of the mutated region is shown, and the mutated bases are listed in order of their approximate peak height and the bases are colored according to their corresponding trace. Underlined bases are those that correspond to the starting sequence. The red arrow in B denotes a simple G > A base substitution. The library shown in C is identical to that shown in B, but the primers and template were heated to 95°C and then allowed to cool before the elongation/ligation step. Positions where the templated base is strongly favored are marked with blue triangles.

Figure 4.

Design of library via SLUPT for anti CTLA-4 scFv antibody. (A) Superimposed crystal structures of the scFv ipilimumab (green, pdb code 5XJ3) and the Fab fragment of tremelimumab (teal, pdb code 5GGV). CTLA-4, seen in both structures is magenta and red, respectively. The locations of the mutated residues within the library are shown as blue spheres. (B) SLUPT donor primers used to create the DNA libraries of the scFv gene in a single reaction. The donor primers are in a purple box aligned to the wt scFv sequence. The overlapping primers were not used together, and there are six donor primers per library. The primers span ∼650 nucleotides, the closest pair is 5 bp, the farthest pair is ∼130 bp. The scFv PCR product is ∼900 base pairs.