Parallel analysis of tetramerization domain mutants of the human p53 protein using PCR colonies

Abstract

A highly-parallel yeast functional assay, capable of screening approximately 100-1,000 mutants in parallel and designed to screen the activity of transcription activator proteins, was utilized to functionally characterize tetramerization domain mutants of the human p53 transcription factor and tumor suppressor protein. A library containing each of the 19 possible single amino acid substitutions (57 mutants) at three positions in the tetramerization domain of the human p53 protein, was functionally screened in Saccharomyces cerevisiae. Amino acids Leu330 and Ile332, whose side chains form a portion of a hydrophobic pocket that stabilizes the active p53 tetramer, were found to tolerate most hydrophobic amino acid substitutions while hydrophilic substitutions resulted in the inactivation of the protein. Amino acid Gln331 tolerated essentially all mutations. Importantly, highly parallel mutagenesis and cloning techniques were utilized which, in conjunction with recently reported highly parallel DNA sequencing methods, would be capable of increasing throughput an additional 2-3 orders of magnitude.

Fig. 1

Construction of tagged mutant p53 expression plasmids. Plasmids were constructed in a three step process. (A) 200 bp tags containing six variable positions were synthesized using overlapping oligonucleotide PCR; tags were cloned into the expression plasmid using homologous recombination mediated gap repair. (B) Mutant p53 genes were constructed using mutagenic crossover PCR; both degenerate (–NNN–) and mutant specific primers were used. (C) Final plasmid assembly was done by ligating SpeI/HindIII digested tagged vector and mutant p53 genes from (A) and (B)

Fig. 2

(A) Polony method used to quantify mutants during p53 growth competition. Plasmid DNA was initially isolated from the culture and used as template in a polony PCR. Common primers amplified all STs, each associated with a different p53 mutant and resulting in an individual polony. Polonies were identified by conducting six sequential single-base extensions using fluorescently-labeled nucleotides. The position of each polony was manually logged after sybr green staining using Metamorph software (Universal Imaging). Data from each extension was assembled into the final code sequence using a simple routine developed in our laboratory. In the representative frames above, two polonies are traced through all six extensions. Upper: tag sequence “AACAAA” corresponds to Gln331Gly. Lower: tag sequence “TCCCAA” corresponds to Gln331Met. (B) Expression plasmids were constructed using the p415CYC1 vector which carries the constitutive CYC1 promoter

Fig. 3

Human p53 tetramerization domain. Interacting regions of two p53 monomers (peptide backbones visualized in green and blue) which form a dimer. A second identical dimer (not shown) mates with the first to form the final assembled protein. Side chains of the three residues tested in this work, Leu330, Gln331 and Ile332, and the side chain of Phe341 are highlighted. R-groups of Leu330, Ile332 and Phe341 are oriented toward each other and form a portion of the hydrophobic pocket, which stabilizes the p53 tetramer

Fig. 4

Possible high throughput mutant construction and analysis. Only minor modifications to the procedures for mutant construction and analysis used in this work are required to greatly increase throughput and compatibility with ultrahigh throughput sequencing methods. (A) Random mutagenesis of portion of gene using mutagenic crossover PCR, error prone PCR, annealing degenerate single-stranded synthetic oligonucleotides, etc.; length of mutant portion of gene limited by sequencing method employed. (B) Construction of growth competition-ready mutant strain library using gap repair cloning. (C) Pooled mutant growth competition. (D) Mutant segment isolation and sequencing preparation using PCR. (E) Chip-based ultrahigh throughput sequencing