Modified version of pACT-2 that simplifies cloning with NdeI.

Since its description in 1989 by Fields and Song (3), the yeast two-hybrid system has proven invaluable for the characterization of protein-protein interactions and the isolation of new proteins that interact with given proteins (for an introduction to many of the aspects of the yeast two-hybrid system, see Reference 1). Although the original method has been modified and new yeast strains have been engineered to improve protein expression and decrease false positives, the rationale remains the same (Matchmaker two-hybrid system, Matchmaker two-hybrid system 2, system 3, and LexA system; Clontech Laboratories, Palo Alto, CA, USA). In addition, variants have been created such that it is now possible to perform a one-hybrid screen to clone cDNA sequences coding for specific DNA binding proteins and a three-hybrid screen to investigate complexes that involve interactions between three partners (2,7,9,11). The field of intermediate filament (IF) biology has made use of the yeast two-hybrid system to investigate interactions between individual protein monomers that are the initial step of IF assembly (6,8,10). Interactions between subdomains of the central rod domain have also been characterized by the yeast two-hybrid system (8). We have used the yeast two-hybrid system to investigate the specificity of interaction between the components of an ocular lens specific filament termed the “beaded filament” (4). As part of the experimental design, it was necessary to construct positive and negative controls by subcloning different IF cDNA sequences into two-hybrid vectors. The most straightforward method of subcloning these cDNAs was as NdeIEcoRI fragments, and, unfortunately, pACT-2 contains NdeI restriction sites at position 235 and 6131 in addition to the polylinker. Therefore, pACT-2 is not suitable for directional cloning with NdeI (Figure 1). The companion vectors pAS2 and pAS2-1 each contain a single NdeI site and were thus suitable for cloning of NdeI restriction fragments without modifications. Because of the number of the clones we wished to construct and the prospect of performing blunt end ligations if we could not use the NdeI site in the polylinker of pACT2, we chose to use in vitro mutagenesis to remove the NdeI sites outside the polylinker. Removal of the NdeI sites outside the polylinker would create a vector that we could use to directionally clone restriction fragments that contain NdeI-cohesive ends at the 5′ end. Oligonucleotides encompassing each NdeI site were designed such that mutagenesis using the QuikChange kit (Stratagene, La Jolla, CA, USA) would result in the change of a single nucleotide, from CATATG to C ACATG, eliminating the NdeI recognition sequence. Oligonucleotides (Life Technologies, Rockville, MD, USA) used to remove the site at 235 were (top strand) 5′-CTCGGTACTATGCACATGATCCAATATCAAAGG-3′ and 5′CCTTTGATATTGGATCATGTGCATAGTACCGAG-3′ (bottom strand); the location of the destroyed NdeI site is underlined. An oligonucleotide complimentary to pACT2 sequences 126–149 was used for DNA sequencing to verify the mutation. DNA sequencing was performed by the University of California Davis Division of Biological Sciences DNA Sequencing Facility. Oligonucleotides used to remove the NdeI site at 6131 were (top strand, 5′-GTACTGAGAGTGCACCACATGCGGTGTGAAATACCGC-3′ and bottom strand, 5′GGTATTTCACACCGCATGTGGTGCACTCTCAGTACAATCTGC-3′; the location of the destroyed NdeI site is underlined). Verification of the mutation Benchmarks