A simple, robust, broadly applicable insertion mutagenesis method to create random fluorescent protein: target protein fusions

Abstract

A simple, broadly applicable method was developed using an in vitro transposition reaction followed by transformation into Escherichia coli and screening plates for fluorescent colonies. The transposition reaction catalyzes the random insertion of a fluorescent protein open reading frame into a target gene on a plasmid. The transposition reaction is employed directly in an E. coli transformation with no further procedures. Plating at high colony density yields fluorescent colonies. Plasmids purified from fluorescent colonies contain random, in-frame fusion proteins into the target gene. The plate screen also results in expressed, stable proteins. A large library of chimeric proteins was produced, which was useful for downstream research. The effect of using different fluorescent proteins was investigated as well as the dependence of the linker sequence between the target and fluorescent protein open reading frames. The utility and simplicity of the method were demonstrated by the fact that it has been employed in an undergraduate biology laboratory class without failure over dozens of class sections. This suggests that the method will be useful in high-impact research at small liberal arts colleges with limited resources. However, in-frame fusion proteins were obtained from 8 different targets suggesting that the method is broadly applicable in any research setting.

Keywords: insertion mutagenesis; protein fusions.

Graphical Abstract

Fig. 1.

Tn5 Transposase (dumbbell shape) binds to FP ORF amplicons (black and hashed rectangles) at 19-nucleotide Tn5 binding sites (black). This complex interacts with target DNA (dashed circles) at random sites. Following transformation and incubation at 30°C, fluorophore activation allows identification of fluorescent colonies (black spots and gray spots are nonfluorescent colonies).

Fig. 2.

The cumulative data obtained from undergraduate laboratory class are shown. From 2014 to 2020, a single target ORF (EF 4) was employed in transposition reactions with the FP ORF GFPmut3. The amino acid sequence of the EF 4 ORF is shown. The colored amino acid residues indicate EF 4 domains: domain 1 (tan residues 1–188), domain 2 (purple residues 189–281), domain 3 (blue residues 291–371), domain 4 (gray residues 398–486), and the C-terminal disordered domain (brown residues 487–599). The underlined numbers above the sequence indicate insertion sites that were confirmed by student researchers and the number of times each insertion site was recovered. ** indicates the insertion at codon 128 was recovered 20 times and codon 129 twice. *** indicates a cluster of insertions at codons 167 (twice), 168 (13 times), and 169 (once). The cluster at 282, 283, and 284 was recovered once, 5 times, and once, respectively, and at 292, 293, and 294, twice, once, and 4 times, respectively. These data were derived from 159 separate student experiments.

Fig. 3.

Tn5 Transposase (gray dumbbell color) binds randomly to the target DNA (dashed circle) and cuts target DNA strands at sites 9 nucleotides apart generating a 9-nucleotide complimentary overhang. The transposase ligates the inserted DNA generating a double-strand structure with 5′ and 3′ 9-nucleotide gaps. The gaps in the resultant double-strand circle are filled in by host DNA repair polymerases following transformation. Codon N is the codon that immediately precedes the Tn5 cut site. Codon N + 1, N + 2, and N + 3 represent the 3 codons following N, which comprise the 9-nucleotide overhang. The result of ligation by Tn5 and host cell repair is that codons N + 1, N + 2, and N + 3 were repeated on the 3′ side of the FP ORF (hashed rectangle). The 19-nucleotide recognition sequence for Tn5 is represented by the black rectangle. Only cleavages between codons are recovered because cleavage within a codon produces an out-of-frame ORF. Out-of-frame ORF fusions fail the colony screen for fluorescence.

Fig. 4.

Testing effects of altering linker sequence (black bold capital letters) between the target ORF and the FP ORF (hashed rectangle). a) The 19-nucleotide Tn5 recognition element is a defined sequence that must always be present at the 5′ and 3′ ends of the inserted DNA as inverted repeats. The first 18 nucleotides would encode the amino acid sequence LSLIQI, and the 19th nucleotide would become the first base of the first codon of the FP ORF. At the 3′ end, the first nucleotide of the inverted repeat would replace the third nucleotide of the ORF's stop codon creating a read-through ORF. The subsequent 18 nucleotides encode the amino acids DVYKRQ. Due to the staggered cut of the Tn5 transposase, codons N + 1, N + 2, and N + 3 were repeated at the 3′ end of an insertion (Fig. 3) comprising additional nonnative sequence. b) To make the number of 5′ and 3′ inserted codons symmetrical, 3 additional codons were inserted at the 5′ end (underlined letters represent the amino acids encoded by this modification). c) To increase the length of the linker region to optimal 12 codons and insert codons encoding more favorable linker amino acids, codons were added, which would encode the underlined amino acids. George and Heringa (2002) and Suyama and Ohara (2003) were used as a guide to design optimal linker length and favorable interdomain amino acid linker sequence. PCR primers employed to insert these modifications are listed in Table 1.

Fig. 5.

Schematic of donor amplicons generated as substrates for MORFIN mutagenesis. The black rectangle indicates the position of 5′ and 3′ linker sequence. “No extra linker a.a.,” “3 extra linker a.a.,” and “6 extra linker a.a.” are explained in Fig. 4. The diagonal filled rectangle represents the ORF for GFPmut3, the dotted rectangle represents the ORF for mCherry2, and the hashed rectangle represents the ORF for mYPET.

Fig. 6.

The unexpected sensitivity of mYPET to linker sequence. Transposition reactions were transformed directly into E. coli and plated at high colony density. After a 36-h incubation at 30°C, the plates were imaged using a laser plate scanner. Nonfluorescent colonies are light gray, and fluorescent colonies are black. Examples of prominent black fluorescent colonies are indicated by the arrowheads. The results from 3 different amplicons are shown: (1) mYPET with no extra linker, (2) mYPET with 3 extra linkers, and (3) mYPET with 6 extra linkers. The meaning of “no extra linker,” “3 extra linkers,” and “6 extra linkers” is explained in Fig. 3. The target ORF in this case expressed EF G. All black colonies were restruck to purify clones that produced FPs. Purified clones were subjected to plasmid purification, and pure plasmids were sequenced to confirm insertion sites. The tiny black specks found on plates numbered 1 were not fluorescent bacterial colonies.

Fig. 7.

An example of the restrictive nature of the FtsZ ORF. Transposition reactions were transformed directly into E. coli and plated at high colony density. After a 36-h incubation at 30°C, the plates were imaged using a laser plate scanner. Nonfluorescent colonies are light gray, and fluorescent colonies are black. The results from 3 different FPs are shown. In each case, the linker was the 6 extra linker (Fig. 3). Fluorescent colonies were only found when mCherry or mYPET was employed as the FP ORF. All of the mCherry candidates were off target. In-frame FP–FtsZ fusions were only obtained from mYPET transposition reactions. The same result was obtained for the MreB ORF. All black colonies were restruck to purify clones that produced FPs. Purified clones were subjected to plasmid purification, and pure plasmids were sequenced to confirm insertion sites.