Elsevier

Plasmid

Volume 114, March 2021, 102554
Plasmid

Functional analysis of the catalytic triad of the hAT-family transposase TcBuster

https://doi.org/10.1016/j.plasmid.2021.102554Get rights and content

Highlights

  • The predicted catalytic triad of TcBuster transposase was mutated (D223N, D289N, and E589Q).

  • As expected, the catalytic mutants could not integrate a drug-resistance gene into human cells in tissue culture.

  • A modified ChIP assay revealed that the catalytic mutants maintained binding to the TcBuster inverted repeat sequences.

  • Wild-type TcBuster transposase forms filamentous aggregate rodlet structures that average 3.3 μm in length.

  • In contrast, catalytic mutants had low transposase levels and did not form rodlet structures.

Abstract

TcBuster is a hAT-family DNA transposon from the red flour beetle, Tribolium castaneum. The TcBuster transposase is of interest for genome engineering as it is highly active in insect and mammalian cells. To test the predicted catalytic triad of TcBuster, each residue of the catalytic triad of a haemagglutinin-tagged TcBuster transposase was individually mutated to a structurally conserved amino acid. Using a drug-resistant colony assay for transposon integration, we found that the D223N, D289N, and E589Q mutants of TcBuster transposase were inactive in human cells. We used a modified chromatin immunoprecipitation assay to determine that each mutant maintained binding to TcBuster transposon inverted repeat elements. Although the catalytic mutants retained their transposon binding properties, mutants displayed altered expression and localization in human cells. None of the catalytic mutants formed characteristic TcBuster transposase rodlet structures, and the D223N and D289N mutants were not able to be detected by immunofluorescence microscopy. Immunoblot analysis demonstrated that the E589Q mutant is less abundant than wild-type TcBuster transposase. Cells transfected with either TcBuster or TcBuster-E589Q transposase were imaged by structured illumination microscopy to quantify differences in the length of the transposase rodlets. The average length of the TcBuster transposase rodlets (N = 39) was 3.284 μm while the E589Q rodlets (N = 33) averaged 1.157 μm (p < 0.0001; t-test). The catalytic triad mutations decreased overall protein levels and disrupted transposase rodlet formation while nuclear localization and DNA binding to the inverted repeat elements were maintained. Our results may have broader implications for the overproduction inhibition phenomenon observed for DNA transposons.

Introduction

DNA transposons are mobile genetic elements that move through a DNA intermediate. Some transposons that use a “cut and paste” mechanism have transposases possessing a catalytic triad comprised of D, D, and E or D residues (Yuan and Wessler, 2011). This catalytic triad is required for the transposase to cut the DNA at the terminal inverted repeat sequences of the transposon and paste the transposon DNA into the target site (Yuan and Wessler, 2011). The DDE triad was first discovered within a motif in the bacterial IS4 family of insertion sequences (Rezsohazy et al., 1993) and is complexed with a divalent metal cation (Leschziner et al., 1998).

DNA transposon systems have been developed for genome engineering in mammalian cells. These plasmid-based integrating vectors have enormous potential for gene therapy and transgenesis applications (Tipanee et al., 2017). TcBuster and AeBuster1 were first reported in 2011 as having activity in Drosophila melanogaster and Ae. aegypti insects (Arensburger et al., 2011). Both of these Buster transposases were more similar to the human Buster domesticated transposases than to the other hAT elements (Arensburger et al., 2011). Therefore, the hAT superfamily can be divided into two distinct families: Buster (including TcBuster, AeBuster, and mammalian SPIN transposons) and Ac (including the Hermes, hobo, Tam3, and Tol2 transposons) (Arensburger et al., 2011). In vitro reactions with purified protein indicate that TcBuster transposition occurs without the need for host cofactors, allowing this transposon to be active in a variety of species (Li et al., 2013). In vitro cleavage and strand transfer experiments as well as excision product sequence analysis support a mechanism of TcBuster transposition in which the transposon is excised via a hairpin intermediate followed by 3’OH transposon end joining (Li et al., 2013; Woodard et al., 2012). The hairpin mechanism is shared by Ac-family transposon Hermes and other RNase H-family transposases (Hickman et al., 2018). TcBuster was the most active transposon studied for insertional mutagenesis and Tn-seq in the yeast Pichia pastoris (Zhu et al., 2018). The TcBuster transposon from Tribolium castaneum is highly active in cultured human cells and in mice (Li et al., 2013; Woodard et al., 2012). We found exogenously expressed human Buster proteins Buster1, Buster3, and SCAND3 did not mobilize a TcBuster transposon in human cells (Woodard et al., 2012).

Our previous work demonstrated that the hemagglutinin (HA)-tagged TcBuster transposase protein maintained high activity and was localized to intranuclear rodlets (Woodard et al., 2017). Another member of the hAT transposon superfamily, Ac transposase, forms similar structures in insect (Hauser et al., 1988), petunia (Heinlein et al., 1994), maize (Heinlein et al., 1994; Essers et al., 2000), tobacco (Scofield et al., 1993), and zebrafish (Emelyanov et al., 2006). Although purified hAT transposase proteins are often insoluble, not all of them form rodlet structures (Hickman et al., 2005; Iida et al., 2004; Michel and Atkinson, 2003; Xu et al., 2015). TcBuster rodlets have been observed in several human cell lines and in mouse liver cells after hydrodynamic tail vein injection (Woodard et al., 2017). The formation of TcBuster rodlets is concurrent with the end of active transposition, suggesting that rodlet formation may correlate with overproduction inhibition, which is known to affect DNA transposons (Woodard et al., 2017). Since transfection with too much transposase lowers transposition, transposase dose must be optimized for gene therapy development.

The DDE domain has been experimentally confirmed via mutation analysis for the related hAT-family Hermes and Activator transposases (Michel et al., 2002; Lazarow et al., 2012). Michel and colleagues made amino acid substitutions (D402N and E572Q) that mutated the catalytic site of Hermes while maintaining the transposase protein structure (Michel et al., 2002). The DDE domain of TcBuster has been computationally predicted by Arensburger and colleagues (Arensburger et al., 2011). The published alignment of TcBuster with 21 similar transposases further supports the predicted DDE catalytic triad (Atkinson, 2015). This study confirms these alignments by showing that loss of each residue of the catalytic triad resulted in an inactive TcBuster transposase mutant. We sought to determine how mutation of the catalytic triad of TcBuster would affect its activity and rodlet formation. Therefore, we examined the activity, localization, and binding to inverted repeat DNA for each catalytic mutant of TcBuster transposase.

Section snippets

Generation of catalytic mutants and colony assay

A list of all plasmids and references is provided (Table 1). Catalytic mutants were chosen based on the alignment of the TcBuster transposase to other hAT-family transposases in previously published studies (Arensburger et al., 2011; Atkinson, 2015). Each mutant is identical to pCMV-HA-TcBuster (Woodard et al., 2017) except for the relevant codon as confirmed by sequencing (Genewiz), GAC- > AAC for D to N mutations or GAA - > CAA for E to Q. Overlap PCR was used to make D223N from pCMV-HA-

Single substitutions in the predicted catalytic triad residues of TcBuster transposase yield inactive mutants

We made three catalytic mutants containing one amino acid change each in the codon-optimized and HA-tagged pCMV-HA-TcBuster plasmid for transposase expression in human cells (Fig. 1). The N-terminus of the wild type transposase and each catalytic mutant has an HA tag to enable antibody detection of the transposases. We termed the mutations D223N, D289N, and E589Q based on the numbering of amino acids in the publicly deposited wild-type TcBuster transposase coding sequence, GenBank accession

Discussion

In this study, we first confirmed the predicted DDE catalytic domain for the TcBuster transposase, demonstrating that each of the three TcBuster catalytic mutants was inactive (Fig. 1). We engineered the most conservative amino acid changes possible to preserve the structural integrity of the transposases. We performed a modified ChIP assay to assess the ability of the catalytic mutants to bind TcBuster inverted repeat sequences and showed that both wild-type and inactive catalytic mutants were

Funding

L.E.W. was supported by grants from the National Institutes of Health [5T32DK062706 and 5T32DK060445-08], the Department of Veterans Affairs [BX002797 and BX004845], and the Vanderbilt O'Brien Kidney Center [5P30DK114809-02], in addition to generous support from Dr. and Mrs. Harold M. Selzman. The National Institutes of Health [DK093660], Department of Veterans Affairs [BX002190], and the Vanderbilt Center for Kidney Disease supported M.H.W. This material is the result of work supported with

Authors' contributions

L.E.W. conceived of the experiments and methods. L.E.W., F.M.W., and I.C.J. performed the experiments. L.E.W. and I.C.J. analyzed the data. L.E.W. wrote the original manuscript draft. I.C.J. and M.H.W. edited the manuscript. M.H.W. and L.E.W. provided supervision, funding acquisition, project administration, and resources.

Acknowledgements

We thank Peter Atkinson and Nancy Craig for providing the TcBuster transposon and helpful discussions. We thank Wentian Luo for generous discussion of experimental procedures. We thank Ruth Ann Veach for critical reading of the manuscript. Data collection and analysis was performed in part through the use of the Vanderbilt Cell Imaging Shared Resource.

References (27)

  • C. Hauser et al.

    Overproduction of the protein encoded by the maize transposable element ac in insect cells by a baculovirus vector

    Mol. Gen. Genet.

    (1988)
  • M. Heinlein et al.

    In vivo aggregation of maize activator (Ac) transposase in nuclei of maize endosperm and Petunia protoplasts

    Plant J.

    (1994)
  • A.B. Hickman et al.

    Molecular architecture of a eukaryotic DNA transposase

    Nat. Struct. Mol. Biol.

    (2005)
  • Cited by (0)

    View full text