Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Comparative Study
, 34 (8), e57

Genomic Analysis of Insertion Behavior and Target Specificity of mini-Tn7 and Tn3 Transposons in Saccharomyces Cerevisiae

Affiliations
Comparative Study

Genomic Analysis of Insertion Behavior and Target Specificity of mini-Tn7 and Tn3 Transposons in Saccharomyces Cerevisiae

Michael Seringhaus et al. Nucleic Acids Res.

Abstract

Transposons are widely employed as tools for gene disruption. Ideally, they should display unbiased insertion behavior, and incorporate readily into any genomic DNA to which they are exposed. However, many transposons preferentially insert at specific nucleotide sequences. It is unclear to what extent such bias affects their usefulness as mutagenesis tools. Here, we examine insertion site specificity and global insertion behavior of two mini-transposons previously used for large-scale gene disruption in Saccharomyces cerevisiae: Tn3 and Tn7. Using an expanded set of insertion data, we confirm that Tn3 displays marked preference for the AT-rich 5 bp consensus site TA[A/T]TA, whereas Tn7 displays negligible target site preference. On a genome level, both transposons display marked non-uniform insertion behavior: certain sites are targeted far more often than expected, and both distributions depart drastically from Poisson. Thus, to compare their insertion behavior on a genome level, we developed a windowed Kolmogorov-Smirnov (K-S) test to analyze transposon insertion distributions in sequence windows of various sizes. We find that when scored in large windows (>300 bp), both Tn3 and Tn7 distributions appear uniform, whereas in smaller windows, Tn7 appears uniform while Tn3 does not. Thus, both transposons are effective tools for gene disruption, but Tn7 does so with less duplication and a more uniform distribution, better approximating the behavior of the ideal transposon.

Figures

Figure 1
Figure 1
Tn3 and Tn7 Sequence Logos: Whole genome and hotspots. (a) Tn3 hotspot Sequence Logo, calculated using the 1202 target sequences that harbor multiple insertions. (b) Modified Sequence Logo for Tn3 insertion site, calculated using all 12 109 genomic insertions. Consensus sequence: [T/C]xxTA[A/T]TA. Nucleotides colored red are enriched more than 4% in hotspot consensus, while nucleotides colored grey do not differ significantly. (c) Tn7 hotspot Sequence Logo, calculated using the 738 target sequences that harbor multiple insertions. (d) Modified Sequence Logo for Tn7 insertion site, calculated using all 9134 genomic insertions. Consensus sequence: [A/T]. No nucleotides are enriched more than 4% in Tn7 hotspot consensus.
Figure 2
Figure 2
Plot displaying distribution of average insertions per 10 kb for each of the sixteen S.cerevisiae chromosomes. An equal number (8556) of Tn3 and Tn7 insertions were considered. Error bars represent ±σ.
Figure 3
Figure 3
Comparison of normalized (8556 insertions each) sets of Tn3 and Tn7. Plots show number of windows (log scale) at each size (a) 3 kb, (b) 1 kb, (c) 0.5 kb that contain a given number of insertions, overlaid with a Poisson distribution.
Figure 4
Figure 4
Cumulative insertions across entire genome (12.1 Mb). Distribution of all Tn3 (red, 10 776 insertions) and Tn7 (green, 8556 insertions) insertions across all sixteen chromosomes: (a) Insertions were tallied and the cumulative count (y) was plotted against chromosomal coordinate (x). Horizontal plateaus represent cold-spots, long stretches of genomic material with no scored insertions; steep vertical areas represent hotspots, with many insertions clustered in a localized area. For the region shown, the Tn3 insertion patterns differs significantly from expected (uniform, shown as diagonal) with QK–S = 0.00, whereas Tn7 is indistinguishable from uniform (QK–S = 0.05).
Figure 5
Figure 5
Windowing procedure for K–S. (a) Transposon insertions (green) lie along a stretch of genomic sequence (blue). (b) Dividing this sequence into eight equally-sized windows and assigning a binary value to insertions in this region yields a more uniform insertion density; half of all possible windows contain at least one insertion. (c) Region divided into four windows; now, only one window contains no insertions (‘cold-spot’). (d) As windows grow in length, the insertion distribution approaches uniform.
Figure 6
Figure 6
Insertion behavior at various window sizes, chromosome I (230 kb). Distribution of all Tn3 (red) and Tn7 (green) insertions across chromosome I, at different window sizes as shown. Notice the y-axis of total counts per chromosome is changed between windows, since multiple insertions within any window count as a single insertion event. As window size is increased, both Tn3 and Tn7 distributions appear more uniformly distributed. (a) No windowing: QK–S(Tn3) = 0.00, QK–S(Tn7) = 0.00. (b) Window = 100 bp: QK–S(Tn3) = 0.00, QK–S(Tn7) = 0.00. (c) Window = 200 bp: QK–S(Tn3) = 0.00, QK–S(Tn7) = 0.00. (d) Window = 500 bp: QK–S(Tn3) = 0.00, QK–S(Tn7) = 0.01. (e) Window = 1000 bp: QK–S(Tn3) = 0.06, QK–S(Tn7) = 0.10. (f) Window = 3000 bp: QK–S(Tn3) = 0.56, QK–S(Tn7) = 0.53.
Figure 7
Figure 7
Whole genome QK–S versus window size. K–S tail probabilities (QK–S) for Tn3 (red) and Tn7 (green) at various window sizes. Points correspond to the genomic QK–S values at each window size. The QK–S = 0.05 threshold is shown as a dotted line; points above this line are distributions indistinguishable from uniform, whereas points below this line correspond to distributions significantly different from uniform.
Figure 8
Figure 8
(a) Distribution of dinucleotide frequencies assessed across 1 kb windows. Four genomic subsets are shown. From left to right: none (windows in which neither Tn3 nor Tn7 insertions are found: 7225 windows total); Tn7 only (windows in which only Tn7 insertions are found: 2045 windows total); Tn3 only (windows in which only Tn3 insertions are found: 1258 windows total); Both (windows in which both Tn3 and Tn7 insertions are found: 1547 windows total). (b) Table of GC content and incidence of TA[A/T]TA 5mers/1 kb in these subsets and in the genome as a whole. (c) Venn diagram illustrating complementarity in coverage of 1 kb windows (12 075 windows total). (d) Results of X2 test on dinucleotide and trinucleotide counts in each of the four main subsets and three randomly selected subsets.
Figure 9
Figure 9
Schematic illustration of excluded regions. A schematic illustration of the first ∼100 kb of chromosome XII: regions possibly underrepresented in our genomic library were identified as contiguous, extended cold-spots of at least 6 kb where neither Tn3 nor Tn7 insertions occurred. These regions are shown in yellow, above. Tn3 insertions are shown in red, Tn7 insertions are green.

Similar articles

See all similar articles

Cited by 11 PubMed Central articles

See all "Cited by" articles

References

    1. Des Etages S.A., Kumar A., Snyder M. Transposons as tools. In: Brenner S., Miller J.H., editors. Encyclopedia of Genetics. San Diego, CA: Academic Press; 2001. pp. 2034–2040.
    1. Davies C.J., Hutchison C.A., IIIrd Insertion site specificity of the transposon Tn3. Nucleic Acids Res. 1995;23:507–514. - PMC - PubMed
    1. Pribil P.A., Haniford D.B. Target DNA bending is an important specificity determinant in target site selection in Tn10 transposition. J. Mol. Biol. 2003;330:247–259. - PubMed
    1. Pribil P.A., Wardle S.J., Haniford D.B. Enhancement and rescue of target capture in Tn10 transposition by site-specific modifications in target DNA. Mol. Microbiol. 2004;52:1173–1186. - PubMed
    1. Liu G., Geurts A.M., Yae K., Srinivasan A.R., Fahrenkrug S.C., Largaespada D.A., Takeda J., Horie K., Olson W.K., Hackett P.B. Target-site preferences of sleeping beauty transposons. J. Mol. Biol. 2005;346:161–173. - PubMed

Publication types

Substances

Feedback