General Information/IS History

IS History

It is now over 40 years since the first IS were described. They were identified as short DNA segments found repeatedly associated with mutations in the gal operon and bacteriophage λ . Shortly after, it was established that IS were normal residents of the Escherichia coli chromosome sometimes present in multiple copies. They were shown to be involved in generating deletions and in activating gene expression . They were also identified as constituents of bacterial plasmids . At about the same time, it was observed that antibiotic resistance genes could also be transferred or “transposed” from one plasmid to another (Fig.2.1) and it was recognized that IS and “transposons” (a term originally coined by Hedges and Jacob ) were both members of a group of genetic entities: Transposable or Mobile Genetic Elements (TE or MGE).

Fig. 2.1. The first Transposon discovered in 1974. This figure illustrates the independent movement (transposition) of the Ampicillin resistance transposon, transposon A (later renamed as Tn1), from its original plasmid RP4 to three different plasmids R64, JR66a and Sa. It also shows secondary transposition from Sa to R64 (bottom).

Transposons and IS were identified by electron microscopy of heteroduplexes between plasmids or phage with and without the insertion or by homoduplexes (Fig.2.2 and Fig.2.4; e.g ). This relationship between IS and transposons was reinforced by the observation that different DNA segments carrying different genes could be translocated by two flanking IS (compound transposons) (Fig.2.3and Fig.2.4). In many compound transposons one of the IS is inactivated or has reduced activity due to mutation. This tends to increase the coherence of the transposon in comparison to the autonomous activity of the individual IS. For example, in Tn5, the inside end (IE) of the left IS50 carries a mutation which both creates a better promoter sequence to drive expression of the resistance genes and introduces a UAA nonsense codon at the 3’ end of the transposase gene . In the case of Tn10, the left IS10 differs from the functional right IS10 at a number of nucleotide positions in both in its regulatory region and in the transposase . On the other hand, not all composite transposons possess an inactive copy of the flanking IS. In Tn9, flanked by two directly repeated copies of IS1, both IS are active .

Fig. 2.2. Electron microscopy showing Insertions with short stems. Part of a heteroduplex molecule formed between plasmids R100.1 and Rldrdl9 showing a single-stranded insertion loop of the ampicillin resistance transposon Tn3 carried by plasmid Rldrd19 but not by R100.1. In this type of experiment, both plasmids are denatured into their single-strand DNA form, mixed and renatured to form double-stranded DNA where they are complementary and single-strand regions where they do not share sequence homology. (Electron Microscopy by Edouard Boy de la Tour and Lucien Caro).

Fig. 2.3. Compound Transposons structure schema with examples of transposons Tn5, Tn10, Tn903, Tn602, and Tn9. The figure is not to scale. The TE is shown as a pink box; purple arrows, Tpase genes; red boxes, passenger genes; thin arrows, promoters, and transcription direction; dark blue arrows, terminal inverted repeats; white asterisks, GATC Dam methylation sites which are present in some TE and control their level of activity; OE and IE: outside ends and inside ends of the IS with respect to the transposon. This is a generic cartoon showing the general arrangement of flanking IS in several different Tn. Tn5 (5818 bp) carries inverted copies of IS50 (1534 bp); Tn10 (9147 bp) carries inverted copies of IS10 (1329 bp); Tn903 (3094 bp) carries inverted copies of IS903 (1057 bp,); Tn602 (3064 bp) carries direct flanking copies of IS903. Tpase, Tpase; kan, tet, cam: resistance genes for kanamycin, tetracycline, and chloramphenicol respectively; Tn9 (2638 bp) carries direct flanking copies of IS1 (768bp).

Fig. 2.4. Compound Tn5. The Left-hand panel shows the structure of Tn5 and the way in which the flanking inverted IS50 insertion sequences would form a double-strand stem on denaturation and renaturation. The Right-hand panel shows the structure of plasmid pOX38:: Tn5 as revealed by electron microscopy. Plasmid DNA was denatured and renatured o form a snap-back structure that locates the Tn5 insertion. (Electron Microscopy by Edouard Boy de la Tour and Lucien Caro).

It was also realized that IS might be related to the controlling elements discovered by genetic analysis of maize several decades previously . However, in spite of the observation that IS can be present in some bacterial species in extremely high copy numbers , little at the time prepared us for the subsequent recognition of the preponderant role they play in shaping genomes, of their extreme diversity and their widespread distribution (see reference )(Fig.2.5).

Fig. 2.5 Transposases: the most abundant and ubiquitous genes in nature. A "word cloud" (http://www.wordle.net) representing the 100 most abundant functional roles in metagenomes. The font size of each functional role is proportional to its abundance (Taken from Aziz et al. 2010).

Early Transposition Models

The unexpected recombination properties of transposable elements which were accumulating in the mid- to late-1970s led to a flurry of molecular models published between 1978 and 1981. Many of these independently derived models shared common features, a result that reflects the delays in communicating ideas, manuscripts and journals in a pre-internet age. They were developed to explain properties such as the absence of a requirement for extensive homology between donor and target and independence from RecA , generation of short direct flanking repeats of target DNA (e.g.), adjacent deletions and other transposon-induced DNA rearrangements and, in the case of Tn3 and bacteriophage Mu, transposition-associated replication.

The first model published in 1979 in a volume of Cold Spring Harbor Symposia on Quantitative Biology from a symposium in 1978 . This model was designed to explain the 9 base pair target duplication observed following insertion of IS1. It included both non-replicative and replicative proposals. In the replicative model (Fig.2.6 A), the target undergoes a double strand staggered cleavage to prepare for generating the direct flanking target repeats and one transposon strand is transferred and copied into the target. Following completion, replication of the second transposon strand occurs in the target. The model also showed how deletion intramolecular transposition could generate adjacent deletions (Fig.2.6 B).

Fig. 2.6. The Grindley-Sherratt Model for IS1 Transposition. Transposon DNA is represented as a zig-zag line and newly replicated DNA is in bold. Model for transposition (Top). A) “An insertion-element-specific enzyme”, the transposase, binds to “the terminal sequence of the element”, the IR, and catalyzes single-strand cleavage at one end while a host enzyme was proposed to introduce a staggered double-strand break, either by concomitant cleavage of both strands or consecutive cleavage of each strand, at the target site. B) The single strand with a 3’ end is transferred to a 5’ end at the cleaved target site. The authors also propose that the transposase might remain bound to protect its activity. C) the (recessed) 3’ end of the target DNA is then used as a primer for replication and generates one of the targets DR in doing so. D) the extended new strand is joined to the free 5’ target end and replication proceeds from the end free 3’ target end, displacing the invading TE strand in the process and completing the second copy of the DR. D) Ligation displaces the second half of the host dimeric enzyme. Replication from the resultant free 3' end gives a second copy of the target sequence. Two outcomes are shown: an entire copy choice pathway leaving the copied TE strand intact (left) or a breakage rejoining reaction (right). Formation of adjacent deletions (Bottom). The deletion occurs by intramolecular transposition. A) Using a target site on the same molecule B) Breakage of the transferred strand occurs during replication and C) Rejoining both ends and elimination of the intervening DNA resulting in the adjacent deletion.

This model was soon followed by a replicative transposition model which included a step in which both donor and target replicons underwent fusion to form a “cointegrate” (Fig.2.7 A). This is generated by cleavage of a single strand at each transposon end, attack of the target in a staggered way and re-joining to generate an intermediate (often called a “Shapiro” intermediate) in which both donor and target replicons are joined via a single Tn copy in a branched structure. The branched DNA forms two potential replication forks which are used to duplicate the transposon generating an intermediate in which a circular fused donor-target carries a directly repeated transposon copy at each junction. This was then proposed to “resolve” by recombination between the two transposon copies resulting in regeneration of the original donor and creating a target molecule containing a transposon copy. The model also explained how this type of transposition, if it occurs into an intramolecular target, could result in deletions and inversions (Fig.2.7 B).

Fig. 2.7. Shapiro Replicative transposition Model. A) Transposition and replication. Donor DNA is represented by solid lines with the transposon shown in bold. Target DNA is shown as a dashed line with the bases which form the target duplication indicated as small black squares. Newly replicated DNA is shown as a white box. Lower case letters a and b indicate sequences flanking the transposon in the donor, c and d indicate those that flank the target integration site. The top drawing shows an overview of donor and target to indicate how they might associate to initiate transposition. The positions at which strand cleavage will occur are shown by small arrows (I). Below, the figure shows simultaneous single-strand cleavage at both ends; strand transfer of these ends in a staggered manner into the target site resulting in a branched molecule in which donor and target are joined with the formation of a (replication) fork at each end (II); replication of the transposon to generate a cointegrate in which donor and target molecules are separated at each junction by a copy of the transposon (III); and recombination between the directly repeated transposons to separate donor and target DNA (IV). B) The consequences of intramolecular transposition. Here, the target site is located on the same DNA molecule as the transposon, and the duplicated target sequence is shown as white squares. The top diagram shows transposition into the double-strand target in one orientation leading to the deletion of a circular form of the DNA located between the transposon and its target site. The bottom diagram shows the transposon attack of the target in the opposite orientation leading to inversion of the intervening DNA. Note that in both cases, the duplicated target sequences do not appear flanking a single transposon copy but are associated with separate transposon copies.

Although the 1978 Symposium included only a short section on transposable elements, the 1980 Symposium was dedicated entirely to Moveable Genetic Elements. One proposition published in this volume included a simple modification of the replicative cointegrate model, single second strand cleavage of the non-transferred strand at each end, to explain conservative (non-replicative) transposition .

Another model, published in 1980, was designed to take into account the observation that bacteriophage Mu appeared to target replication forks . In this model (Fig.2.8)., phage Mu was proposed to replicate in situ (i.e. from its integrated location as a prophage following cleavage at one end and replication using the flanking 3’OH host DNA as a primer. A single strand phage intermediate was then proposed to invade a nascent replication fork (possibly by using interactions between a phage encoded protein and a component of the fork – as transposon Tn7 has subsequently been shown to do).

Fig. 2.8. The Fitts-Taylor Model for Mu Transposition. On the left, flanking host DNA is represented as a zig-zag line and newly replicated DNA is in bold. The top cartoon shows a copy of phage Mu inserted into host DNA. A) cleavage occurs at one end to generates a 5’ phosphate (P) and a 3’OH on the host DNA. B) replication of phage DNA proceeds using the host 3’OH as a primer. C) multiple rounds of replication occur producing linear single strand phage DNA. D) the single strand intermediate is aligned into an active host replication fork E) insertion occurs by ligation of the host 3’OH with the phage 5’P and replication continues using the 3’OH pf the phage strand. F) the inserted Mu is subsequently replicated as a double-strand insertion by the second round of host replication.

This volume also included, as an appendix, another replicative transposition model to explain transposition of bacteriophage Mu . This was later published as a full article . The essential elements of this model were based on the results of electron microscopy analysis of DNA following induction of Mu transposition in vivo. This model (Fig.2.9 A) proposed an initial protein-mediated association between the target molecule and a circular Mu copy, followed by, what the authors called “attachment” (now called strand transfer) involving a single strand nick on the transposon and transfer of the end to a target DNA which had undergone double strand cleavage. The free 3’OH on the cleaved target molecule could then serve as a primer for “roll-in” replication and, when reaching the second transposon end, “roll-in” termination then occurs. The model also provided an explanation of the formation of cointegrates and simple transposition by the choice of strand cleavage and transfer at the termination step (Fig.2.9 B); for inversion (Fig.2.10 C) and deletion (Fig.2.10 D) of DNA by intermolecular transposition events; and for the occurrence of tandem multimer inserts when roll-in termination fails to occur and replication continues for another round (Fig.2.11 E).

Fig. 2.9. Harshey Bukhari roll-in replication model. Transposon DNA is shown as a zig-zag line in bold. The donor molecule is shown as a circle carrying the phage. DNA polarity is indicated by arrow heads (3’ OH) and small black circles (5’ P). L and R represent the left and right Mu ends. Transposase is shown as an unfilled oval. Newly replicated DNA is shown as a dashed line. A) 1) Shows the transposon and its target. 2) Shows Association of transposon and target using transposase bound at both transposon ends. The arrows indicate transposase-catalyzed cleavage. The transposon is cleaved on one strand only at one end, while both target strands are cleaved. 3) Shows transfer of a 3’ OH end of the transposon to a 5’ end of the target (Attachment). This creates a replication fork at the point of transfer. 4) Shows roll-in replication of the transposon using the 3’OH group of the target as a primer. 5) The accompanying transposase is shown to recognize the other end, which results in cleavage and roll-in termination. B) The model explains cointegrate formation (a) and simple transposition (b) simply by the choice of strand cleavage and transfer at termination.

Fig. 2.10. Hershey Bukhari roll-in replication model. Transposon DNA is shown as a zig-zag line in bold. The donor molecule is shown as a circle carrying the phage. DNA polarity is indicated by arrow heads (3’ OH) and small black circles (5’ P). L and R represent the left and right Mu ends. Transposase is shown as an unfilled oval. Newly replicated DNA is shown as a dashed line. A, B, C, D, and E represent genetic markers which are subject to inversion or deletion. C) Intramolecular transposition leading to DNA inversion. D) Intramolecular transposition leading to DNA deletion.

Fig. 2.11. Harshey Bukhari roll-in replication model. Transposon DNA is shown as a zig-zag line in bold. The donor molecule is shown as a circle carrying the phage. DNA polarity is indicated by arrow heads (3’ OH) and small black circles (5’ P). L and R represent the left and right Mu ends. Transposase is shown as an unfilled oval. Newly replicated DNA is shown as a dashed line. A, B, C, D, and E represent genetic markers that are subject to inversion or deletion. D) A circular form of the transposon is shown. Continued replication through the second transposon end and “random” termination id proposed to result in tandem duplications of inserted DNA

Another very similar model was published in the same journal later that year (Fig.2.12). This was built to explain observations that the frequency of transposition of IS1-based compound transposons appeared to decrease exponentially as a function of increasing size . This model provided a mechanism for alternative pathways of cointegrate formation and direct transposition which simply depended on which of the two strands at termination were exchanged (Fig.2.12 A and B), how individual IS transposition or transposition of the entire compound transposon might occur (Fig.2.13 C) and with the same idea (failure to terminate), how tandem transposon copies could be inserted (Fig.2.14 D and E).

Fig. 2.12. Galas Chandler rolling circle transposition model Interreplicon transposition pathways. A) An overall view of the model for transposition and cointegrate formation is shown in this series of drawings. The initial transposon is represented as the purple section of the donor plasmid (green), with the ends labeled A and B. The target site is indicated in the first drawing as a small red segment of the recipient plasmid (yellow) flanked by labels C and D. The three basic steps are indicated in the drawings labeled a (cleavage and single-strand transfer), b (formation of a replication fork and replication of the transposon), and c (termination). Strands of DNA synthesized during the process of replication are indicated as black strips. The two possible final states are shown in f and g, labeled to correspond to the drawings of B). B) Details of the transposition model. This figure shows the steps of the model in more detail, illustrating the fate of each DNA strand. The strands of the transposon are indicated as in A); the donor plasmid strands are shaded in this figure for added clarity. In d and e the circled number indicates nicks to be made in the donor strand and the heavy, curved arrow shows the ligation to be made. In f and g the arrows point out this newly formed joint.

Fig. 2.13. Galas Chandler rolling circle transposition model. C) Model pathways for a compound transposon (a section of DNA flanked by two transposable elements), showing the effects of the branching between termination and continued replication and between cointegrate formation and direct transposition. The possibility that a cointegrate can be resolved by recombination is shown in parentheses to indicate the observed stability of these structures. In the case of resolution of a cointegrate formed by duplication of TnX, there are several possible ways that resolution can occur. These pathways are not shown.

Fig. 2.14. Galas Chandler rolling circle transposition model. D) Artist impression of transposase emotions on successful or unsuccessful transposition termination. E) The consequences of unsuccessful termination in the case of a compound transposon.

Formally, these models can all be simply classed by a series of reactions whose order can be varied (Fig.2.15). Note that these did not foresee the large diversity in transposition reactions to generate circular transposon intermediate or the detailed cleavages and interstrand reactions discovered later (see section "Reaction Mechanisms").

Fig. 2.15. Flow diagram of events leading to transposition products according to standard hypotheses. The sites in the transposon (open bar) a and b and target c and d are shown at the bottom of the figure and referred to in the diagram.

Acknowledgements

We would like to thank David Sherratt (University of Oxford), Nigel Grindley (Yale University), Rasika Harshey (University of Texas at Austin) and Jim Shapiro (University of Chicago) for permission to use their original figures. We would also like to thank David Galas (Pacific Northwest Research Institute) for permission to use his unpublished drawings.