TnphoA’s transposition activity derives from that of this Tn5 transposon by a conservative mechanism in a relatively rare but highly regulated process [2]. The required elements for Tn5 transposition are the transposase protein (encoded from the transposase gene on the Tn5 itself), specific end sequences (which define the termini of the transposon), and the target DNA [3]. Host proteins and conditions, such as the degree of DNA dam methylation, also play a role in the regulation and frequency of transposition, as does the transposon-encoded Inh (Inhibitor) protein.

Although the exact mechanism of Tn5 transposition is not completely know. The following model is useful for understanding the system. First, there is sequence-specific binding of the transposase protein to the 19 bp specific end sequences of Tn5 [3]. A synaptic complex then forms in which the ends of the transposon are brought together by transposase oligomerization [3]. Cleavage of the DNA adjacent to the transposon is mediated by the transposase protein when then binds the target DNA sequence [3]. Next both strands of the target DNA are cleaved and strand become covalently linked to the target’s 5’-PO₄ ends [3]. Lastly there is a repair process which resolves the transposition [3].

The phoA portion of the TnphoA construct came from Escherichia coli K12 where it codes for the periplasmic protein alkaline phosphatase [4]. This enzyme must be secreted from the cytoplasm in order to demonstrate activity and its secretion is determined by the presence of a signal sequence at its amino-terminal end [4]. Once the promoter and signal sequence are deleted from the phoA gene, its activity can only be detected if its transcription is placed under the control of (and in the same reading frame as) another gene’s promoter that has a signal sequence at its amino terminus. This effectively makes TnphoA a probe for export signals in proteins into whose genes it inserts [1].

In order to confirm that the signal sequence of the other protein in the fusion is required to obtain alkaline phosphatase activity, an experiment was performed by [1] in their original paper on the construction of TnphoA. A deletion mutant of bla (the gene which codes for β-lactamase), missing the signal sequence, had been previously constructed [5]. Insertion into the wild type bla gene resulted in alkaline phosphatase activity whereas insertion into the deletion mutant of bla did not. Therefore, the activity of the bla-phoA hybrid protein clearly depends on the existence of this signal sequence in β- lactamase [1].

Other experimental data also points to the utility of the TnphoA construct and confirms our understanding of its mechanism of action. For example, approximately 75% of 57 independently isolated insertions were located in different positions of the target DNA (shown by restriction enzyme analysis) [1]. This suggests that TnphoA maintains the same low insertional specificity of Tn5. Additionally; cell fractionation confirmed that hybrid proteins located to the cell periplasm [1]. Finally, it was also shown that the hybrid proteins possessed substantial stability (with a half life of 30 mutants in the case of the bla-phoA fusions). Although most eventually broke down to give alkaline phosphatase-sized fragments [1].

Several applications of TnphoA mutagenesis have already been outlined in brief. More thorough descriptions in the contexts of specific examples will now be explored. The ability of TnphoA to give insights into the membrane topology structure of proteins will be addressed with consideration given to Tnlacz-another Tn5 derivative, which is complementary to TnphoA. Along similar lines, an example of protein characterization in Haemophilus influenza will be described. Finally, the utilization of TnphoA mutagenesis to identify new virulence genes in Agrobacterium tumefaciens will also be reviewed briefly.

Since the enzymatic activity of alkaline phosphatase depends on its subcellular location, with export to the periplasm corresponding to high levels of activity and presence in the cytoplasm corresponding to low levels of activity, this can be a useful tool to analyze membrane protein topology. Other methods to ascertain information about membrane protein structure do exist, such as applying proteases or other reactive molecules, but this genetic approach has the advantage of not depending directly on exposure to amino acid side chains [1].

One example of how this approach has been successfully used can be found in the case of bex locus in encapsulated Haemophilus influenzae. A 5.1 kb region in the capsulation locus of H. influenzae was determined using Sanger dideoxy-sequencing and found to contain four open reading frames, called bexd, bexC, bexB, and bexA [6]. These genes (excluding the 3’-end of bexD) were cloned into a plasmid, pLNB24, and mutagenized with TnphoA [6]. Mutants were selected through their production of the blue-colony phenotype on X-phos media and were then examined further [6]. All three of the clones that were looked at were located in the 5’ half of bexC, suggesting that its gene product is secreted into the periplasm [6]. Based on this and other information it was postulated that this protein may participate in a pore or carrier mechanism which might assist in the establishment and maintenance of the polysaccharide capsule of H. influenzae, which is a major virulence determinant [6].

A major advance in membrane protein topology studies came with the development by [7] of Tnlacz, which is also a Tn5 derivative and complementary to TnphoA. The lacz gene codes for the enzyme β-galactosidase which is only detectable in the cytoplasm, meaning that Tnlacz must insert into a gene which is not secreted from the cytosol frequently substantial in order to be detected. Even if the Tnlacz inserts into a different reading frame from the gene it has inserted into, there is still frequently substantial β-galactosidase activity, possibly due to translational [7]. Due to its complementary function with TnphoA, the Tnlacz construct can aid in the determination of where a protein or protein domain is normally within the cell. This was shown by [7] with the pBR325 plasmid which encodes chloramphenicol transacetylase, β-lactamase and tetracycline resistance protein [7].

Further enhancing the utility of the TnphoA and Tnlacz contructs is the ability to convert one type of fusion to another through a process known as fusion switching which uses recombination or in vitro manipulation [7]. This is possible because both constructs share the same 48 bp left-end sequence from the Tn5 transposon from which they were each derived. Since it is possible to generate alkaline phosphatase and β-galactosidase fusions at the same location in a protein, a comparison of their enzymatic activities can be performed to better establish the subcellular location of the protein site [7].

One of the major applications of TnphoA mutagenesis is the identification of new genes that code for transmembrane or secreted proteins. An excellent example of this application is found in the case of Agrobacterium tumefaciens, gram negative bacteria that causes crown gall tumors on plants through a unique DNA transfer process [8] undertook an extensive mutational analysis of the organism using TnphoA mutagenesis. Advantages of this approach included the transposon’s (much like other derivatives of Tn5) relatively random insertions into the genome, its relatively high degree of stability and the likelihood that may genes involved in the transfer process code for transmembrane or periplasmic proteins.

Before the transposon mutagenesis could begin, however, it was first necessary to derive phoA minus mutants. This was because A. tumefaciens expresses endogenous alkaline phosphatase activity that would completely mask the mutant phenotype. Chemical mutagenesis was used successfully to obtain multiple strains which had stable pho minus phenotypes [8] and continued virulence was confirmed to ensure that virulence related genes remained functional [8].

TnphoA was introduced into a phoA minus, virulent strain on a plsmid which is not maintained in A. tumefaciens [8]. The advantage of using such a “suicide” vector which, in this case, contains the transposase gene outside of the mobilized transposon, is that the likelihood of the insert being maintained at that site is increased since the transposase gene would be required for future transposition activity. This procedure was pioneered in Rhizobium (a relative of Agrobacterium) by [9]. Approximately 2000 mutants of A. tumefaciens with phoA activity were screened for virulence from which 30 avirulent or weakly virulent mutants were drawn [8]. Characterization of the mutants using Southern analysis showed which of the mutants located were previously known and which were newly identified genes [8].