DNA duplication followed by divergent evolution of the two copies is a major method postulated for the formation of new genes. There are only a few cases in T-even phages for which there is evidence for such duplication. These include the fibrous products of genes 12, 34 and 37, as well as head proteins gp23 and gp24. We have found an unusual and apparently ancient repeat involving the first 460 nucleotides of alt and a pair of ORFs, alt.-1 and alt.-2.

Alt.-1 and alt.-2 are located between the T4 map positions 123.254 and 123.437 and between 123.054 and 123.255 respectively (Kutter, Guttman, Mosig, Rüger, 1992). This gives the two ORFs a total length of 383 base pairs. Alt.-1 and Alt.-2 are flanked by the alt gene on one side and alt.-3 on the other (Alt.-3 has its own early promoter). It is likely that the alt.-1 and alt.-2 genes are expressed since there is no terminator before them and they seem to be a part of the same transcript as alt. There is a good Shine-Dalgarno (AAGGA) before alt.-1 and a partial Shine-Dalgarno (AGG) properly spaced before alt.-2.

There is some question about whether alt.-1 and alt.-2 are two separate ORFs or only one. The end of alt.-1 overlaps the beginning of alt.-2 by one base pair. If the sequences of the two orfs are arranged serially so that they overlap by that one base pair, they can be aligned with the first 388 base pairs of the alt gene sequence with 65% homology in nucleotide sequence and a 59% homology in amino acid sequence. The current sequence of T4 alt.-1, alt.-2 has two separate ORFs. If this is the case, then alt.-2 would contain a frame shift. We are resequencing the region to confirm the current recorded sequence. If alt.-1 and alt.-2 are two separate orfs, it is possible that they are the remenants of an ancient duplicate of the alt gene that had a frame shift mutation which produced a premature stop codon. This raises the possibility that the RNA polymerase may execute a translational frame shift to read the two orfs as one protein. This process is known to happen in other T4 and E-coli genes such as T4 gene 60 and E. coli release factor 2 (Gesteland, Weiss and Atkins, 1992).

gpAlt is the protein that is known to be responsible for the ADP ribosylation of one of the alpha subunits of the host RNA polymerase, causing increased recognition of most T4 early promoters (Koch, Raudonikiene, et al, 1995). The NAD⁺ binding site of gpAlt is located at the NH₂ terminus (Koch and Rüger, 1994). This is the same area that is homologous to the alt.-1, alt.-2 sequences. Therefore, it has been hypothesized that the alt.-1 and alt.-2 may code for a NAD⁺ binding protein and/or an ADP-ribosyl transferase which may act independently or as a subunit of a larger enzyme.

We are also comparing the T4 alt.-1,alt.-2 region to related phages in an effort to see how conserved this region is and therefore, get a better understanding of the history and potential function of these orfs. We have carried out PCR on 23 different T-even phages using as many as five different pairs of primers predominantly located in highly conserved regions flanking the alt.-1, alt.-2 orfs. Our preliminary PCR preparations produced bands in 17 of the 23 phage strains tested. In most cases the the PCR fragments were of similar sizes, however, a few strains produced fragments of variable size. Poland's sequence has a 97% homology to T4 alt.-1,alt.-2 region. However, RB69 has a completely different stretch of DNA in this region with a good Shine-Dalgarno (GAGG) properly spaced before a start codon. This appears to be a previously unknown gene, which has been inserted in place of alt.-1 and alt.-2.

We are comparing the predicted secondary structure of alt.-1,alt.-2 to the secondary structures of other ADP-ribosyltransferases such as diphtheria toxin, T4 Alt, T4 ModA and T4 ModB. ModA and ModB have been shown to bind NAD⁺ and both of their predicted secondary structures are compatible with forming a Rossmann fold. Rossmann folds are known to bind NAD⁺ in several dehydrogenase enzymes (Rossmann et al., 1975). The predicted secondary structure of alt.-1alt.-2 does not seem to have a Rossmann fold pattern. We expected to see a repeated pattern of beta sheet of about 7 residues followed by an alpha helix of about 10 residues (Branden and Tooze, 1991).

T4 Alt has an NAD⁺-binding site within the first 150 amino acids on its NH₂ terminal region but its predicted secondary structure is not compatible with that of a Rossmann fold as predicted by Ruger (personal communication). Rather, it fits a pattern of the NAD⁺-binding site found in the diphtheria toxin. The diphtheria NAD⁺-binding site is a fold on the C-domain at a junction of two approximately orthogonal, antiparallel beta-sheet domains. Both the overall C-domain and the NAD⁺-binding site are distinct and unrelated to the Rossmann fold (Bell and Eisenberg, 1995). When comparing predicted secondary structure of alt.-1, alt.-2 to that of the first 128 amino acids of Alt using DNAstar, we see a strong similarity between the two.

(Side thought) While on the subject of toxins, with the alt.-1,alt.-2 region being as small as it is, it has been asked whether such a small protein could act as a ribosyltransferase. The smallest ADP ribosyltransferase found is the pertussis toxin which is only 180 amino acids. It catalyzes the NAD⁺ glycohydrolase reaction and ADP-ribosylates its target protein (Finck-Barbancon and Barbieri, 1994).

We are also cloning alt.-1,alt.-2 as a single fragment into the pET16b expression vector in order to determine the activity of the gene products. Once expressed and purified, we will be determining activity by column chromatography using a histidine tail and/or Sepharose blue gel.

references on their way!

Return to top.

Science Articles from CPJ

The Evergreen State College
2700 Evergreen Parkway NW
Olympia, WA 98505
(360) 867-6000