The Mycobacteriophage D29 Gene 65 Encodes an Early-Expressed Protein That Functions as a Structure-Specific Nuclease

The genomes of mycobacteriophages of the L5 family, which includes the lytic phage D29, contain several genes putatively linked to DNA synthesis. One such gene is 65, which encodes a protein belonging to the RecA/DnaB helicase superfamily. In this study a recombinant version of the mycobacteriophage D29 gp65 was functionally characterized. The results indicated that it is not a helicase as predicted but an exonuclease that removes 3′ arms from forked structures in an ATP-dependent manner. The gp65 exonuclease acts progressively from the 3′ end, until the fork junction is reached. As it goes past, its progress is stalled over a stretch of seven to eight nucleotides immediately downstream of the junction. It efficiently acts on forked structures with single stranded arms. It also acts upon 5′ and 3′ flaps, though with somewhat relaxed specificity, but not on double-stranded forks. Sequence comparison revealed the presence of a KNRXG motif in the C-terminal half of the protein. This is a conserved element found in the RadA/Sms family of DNA repair proteins. A mutation (R203G) in this motif led to complete loss of nuclease activity. This indicated that KNRXG plays an important role in the nuclease function of not only gp65, but possibly other RadA/Sms family proteins as well. This is the first characterization of a bacteriophage-derived RadA/Sms class protein. Given its mode of action, it is very likely that gp65 is involved in processing branched replication intermediates formed during the replication of phage DNA.Fork structures are intricately associated with DNA replication. Such structures result due to unwinding of the DNA ahead of the replicating machinery. The unwound single strands are then used as templates for the synthesis of the new strands, either continuously (leading strand) or discontinuously (lagging strand). Repair of stalled forks involve complex mechanisms which may vary from one organism to another (5). However, in most cases the process requires nucleases that recognize stalled fork structures and cleave them specifically. Such nucleases are generally referred to as structure-specific nucleases (25). One such nuclease named FEN1 found in eukaryotes has been studied fairly extensively, and it is believed that this nuclease is involved in the removal of 5′ flaps from Okazaki fragments (11, 23). FEN1 belongs to a larger family of structure-specific nucleases, which includes human XPG (17), an endonuclease related to the disease xeroderma pigmentosa. Although the XPG family is associated with the removal of 5′ flaps the XPF type proteins are needed for removing the 3′ flaps (3). Similar proteins have been found in several Archaea (28). In Escherichia coli, the Holliday junction resolving enzyme system RuvABC is believed to be involved in resolving stalled forks by creating double-stranded breaks, which may be repaired through homologous recombination (29). Studies in E. coli have revealed that there are multiple redundant pathways that are capable of repairing stalled forks. One such pathway involves a protein named RadA/Sms, the absence of which results in partial increase in sensitivity to radiation in E. coli (2). Genes encoding RadA/Sms family proteins are present in many bacteria, including mycobacteria. Most of these members carry a conserved element KNRFG. It is believed (2) that RadA/Sms family of proteins may generate double-stranded breaks at fork junctions, although this has not been specifically demonstrated.Mycobacteriophages of the L5 family, which includes D29, BxB1, may be either temperate or potentially temperate (D29) (14, 15, 27). Despite their temperate character these phages share a strong resemblance with lytic phages. An important feature shared by lytic phages in general is their ability to synthesize DNA using phage-encoded DNA polymerases (13). They also possess many genes linked to nucleotide metabolism. It appears that as far as DNA replication is concerned, lytic phages prefer to be self-sufficient. This is apparently an important issue since lytic phages inactivate their host and therefore host-specific functions cannot be used to support phage growth.Following the availability of the genome sequence, many interesting aspects of mycobacteriophages have come to light. The central region of mycobacteriophage L5/D29 genome has been predicted to harbor several genes whose products may contribute directly or indirectly toward synthesis of new DNA strands. In a recent investigation from this laboratory it has been demonstrated (4) that at least some of the genes in this region are involved in the production of deoxyribonucleotide precursors which are probably needed at increased levels during phage replication. Apart from these genes there are several others which probably encode DNA polymerization related functions. One such gene that drew our interest was gene 65, which appears to encode a RecA/DnaB helicase superclass protein (22). The N-terminal region of this protein contains the Walker motifs A and B, which are characteristically present in the members of the RecA/DnaB superfamily. Walker motifs A and B (30) are found in proteins that hydrolyze ATP for executing their respective functions. To investigate the possible function of gp65, its gene was overexpressed in E. coli, and the recombinant protein was purified. Assays performed with the recombinant gp65 revealed that it is a structure-specific nuclease that acts exonucleolytically on fork structures, resulting in truncated forms lacking the 3′ arm. This function was demonstrated to require a particular motif KNRXG that is omnipresent in the RadA/Sms family of proteins (2). This characterization of D29 gp65 could give us better insight into how mycobacteriophages replicate their DNA within their hosts.