
Full text loading...
Category: Clinical Microbiology
Diversity-generating Retroelements in Phage and Bacterial Genomes, Page 1 of 2
< Previous page | Next page > /docserver/preview/fulltext/10.1128/9781555819217/9781555819200_Chap53-1.gif /docserver/preview/fulltext/10.1128/9781555819217/9781555819200_Chap53-2.gifAbstract:
Mobile genetic elements have repeatedly been called to duty in life-and-death struggles between hosts and their pathogens ( 1 , 2 , 3 , 4 ). One of their greatest utilities is the capacity to create DNA sequence diversity in protein-encoding genes, thereby generating protective shields to defend against enemies, or to create arsenals of weapons to exploit potential hosts. After decades of research, considerable evidence now suggests that the V(D)J recombination system, which is essential for generating adaptive immunity in vertebrates, has evolved from an ancestral DNA transposon ( 2 , 3 , 4 ). The site-specific recombinases responsible for V(D)J recombination, RAG1 and RAG2, are able to catalyze DNA transposition in a manner analogous to DNA transposons ( 2 ), and the RAG1 core and V(D)J recombination signals are likely derived from the transposase and terminal repeats of an ancient DNA transposon similar to Transib ( 3 , 4 ). Ironically, pathogens also exploit mobile genetic elements to generate protein diversity, altering their antigenic characteristics to evade host immunity ( 1 ). This process of antigenic variation is employed by Borrelia species, Neisseria gonorrhoeae, and other pathogens. Bacterial antigenic variation often involves a single, highly expressed gene encoding an abundant surface protein and dozens of archived ones that are homologous but different from each other. Replacing all or part of the expressed copy by DNA transposition leads to antigenic variation on the surface of the pathogen.
Full text loading...
BPP-1 phage and its diversity-generating retroelement (DGR). (A) The BPP-1 genome is represented in the prophage form flanked by a duplication of the His-tRNA gene formed during integration. Functional assignments for most gene clusters are indicated, along with the cI-like repressor and the DGR cassette. (B) Schematic representation of the DGR cassette and its function in phage tropism switching. The cassette contains three genes (mtd, avd, and brt) and two 134-bp repeats (template and variable repeats, or TR and VR, respectively). The VR is located at the 3′ end of the mtd gene, which encodes the distal tail fiber protein responsible for receptor recognition. Located at the 3′ ends of the VR and TR are IMH (initiation of mutagenic homing) and IMH* elements, respectively, in addition to a GC-rich element. Phage tropism switching occurs through DGR-mediated mutagenic homing, in which TR sequence information is transferred to the VR with adenine residues in the TR appearing as random nucleotides in the VR. Shown on the bottom are electron micrographs of the BPP-1 phage; globular structures at the distal ends of tail fibers are Mtd trimers (two per fiber). (C) Comparison of the BPP-1 TR and VR. The TR and VR sequences are shown in bold. The VR variable positions and the corresponding adenine residues in the TR are shown in red. IMH, IMH*, and GC-rich elements are also indicated. There are 23 adenines in the TR, which can theoretically generate ∼1014 different DNA sequences, or ∼1013 different peptides. (Adapted from 7 and 9 .)
Diversification of a surface-displayed lipoprotein by a Legionella DGR. (A) The L. pneumophila strain Corby DGR is encoded on a genomic island that differs in G+C content from the rest of the genome. The VR sequences at the 3′ end of the diversified locus, ldtA, are flanked by tandem hairpin/cruciform structures that are essential for efficient homing. The TR contains 43 adenine residues, which can create a potential repertoire of 1026 different VR DNA sequences. (B) LdtA contains atypical TAT (twin arginine transport) and Lpp (lipobox, lipid modification) signals at the N terminus. (C) Cellular localization studies demonstrated that LdtA is exported through the inner membrane via the TAT pathway, lipid modified, and anchored on the outer surface of the outer membrane via an Lpp-like lipoprotein processing pathway. VR-encoded residues are surface displayed by a C-terminal CLec fold. (Adapted from 17 .)
The CLec fold as a scaffold for display of DGR-generated protein diversity. (A) Left: The BPP-1 Mtd protein forms a pyramid-shaped homotrimer with VR-encoded residues exposed on the bottom surface. Right: An Mtd monomer containing β-prism, β-sandwich, and VR-encoded CLec domains, from the N to the C terminus. (B) Backbone structures of the VR regions of five Mtd variants with different ligand specificities are shown. Despite side-chain variations in diversified VR residues, the backbone structures are nearly superimposable. (C) Comparison of the CLec VR regions of BPP-1 Mtd and a T. denticola variable protein, TvpA. For the Mtd VR, the β2β3 loop of a second monomer is also shown (blue). (D) Superposition of the VR regions of BPP-1 Mtd (light orange) and T. denticola TvpA (blue). (E) Interaction of an Mtd homotrimer with the receptor protein pertactin. See text for details. (Adapted from 14 , 20 , and 22 .)
The TPRT model of BPP-1 DGR-mediated mutagenic homing. (A) Mutagenic homing occurs through a TR RNA intermediate and is RecA independent, similar to group II intron homing. A marker co-conversion assay mapped the cDNA transfer boundary to a narrow region within the GC-rich element at the 3′ end, which may represent 3′ cDNA integration site(s). The marker transfer boundary at the 5′ end was more heterogeneous. A target DNA-primed reverse transcription model, similar to that of group II intron homing, has been proposed to explain these observations. The DNA target site was hypothesized to be nicked within the GC-rich element, with the exposed 3′-hydroxyl group serving as a primer for adenine-specific error-prone reverse transcription of the TR RNA. Integration of cDNA products at the 5′ end requires short stretches of homology between the VR and the cDNA and may occur through strand displacement or template switching followed by break repair. Subsequent DNA replication would then create progeny genomes with mutagenized variable regions. (B) Deletion of the VR sequence upstream of GC and IMH elements appeared to block 5′ cDNA integration but not 3′ cDNA integration, as analyzed by PCR with primer sets 1/4 and 2/3, respectively. Sequence analysis showed adenine mutagenesis in PCR products generated with primers 2&3. (C) 5′ cDNA integration in ΔVR1-99 was restored by inserting a 50 bp mtd sequence, which is homologous to the region upstream of the deletion junction, in the TR. (Adapted from reference 7 )
Role of a DNA secondary structure in DGR target recognition. (A) A DNA hairpin/cruciform structure downstream of the VR is required for BPP-1 DGR target recognition. The wild-type (WT) structure contains an 8-bp GC-rich stem and a 4-nt GAAA loop and is located 4 bp downstream of the VR. Mutating the 3′ half of the stem (StMut) dramatically reduced DGR mutagenic homing (B) and phage tropism switching (not shown), while complementary changes to the 5′ half of the stem (StRev) restored DGR activity in both assays. (B) PCR-based DGR homing assays with sequence-tagged TRs and VRs flanked by WT or mutant stem sequences. Shown on the right is a diagram of the PCR assay. Green represents the tag sequence transferred from the TR to the VR. P1 to P4 are primers annealing to the tag or flanking regions. (C) Similar DNA structures are found at analogous positions in a number of other phage (two shown) and bacterial (one shown) DGRs. The phage stems are GC rich and range from 7 to 10 bp, and loops have a conserved 4-nt sequence, G(A/G)NA. The L. pneumophila Corby DGR has a more complex tandem structure that is required for homing. (D) BPP-1 DGR target recognition at the 3′ end is both sequence and structure dependent, requiring GC, IMH, and a hairpin/cruciform structure. Target recognition at the 5′ end is homology mediated. By inserting a gene of interest (GOI) upstream of GC, IMH, and the DNA structure, the heterologous gene can be diversified by the BPP-1 DGR through appropriate engineering of the TR. (Adapted from 41 .)
Avd and Brt. (A) Left: The BPP-1 Avd protein forms a homopentameric structure, with each monomer containing four helices running up and down (side view). The pentamer is highly positively charged (top view; blue, positively charged; red, negatively charged). Right: Amino acid residues on the side, top, and bottom of the Avd pentamer that were tested for Avd–Brt binding and/or DGR homing ( 27 ). (B) DGR RT domains and the sequence logo of its highly conserved domain R4. R1 to R7 are conserved sequence blocks found in the finger and palm domains of retroviral RTs, such as human immunodeficiency virus type 1 (HIV-1) RT (bottom). DGR RTs contain sequence insertions between R2 and R3 (R2a), and between R3 and R4 (R3a), as well as divergent N and C termini. They do not contain the thumb (Th) and RNase H (RH) domains that are found in HIV-1 RT. The domain R4 sequence logo of 155 DGR RTs was generated by Schillinger et al. using WebLogo ( 13 , 42 , 43 ). Comparison with the domain R4 sequence logo that we generated from 93 bacterial group II intron RTs (group II intron database: http://webapps2.ucalgary.ca/∼groupii/orf/orfalignment.html; [ 44 , 45 , 46 ]), which are most closely related to DGR RTs, showed several characteristic differences, including the two highly conserved positions labeled with an asterisk (*). Also included for comparison is the corresponding amino acid sequence block of HIV-1 RT (strain BRU; GenBank accession no. K02013). The glutamine residue at position 151, which plays a role in nucleotide and template preference during reverse transcription, is highlighted in blue. (Adapted from 13 .)