Grantee Research Project Results
Final Report: Phage-Mediated Transfer of Genes Between Bacterial Species
EPA Grant Number: R825348Title: Phage-Mediated Transfer of Genes Between Bacterial Species
Investigators: Cohan, Frederick M. , Berger, Evelyn , Mitrica, Ionel , Majewski, Jacek , Libsch, Jason , Palmisano, Margaret , Feldgarden, Michael , Palys, Thomas
Institution: Wesleyan University
EPA Project Officer: Hahn, Intaek
Project Period: January 5, 1997 through January 4, 2000
Project Amount: $443,966
RFA: Exploratory Research - Environmental Biology (1996) RFA Text | Recipients Lists
Research Category: Biology/Life Sciences , Aquatic Ecosystems
Objective:
The objectives were to: (1) determine how the potential for genetic exchange between ecologically distinct groups of bacteria is influenced by sequence divergence; and (2) develop improved methods for identifying ecologically distinct groups of bacteria from sequence data.Summary/Accomplishments (Outputs/Outcomes):
Sexual Isolation and Sequence Divergence. Our earlier work (supported by a previous EPA grant, R821388) demonstrated that resistance to genetic transformation across Bacillus species increases exponentially with the sequence divergence between donor and recipient. That is, transformation frequencies decline the same amount (approximately 10-fold) with each additional 5-percent sequence divergence. We further found that the exponential function is robust with respect to the genotype and species of the recipient, as well as the conditions of transformation. In the present grant, we investigated the molecular mechanisms underlying the exponential relationship between sexual isolation (i.e., resistance to recombination) and sequence divergence. The motivation is that with a better understanding of the mechanisms underlying sexual isolation, we should be able to predict how sexual isolation might be affected by the mode of recombination (i.e., transformation, transduction, and conjugation) and the size of the DNA segment that is recombined.We tested the effect of mismatch repair by comparing a wild-type strain and an isogenic mismatch-repair mutant for the relationship between sexual isolation and sequence divergence in Bacillus transformation. Mismatch repair was shown to contribute to sexual isolation, but was responsible for only a small fraction (16 percent) of the sexual isolation observed. Another possible mechanism of sexual isolation was that more divergent recipient and donor DNA strands have greater difficulty forming a heteroduplex, because a region of near-perfect identity between donor and recipient is required for initiation of the heteroduplex. Under a "one-chance" model, a given donor fragment is required to form a perfect match with the recipient at both ends. If matching is successful at both ends, recombination proceeds; but, if matching fails, no more opportunities for heteroduplex formation are allowed. A mathematical analysis showed that this one-chance model yields an exponential relationship between sexual isolation and sequence divergence. Moreover, this model yields an estimate of the size of the region of perfect identity that is comparable to independent estimates for Escherichia (E.) coli. For these reasons, and because all other mechanisms of sexual isolation may be ruled out, we concluded that resistance to heteroduplex formation is predominantly responsible for the exponential relationship between sexual isolation and sequence divergence in Bacillus transformation.
This conclusion was further supported by transformation experiments using artificially constructed donor DNA segments. In these constructs, a very divergent segment of the rpoB gene from a donor strain was sandwiched between rpoB segments from the recipient strain. By providing flanking sequences that were identical to the recipient, these constructs facilitated recombination of the intervening divergent segment. The constructs completely eliminated sexual isolation between extremely divergent species (provided mismatch repair was eliminated by using mismatch repair mutants as recipient). These experiments demonstrated that it is not the overall divergence between donor and recipient strands that hinders recombination. Instead, divergent strains fail to recombine because they lack small regions of identity required as anchors for strand initiation.
We investigated whether regions of identity are required at one or both ends of each recombining segment. We analyzed the junctions of recombination events in 80 recombinants, and found that every junction, at both ends of each integrated segment, corresponded to a region of near identity between recipient and donor. This demonstrated that near identity between recipient and donor is required at both ends. This is in contrast to studies with E. coli, where only one junction of recombination needs to be nearly identical.
By analyzing the sequences of recombination junctions, we explored the criteria of sequence similarity that allow a region to act as a junction in recombination. We found that the region must be at least 20 bp in length, with complete identity or at most one mismatch, and high GC content is required when a mismatch is present. All these criteria are related to the stability of the donor recipient heteroduplex, and one single stability criterion takes into account all three criteria: that the junction region at each end of the recombining segment must be stable enough to have a predicted melting temperature of 51.2?C or more.
These findings have important implications for the spread of new adaptations, such as antibiotic resistance or artificially introduced foreign genes, in the bacterial world. Once a gene for a novel adaptation transfers into a species from a divergent donor, the gene becomes flanked by recipient DNA, and, in the absence of barriers other than resistance to strand invasion, this allows efficient homologous recombination of the new adaptation into related species. Our results suggest that even very divergent and heterologous genes may transfer between species at very high rates via homologous recombination.
It is striking how much more mismatch repair contributes to recombination barriers in Escherichia than in Bacillus: in E. coli, mismatch repair is almost entirely responsible for the resistance to interspecies recombination. This raised the question of whether the role of mismatch repair in Escherichia or Bacillus is the more typical. We approached this issue by investigating another system of microbial recombination: natural transformation in Streptococcus (S.) pneumoniae. We transformed three recipient strains, two wild type and one deficient for the mismatch repair system (Hex-), with DNA from a diversity of related species with varying degrees of DNA sequence divergence.
Interspecies transformation of the two wild type S. pneumoniae strains followed an exponential decay function of sequence divergence that was very similar to that for Bacillus. The slope of the exponential decay function was much less steep for the mismatch repair mutant. Averaged over all donors, mismatch repair was responsible for 34 percent of the observed sexual isolation in Streptococcus (compared to 16 percent in Bacillus). Because the DNA uptake step in Streptococcus is not sequence specific and all of the recipient strains are nonrestricting, we have proposed that the remaining 66 percent of the barrier to recombination is due to difficulty in forming the donor-recipient DNA heteroduplex molecule, as in Bacillus.
While Bacillus transformation frequencies were reasonably well predicted by the one-chance model's exponential function, the transformation frequencies deviated from the purely exponential function in a consistent manner. All strains tested showed less resistance to transformation with closely related donors than expected under the one-chance model. We hypothesized that a "multiple-chance" model might better explain the pattern of transformation data. For example, under a two-chance model, if a donor segment's ends fail to form a perfect match with the recipient, one or both donor ends could be digested to a random extent, and the new donor ends would then be available for a second chance at matching the recipient. For 10 out of 10 strains tested, a multiple-chance model (either two-chance or three-chance) fits the transformation frequency data more closely than the purely exponential function of the one-chance model. We concluded that a transforming fragment has up to two or three chances to successfully form a heteroduplex with the recipient, with random digestion of the ends before the second and third chances.
We then hypothesized that a very long donor segment might have a greater number of chances to successfully match the recipient than does a shorter donor segment. This would be the case if after each failure to match the recipient, a short segment is digested away, and further chances for matching keep recurring until the segment becomes too small to recombine. Alternatively, there may be a constant number of chances for matching, regardless of the donor fragment length. If longer segments are indeed allowed more chances for successful integration, then longer segments are predicted to be less sensitive to donor-recipient sequence divergence. Under this hypothesis, the various modes of recombination should yield different sensitivities to sequence divergence. Transformation, where very small fragments (~ 8 kb) are transferred, should be most sensitive to sequence divergence; transduction (which transfers long fragments, ~ 200-300 kb) would be less sensitive to sequence divergence, and conjugation would be less sensitive still.
We have tested this prediction by comparing the relationship between sexual isolation and sequence divergence for transduction and transformation using our model system of Bacillus (B.) subtilis. Remarkably, transduction and transformation were statistically indistinguishable in their sensitivity to sequence divergence. Sexual isolation was predicted reasonably well with an exponential model, except that, like transformation, sexual isolation was lower than expected for very closely related donors. Indeed, like transformation, transduction frequencies showed the best fit to a model in which multiple opportunities for donor-recipient matching are allowed. Perhaps the number of opportunities for matching is determined less by the size of the donor fragment than by a fixed amount of time that a donor strand can survive nuclease activity.
It appears that the near-exponential relationship between sexual isolation and sequence divergence discovered in Bacillus transformation applies not only to transformation systems in other taxa, but to other modes of recombination as well. The possible hazard of an artificial gene spreading between closely related species appears to be predicted well by the sequence divergence between species, regardless of the mechanism of recombination.
Rates of Recombination, Migration, and Horizontal Transfer in Bacillus Group III Temperate Bacteriophages. Bacillus Group III temperate bacteriophages are harbored as prophage by B. subtilis and closely related species. We investigated the sequence diversity of Bacillus Group III prophages within and between several Bacillus taxa, and from several deserts, to infer rates of horizontal transfer, migration, and intragenic recombination in nature. The phage-borne immunity protein was amplified from host strains of B. vallismortis, B. mojavensis, B. licheniformis, B. subtilis subsp. subtilis, and B. subtilis subsp. spizizenii. Group III prophages were found to be present in moderately high frequencies within most Bacillus taxa and in all desert environments surveyed. Recombination among Bacillus prophages has occurred at the same low rate as previously found for their host bacteria ( 10-8). Bacillus Group III prophages formed one large sequence cluster, showing no evidence of clustering by host taxon or desert source. Therefore, rates of horizontal transfer of prophages between host taxa must be high, indicating no significant barriers to transmission among the taxa surveyed. Rates of intracontinental and intercontinental migration were high, suggesting that Group III bacteriophages are not isolated by distance or differences in local adaptations.
These results suggest that phage-mediated transduction has the potential to transfer host genes in nature, across host species as well as across continents.
Discovery, Classification, and Quantification of Ecological Diversity in the Bacterial World. Here we have developed an improved means to identify ecologically distinct bacterial populations that are very closely related.
All living organisms fall into discrete clusters of closely related individuals on the basis of gene sequence similarity. We developed an evolutionary genetic theory that predicts that in bacteria, each such sequence similarity cluster should correspond to an ecologically distinct population. Surveys of sequence diversity in protein-coding genes corroborated this theory by showing that sequence clusters correspond to ecological populations. We have predicted that future population surveys of protein-coding gene sequences can be expected to disclose many previously unknown ecological populations of bacteria. Moreover, sequence clusters should provide a useful tool for characterizing the ecological diversity within a habitat. Sequence similarity clustering in protein-coding genes is recommended as a primary criterion for demarcating taxa.
We have distinguished two very different uses for sequence data in the systematics of closely related species. One use is to diagnose and classify new isolates into existing species whose sequences have already been characterized. The second use is to discover cryptic populations and species within existing taxa. Strain diagnosis can be accomplished by either protein-coding genes or the more slowly evolving 16S rRNA genes. Others have shown, previously, that 16S rRNA sequences can distinguish the most closely related species, because only one diagnostic nucleotide site is required for this purpose. However, we have shown that the extremely low rate of 16S rRNA evolution frequently prevents this gene from allowing discovery of cryptic populations. We have shown that the greater rate of evolution in protein-coding genes makes these genes useful in both diagnosis of new strains and discovery of new taxa.
We have recently improved our sequence-based approach for discovering the ecological diversity within a bacterial species. Our "star method" is based on theory demonstrating that a phylogeny of strains from the same ecotype should appear as a star radiation, with all strains equally related. With this premise, we have concluded that when a clade of strains within a taxon is not consistent with a star phylogeny, that clade must contain more than one ecotype. We have developed a method for identifying the most inclusive clades that are each consistent with a star phylogeny. Our interpretation is that each such clade is a distinct ecotype.
Application of this method shows that most bacterial species tested cannot be resolved into a star phylogeny. This implies that these species must each contain more than one ecotype. Analysis of two bacterial species (Neisseria meningitidis and Staphylococcus aureus) has demonstrated the existence of dozens of ecotypes in each species, each consistent with a star phylogeny. We have concluded that the bacterial world has at least an order of magnitude more ecologically distinct and coexisting populations than was previously expected.
Because the various ecological populations within a named species are extremely similar in their DNA sequences, we can predict that there should be very little sexual isolation between these populations. Therefore, a genetically engineered microorganism may easily transfer artificial genes into a great diversity of ecologically distinct populations, all within a named bacterial "species."
Journal Articles on this Report : 6 Displayed | Download in RIS Format
| Other project views: | All 9 publications | 9 publications in selected types | All 6 journal articles |
|---|
| Type | Citation | ||
|---|---|---|---|
|
|
Majewski J, Cohan FM. The effect of mismatch repair and heteroduplex formation on sexual isolation in Bacillus. Genetics 1998;148(1):13-18. |
R825348 (1999) R825348 (Final) |
Exit Exit |
|
|
Majewski J, Cohan FM. Adapt globally, act locally: The effect of selective sweeps on bacterial sequence diversity. Genetics 1999;152(4):1459-1474. |
R825348 (Final) |
not available |
|
|
Majewski J, Cohan FM. DNA sequence similarity requirements for interspecific recombination in Bacillus. Genetics 1999;153(4):1525-1533. |
R825348 (1999) R825348 (Final) |
Exit |
|
|
Majewski J, Zawadzki P, Pickerill P, Cohan FM, Dowson CG. Barriers to genetic exchange between bacterial species: Streptococcus pneumoniae transformation. Journal of Bacteriology 2000, Volume: 182, Number: 4 (FEB), Page: 1016-1023. |
R825348 (1999) R825348 (Final) |
Exit Exit |
|
|
Palys T, Nakamura LK, Cohan FM. Discovery and classification of ecological diversity in the bacterial world: the role of DNA sequence data. International Journal of Systemic Bacteriology 1997;47(4):1145-1156. |
R825348 (1999) R825348 (Final) |
Exit |
|
|
Palys T, Berger E, Mitrica I, Nakamura LK, Cohan FM. Protein-coding genes as molecular markers for ecologically distinct populations: the case of two Bacillus species. International Journal of Systemic and Evolutionary Microbiology 2000;50(Part 3):1021-1028. |
R825348 (1999) R825348 (Final) |
Exit |
Supplemental Keywords:
genetically engineered microorganisms, population biology, evolution, genetics., RFA, Scientific Discipline, Ecosystem Protection/Environmental Exposure & Risk, Ecology, Ecosystem Protection, exploratory research environmental biology, Chemical Mixtures - Environmental Exposure & Risk, Ecosystem/Assessment/Indicators, Environmental Chemistry, Genetics, Chemistry, Ecological Effects - Environmental Exposure & Risk, Ecological Effects - Human Health, Biology, Ecological Indicators, risk assessment, transduction, ecological effects, bacteria, gene-environment interaction, Bacillus subtilis, phylogenetics, microorganisms, soil, gene transfer, DNA, phage mediatedRelevant Websites:
http://www.wesleyan.edu/bio/cohan/cohan.html
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.