Regulation of hypercompetence in Legionella pneumophila

Abstract

Although many bacteria are known to be naturally competent for DNA uptake, this ability varies dramatically between species and even within a single species, some isolates display high levels of competence while others seem to be completely nontransformable. Surprisingly, many nontransformable bacterial strains appear to encode components necessary for DNA uptake. We believe that many such strains are actually competent but that this ability has been overlooked because standard laboratory conditions are inappropriate for competence induction. For example, most strains of the gram-negative bacterium Legionella pneumophila are not competent under normal laboratory conditions of aerobic growth at 37 degrees C. However, it was previously reported that microaerophilic growth at 37 degrees C allows L. pneumophila serogroup 1 strain AA100 to be naturally transformed. Here we report that another L. pneumophila serogroup 1 strain, Lp02, can also be transformed under these conditions. Moreover, Lp02 can be induced to high levels of competence by a second set of conditions, aerobic growth at 30 degrees C. In contrast to Lp02, AA100 is only minimally transformable at 30 degrees C, indicating that Lp02 is hypercompetent under these conditions. To identify potential causes of hypercompetence, we isolated mutants of AA100 that exhibited enhanced DNA uptake. Characterization of these mutants revealed two genes, proQ and comR, that are involved in regulating competence in L. pneumophila. This approach, involving the isolation of hypercompetent mutants, shows great promise as a method for identifying natural transformation in bacterial species previously thought to be nontransformable.

FIG. 1.

Reporter plasmids and strains used to assay Lp02 competence. (A) Reporter plasmids used for transforming L. pneumophila were constructed by first cloning a 3-kb piece of L. pneumophila DNA into pBluescript II KS⁺ to generate plasmid pJB955. Three pJB955 derivatives (pJB1389, pJB964, and pJB957) were constructed by cloning a gentamicin, chloramphenicol, or kanamycin resistance cassette, respectively, into the unique HindIII site in the 3-kb insert. (B) Two Lp02-derived reporter strains were used to assay natural competence. JV1103, a Cm^r-marked version of Lp02, was constructed by natural transformation using plasmid pJB964. JV1160, a Kan^r-marked version of Lp02, was constructed by natural transformation using plasmid pJB957. With these strains and plasmids, successful recombination can be assessed not only by gain of a drug resistance marker from the transforming reporter plasmid but also by loss of a drug resistance marker from the reporter strain.

FIG. 2.

Natural transformation of Lp02 is temperature dependent. (A) Transformation frequency as a function of growth temperature. L. pneumophila reporter strain JV1103 was grown on solid medium at 26, 30, 34, and 37°C in the presence of 200 ng of DNA from the reporter construct pJB957 (Kan^r). After 48 h for the 30, 34, and 37°C samples or 120 h for the 26°C sample, bacteria were plated on selective media versus nonselective media in order to determine the number of Kan^r transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁹; error bars indicate standard deviations. (B) Plate photograph of Kan^r L. pneumophila transformants. Approximately 10⁶ cells from each transformation reaction represented in panel A were plated on media containing kanamycin.

FIG. 3.

The frequency of Lp02 transformation is dependent on DNA source and quantity. (A) Transformation frequency as a function of plasmid DNA quantity. Reporter strain JV1103 was grown on solid media at 30°C in the presence of various quantities of pJB957 (Kan^r) plasmid DNA. After 48 h, bacteria were plated on selective media versus nonselective media in order to determine the number of kanamycin-resistant transformants in the total cell population. (B) Transformation frequency as a function of chromosomal DNA quantity. Transformation was assayed as in panel A using JV1160 (Kan^r) chromosomal DNA instead of plasmid DNA. The limit of detection for transformation frequency was 10⁻⁹; error bars indicate standard deviations.

FIG. 4.

The frequency of Lp02 transformation is dependent on the amount of homologous DNA available for recombination. The figure shows transformation frequency as a function of the length of homologous DNA flanking a selectable marker. Reporter strain JV1103 was grown on solid media at 30°C in the presence of 1 μg of a DNA fragment consisting of a Kan^r marker flanked by 80, 250, 500, or 1,500 bp of L. pneumophila DNA on either side. Fragments were used as linear molecules (black bars) or were present as part of a plasmid (gray bars). After 48 h, bacteria were plated on selective media versus nonselective media in order to determine the number of kanamycin-resistant transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁹; error bars indicate standard deviations.

FIG. 5.

Kinetics of transformation in a broth assay. The figure shows transformation frequency as a function of time. Reporter strain JV1103 was grown in AYE at 30°C with low-speed shaking to early exponential phase. One-milliliter aliquots were removed, and 1 μg of reporter construct pJB957 (Kan^r) DNA was added for 0, 5, 10, 20, 40, 60, or 120 min, followed by a 15-min incubation at 37°C with DNase I at 0.1 mg/ml. Bacteria were plated on selective media in order to determine the number of kanamycin-resistant transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁸; error bars indicate standard deviations.

FIG. 6.

The frequency of Lp02 transformation is dependent on bacterial growth phase. (A) Transformation frequency as a function of patch age. Reporter strain JV1103 was grown on solid media at 30°C for 24, 48, 72, or 96 h. One microgram of reporter construct pJB957 (Kan^r) DNA was added and incubated for an additional 4 h, and then cells were plated on selective versus nonselective media in order to determine the number of kanamycin-resistant transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁹; error bars indicate standard deviations. (B) Transformation frequency and growth phase as a function of time. Reporter strain JV1103 was grown to stationary phase, back-diluted into fresh AYET medium, and cultured at 30°C with gentle shaking to stationary phase. Periodically, 1-ml aliquots were removed, the optical density was measured, and a portion of the cells were incubated for 2 h with 1 μg of reporter construct pJB957 DNA. These bacteria were then plated on selective versus nonselective media in order to determine the number of transformants in the total cell population. The optical density of the culture at a given time is indicated by open squares connected by a gray line, and the scale is shown on the right-hand y axis. The log transformation frequency of the culture at a given time is indicated by filled squares connected by a black line, and the scale is shown on the left-hand y axis. The limit of detection for transformation frequency was 10⁻⁸; error bars indicate standard deviations.

FIG. 7.

Frequency of transformation versus pH and nutrient availability. (A) Transformation frequency as a function of medium source. Repression of competence by conditioned media can be reversed by addition of nutrients and adjustment of pH. Reporter strain JV1103 was grown in AYET at 30°C with shaking to early exponential phase. One-milliliter aliquots were removed, and cells were pelleted and resuspended in 1 ml of the original medium (pH 6.8), fresh medium (pH 6.8), conditioned medium (pH 7.2), conditioned medium with the pH adjusted (pH 6.8), conditioned medium supplemented with nutrients (pH 7.2), or conditioned medium with the pH adjusted and supplemented with nutrients (pH 6.8) and incubated at 30°C with shaking for 2 h. One microgram of pJB957 (Kan^r) DNA was added, and cells were further incubated for 2 h. Bacteria were plated on selective media in order to determine the number of transformants in the total cell population. (B) Transformation frequency as a function of medium pH. Strain JV1103 was grown in AYE at 30°C with low-speed shaking to early exponential phase. One-milliliter aliquots were removed, and cells were gently pelleted and resuspended in 1 ml of AYET at pH 6.0, 6.3, 6.6, 6.9, 7.2, 7.5, 7.8, 8.1, or 9.0 or in conditioned AYET (pH 7.4) and incubated at 30°C with shaking for 2 h. One microgram of reporter construct pJB957 (Kan^r) DNA was added, and cells were further incubated for 2 h. Bacteria were plated on selective media in order to determine the number of kanamycin-resistant transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁸; error bars indicate standard deviations.

FIG. 8.

Lp02 is significantly more competent than other L. pneumophila serogroup 1 strains. (A) Relationship among the three commonly studied L. pneumophila serogroup 1 strains Lp02, JR32, and AA100. Strains Lp02 and JR32 are derivatives of the same parent strain, L. pneumophila Philadelphia-1. AA100 is a derivative of an unrelated serogroup 1 strain, L. pneumophila Wadsworth. (B) Transformation frequency (plate assay) as a function of strain origin. Four L. pneumophila serogroup 1 strains were assayed using the 30°C plate assay. Strains were grown on solid media at 30°C in the presence of 1 μg of reporter construct pJB957 (Kan^r) plasmid DNA. After 48 h, bacteria were plated on selective versus nonselective media in order to determine the number of kanamycin-resistant transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁹; error bars indicate standard deviations. (C) Transformation frequency (broth assay) as a function of strain origin. Four L. pneumophila serogroup 1 strains were assayed using the 30°C liquid assay. The four strains were inoculated into AYET broth to an OD₆₀₀ of 0.02 and cultured at 30°C with low-speed shaking (100 rpm) in the presence of 1 μg of reporter construct pJB957 (Kan^r) plasmid DNA. After 72 h, bacteria were plated on selective versus nonselective media in order to determine the number of kanamycin-resistant transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁹; error bars indicate standard deviations.

FIG. 9.

Inactivation of proQ or comR induces competence of strain AA100. (A) Transformation frequency and complementation of strain AA100 containing a proQ mutation. AA100 containing the empty vector pJB1653 (AA100 + vector), AA100 proQ::mini-Tn10kan (JV1729) containing the empty vector pJB1653 (Q⁻ + vector), JV1729 containing the AA100 proQ complementing clone pJB1659 (Q⁻ + Q_AA100), and JV1729 containing the Lp02 proQ complementing clone pJB1661 (Q⁻ + Q_Lp02) were assayed for the ability to take up pJB964 (Cm^r) DNA by the 30°C plate assay. (B) Transformation frequency and complementation of strain AA100 containing a comR mutation. AA100 containing the empty vector pJB1653 (AA100 + vector), AA100 comR::mini-Tn10kan (JV1727) containing the empty vector pJB1653 (R⁻ + vector), JV1727 containing the AA100 comR complementing clone pJB1665 (R⁻ + R_AA100), and JV1727 containing the Lp02 comR complementing clone pJB1663 (R⁻ + R_Lp02), were assayed for the ability to take up pJB964 DNA by the 30°C plate assay. (C) Transformation frequency of strain Lp02 in the presence of proQ complementing clones. Lp02 containing either the empty vector pJB1653 (Lp02 + vector), the AA100 proQ complementing clone pJB1659 (Lp02 + Q_AA100), or the Lp02 proQ complementing clone pJB1661 (Lp02 + Q_Lp02) was assayed for the ability to take up pJB964 DNA by the 30°C plate assay. (D) Transformation frequency of strain Lp02 in the presence of comR complementing clones. Lp02 containing either the empty vector pJB1653 (Lp02 + vector), the AA100 comR complementing clone pJB1665 (Lp02 + R_AA100), or the Lp02 comR complementing clone pJB1663 (Lp02 + R_Lp02) was assayed for the ability to take up pJB964. In each transformation reaction, bacteria were exposed to 1 μg of reporter plasmid DNA for 48 h and then plated on selective versus nonselective media in order to determine the number of chloramphenicol-resistant transformants in the total cell population. The limit of detection for transformation frequency was 10⁻⁹; error bars indicate standard deviations.