Motility characteristics are altered for Rickettsia bellii transformed to overexpress a heterologous rickA gene

Abstract

The rickettsial protein RickA activates host cell factors associated with the eukaryotic actin cytoskeleton and is likely involved with rickettsial host cell binding and infection and the actin-based motility of spotted fever group rickettsiae. The rickA gene sequence and protein vary substantially between Rickettsia species, as do observed motility-associated phenotypes. To help elucidate the function of RickA and determine the effects of species-specific RickA variations, we compared extracellular binding, intracellular motility, and intercellular spread phenotypes of three Rickettsia bellii variants. These included two shuttle vector-transformed R. bellii strains and the wild-type isolate from which they were derived, R. bellii RML 369C. Both plasmid shuttle vectors carried spectinomycin resistance and a GFPuv reporter; one contained Rickettsia monacensis-derived rickA, and the other lacked the rickA gene. Rickettsia bellii transformed to express R. monacensis rickA highly overexpressed this transcript in comparison to its native rickA. These rickettsiae also moved at higher velocities and followed a more curved path than the negative-control transformants. A lower proportion of R. monacensis rickA-expressing bacteria ever became motile, however, and they formed smaller plaques.

FIG 1

Map of the pRAM18dSGA[RmRickA] shuttle vector. The shuttle vector pRAM18dSGA[RmRIckA] was constructed using the pGEM vector backbone (Promega, Madison, WI), which is selectable by ampicillin (Amp) and thus enables production of pRAM18dSGA[RmRickA] shuttle vectors in Escherichia coli. It includes the region containing the parA and dnaA genes of the R. amblyommii plasmid pRAM18 and the intervening sequence (23), and it carries the genes for spectinomycin and streptomycin resistance (labeled Spec) through insertion of the aadA gene and a GFPuv reporter, both driven by ompA promoters. The rickA gene from R. monacensis, along with the region 5′ of the gene, were PCR amplified and inserted into the multiple-cloning site at the Nhel restriction site.

FIG 2

R. bellii transformants undergoing actin-based motility. Time-lapse micrographs in panels A and D to F were compiled as t-projections. (A) R. bellii_MCS at 34°C. (B) Dividing rickettsiae may present actin tails on both joined bacteria, resulting in spinning. Measurements from dividing bacteria were not included in mean values of velocity or path curvature. Dividing rickettsiae are indicated by white arrowheads. (C) Twinned tails on single bacteria were occasionally visible, in this case on R. bellii_MCS at 34°C. The red channel gamma value of this image was adjusted to accentuate the brightness of the actin tails against the background. (D) The R. bellii_RmRickA mean velocity at 23°C is just over half that of the rickettsiae pictured in panel A. (E) Green channel fluorescent 10-movement t-stack of an R. bellii_MCS variant at 34°C. (F) The t-stack pictured in panel E was measured to determine the velocity and path curvature. The path length (yellow line) was 43.7 μm and the elapsed time of the 10-movement stack was 118.25 s, yielding a velocity of 0.37 μm/s. The absolute distance traveled over the course of the t-stack was 37.2 μm, as indicated by the red line. Division of the absolute distance traveled (37.2 μm) by the path length (43.7 μm) yielded a path curvature value of 0.8513, which is very close to the mean value for R. bellii_MCS at 34°C. Bars, 10 μm.

FIG 3

Relative copy number and expression of native and plasmid-associated rickA in transformed and wild-type R. bellii. Plasmid rickA was derived from R. monacensis, which allowed for differentiation of plasmid sequence from the native sequence. (A) Numbers of genetic copies of each type of rickA, normalized to the number of gltA copies. (B) Expression of native and plasmid rickA relative to gltA expression. Plasmid-derived rickA was only present above the threshold of detection in R. bellii_RmRickA. Relative gene copy number and expression of native rickA were comparable between all genotypes of R. bellii. Error bars represent standard errors.

FIG 4

Change in the proportion of bound and motile transformant rickettsiae over time. Asterisks indicate significant differences between the transformants at the time points marked. (A) At 2 h p.i. the greatest difference between the binding of transformants was observed, as three times more R. bellii_RmRickA adhered to the host cells than R. bellii_MCS. Rickettsiae were inoculated at a rate of 200 bacteria per Vero cell. (B) A significant difference in the proportion of motile bacteria was also present at 2 h p.i., corresponding to the disparity in binding observed at that time point. The proportion of motile rickettsiae increased through time until, beginning at 16 h p.i., the proportion of motile R. bellii_MCS was substantially higher than the proportion of motile R. bellii_RmRickA. Error bars show standard errors.

FIG 5

Comparison of mean velocity and mean path curvature values of the rickettsial transformants. Asterisks indicate significant differences between transformants for the same temperature and phenotype observed. (A) Comparisons of mean velocities of R. bellii transformants. Results showed a substantial increase in the mean velocity of both groups at 34°C. (B) Mean path curvature values of transformants. As a path curvature value approaches 1, the straighter the path the bacteria have followed. At 23°C, the mean path of R. bellii_RmRickA was marginally, though significantly, more curved than that of R. bellii_MCS. At 34°C, however, the path curvature value was much lower in R. bellii_RmRickA, corresponding to more curved and convoluted paths. The exacerbation of phenotypic differences at 34°C relative to 23°C indicates the temperature dependence of the response. Error bars show standard errors.

FIG 6

Representative plaques produced by transformant and wild-type R. bellii. The area of infection formed by the WT is visible as a region with less stain uptake into cells, rather than as a plaque. Because plaques were visualized with the vital stain neutral red, dead cells were unable to endocytose stain. The lighter region associated with the WT demonstrated that the infection did not kill the host cells within the 10-day incubation period, though the lighter color may indicate reduced cell viability.