Figure 1.
Domain formation in B. burgdorferi as determined by immunogold TEM analysis requires raft-supporting sterols.
A. Representative negative-stain TEM images of B. burgdorferi substituted with the indicated sterols and probed for sterol glycolipids using a rabbit antibody to asialo-GM1 followed by a secondary anti-rabbit antibody conjugated to 6 nm colloidal gold. Micrographs show electron dense regions, which are portions of B. burgdorferi and show associated gold particles. Two images (of about 400 nm long segments) are shown for each sterol. Top row, sterols strongly supporting ordered domain formation; middle row, sterols with an intermediate ability to form ordered domains; bottom row, sterols that inhibit ordered domain formation [22], [26]–[28]. Boxes highlight sterol glycolipid clusters. Bars = 100 nm. TEM micrographs for additional sterols are shown in Fig. S3 in Text S1. B. Pooled “K-function” analysis of TEM experiments. The spatial distribution of gold particles is presented as curves representing the mean values of L(r)-r from images of three different bacteria from three independent sterol substitution experiments. Values of L(r)-r above the CI (95% confidence level, dashed line) indicate clustering (i.e. domain formation) of the sterol glycolipid at that specific length scale.
Figure 2.
FRET detection of ordered domain formation as a function of temperature in B. burgdorferi.
A. Demonstration of FRET assay performance in model membranes. Samples of small unilamellar vesicles contained 100 µM lipid and (in Fo samples) TMADPH or (in F samples) both TMADPH and ODRB. Vesicles were composed of 2∶1 (mol∶mol) POPC/chol (open circles), 2∶1 DOPC/chol (open triangles), or 1∶1∶1 DPPC/DOPC/chol (filled circles). F/Fo is the fraction of donor fluorescence unquenched by FRET. Average of duplicates and range are shown. B. Detection of ordered domain formation in B. burgdorferi by FRET. F and Fo samples contained B. burgdorferi (4×108 cells/ml) with TMADPH (Fo samples) or both TMADPH and ODRB (F samples). Symbols: untreated cells (diamonds), cholesterol depleted cells (bold plus sign), or cells substituted with cholesterol (filled circles), dihydrocholesterol (open triangles), ergosterol (inverted, vertex down, triangles), lanosterol (plus sign), zymosterol (squares), coprostanol (filled triangles) or androstenol (open circles). Mean F/Fo values from four samples, or two in the case of untreated cells and lanosterol, are shown. For clarity, error bars are omitted in B. (Summary FRET data with error bars is shown in Figure S5 in Text S1.)
Figure 3.
Sterol-dependence of ordered domain formation in B. burgdorferi as determined by detergent (TX-100) resistance.
Levels of cholesterol glycolipids in: A. DRM fractions, or B. soluble (SOL) fractions based on ELISA assay. Mean values and standard error of the mean from three experiments are shown. One-way ANOVA, *** P<0.001, ** P<0.01.
Figure 4.
Only lipids associating with ordered domains colocalize with B. burgdorferi sterol glycolipid domains.
A. Representative negative-stain TEM images showing localization of biotin-PEG-DPPE (left) and biotin-PEG-DOPE (right) as detected by anti-biotin and second antibody conjugated to 15 nm gold particles, in relation to cholesterol glycolipids, detected by antibody conjugated to 6 nm gold particles as in Figure 1. Electron dense areas in micrographs are sections of bacteria. Boxes denote clusters of sterol glycolipids. Notice the difference in large gold particle proximity to small gold particle clusters for biotin-PEG-DOPE and biotin-PEG-DPPE. Bars = 100 nm. B. Quantification of biotinylated lipid co-localization with cholesterol glycolipids. The co-localization parameter C40 (see Materials and Methods) indicates the percent of 6 nm gold particles within 40 nm of a 15 nm gold particle. C40 was calculated for both biotin-PEG-DOPE and biotin-PEG-DPPE samples from the mean of three different images. C. Representative immunoblots of B. burgdorferi fractions following TX-100 treatment and density gradient separation probed with anti-biotin antibodies.
Figure 5.
Effect of sterol substitution on the morphology of B. burgdorferi.
Representative negative-stain TEM images of untreated B. burgdorferi or B. burgdorferi incubated 5 h at 33°C after substitution with the indicated sterols. B = Attached membrane vesicles (“blebs”). F = .flagella. Bars = 500 nm.
Figure 6.
Sterols that differ in raft-forming ability have differential effects on B. burgdorferi membrane permeability.
A. Propidium iodide staining of B. burgdorferi substituted with the indicated sterols. The x axis shows incubation time at 33°C after sterol substitution. The mean of three experiments each of which had three samples is shown. Standard error of the mean values (not shown to enhance figure clarity) were typically ±1 fluorescence units and did not exceed ±2 fluorescence units for any data point. B. Representative Coomassie-stained SDS-PAGE gels of supernatants from B. burgdorferi 5 h after substitution with the indicated sterols. Approximate M.W. shown at left of gels. C. Representative immunoblots showing the release of the cytosolic chaperone DnaK and periplasmic flagella subunit FlaB into the supernatants 5 h after substitution with the indicated sterols. Key; depl = sterol-depleted, ergo = ergosterol, chol = cholesterol, lano = lanosterol, zymo = zymosterol, copr = coprostanol, andr = androstenol, chol form = cholesterol formate, stig = stigmasterol, dihy = dihydrocholesterol, desm = desmosterol.