Pleiotropic effects of rfa-gene mutations on Escherichia coli envelope properties

Mutations in the rfa operon leading to severely truncated lipopolysaccharide (LPS) structures are associated with pleiotropic effects on bacterial cells, which in turn generates a complex phenotype termed deep-rough. Literature reports distinct behavior of these mutants in terms of susceptibility to bacteriophages and to several antibacterial substances. There is so far a critical lack of understanding of such peculiar structure-reactivity relationships mainly due to a paucity of thorough biophysical and biochemical characterizations of the surfaces of these mutants. In the current study, the biophysicochemical features of the envelopes of Escherichia coli deep-rough mutants are identified from the molecular to the single cell and population levels using a suite of complementary techniques, namely microelectrophoresis, Atomic Force Microscopy (AFM) and Isobaric Tag for Relative and Absolute Quantitation (iTRAQ) for quantitative proteomics. Electrokinetic, nanomechanical and proteomic analyses evidence enhanced mutant membrane destabilization/permeability, and differentiated abundances of outer membrane proteins involved in the susceptibility phenotypes of LPS-truncated mutants towards bacteriophages, antimicrobial peptides and hydrophobic antibiotics. In particular, inner-core LPS altered mutants exhibit the most pronounced heterogeneity in the spatial distribution of their Young modulus and stiffness, which is symptomatic of deep damages on cell envelope likely to mediate phage infection process and antibiotic action.


Mi nD
Site-determi ni ng protei n ≥20 Methyl -a ccepting chemota xis protei n II , pri mary receptor of a s pa rtate, a s pa ra gi ne a nd glutama te Ts r Methyl -a ccepting chemota xis protei n I, receptor of s eri ne s ens or and rel ated ami no a ci ds s uch as cys tei ne, L-a la nine a nd glyci ne n = 17

SI-3. Western blot analysis for iTRAQ data validation.
Cell wall proteins were extracted as described in the main text. Cytoplasmic proteins were extracted as follows: cells were broken in a French press twice at 1 kBar and centrifuged at 2500 × g for 10 min. Then, supernatant was centrifuged at 12000 × g for 20 min and diluted v/v in NP40 buffer (50 mM Tris/HCl pH 7.3, 150 mM NaCl, 1 mM EDTA, 0.1% SDS and 1% NP40). Both cell wall and cytoplasmic proteins (10 µg) were separated by SDS-PAGE, transferred to a nitrocellulose membrane and stained with Ponceau S 0.1 % (w/v), and acetic acid 5% (v/v) for verification. Membranes were probed with rabbit polyclonal antibodies specific for Flagellin (Abcam, dilution 1/15000) or for Methyl-accepting chemotaxis protein II TAR (antibodies-online GmbH, dilution1/10000) for confirmation of ITRAQ results. Membranes were also probed with mouse antibody specific for E.coli DnaK (clone 8E2/2, dilution 1/500). DnaK is a cytoplasmic protein detected only in cytoplasmic extract, which therefore helped in confirming the purity of the membrane extract. Secondary antibodies were horseradish peroxidase coupled anti-rabbit IgG (dilution 1/10000) or anti-mouse (dilution 1/2000) IgG and the signal was detected upon chemoluminescence measurement (BioRad) and use of Imager 600 (Amersham).

SI-4. Outer membrane protein extraction.
Specific outer membrane extraction was performed as described previously by Yethon et al. (2000). A given amount of cells (fixed upon measurement of the optical density at 600 nm) was harvested by centrifugation of 100 ml exponential growth cultures (OD600nm= 0.6), resuspended in 20 mM Tris buffer (pH 8), and lysed by passage through a French pressure cell. After removal of cell debris (5000 × g for 10 min), the total membrane fraction was collected by centrifugation (100.000 × g for 2 h) and then resuspended in 2% Sarkosyl. The Sarkosyl-insoluble outer membrane fraction was collected by centrifugation, washed a second time in 2% Sarkosyl, and centrifuged again. The resulting pellet was resuspended in 1 ml of 20 mM Tris buffer (pH 8), denatured in Laemmli buffer, analyzed on a SDS-PAGE 12% and Coomassie stained.
Indicated bands ( Figure S3) were manually excised from gels and were cut into cubes. Tryptic digestion and mass spectrometry analysis were performed by the proteomic platform 3P5 (Université Carbamidomethylation of cysteins was set as constant modification and oxidation of methionines was set as variable modification. Figure S3 shows the SDS-PAGE analysis of the outer membrane fractions of BW25113 (WT), JW3596, JW3601 and JW3606 (in that order from left to right) and reveals that the dramatic decrease (more than 90%) of porins OmpA, OmpF and OmpC argued by several authors (Nikaido, 1979;Nikaido and Vaara, 1985;Parker et al., 1992;Schnaitman and Klena, 1993) is not supported by our experimental data. OMP extractions and SDS-PAGE experiments were performed in triplicates and no significant difference was observed between OmpA, OmpF and OmpC (only one representative SDS-PAGE gel is presented in Fig. S3).

SI-5. Strain complementation experiments.
Complementation experiments were performed using the ASKA plasmids provided by the National BioResource Project (NIG, Japan): E. coli. Plasmids pCA24N::rfaJ, pCA24N::rfaG, pCA24N::rfaC were extracted from the ASKA clone(-) JW3601, JW3606 and JW3596 using the Macherey-Nagel NucleoSpin Plasmid kit (Fisher Scientific, France) and transformed to the respective Keio stains JW3601, JW3606 and JW3596. As a control, the empty plasmid pCA24N was transformed in the WT strain BW25113. Thermo-competent cells were prepared and transformed according to standard procedures (Green and Sambrook, 2012) and spread on LB plates containing 25 μg/ml chloramphenicol and 30 µg/ml Kanamycin (only chloramphenicol for BW25113 pCA24N). Clones were selected, verified by PCR, and cultured for further analyses in M9 broth supplemented with 1.0 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) and the appropriate antibiotics.  pCA24N. These values correspond to the lowest concentrations of SDS that inhibit the growth of strains. The cell growth conditions adopted for BW25113 and rfa-mutants are as specified in the strains and culture conditions section of the main text (M9 medium growth, kanamycin for mutants).
For complemented strains, growth conditions are as specified above (M9 medium growth, chloramphenicol, kanamycin for mutants, and IPTG). Results obtained reveal that complementation totally suppresses the SDS sensitivity of JW3606 and JW3596, which is a peculiar feature of inner core truncated LPS-mutants. These data are in good agreement with the expected results and confirm that for these two strains complementation restores the wild-type phenotype. Following the methodology detailed in the main text, Figure S4A displays surface roughness (Rsurface) of cells of the complemented strains (displaying a morphology similar to that given in Fig. 3A of the main text) denoted hereafter as JW3601 * , JW3606 * and JW3596 * carrying the plasmids pCA24N::rfaJ, pCA24N::rfaG, pCA24N::rfaC, respectively, and of the reference strain BW25113 * (WT * ) transformed with the empty plasmid pCA24N. The cell growth conditions adopted for each strain is specified above (i.e. M9 growth medium, presence of chloramphenicol, of kanamycin for mutants and IPTG). Adopting the procedure detailed in the main text (Fig. 4), Young modulus (E) and cell spring constant (kcell) were further evaluated and results are reported in Figures S4B and S4C

The number of cells examined for evaluation of cells surface roughness and nanomechanical
properties is here 10 to 15, in detail 2 to 8 cells per cell culture and 2 to 4 different cell cultures were considered. Altogether, the AFM-derived results show that the surface roughness and the nanomechanical features of JW3601 * , JW3606 * , JW3596 * and WT * are comparable, and that they significantly differ from those discussed in the main text for the WT and deep-rough mutants cells pictured in Fig. 3A. It is not unexpected that the similar cell surface properties of JW3601 * , JW3606 * , JW3596 * and WT * are different from those of the BW25113 control strain lacking the empty plasmid pCA24N, recalling that the different growth conditions adopted with and without pCA24N (presence/absence of chloramphenicol and IPTG) impact on e.g. the rate of cells division, the ductility of their envelop, and their surface elasticity. The key result here is that the surface characteristics of JW3601 * , JW3606 * , JW3596 * and WT * are of the same magnitude and that there is no clear sign of dependence on cell strain. In particular, the values of Rsurface obtained for JW3601 * , JW3606 * ,