Bacterial strains are listed in Table 1. Bacterial preinocula were grown in Luria Bertani (LB) broth as static cultures at 37°C for 7-8 hours, reaching an optical density of 1.5-2 at 600 nm. Cultures were then diluted 1:100 into 40 ml of LB broth and subsequently introduced into two 10 ml High-Aspect-Ratio Vessels (HARV, Synthecon, Inc. Houston, Tex), a special kind of rotating bioreactors designed at the Johnson Space Center (Houston, Tex.) [19]. Care was taken to ensure that the bioreactors were completely filled with culture medium, avoiding any bubbles, to create a low-shear environment. The bioreactors were oriented to grow cells under conditions of modeled microgravity (its axis of rotation is perpendicular to gravity) or normal gravity (its axis of rotation is parallel to gravity vector). All incubations in the HARV were done at 37°C with a rotation rate of 25 rpm for 20 hours. Gas exchange occurs by perfusion through a permeable membrane and is sufficient to maintain an aerobic environment. Cell growth was monitored spectrophotometrically at 600 nm.

Mutations were constructed in S. Typhimurium ATCC14028 using the method of Datsenko and Wanner [20]. The coding sequences of katG, katE, oxyR and rpoS genes were replaced with the chloramphenicol resistance cassette amplified with the primers listed in Table 2, using the pKD3 plasmid [20] as a template for PCR amplification. The coding sequences of katN and rpoE genes were replaced with the kanamycin resistance cassette amplified with the primers listed in Table 2, using the pKD4 plasmid [20] as a template for PCR amplification. The constructs were verified by PCR analysis. Mutations were transferred into a clean S. Typhimurium ATCC14028 background by P22 transduction. The new strains are listed in Table 1. The double mutants katEkatN and katGkatN were generated by transduction of katN::kan allele from strain PF134 into PF126 (katE::cat) and PF114 (katG::cat), obtaining strains PF139 and PF132, respectively. To construct the double mutant katEkatG, the resistance cassette was eliminated from strain PF114 (katG::cat) by homologous recombination between the two flanking FRT sites using plasmid pCP20 [20, 21], thereby obtaining strain PF140 (katG::scar). Subsequently, the katE::cat was introduced in the PF140 by P22 transduction, to obtain strain PF141 (katE::cat, katG::scar).

Epitope tagging of KatE, KatG, KatN and SodA was obtained by adding a 3xFLAG at the 3’-terminus of the genes, following a described procedure [22]. DNA fragments were amplified on pSUB11 with primers oPF10/11 (for katG::3xFLAG), oPF14/15 (for katN::3xFLAG), oPF18/19 (for katE::3xFLAG) and oli105/106 (for sodA::3xFLAG) listed in Table 2. DNA was electroporated in S. Typhimurium 14028 containing plasmid pKD46 [20]. Transformants were selected and recombination was confirmed by PCR. Subsequently, the 3xFLAG mutations were transferred into a clean 14028 background by P22 transduction. To have an internal standard for western blot analysis, a 3xFLAG epitope-tagged chloramphenicol acetyl transferase gene (cat) constitutively expressed under control of the tac promoter, was transduced from strain MA7223 [23], obtaining the doubly tagged strains listed in Table 1.

Strains carrying the epitope-tagged genes were grown in the bioreactors as described above. After 20 hours of growth, 10⁸ cells were removed immediately and harvested by centrifugation for the analysis of protein accumulation by Western blotting. The pellets were boiled for 10 min in Laemmli lysis buffer and the proteins were separated by 12% SDS-PAGE and blotted onto nitrocellulose membrane (Hybond ECL, Amersham). The epitope flagged proteins were revealed by the use of an anti-FLAG M2 monoclonal antibody (Sigma-Aldrich) as primary antibody, and an anti-mouse HRP-conjugated IgG (Sigma-Aldrich) as secondary antibody. Detection was performed by enhanced chemiluminescence (ECL Advance, Amersham). Densitometric analysis of the western blots was carried out with the Gel-Pro Analyzer software (Media Cybernetics) to evaluate differences in protein accumulation in NG and LSMMG.

Bacteria grown under LSMMG or NG were evaluated for their ability to survive to hydrogen peroxide stress. After 20 hours of growth under conditions described above, a period sufficient to reach the stationary phase, bacterial cultures were immediately (within 5 min) diluted to 10⁷ cells/ml in phosphate buffered saline (PBS) and incubated statically at room temperature with 17.4 mM hydrogen peroxide, accordingly to a previous investigation [6]. Aliquots were removed after 30 minutes and the number of viable cells was determined by serial dilution and plating onto LB agar. The percent survival following hydrogen peroxide exposure was calculated for each strain by dividing the number of CFU (Colony Forming Units) obtained upon incubation in hydrogen peroxide by the number of CFU obtained upon incubation in PBS alone. Each assay was repeated at least three times, and standard deviations were calculated. A time course of resistance to 17.4 mM hydrogen peroxide of LSMMG and NG grown cells was carried out under the same conditions, by removing bacterial aliquots after 15, 30 and 45 minutes.

To determine susceptibility to acid stress, bacteria grown in LSMMG and NG conditions for 20 hours were diluted to 2x10⁷ cells/ml in LB medium and subsequently to 10⁷ cells/ml by adding a citrate buffer adjusted to pH 3.0. and incubated statically for 1 hour at room temperature. Aliquots were removed after 1 hour and the number of viable cells was determined by serial dilution and plating onto LB agar. The percent survival following acid exposure was calculated for each strain by dividing the number of CFU obtained upon incubation in acid citrate buffer (pH 3.0.) by the number of CFU obtained upon incubation in citrate buffer (pH 7.12). Each assay was repeated at least three times, and standard deviations were calculated.

Bacteria grown in LSMMG and NG conditions were immediately pelleted at 4°C, resuspended in a phosphate buffer (5 mM potassium phosphate pH 7,5 mM disodium EDTA and 10% glycerol) and lysed by sonication. To avoid inactivation of heat-labile catalases during sonication, cells were given four rounds of 15-s pulses with a Branson B-12 sonicator and a microtip at a power setting of 6. Tubes were kept on ice between rounds. Debris was pelleted by centrifugation at 4°C for 20 minutes at 17000 x g. Total protein concentration was determined by the method of Lowry et al. [24].

Catalases present in crude extracts were separated by electrophoresis through a 9% nondenaturing polyacrylamide gel and visualized according to the method of Woodbury et al. [25]. Briefly, the gels are washed in distilled water and then incubated with 0.003% hydrogen peroxide for 10 minutes. The gels are next briefly rinsed with distilled water and subsequently placed in a solution containing 1% ferric chloride and 1% potassium ferricyanide (III) for 5 minutes. Achromatic zones against the dark background are due to the activity of catalase, which prevent the H₂O₂-dependent reduction of ferricyanide to ferrocyanide. Densitometric analyses of gels were carried out to obtain a semiquantitative evaluation of the activity of the different catalase in NG and LSMMG. Measurements of total catalase activity in cell extracts were carried according to described procedures [26]. The peroxidase activity of KatG was visualized on native gel by a previously described procedure [27].

To understand the effects of microgravity on Salmonella enterica, we have tested the susceptibility to hydrogen peroxide of Salmonella enterica serovar Typhimurium ATCC-14028 cultures grown in low shear modeled microgravity (LSMMG) and in normal gravity (NG) conditions. Bacteria were grown for 20 hours in HARV bioreactors rotating on a horizontal axis to obtain a LSMMG environment or on a vertical axis to grow bacteria under NG, accordingly to established procedures [28]. The harvested cultures were immediately diluted in PBS and challenged with the addition of 17.4 mM hydrogen peroxide for 30 min at room temperature. Fig. (1) shows that bacteria grown under LSMMG are significantly more resistant to hydrogen peroxide than bacteria grown under NG. Similar results were obtained also with bacteria cultivated for 10 hours in bioreactors. Even though under these conditions the difference in hydrogen peroxide-resistance of bacteria cultivated in LSMMG and NG was less striking, bacteria grown under reduced gravitational force still exhibited higher resistance to H₂O₂ (data not shown). The inset of Fig. (1) shows that up to 30 min incubation with hydrogen peroxide (which corresponds to less than 35 min from the moment of extraction of bacteria from the bioreactors) there is a sharp difference in bacterial survival between bacteria grown in LSMMG and NG. At 45 min of incubation with hydrogen peroxide bacteria grown in LSMMG show a significant increase in mortality, which may be explained by the transient nature of the phenotypic changes induced by microgravity [6].

Fig. (1). Resistance of different S. enterica strains to hydrogen peroxide. Results are expressed as percent survival after 30 minutes incu-bation with 17.4 mM hydrogen peroxide at room temperature. The S. Typhimurium LT2 strain was incubated with 8.7 mM hydrogen perox-ide, due to its higher susceptibility to this agent. The different serovars were grown in both LSMMG (solid bars) and NG (open bars) condi-tions. The percent survival was calculated for each strain by dividing the number of CFU/ml obtained after incubation in the presence of hydrogen peroxide by the number of CFU/ml obtained after incubation in PBS alone. The results are the average of three trials and represent mean ± standard errors. The inset shows a time course of S. Typhimurium ATCC14028 survival to 17.4 mM hydrogen peroxide, grown in LSMMG (diamond) and NG (square) conditions. The results are the average of three samples and represent mean ± standard errors.

Fig. (2). Resistance of different S. enterica strains to acid stress. Results are expressed as percent survival after 1 hour incubation with citrate buffer (pH 3.0.) at room temperature. The different serovars are grown under conditions of LSMMG (solid bars) and NG (open bars). The percent survival was calculated for each strain by dividing the number of CFU/ml obtained after incubation in citrate buffer (pH 3.0.) by the number of CFU/ml obtained after incubation in citrate buffer at pH 7.12. The results are the average of three independent experiments and represent mean ± standard errors.

Fig. (3). Visualization of catalase activity on polyacrylamide gels. The activity bands due to the three different Salmonella catalases were detected in native 9% polyacrylamide gels using the method of Woodbury et al. (1971). 50 µg of proteins from total cellular extracts were loaded in each lane: 1 and 2, ATCC 14028; 3 and 4, PF132 (katG::cat katN::kan); 5 and 6, PF139 (katE::cat katN::kan); 7 and 8, PF141 (katE::cat katG::scar); 9 and 10, DT104; 11 and 12, LT2; 13 and 14, S. Choleraesuis A50; 15 and 16, S. Enteritidis LK5. The odd lanes contain extracts from LSMMG grown bacteria. In the even lanes extracts from NG grown bacteria were loaded.

Fig. (4). KatG accumulation in S. Typhimurium ATCC14028. A) Strain PF114 was grown in both LSMMG and NG conditions for 20 hours and then 108 cells were immediately removed and treated as described in Materials and Methods for protein immunodetection. Cat (Chloramphenicol acetyl transferase) accumulation is shown as an internal standard. B) densitometric analysis of the Western blot presented in panel A; relative OD were calculated as Cat/KatG ratios.

Fig. (5). Hydrogen peroxide susceptibility of wild type and mutant S. Typhimurium ATCC14028 strains lacking the kat genes.. Results are expressed as percent survival after 30 minutes incubation with 17.4 mM hydrogen peroxide at room temperature. The different strains [wt; PF139 (katE::cat katN::kan); PF141 (katE::cat katG::scar); PF132 (katG::cat katN::kan)] were grown in both LSMMG (solid bars) and NG (open bars) conditions. The percent survival was calculated for each strain by dividing the number of CFU/ml obtained after incubation with hydrogen peroxide by the number of CFU/ml obtained after incubation in PBS alone. The results are the average of three experiments and represent mean ± standard errors.

Fig. (6). Susceptibility of wild type and mutant S. Typhimurium ATCC14028 strains deleted of the global gene regulatores rpoS, rpoE, oxyR and hfq to hydrogen peroxide and acid stress. Strains PF148(rpoS), PF137 (rpoE), PF133 (oxyR) and PF125 (hfq) were grown in both LSMMG (solid bars) and NG (open bars) conditions and then exposed to hydrogen peroxide (upper panel) or to acid stress (lower panel). Wild type and PF137 strains were incubated with 17.4 mM hydrogen peroxide whereas the PF148, PF133 and PF125 strains were incubated with 5 mM hydrogen peroxide. Results are expressed as percent survival after 30 minutes incubation with hydrogen peroxide at room tem-perature (upper panel) and after 1 hour incubation with citrate buffer at pH 3.0. (lower panel). The percent survival was calculated as de-scribed in the legends to figure 1 and 2. The results are the average of three independent experiments and represent mean ± standard errors.

To evaluate the general significance of these results, we analyzed the susceptibility of some other Salmonella strains to hydrogen peroxide and acid stress after growth in LSMMG and NG. We tested a virulent DT104 S. Typhimurium strain from our laboratory collection [29], S. Typhimurium LT2, an attenuated strain characterized by low levels of RpoS production [30, 31], S. enterica serovar Choleraeasuis A50 [32] and S. enterica serovar Enteritidis LK5 (lab collection).

Fig. (1) shows that these strains exhibited differential susceptibility to hydrogen peroxide, S. Typhimurium DT104 showing the highest resistance. S. Typhimurium LT2 proved to be particularly sensitive to this agent and, therefore, the data concerning this strain reported in Fig. (1) were obtained by incubating bacteria with a lower concentration of hydrogen peroxide. Interestingly, cultivation under LSMMG enhanced resistance to hydrogen peroxide of all these strains, suggesting that this is a common feature of most Salmonella strains.

As previous studies carried out with S. Typhimurium SL1344 have indicated that LSMMG enhances Salmonella resistance to acid stress [2], we have also tested the acid stress resistance of all the Salmonella strains employed in this work. To this aim, we grew bacteria in HARV bioreactors in both LSMMG and NG for 20 hours and then exposed the cells to pH 3.0 for 1 hour at room temperature. In agreement with previous studies, all the strains examined displayed a marked increase in survival at pH 3.0 when grown in LSMMG condition (Fig. 2). The only exception is represented by the S. Typhimurium LT2 strain, which did not show a significant difference (according to a Student's t-test analysis) in survival between cells cultivated in LSMMG and NG.

In consideration of the higher resistance to hydrogen peroxide of Salmonella strains grown in LSMMG, we looked for alterations in the activity of bacterial enzymes involved in H₂O₂ detoxification. Three different catalases, KatG and the RpoS-regulated catalases KatE and KatN, have been identified in Salmonella [33]. KatE and KatG are heme-containing catalases, whereas KatN is likely a manganese-containing enzyme. The activity of these proteins in bacterial extracts was assayed in non-denaturing 9% polyacrylamide gels stained for catalase activity (Fig. 3). Bacterial extracts were prepared following a previously described method that prevents KatG and KatN inactivation [26]. In fact, high temperatures and prolonged sonication readily inactivate KatG [26] and we have found that similar conditions disrupt KatN activity (data not shown). Moreover, we have also observed that this catalase is inactivated by extraction procedures involving the use of a detergent such as deoxycholic acid. Fig. (3) shows that KatN activity is increased in the extracts from S. Typhimurium ATCC14028 cells cultivated in LSMMG conditions (see lanes 1 and 2). Such an increase in KatN activity is clearly appreciable also in the isogenic mutant strain PF141, which is deleted of the katE and katG genes (Fig. 3, lanes 7 and 8). A densitometric analysis of the gel indicates that KatN activity in S. Typhimurium is 5-10 folds higher in bacteria cultivated in LSMMG. Moreover, Fig. (3) also shows a decrease (close to the 50% according to the densitometric analysis) in the intensity of the two KatG bands in extracts from bacteria cultivated in LSMMG, compared to extracts from bacteria grown in NG. A comparable modification in KatG activity can be observed also in the mutant strain PF139, which lacks the katE and katN genes (lanes 5 and 6). As KatG is a bifunctional catalase-peroxidase [34] we also stained gels for peroxidase activity, confirming the decrease in KatG activity in LSMMG (data not shown). The microgravity-induced changes in KatG and KatN activity were confirmed by quantitative spectrophotometric assays carried out on soluble protein extracts on S. Typhimurium ATCC 14028 strains (Table 3). Although the absolute values of catalase activity showed significant variations from experiment to experiment, we were never able to appreciate significant changes in total catalase activity in cell extracts from bacteria grown in LSMMG and NG deriving from the same preinoculum. Similarly, we did not observe differences in KatE activity between the mutant strain PF132 grown in LSMMG and NG conditions. In contrast, in good agreement with the densitometric analysis, the LSMMG environment induced a 2-fold decrease in the KatG activity of the mutant strain PF139. We were unable to measure the activity of KatN in NG condition, possibly because it is below the lower limit of the assay (0.05 units) [35]. These observations suggest that the increase in KatN activity largely compensates for the reduction of KatG activity observed in extracts from LSMMG-cultivated bacteria.

A comparable variation in the catalase activity pattern between bacteria grown in LSMMG and NG was observed in extracts from S. Typhimurium DT104 (lanes 9 and 10) and S. Enteritidis LK5 (lanes 15 and 16). In contrast, the KatN activity band was not observed in extracts from S. Typhimurium LT2 (lanes 11 and 12). This observation, together with an apparent reduction of KatE activity, is likely related to the mutation in the rpoS allele that characterizes this strain [36]. Nevertheless, also in S. Typhimurium LT2 KatG activity is higher in bacteria grown in NG than in LSMMG. In S. Choleraesuis, where KatG is expressed at much higher levels than in the other strains and the activity band corresponding to KatN is hardly visible (lanes 13 and 14), we were not able to identify alterations in the activity of catalases between bacteria grown in the two conditions. It should be noted that this strain is less resistant to hydrogen peroxide stress than S. Typhimurium DT104 and S. Typhimurium ATCC 14028, despite its catalase activity is quite high (25,7 U/mg in LSMMG and 26,2 U/mg in NG). This observation indicates that additional factors are responsible for the intrinsic susceptibility of the different strains to hydrogen peroxide.

To analyze the possibility that changes in catalase activity bands reflect modifications in the intracellular concentration of proteins, we inserted a sequence encoding the 3xFLAG epitope at the 3’-terminus of katN, katG and katE genes and of a few other genes involved in oxidative stress response (dps, sodCI, sodCII, sodA) producing, respectively the ferritin-like protein Dps (DNA-binding proteins from starved cells), Cu,Zn superoxide dismutases I and II (SodCI and SodCII) and Mn superoxide dismutase (SodA). We observed a repeatable 15-20% reduction in the accumulation of KatG in LSMMG (Fig. 4), which partially accounts for the above described reduction in enzyme activity. In contrast, we did not observe appreciable variations in the intracellular concentration of Dps, SodCI, SodCII, SodA, KatE and KatN in bacteria cultivated in LSMMG and NG conditions (data not shown).

To evaluate the effective contribution of genes coding for catalases to the enhanced hydrogen peroxide resistance of Salmonella strains grown in LSMMG, we constructed mutant strains of S. Typhimurium ATCC14028 lacking the single katG, katE or katN genes and the double mutants katEkatG, katGkatN and katEkatN. All these mutants were grown either in LSMMG or in NG conditions and the harvested bacteria were incubated with hydrogen peroxide. The single katE, katG and katN mutants were more resistant to killing by hydrogen peroxide when grown in LSMMG, although their survival was slightly lower than that of the wild type strain (data not shown). A similar behaviour was observed for the double katEkatN (PF139) and katEkatG mutants (PF141) (Fig. 5). In contrast, the resistance to hydrogen peroxide of the katGkatN (PF132) mutant strain was identical in bacteria cultivated in LSMMG or NG conditions. This result demonstrates that both KatG and KatN concur to the higher hydrogen peroxide resistance of the Salmonella strains grown in LSMMG. Interestingly, the katGkatN (PF132) mutant strain still exhibited higher resistance to acid stress when cultivated in LSMMG than in NG conditions (data not shown). This observation indicates that KatG and KatN are not involved in the enhanced resistance to low pH induced by microgravity.

Recent studies have suggested that the global transcriptional regulators Hfq could play a major role in the alterations in gene expression induced by LSMMG or by spaceflights in S. Typhimurium [9] as well as in P. aeruginosa [37]. Therefore, we have analyzed the effect on catalases expression of mutations in hfq and other global regulators: rpoE, which modulates Salmonella virulence and resistance to oxidative stress through the regulation of several genes involved in the envelope stress response [38], oxyR, a global regulator of genes involved in bacterial response to hydrogen peroxide [39] and rpoS, the master regulator of a variety of stationary-phase genes, required for Salmonella virulence in mice [40-42].

As expected, Salmonella strains unable to produce Hfq, OxyR, RpoE or RpoS were much more susceptible to hydrogen peroxide than the wild type strain (Fig. 6). In particular, the hfq,oxyR and rpoS mutants were almost completely killed upon treatments with 17.4 mM hydrogen peroxide (data not shown). Therefore, the survival data reported in Fig. (6) for these three mutant strains were obtained upon incubation with a lower amount of hydrogen peroxide (5 mM). Independently of the higher susceptibility to oxidative stress, all these mutant strains proved to be more resistant to hydrogen peroxide when cultivated in LSMMG than in NG. This observation indicates that such global regulators are not responsible of the increased resistance to peroxide stress observed in Salmonella strains grown in microgravity conditions. Finally, we assayed the susceptibility of these mutants to pH acid, after a growth in NG and in LSMMG condition (Fig. 6). Interestingly, deletion of rpoE reduces the LSMMG-induced resistance to acid pH.