Ed by Triton X-114 fractionation, but a significant 
amount of LipL32 found in protoplasmic cylinder fraction byanti-FlaA2 serum was utilized to assess permeabilization efficiency, demonstrating that while methanol appears to be the most effective permeabilization agent, the three other methods also resulted in OM disruption (Fig. 4).LipL32 is associated with the leptospiral membraneMembrane affinity analysis was performed to determine whether LipL32 is associated with the lipid bilayer. Treatment of bacterial cells with lysozyme and several freeze-thaw cycles, followed by centrifugation 25033180 separates proteins into soluble (cytoplasmic and periplasmic) and pellet (total membrane) fractions [28]. The membrane fraction was treated with high pH (0.1 M Na2CO3), high salt (0.6 M NaCl), 1326631 or urea (1.6 M), to release peripheral membrane proteins not anchored in the lipid bilayer [21,26,29?1]. Immunoblot analysis of the soluble (supernatants) and Pentagastrin custom synthesis insoluble (pelleted) membrane fractions revealed that the bulk of LipL32 remained associated with the membrane fraction after all treatments (Fig. 5). Integral outer membrane protein OmpL1, and two OM-lipoproteins; LipL46, and LipL41 were included as positive controls and could not be released from the membrane by any treatment (Fig. 5;[26,30]). As a positive control for release from the membrane, the effect of treatments on the peripheral membrane protein, P31LipL45, also known as Qlp42 [32] was also assessed. Substantial release from the membrane by urea and Na2CO3 was observed (data not shown), as previously described [21,30].LipL32 Is a Subsurface Lipoprotein of LeptospiraFigure 6. Reused from: PLoS One. 2011; 6(7): PD 168393 e21962. Confocal microscopy was performed with live L. interrogans using antisera specific for LIC10258, LIC12880, LIC12238, LipL32 (surface-exposed lipoprotein) and GroEL (protoplasmic cylinder marker). FITC-conjugated secondary antibodies were used to detect the surface-bound antibodies (B). Leptospires were identified by propidium iodide (A) staining of the DNA. Co-localization is shown in the merged images (C). doi:10.1371/journal.pone.0051025.gmembrane vesicle fractionation [12], most likely due incomplete separation of outer membrane from inner membrane vesicles rather than inner membrane localization. Our results showing a subsurface location for LipL32 appear to contradict previous studies. This prompted us to reexamine the evidence for LipL32 surface localization presented in previous studies. Immunoelectron microscopy of intact leptospires was presented as evidence for LipL32 surface-exposure [18]. However, given the abundance of LipL32, significantly more immunogold staining should have occurred than what was observed. For example, immunoelectron microscopy of Borrelia burgdorferi using OspC antibodies results in dense staining of the surface of the organism with gold particles [37]. When surface immunofluorescence was performed with rabbit serum recognizing LipL32 [18], much weaker and irregular antibody labeling was obtained in intact cells when compared to permeabilized cells. One possible explanation is that this labeling resulted from damaged organisms presented in that particular microscopic field. When LipL32 was used as a positive control in previously published IFA experiments [19,38], LipL32 surface-exposure was inconclusive as only one of two cells was labeled by antibodies in one study (Fig. 6) [19], while only one cell per microscopic field was shown in the 
other study.Ed by Triton X-114 fractionation, but a significant amount of LipL32 found in protoplasmic cylinder fraction byanti-FlaA2 serum was utilized to assess permeabilization efficiency, demonstrating that while methanol appears to be the most effective permeabilization agent, the three other methods also resulted in OM disruption (Fig. 4).LipL32 is associated with the leptospiral membraneMembrane affinity analysis was performed to determine whether LipL32 is associated with the lipid bilayer. Treatment of bacterial cells with lysozyme and several freeze-thaw cycles, followed by centrifugation 25033180 separates proteins into soluble (cytoplasmic and periplasmic) and pellet (total membrane) fractions [28]. The membrane fraction was treated with high pH (0.1 M Na2CO3), high salt (0.6 M NaCl), 1326631 or urea (1.6 M), to release peripheral membrane proteins not anchored in the lipid bilayer [21,26,29?1]. Immunoblot analysis of the soluble (supernatants) and insoluble (pelleted) membrane fractions revealed that the bulk of LipL32 remained associated with the membrane fraction after all treatments (Fig. 5). Integral outer membrane protein OmpL1, and two OM-lipoproteins; LipL46, and LipL41 were included as positive controls and could not be released from the membrane by any treatment (Fig. 5;[26,30]). As a positive control for release from the membrane, the effect of treatments on the peripheral membrane protein, P31LipL45, also known as Qlp42 [32] was also assessed. Substantial release from the membrane by urea and Na2CO3 was observed (data not shown), as previously described [21,30].LipL32 Is a Subsurface Lipoprotein of LeptospiraFigure 6. Reused from: PLoS One. 2011; 6(7): e21962. Confocal microscopy was performed with live L. interrogans using antisera specific for LIC10258, LIC12880, LIC12238, LipL32 (surface-exposed lipoprotein) and GroEL (protoplasmic cylinder marker). FITC-conjugated secondary antibodies were used to detect the surface-bound antibodies (B). Leptospires were identified by propidium iodide (A) staining of the DNA. Co-localization is shown in the merged images (C). doi:10.1371/journal.pone.0051025.gmembrane vesicle fractionation [12], most likely due incomplete separation of outer membrane from inner membrane vesicles rather than inner membrane localization. Our results showing a subsurface location for LipL32 appear to contradict previous studies. This prompted us to reexamine the evidence for LipL32 surface localization presented in previous studies. Immunoelectron microscopy of intact leptospires was presented as evidence for LipL32 surface-exposure [18]. However, given the abundance of LipL32, significantly more immunogold staining should have occurred than what was observed. For example, immunoelectron microscopy of Borrelia burgdorferi using OspC antibodies results in dense staining of the surface of the organism with gold particles [37]. When surface immunofluorescence was performed with rabbit serum recognizing LipL32 [18], much weaker and irregular antibody labeling was obtained in intact cells when compared to permeabilized cells. One possible explanation is that this labeling resulted from damaged organisms presented in that particular microscopic field. When LipL32 was used as a positive control in previously published IFA experiments [19,38], LipL32 surface-exposure was inconclusive as only one of two cells was labeled by antibodies in one study (Fig. 6) [19], while only one cell per microscopic field was shown in the other study.
