Effects of the temperature on the recombinant display system
CotB, CotC and CotG are abundant coat proteins widely used as carriers to display heterologous proteins on the spore surface . All three proteins have been recently found differentially represented in spores produced at 25, 37 or 42°C, with CotB and CotG more abundant in spores prepared at 25°C and CotC more abundant in 42°C spores . We used isogenic B. subtilis strains carrying DNA coding for the model antigen TTFC (tetC) fused to the gene coding for either CotB (cotB)  or CotC (cotC)  to evaluate the abundance of the fusion proteins at the different temperatures. Spores of strains RH103 (cotB::tetC) and RH114 (cotC::tetC) were produced at 25, 37 and 42°C and purified, as previously reported . Surface proteins were extracted from RH103 and RH115 spores by the SDS-DTT or NaOH treatments, respectively and used for western blotting analysis with anti-CotB  or anti-CotC  antibodies. As shown in Figure 1, the fusion protein CotB-TTFC was more abundantly extracted from 25°C spores than from 37 or 42°C spores (upper panel), while the fusion CotC-TTFC was more represented in 42°C spores than in 37 or 25°C spores (lower panel).
A flow cytometry approach was used to quantify the differences in the display of CotB-TTFC and CotC-TTFC at the various temperatures and evaluate their surface exposure. Spores of strains RH103 and RH114 were reacted with anti-TTFC  antibodies, then with fluorescently labeled secondary antibody and analyzed by flow cytometry as previously reported . The threshold of positive events was set at 1x103 fluorescence intensity and the percentages of fluorescent events for each temperature are indicated in red in each panel. The flow cytometry analysis indicated that CotB-TTFC was displayed with the highest efficiency in spores prepared at 25°C (86.9% positive events) and that such efficiency decreased in 37 and 42°C spores (Figure 2). The efficiency of display was opposite with CotC-TTFC with the highest levels observed with 42°C spores (90.0% of positive events) and lower levels with 37 and 25°C spores (Figure 2).
Results of Figures 1 and 2 indicated, respectively, the amounts of fusion proteins extracted and exposed on the spore surface but did not allow to exclude that other amounts of each fusion were actually present (but not extracted or not exposed) on spores produced at different temperatures. To address this issue, we used different isogenic strains of B. subtilis RH238, carrying the Green Fluorescent Protein (GFP) fused to CotC , and RH296, carrying the Red Fluorescent Protein (RFP) fused to CotG . A fluorescence microscopy analysis on spores prepared at 25, 37 or 42°C and the quantification of the fluorescence signals performed by the ImageJ software, as previously reported , indicated that the CotG-based fusion was more abundant at 25°C, less abundant at 37°C and almost undetectable at 42°C while the CotC-based fusion showed an opposite pattern (Figure 3).
Results of Figure 3, confirming results of Figures 1-2, allow to conclude that the CotB- and CotG-based fusions are efficiently displayed when spores are produced at 25°C, while CotC-based fusions are better displayed when spores are produced at 42°C and, therefore, that is possible to modulate the amount and the surface exposure of fusion proteins displayed on the spore by changing the temperature of spore production on the base of the carrier protein used for the display.
Effects of the temperature on recombinant spores displaying two fusion proteins
An extension of the recombinant spore-display technology is the use of spores carrying more than one heterologous protein. By chromosomal DNA-mediated transformation , the gene fusion carried by strains RH238 (cotC::gfp) was moved into strain RH296 (cotG::rfp) obtaining strain RH406 that carried both fusions. As shown in Figure 4, spores of strain RH406 presented both fluorescent proteins on their surfaces in similar amounts when spores were grown at 37°C. When spores were produced at 25°C the red fluorescent signal (CotG-RFP) was more abundant than the green one (CotC-GFP) that was instead predominant when spores were grown at 42°C.
Results of Figure 4 highlight an important improvement for the spore-display technology, showing that it is possible to produce spores that simultaneously display two heterologous proteins and to control which displayed protein has to be more abundantly represented by selecting the temperature of spore production.
Effects of the temperature on the non-recombinant display system
To evaluate the effects of the temperature on non-recombinant spore-display (adsorption) we used three model proteins: the pentapeptide HPHGH (herein PPT) of 0.77 kDa , the commercially available lysozyme (herein LYS) of 14.4 kDa (Sigma) and the commercially available bovine serum albumin (herein BSA) of 66.4 kDa (New England-Biolabs). All three proteins were fluorescently labeled with rhodamine as previously described  and 10 mM of each model protein independently used for adsorption with 5.0x108 purified spores of the B. subtilis strains PY79  produced at 25, 37 or 42°C. The adsorption reactions were carried out for 1 hour at 25°C in 50mM Sodium Citrate buffer, pH 4.0, as previously described . Adsorbed spores were collected by centrifugation and analysed by fluorescence microscopy and flow cytofluorimetry, as previously described . As shown in Figure 5, all three proteins were adsorbed to the spores and the relative fluorescence signals were analyzed by the ImageJ software (NIH), as previously reported . Since the proteins were fluorescently tagged with rhodamine, an amine-specific label, the number of fluorophore molecules attached to each protein was different, impairing a comparison of fluorescence levels between different proteins. However, the analysis allowed to conclude that: i) PPT adsorbed with similar efficiency to 37°C and 42°C and slightly less efficiently to 25°C spores (3742>25); ii) LYS had a pattern of adsorption similar to that described for PPT (3742>25); and iii) BSA adsorbed at similar levels to 25, 37 or 42°C spores (25=37=42) (Figure 5). Adsorbed spores were analyzed by flow cytometry and the percentage of positive-fluorescent events was obtained as described for Figure 2. This analysis confirmed the fluorescence microscopy results of Figure 5, with PPT and LYS adsorbing less efficiently to 25°C spores than to 37 or 42°C and BSA adsorbing with similar efficiency to all three types of spores (Figure 6).
Although the molecular mechanism of spore adsorption is not known in detail, it is likely that more factors are involved in the process. The negative electric charge and relative hydrophobicity of the spore surface have both been shown to influence the efficiency of adsorption [10, 14]. Since it has been previously reported that 25°C spores are more hydrophobic than 37 and 42°C spores , we hypothesized that the different relative hydrophobicity of spores could explain the reduced efficiency of adsorption of PPT and LYS to 25°C spores. However, the GRAVY value, an estimation of protein hydrophobicity calculated by adding the hydropathy values of each amino acid residue of a protein and dividing by the number of residues in the protein , for PPT, LYS and BSA were -2.32, -0.15 and -0.45, respectively, with increasing positive values indicating an increasing hydrophobicity. Therefore, proteins with the least (PPT) and the highest (LYS) hydrophobicity value showed a similar adsorption pattern (Figures 5-6), making it unlikely that the hydrophobicity is a major determinant of the efficiency of adsorption, in our experiments. Other physical and chemical parameters of the heterologous proteins, including probably the size and the isoelectric point, have to be considered as they may mediate the ability of proteins to cross the outermost spore layers [15-17], resulting in relevant for the efficiency of the process.
Localization of proteins adsorbed on 25, 37 or 42°C spores
A previous report showed that RFP when adsorbed to spores is able to cross the crust and the outer coat, localizing at the inner coat level . In that study, the RFP fluorescence signal was localized by comparison with the signal due to GFP fused to proteins known to be localized in various spore coat layers . A similar approach was used to evaluate whether the temperature of spore production also affected the localization of the adsorbed proteins within the coat. Since the high red fluorescence signal produced by rhodamine-labeled PPT, LYS or BSA overlapped (and caused interference) with the region of detection for the GFP signal, the localization assays were performed adsorbing RFP to spores carrying the cotC:: gfp fusion  and prepared at 25, 37 or 42°C.
As previously reported , in 37°C spores the red fluorescence signal of RFP was internal to the green signal of CotC-GFP (Figure 7). While RFP localization did not change with 25°C spores, it was slightly altered with 42°C spores where the RFP signal was external with respect to the CotC-GFP signal (Figure 7). The different localization of RFP is most likely due to the different coat structure of spore produced at the various temperatures and indicates that the lamellar and highly electron-dense outer coat (CotB-CotG rich) produced at low temperatures  is somehow a more permeable than the granular and thick coat (CotC rich) produced at 42°C , at least with respect to RFP.