Table 1.
Parameters for production of alginate beads containing 1x108 or fewer CFU/ml of F. tularensis bacteria.
Table 2.
Characteristics of alginate beads produced throughout the study.
Fig 1.
Alginate bead containing GFP-expressing LVS.
Alginate beads were created that contained LVS expressing the fluorescent protein GFP. Only viable bacterial cells inside the generally spherical shaped capsule are visible by fluorescence microscopy.
Fig 2.
Effect of production parameters on bead size.
Bead diameter was assessed for 30 beads produced using nozzle sizes from 80 μm to 200 μm and starting bacterial concentrations of 1x106 to 1x108 CFU/ml. Bead sizes statistically differed due to nozzle sizes (p < 0.0001) and due to starting bead concentration (p = 0.0032). The interaction between nozzle size and starting bacterial concentration was statistically important (p = 0.0019). Significant differences between starting cell concentration within nozzle groups were designated as: *: p < 0.05, **: p < 0.01, ***: p < 0.001.
Fig 3.
Bacterial concentration within beads determined by the bacterial starting concentration.
The viable bacterial concentration in a milliliter of beads was determined by dissolution of beads in citrate solution and viable plate count. Bacterial concentration within beads was statistically significant depending on the starting bacterial concentration (p = 0.0097) and was statistically affected by the interaction of starting concentration and nozzle size (p = 0.0206). Nozzle size alone did not statistically affect the bacterial concentration of the beads (p = 0.0646). Differences between nozzle sizes within groups is signified as: **: p < 0.01, ***: p < 0.001.
Fig 4.
Encapsulation efficiency significantly decreases with increasing starting bacterial concentration.
Encapsulation efficiency was defined as the percentage of bacterial cells that became incorporated into the final alginate bead. Encapsulation efficiency decreased with increasing starting bacterial concentrations. Significant differences between starting bacterial concentrations within nozzle sizes are shown as: *: p < 0.05.
Fig 5.
Encapsulation increases survival of serum-sensitive WbtIG191V incubated in complement-active serum.
WbtIG191V was encapsulated in alginate or APA beads and then incubated in complement-active serum for 1 hour, 24 hours, and 48 hours. A: Viable plate counts were used to determine the percent survival of WbtIG191V freely suspended in serum (dotted bar) compared to WbtIG191V encapsulated in alginate beads (diagonal line bar). WbtIG191V encapsulated in alginate only survived significantly more than freely suspended WbtIG191V after 1 hour (p < 0.001, ***), and increased in number due to growth during that time. Viable WbtIG191V cells in alginate beads were detected after 24 hours of incubation in complement-active serum. However, the difference was not statistically different from freely suspended WbtIG191V. Three independent experiments were run in duplicate per experiment. All values represent the mean +/- the standard error of the mean. B: The presence and absence of growth after incubation of the bacteria in complement-active serum for specified amounts of time was determined for bacteria encapsulated in washed alginate or APA beads containing WbtIG191V. Bacterial growth resulted only after one hour incubation of the bacteria encapsulated in alginate-only beads in complement-active serum. WbtIG191V within APA capsules grew after 48 hours of incubation in complement-active serum, similar to serum-resistant LVS, indicating an increase in protection with the additional coating around the beads. However, APA beads could not be dissolved so quantitative plate count cultures were not possible.
Fig 6.
Anti-LVS antibody titers in immunized mice.
Groups of 4 mice each (2 male and 2 female) were immunized with the described formulations either SQ or IP. Serum samples were taken post-immunization to determine total IgG (heavy and light chain) titers against F. tularensis LVS. Sera were diluted out to 1:12,800. The endpoint titer was determined to be the reciprocal of the dilution with an OD450nm that was 0.100 greater than the OD450nm of mock-immunized mice. Mock-immunized mice were inoculated IP and SQ (SQ not shown), but antibody titers were below the lowest titer measured (1:100). Bars represent the mean titer with error bars representing the standard error of the mean. Error bars are not included for mice immunized with WbtI in APA beads because all of the titers were greater than the endpoint dilution of 1:12,800. Significant antibody titer differences compared to mock-immunized mice are represented as: *: p < 0.05 or **: p < 0.01.
Fig 7.
Survival of mice immunized intraperitoneally and challenged with virulent LVS intranasally.
Groups of 4 mice were immunized SQ (a) or IP (b) with either WbtIG191V (●), WbtIG191V + purified LPS (▲), APA-encapsulated WbtIG191V (▼), APA encapsulated WbtIG191V + purified LPS (♦), LVS (o), or PBS (■). Six weeks post-immunization, mice were challenge with a high dose (1x106 CFU) of virulent LVS IN. Of the SQ-immunized mice, only mice that were immunized with LVS significantly survived the challenge period (p = 0.0014). Of the IP-immunized mice, only mice that were immunized with WbtIG191V and purified LPS, either encapsulated or freely suspended in solution, survived until the end of the study. Survival was 75% (p = 0.0014) and 50% (p = 0.0029), respectively.
Fig 8.
Tissue bacterial loads of mice immunized intraperitoneally and then challenged with virulent LVS intranasally.
BALB/c mice (4 per IP and SQ groups) were immunized with different formulations of bacterial strains within and without APA beads, and then challenged with 1x106 CFU/ml of virulent LVS intranasally. At necropsy, tissue samples were harvested from the liver (a), lung (b), and spleen (c), resulting in an analysis of 42 samples read blindly. Bacterial numbers in these tissues were determined by viable plate count. Significant differences in bacterial load in a gram of tissue are indicated by: *: p < 0.05.
Fig 9.
Organ histopathology scores of mice immunized and then challenged with virulent LVS intranasally.
Four BALB/c mice were immunized with different formulations and then challenged with 1x106 CFU/ml of virulent LVS intranasally. All beads were APA beads. At necropsy, tissue samples were harvested from the liver (a), lung (b), and spleen (c) of all 4 mice with 3–4 samples of each tissue analyzed and read blindly, resulting in an analysis of 42 samples. Samples of tissues were fixed, stained with H&E stain, and graded for histopathological lesions. Infiltration of hepatic lobules in the liver (a) trend higher for mock immunized mice compared to immunized mice, but were not significantly different as a whole, likely due to the low number of animals used (p = 0.0629). Inflammatory infiltrates in the lung (b) were not significantly different between groups (p = 0.6079). Infiltration of red pulp in the spleen (c) were significantly lower in all immunized groups compared to the mock group (p = 0.0110) Significant differences in histopathology scores compared to mock immunized mice are indicated by: *: p < 0.05.
Fig 10.
Innate and adaptive immune response gene expression within lung tissue after challenge with F. tularensis.
Mice (four per group; 2 male and 2 female) were immunized with LVS SQ, WbtIG191V + LPS IP, or WbtIG191V + LPS within APA beads IP, and then challenged with a high dose of LVS IN. At the end of the study RNA was isolated from each lung sample of each mouse that survived challenge, then pooled by immunization group, and analyzed with an RT2 PCR Profiler for the mouse innate and adaptive immune response compared to normal lung tissues. The heat map shows expression levels of each gene. Green indicates minimal expression, and red indicates maximum expression for each group and then grouped by trends in expression.