Engineering Bacteria for Vaccine Production and Stabilization
Engineering Bacteria for Vaccine Production and Stabilization Kelly Shepardson Wells College Aurora, NY Dr. David Putnam Department of Biomedical Engineering Cornell University the cultures were observed until an OD of Introduction Outer membrane vesicles (OMV) are spherical bi-layered structures that spontaneously bud off the surface of gramnegative bacteria. Even though the existence of these OMV have been known for years, their functions in nature are not quite understood [1]. Attention has recently focused on using recombinant DNA technologies to engineer bacteria and use their OMV for drug delivery. The goal of this research is to evaluate the stability of OMV as a function of temperature and buffer formulation for future drug delivery and vaccine applications. This study focused on two different types of vesicles, one type containing a recombinant green fluorescent protein (GFP) and one wild-type (unaltered) control. Overall, twenty different treatment groups were evaluated. Vesicles were prepared in two buffers, a phosphate buffered saline (PBS) or a Tris buffer containing 3% sucrose (TS). The samples were stored at one of five temperatures: 37oC, 4oC, -20oC, -70oC, and lyophilized. The vesicles were characterized over a 12day period using dynamic light scattering (DLS), zeta potential, and transmission electron microscopy (TEM).
Materials and Methods To prepare the OMV samples, overnight cultures of E. coli were transferred at a dilution of 1:100 to 200 mL of fresh LB medium containing 50 μg/mL ampicillin The samples were incubated at 37oC with shaking and the optical density (OD) of
0.4 was reached and then recombinant protein expression was induced by adding L-arabinose to a final concentration of 0.2 percent. Cultures were further incubated at 37oC for 16 hours. After incubation, the culture was centrifuged at 7000rpm to pellet the cells and the supernatant was decanted and filtered through a 0.45μm filter. The filtrate was centrifuged at 28,000rpm to pellet the outer membrane vesicles. The OMV were then suspended in buffer and the protein content was quantified by a BCA protein assay with an albumin calibration curve. Samples were diluted to a final protein concentration of 60μg/mL and stored under the conditions specified above. The diameter of OMV was measured by dynamic light scattering over a twelve day time-course study. Time points were taken on day 0, 1, 3, 6, 9, and 12. For DLS, 500μL of the sample was used for the reading and samples were measured in triplicate. To quantify the net surface charge of the OMV, zeta potential measurements were taken at day 0 and day 12. Seven hundred fifty microliters were used for each sample and measurements were run triplicate. For transmission electron microscopy, the samples were prepared by pipeting 5μL of sample onto a 400-mesh carbon grid and blotting it after ten seconds. Samples were negatively stained with 5μL of uranyl acetate for ten seconds. The grids were visualized using a Technai 12 transmission electron microscope.
Figure 2: Size of OMV (37oC)
Results and Discussion
111
Empty OMV in PBS Empty OMV in TS GFP OMV in PBS GFP OMV in TS 100
Time (days)
Figure 3: Size distribution by intensity of 37oC OMV in PBS Particle size (nm)
105
Empty OMV in PBS Empty OMV in TS GFP OMV in PBS GFP OMV in TS 0
Time (days)
4
8
12
The average particle size remained The initial conditions, (shown in green) are at the normal size for vesicles relatively steady over 12 days, since the particle sizes of both the GFP and wild- (100nm). Figure 4, shows a dramatic effect type OMV remained within 10nm of their initial measurements. Similarly, the of the storage buffer on lyophilized vesicles. There are large differences in samples stored at -70oC and 4oC showed a variation of less than 10nm in the particle particle size for the vesicles in PBS over size over 12 days. For vesicles stored at 12 days, with an increase from 100nm up 4oC, -20oC, and -70oC, there was little to 800nm. effect of the storage buffer on particle size. Vesicle sizes for samples stored at Figure 4: Size of OMV (lyophilized) 37oC, Figure 2, showed a decrease of about 10 nm after one day and then 1000 GFP OMV in PBS remained steady thereafter. Once again, GFP OMV in TS there appeared to be minimal effect of the Empty OMV in PBS storage buffer on vesicle size. A closer Empty OMV in TS look at the particle size distribution, shown 500 in Figure 3, shows that the particles are beginning to form two separate populations on day 12 (shown in red), which consisted of larger aggregates (~300nm) and smaller individual vesicles 0 (~60nm) compared to the distribution on 0 Time (Days) 4 8 12 day zero (green trace). Particle size (nm)
85
12
0
89
8
Figure 1: Sizes of OMV (-20oC) 125
4
Particle size (nm)
The average particle size of vesicles (in nm) were plotted as a function of time for all samples. Figure 1 shows the particle size of GFP and wild-type OMV stored at -20oC. OMV containing GFP were larger than the wild type control at all time points.
The size distribution by intensity of the lyophilized samples, shown in figure 5, also shows that on day 0 for the vesicles in PBS the particle size is around 100nm and on day 12 the vesicles are beginning to form three separate populations of sizes (~120nm, ~800nm, and ~4000nm). Figure 5: Size distribution by intensity of lyophilized OMV in PBS
As shown in Figure 4, the vesicles that were lyophilized in TS stayed constant in size throughout the whole time period and the size distribution, figure 6, shows that at day 12 the size distribution is more uniform with some slight aggregation taking place. The TS buffer is significantly better than phosphate buffered saline for lyophilized vesicle formulations.
Figure 6: Size distribution by intensity of lyophilized OMV in TS
Zeta potential measurements, shown in figure 7, were taken on day zero and day twelve. The results show the net surface charge did not have a significant change over the 12 day period for any of the samples, indicating little observable changes in the vesicle surface characteristics. Figures 8 and 9 display micrographs of vesicles taken by transmission electron microscopy. Figure 8, the lyophilized sample in TS after 12 days, shows single vesicles with lipid bilayers that are individually distributed and approximately 70 nm in diameter.
Figure 7: Zeta Potential measurements (Day 0 and day 12) day 0
d-12 (-70C)
d-12 (-20C)
Zeta Potential (mV)
-10
-24
Empty OMV in PBS Empty OMV in TS GFP OMV in PBS GFP OMV in TS
d-12 (4C)
d-12 (37C)
d-12 (lyophilized)
Figure 9, from the 37oC sample after 12 days, shows individual vesicles that are of smaller size (~25nm), but they clump together to form large aggregates. The micrographs in Figures 8 and 9 are consistent with the particle sizing data obtained by dynamic light scattering. Figure 8: TEM of lyophilized OMV in TS (135000X)
Figure 9: TEM of 37oC OMV in TS (135000X)
great amount of aggregation. Dramatic effects of the buffer were seen in lyophilized samples, where the TS showed no change in particle size and the PBS showed size changes on the order of hundreds of nanometers over the course of 12 days. The lyophilized OMV in PBS could be less stable because a powder was formed after lyophilization compared to lyophilization in TS, which resulted in more visible crystal structures. The more ordered structure of the lyophilized OMV in TS could be the cause for the better stability, but further characterization experiments are required to support this. The zeta potential of the vesicles did not change significantly over the 12 day period, indicating that the surfaces of the vesicles are not dramatically changing. In future work, the results from this experiment can be used to design longer stability tests along the lines of weeks or months. This work will help to determine the optimal buffer formulation and storage conditions of engineered OMV vaccines.
Acknowledgments: I would like to thank the Putnam Group for this experience: David Putnam, Dave Chen, Sara Yazdi, Sharon Wong, Jeisa Mari Pelet-Giraud, Peter Zawaneh, and Julius Korley. Foundation.I would also like to thank NBTC, the DeLisa Group, and John Grazul for instrumentation help. This project was funded by Cornell Center for Materials Research and National Science
Conclusions: The results show that in general, the particle size of outer membrane vesicles is more stable when stored at temperatures of 4oC and below. The 37oC vesicles showed a lot of aggregation and decreased vesicle size. Similarly the lyophilized samples in PBS showed a
References: [1] Kuehn, Meta J. and Kesty, Nicole C. 2005. Genes & Dev., 19, 2645-2655
Materials and Methods To prepare the OMV samples, overnight cultures of E. coli were transferred at a dilution of 1:100 to 200 mL of fresh LB medium containing 50 μg/mL ampicillin The samples were incubated at 37oC with shaking and the optical density (OD) of
0.4 was reached and then recombinant protein expression was induced by adding L-arabinose to a final concentration of 0.2 percent. Cultures were further incubated at 37oC for 16 hours. After incubation, the culture was centrifuged at 7000rpm to pellet the cells and the supernatant was decanted and filtered through a 0.45μm filter. The filtrate was centrifuged at 28,000rpm to pellet the outer membrane vesicles. The OMV were then suspended in buffer and the protein content was quantified by a BCA protein assay with an albumin calibration curve. Samples were diluted to a final protein concentration of 60μg/mL and stored under the conditions specified above. The diameter of OMV was measured by dynamic light scattering over a twelve day time-course study. Time points were taken on day 0, 1, 3, 6, 9, and 12. For DLS, 500μL of the sample was used for the reading and samples were measured in triplicate. To quantify the net surface charge of the OMV, zeta potential measurements were taken at day 0 and day 12. Seven hundred fifty microliters were used for each sample and measurements were run triplicate. For transmission electron microscopy, the samples were prepared by pipeting 5μL of sample onto a 400-mesh carbon grid and blotting it after ten seconds. Samples were negatively stained with 5μL of uranyl acetate for ten seconds. The grids were visualized using a Technai 12 transmission electron microscope.
Figure 2: Size of OMV (37oC)
Results and Discussion
111
Empty OMV in PBS Empty OMV in TS GFP OMV in PBS GFP OMV in TS 100
Time (days)
Figure 3: Size distribution by intensity of 37oC OMV in PBS Particle size (nm)
105
Empty OMV in PBS Empty OMV in TS GFP OMV in PBS GFP OMV in TS 0
Time (days)
4
8
12
The average particle size remained The initial conditions, (shown in green) are at the normal size for vesicles relatively steady over 12 days, since the particle sizes of both the GFP and wild- (100nm). Figure 4, shows a dramatic effect type OMV remained within 10nm of their initial measurements. Similarly, the of the storage buffer on lyophilized vesicles. There are large differences in samples stored at -70oC and 4oC showed a variation of less than 10nm in the particle particle size for the vesicles in PBS over size over 12 days. For vesicles stored at 12 days, with an increase from 100nm up 4oC, -20oC, and -70oC, there was little to 800nm. effect of the storage buffer on particle size. Vesicle sizes for samples stored at Figure 4: Size of OMV (lyophilized) 37oC, Figure 2, showed a decrease of about 10 nm after one day and then 1000 GFP OMV in PBS remained steady thereafter. Once again, GFP OMV in TS there appeared to be minimal effect of the Empty OMV in PBS storage buffer on vesicle size. A closer Empty OMV in TS look at the particle size distribution, shown 500 in Figure 3, shows that the particles are beginning to form two separate populations on day 12 (shown in red), which consisted of larger aggregates (~300nm) and smaller individual vesicles 0 (~60nm) compared to the distribution on 0 Time (Days) 4 8 12 day zero (green trace). Particle size (nm)
85
12
0
89
8
Figure 1: Sizes of OMV (-20oC) 125
4
Particle size (nm)
The average particle size of vesicles (in nm) were plotted as a function of time for all samples. Figure 1 shows the particle size of GFP and wild-type OMV stored at -20oC. OMV containing GFP were larger than the wild type control at all time points.
The size distribution by intensity of the lyophilized samples, shown in figure 5, also shows that on day 0 for the vesicles in PBS the particle size is around 100nm and on day 12 the vesicles are beginning to form three separate populations of sizes (~120nm, ~800nm, and ~4000nm). Figure 5: Size distribution by intensity of lyophilized OMV in PBS
As shown in Figure 4, the vesicles that were lyophilized in TS stayed constant in size throughout the whole time period and the size distribution, figure 6, shows that at day 12 the size distribution is more uniform with some slight aggregation taking place. The TS buffer is significantly better than phosphate buffered saline for lyophilized vesicle formulations.
Figure 6: Size distribution by intensity of lyophilized OMV in TS
Zeta potential measurements, shown in figure 7, were taken on day zero and day twelve. The results show the net surface charge did not have a significant change over the 12 day period for any of the samples, indicating little observable changes in the vesicle surface characteristics. Figures 8 and 9 display micrographs of vesicles taken by transmission electron microscopy. Figure 8, the lyophilized sample in TS after 12 days, shows single vesicles with lipid bilayers that are individually distributed and approximately 70 nm in diameter.
Figure 7: Zeta Potential measurements (Day 0 and day 12) day 0
d-12 (-70C)
d-12 (-20C)
Zeta Potential (mV)
-10
-24
Empty OMV in PBS Empty OMV in TS GFP OMV in PBS GFP OMV in TS
d-12 (4C)
d-12 (37C)
d-12 (lyophilized)
Figure 9, from the 37oC sample after 12 days, shows individual vesicles that are of smaller size (~25nm), but they clump together to form large aggregates. The micrographs in Figures 8 and 9 are consistent with the particle sizing data obtained by dynamic light scattering. Figure 8: TEM of lyophilized OMV in TS (135000X)
Figure 9: TEM of 37oC OMV in TS (135000X)
great amount of aggregation. Dramatic effects of the buffer were seen in lyophilized samples, where the TS showed no change in particle size and the PBS showed size changes on the order of hundreds of nanometers over the course of 12 days. The lyophilized OMV in PBS could be less stable because a powder was formed after lyophilization compared to lyophilization in TS, which resulted in more visible crystal structures. The more ordered structure of the lyophilized OMV in TS could be the cause for the better stability, but further characterization experiments are required to support this. The zeta potential of the vesicles did not change significantly over the 12 day period, indicating that the surfaces of the vesicles are not dramatically changing. In future work, the results from this experiment can be used to design longer stability tests along the lines of weeks or months. This work will help to determine the optimal buffer formulation and storage conditions of engineered OMV vaccines.
Acknowledgments: I would like to thank the Putnam Group for this experience: David Putnam, Dave Chen, Sara Yazdi, Sharon Wong, Jeisa Mari Pelet-Giraud, Peter Zawaneh, and Julius Korley. Foundation.I would also like to thank NBTC, the DeLisa Group, and John Grazul for instrumentation help. This project was funded by Cornell Center for Materials Research and National Science
Conclusions: The results show that in general, the particle size of outer membrane vesicles is more stable when stored at temperatures of 4oC and below. The 37oC vesicles showed a lot of aggregation and decreased vesicle size. Similarly the lyophilized samples in PBS showed a
References: [1] Kuehn, Meta J. and Kesty, Nicole C. 2005. Genes & Dev., 19, 2645-2655
