Team:ETH Zurich/Experiments/Anti Cancer Toxin

• Project Idea
  • Story of CATE
  • CATE in Action
  • Circuit
  • Design
• Engineering
  • Dry Lab
  • Overview
  • Tumor Sensor
  • Heat Sensor
  • 3D Model
  • Final Integration

  • Wet Lab
  • Experiments
  • Lab Journal
  • Protocols
  • Safety
  • Sequences
• Achievements
  • Results
  • Parts
  • Medal Criteria
  • InterLab
  • Collaborations
• Human Practices
  • Overview
  • Integrated Practices
  • Public Engagement
• Our Team
  • Attributions
  • Members
  • Sponsors

Anti-Cancer Toxin Experiments

This is a detailed experiment page dedicated to an individual function. To access other experiments, go to our Experiments page. To get a quick glimpse at all of our achievements, check out Results.

Achievements

• Production of azurin in a controlable manner (via AHL induction).
• Production of an azurin rich bacterial lysate.
• Quantification of azurin in our lysate (corresponding to a 140 µM intracellular concentration of azurin).
• Quantitative assessment of cytotoxic effect of azurin on p53 positive and negative cancer cell lines.

Introduction

Once a tumor environment has been recognized and colonized by our bacteria, the production of an anti-cancer toxin, together with an MRI Contrast Agent agent will be triggered. The module that we engineered to achieve this purpose includes an AND-logic synthetic promoter that recognizes the presence of high lactate and high bacterial density in the tumor environment and allows the expression of the anti-cancer toxin, azurin, and the MRI Contrast Agent, bacterioferritin, from the same operon (Figure 1).

Figure 1. Accumulation of anti-cancer toxin, azurin, into bacterial cells. As long as, the bacteria have colonized the tumor, the synthetic promoter will allow the expression of the operon containing bacterioferritin (bfr) and azurin (azu) genes. The intracellular accumulation, of bacterioferritin as an MRI contrast agent will allow the doctor to decide and give the bacteria the permission to lyse and release the accumulated cytotoxic therapeutic protein, azurin.

Once the Tumor Sensing has been activated, both the Anti-Cancer Toxin azurin and the MRI Contrast Agent bacterioferritin are produced. Azurin will accumulate inside of the bacteria until it is ready for release. Once the doctor verifies the bacterial colonization of the tumor bacteria via MRI, focused ultrasounds will activate Cell Lysis releasing an effective dosage of azurin to the tumor.

For more details about azurin and its role in our system, go to our description of the Anti-Cancer Toxin.

Overview of the Experiments

In our experiments, we replaced the AND-gate promoter in the genetic circuit depicted in Figure 1 with a well-established inducible promoter, Plux, in order to control the expression of the operon. Moreover, we coupled the expression of azurin with a fluorescence output as we exchanged the bacterioferritin gene with a gene encoding for superfolder GFP. In that way we could indirectly assess the functionality of our operon in a controllable manner (Figure 2). This circuit was transformed in E. coli TOP10 which was used for the production of azurin-rich bacterial lysate. Finally, the application of the bacterial lysate on different cancer-cell lines in vitro helped us to emulate the delivery of the anti-cancer payload via cell-lysis against different solid-tumor types.

Figure 2. AHL diffuses into the cell and binds to LuxR. The AHL/LuxR complex activates pLux, which results in transcription of both GFP and azurin. In a further step the cell-lysate containing the products of the operon is applied on different tumor - cells in vitro.

Several preliminary experiments were conducted before the application of the bacterial lysate to different cancer-cell lines. Initially we examined the production of azurin from our engineered bacteria after induction with AHL, indirectly by detecting the fluorescence of the co-expressed sfGFP. An SDS-PAGE analysis confirmed the presence of azurin in the cytosolic fragment of E. coli after AHL induction. Further culture scale-up to 50 mL volumes allowed us the production of azurin-rich bacterial lysate via sonication. A second SDS-PAGE confirmed cell-lysis by sonication as all the overexressed azurin and sfGFP were detected together with the native proteins of E. coli within the lysate. Thanks to our collaborators from Freiburg, we were able to quantify the azurin content of our bacterial lysate, as they provided us with purified azurin, which we used to create a standard curve of purified azurin concentration on the same SDS-PAGE gel.

With our bacterial lysate at hand we had to define the following parameters regarding the experimental set-up of mamalian cell incubation with our lysate:

• Which cancer cell-lines should we use?

• Which is the optimal number of cells to seed in our experiments?

• How can we assess the viability of the mammalian cells?

• How much of azurin-free bacterial lysate can our cells tolerate?

After defining the experimental parameters we tested the cytotoxic potential of our azurin-rich bacterial lysate against different cancer cell lines. Simultaneously, our Freiburg collaborators provided us with a killing curve of purified azurin against HEK-239 cells, enabling the comparison between the efficacy of purified azurin and azurin - rich bacterial lysate.

To read more about each of these experiments, click on the buttons below. For a detailed protocol describing each experiment, visit Protocols.

Phase I - Killing assay development

Co-Expression of Azurin and sfGFP by E. coli TOP10

OBJECTIVE
In this experiment we attempted to co-express azurin and sfGFP in the same operon in a controllable manner, emulating the co-expression of azurin and bacterioferitin in our final genetic circuit. The expression of sfGFP was used as an indirect readout of the operon gene expression. Additionally we tried to express a small azurin-derived peptide, namely p28, which is responsible for azurin's cytotoxic effect on cancer cells [1].

PROCEDURE
E. coli TOP10 were transformed with two plasmids, one responsible for constitutive expression of LuxR transcription factor and one containing the azurin and sfGFP gene in the same operon under an AHL inducible promoter regulated by the activating complex of LuxR::AHL. A second transformant strain was created carrying the gene encoding for p28 instead of the complete azurin protein. Biological triplicates of the engineered bacteria were cultured in shaking flasks at small scale (5 mL), and different AHL induction conditions were applied to them once they reached in the exponential phase (OD_600nm = 0.4), specifically:

• Induction with 10^-4M of AHL

• Mock induction with DMSO

Additionally, bacteria transformed with the same plasmids lacking the Plux_azurin/gfp construct were also cultivated serving as negative control of azurin expression. The cultivation stopped 200 min after induction, a time point where gene expression seemed to have stopped according to separate characterization experiments (see here link to bfr). Samples from each culture were transfered in a 96-well plate and both their absorbanse at 600nm (Abs _600nm) and their sfGFP content was evaluated in a plate reader. In the same plate we also included several concentrations of purified sfGFP allowing us to create a standard curve of sfGFP concentration (Figure 3). This standard curve, eventually allowed us to quantify the expression of sfGFP in the culture samples (Figure 4). A detailed protocol is available in Protocols.

RESULTS
The mean sfGFP concentration per cell was plotted for the biological samples under examination (Figure 2).

Figure 3. A standard curve of sfGFP fluorescence was obtained as the fluorescence of different purified sfGFP concentrations was measured via a spectrophometer. Based on this equation we can convert the fluorescence intensity into actual sfGFP concentration in our sample.

Figure 4. Concentration of sfGFP per cell for different bacterial cultures under different AHL induction conditions in biological triplicates. Upon induction with 10^-4M of AHL (+) both the operons expressing sfGFP are producing almost up to 90 - 100 mg/L of sfGFP in comparison with the DMSO mock-induced cultures (-) which produced less than 10 mg/L due to promoter leakiness. The bacteria carrying the empty vector control showed no fluorescence as expected, due to the lack of the sfGFP expressing operon.

The overexression and accumulation of sfGFP only after AHL induction was also verified via:

• Fluorescent images of the cell cultures exposed to blue-light illuminator (Figure 5, top)

• Images of the bacterial pellets from the bacterial cultures (Figure 5, bottom)

Figure 5. Top: Fluorescent pictures of bacterial cultures containing: 1) the empty vector lacking the whole azurin/sfGFP expressing operon induced with AHL , 2) the vector expressing azurin/sfGFP mock-induced with DMSO and 3) the vector expressing azurin/sfGFP induced with 10^-4M of AHL. Bottom: Pictures of bacterial pellets from different bacterial cultures containg the empty vector lacking the whole azurin/sfGFP expressing operon, the vector expressing azurin/sfGFP and the vector expressing p28/sfGFP. The overexpression and accumulation of sfGFP after AHL induction is verified by the green colour of the bacterial pellet.

CONCLUSION
With this experiment we could indirectly verify the functionality of both the operons expressing azurin and p28 respectively. From this experiments we chose the clone 3 for the creation of azurin-rich and p28-rich bacterial lysates. Samples from these cultures were further analyzed via SDS-PAGE to verify the production of p28 and azurin in a direct way.

Phase II - Test device characterization

SDS-PAGE to Confirm AHL-Induced Expression of azurin and p28

OBJECTIVE
Fluorescence measurements showed indirectly the functionality of the operons expressing azu/sfGFP and p28/sfGFP. To directly confirm that bacterioferritin was indeed co-expressed, an SDS-PAGE analysis was performed.

PROCEDURE
Protein lysates were obtained from bacteria treated with different concentrations of AHL. To determine the samples' protein concentration, needed to prepare the samples for SDS-PAGE, the Bradford protein assay was preformed prior to the SDS-PAGE analysis. A detailed protocol is available in Protocols.

RESULTS
To determine protein concentrations in the lysates via Bradford protein assay, a standard curve was first generated by using a protein standard, bovine serum albumin, and measuring absorbance of different dilutions of the standard. Second, absorbance of the unknown samples was measured and the results were fitted to the curve.

After determining the unknown concentrations, samples were prepared accordingly and subjected to SDS-PAGE analysis. Band sized approximately 14.8 kDa (which corresponds to azurin [3]) were visible in the samples treated with AHL. No bands corresponding to the azurin-derived p28 peptide could be detected, mainly because of its low molecular weight (27 aa, ~4 kDa). There were no bands in untreated samples or the negative control (Figure 7).

Figure 7. SDS-PAGE analysis of azurin and p28 expression upon AHL induction. The samples analyzed were equipped: 1) with a vector expressing azurin/sfGFP, 2) with a vector expressing p28/sfGFP or 3) with an empty vector. Adetectable band corresponding to azurin's molecular weight only appears in the induced sample (AHL +) while it is absent in the mock-induced sample (AHL -). No band could be detected corresponding the p28 on the AHL-induced p28 expressing strain.

CONCLUSION
As expected, a band corresponding to azurin were visible in the sample induced with AHL. With this, to co-expression of azurin and GFP in the test strain for azurin-rich lysate production experiments was confirmed. The production of p28 from our engineered strain couldn't be confirmed and for this reason no further experiments were conducted with this strain.

Determining the Optimal Seeding Cell Concentration for Mammalian Cell Cultures

OBJECTIVE
With this experiment we wanted to define the optimal starting cell population of different cancer cells lines, in order to avoid overgrowth and detachment of the mammalian cells in our experiments.

PROCEDURE
Variable amount of cells were seeded in a 96-well plate and their growth was assessed via microscopy after 24 hour growth in DMEM medium (supplemented with 1% streptomycin, 1% penicillin and 2% FBS).Two different cell - lines were tested, namely HEK-239 and HeLa CCL-2 lc. The use of HeLa CCL-2 lc allowed to monitor the growth of the cells via fluorescence microscopy as these cells are producing constituively citrine. A detailed protocol for this experiment is available in Protocols.

RESULTS
From the microscopy pictures obtained after 24-hours incubation, the optimal cell concentration range was defined between 60'000 and 600'000 cells/mL in a 96-well plate. For higher concetrations the mamallian cell culture seemed to approach confluency.

Figure 8. Pictures of HEK-239 cell cultures with different initial cell-concentrations after 24-hour incubation. The starting concentrations were spanning from 600 cells per well up to 600'000 cells per well, and the total volume loaded was 0.1 mL per well.

Figure 9. Fluorescence pictures of HeLa-CCL2 lc cell cultures with different initial cell-concentrations after 24-hour incubation. This cell line is constitutively producing citrine allowing for growth monitoring via fluorescence microscopy. The starting concentrations were spanning from 600 cells per well up to 600'000 cells per well, and the total volume loaded was 0.1 mL per well.

CONCLUSION
The optimal seeding cell concentration to avoid overgrowth was set to 100'000 cells/mL (between 60'000 and 600'000 cells/mL). The same value was used also in experiments were 1 mL cultures were incubated in 12-well culture plates.

Assessing Different Cell Viability Assays: Killing Effect of 5-Fluouracil on Different HeLa CCL-2 Cell Lines

OBJECTIVE
With this experiment we wanted to determine the efficiency of two different assays in order to evaluate the viability of mamalian cells. The first assay that we tried was the usage of Calcein AM dye. This compound penetrates the cell membrane and upon interaction with active esterases in the mamalian cell cytosol, a fluorescent product is being generated, leading to the staining of living cells containing active esterases in their cytosol. The second assay was based on the HeLa CCL-2 lc cell line ability to produce constitutively citrine. Upon cell death, cells are detaching from the surface and they are being removed via a washing step. In that way, the total fluorescence of the specific sample should be reduced. In this experiment, cell samples with diverse viability are generated by the application of a concetration gradient of a well-known cytotoxic agaent, namely 5-fluouracil [1]. For both assays the goal is to create a killing curve, where the fluorescent signal of the cell sample is dropping in respect to increasing cytotoxic agent concentration.

PROCEDURE
HeLa CCL-2 cells were seeded in a 96-well plate with initial concentration of 100'000 cells/mL. After 24-hour growth the cells are incubated with a different concentrations of 5-fluouracil. After 24 hours the cells are, incubated with calcein AM solution and the viable fraction of the cells is assessed for its fluorescence via spectrophotometer. The use of HeLa CCL-2 lc allowed us monitoring the viability of the cells on a spectrophotometer as these cells are producing constitutively citrine. Decrease in cell viability would result in less cells per well due to the detachment of the apoptotic cells from the well. In terms of fluorescence that would result in the decrease citrine related fluorescence. A detailed protocol is available in Protocols.

RESULTS
Fluorescence data were obtained via a spectrophotometer, for both the cell lines and viability assays used. Calcein AM staining assay gave inconclusive results as medium concentrations of tha killing agent didn't result in cell viability decrease (Figure 11, A). On the other hand, the cytotoxic effect of 5-fluouracil could be reliably monitored via the decrease of citrine fluorescence after the removal of detached non-viable cells (Figure 11, B).

Figure 11. Tecan measurements for viability assessment of HeLa CCL-2 and HeLa CCL-2 lc cells. (A) Detection of fluorescence stemming from calcein staining of viable cells. The results are inconclusive as there is cytotoxic effect detected when high and low amounts of the killing agent are applied to HeLa CCL-2 cells, but medium cocncentrations of the killing agent doesn't seem to affect the viability of the cells. (B) Detection of citrine fluorescence from HeLa CCL-2 lc cells (constitutive production of citrine). In this case the decrease in fluorescence correlates with the increase of the apoptotic agent used for the experiment.

CONCLUSION
The spectrophotometer results couldn't conviencingly prove the expected killing effect of 5-fluouracil on the HeLa CCL-2 cell line. On the other hand, the loss of citrine fluorescence from HeLa CCL-2 lc cell samples due to the removal of the dead cell fraction proved to be a good measure of cell viability. For this reason, we decided to use the HeLa CCL-2 lc cell line in further experiments to assess the cytotoxic effect of our azurin-rich bacterial lysates. However, the Calcein AM assay was still used for qualitative viability assessment of other cancer-cell lines that lack the constitutive citrine expression, via fluorescence microscopy.

Assessing the bacterial lysate tolerance of different cancer-cell lines

OBJECTIVE
With this experiment we wanted to assess the tolerance of the HeLa CCL-2 lc cell line upon incubation with different amount of azurin-free bacterial lysate. The information acquired from this expreriment helped us to define the amount of medium that we can substitute by our bacterial lysate without harming the cells neither because of the native E. coli proteins present in the lysate nor because of nutrient depravation. Additionally, a p53 negative cancer cell line, namely HT-29, was also qualitatively assessed for its tolerance towards azurin-free bacterial via calcein AM assay.

PROCEDURE
For the two different cell lines, namely, HeLa CCL-2 lc and HT-29, cells were seeded in 2 different 12-well culture plates with initial concentration of 100'000 cells/mL. After 24-hour growth the cells are incubated with increasing amounts of azurin-free lysate in order to assess the effect of the lysate on the cancer cell growth . Simultaneously, cell samples were also incubated with increasing amount of PBS, instead of lysate, in order to check the effect of nutrient depravation on the cell growth. After 24 hours the cells, apart from the HeLa CCL-2 lc, were incubated with calcein AM solution. The viability of the cells was assessed via fluorescnce microscopy. The samples of HeLa CCL-2 lc cell-line were directly examined for their viability via fluorescence microscopy as these cells are producing constitutively citrine. A detailed protocol is available in Protocols.

RESULTS
From the microscopy pictures obtained, we could verify that the bacterial lysate can substitute up to 20% of the HeLa CCL-2 lc culture volume without affecting the cell growth (Figure 12).

Figure 12. Pictures of HeLa CCL-2 lc cell cultures after 24-hour incubation with different amount of azurin-free lysate or PBS. The upper row of images depicts the fluorescence stemming from HeLa lc cells incubated with 0%, 5%, 10%, 20%, 30% bacterial lysate, while the lower row depicts the cell samples incubated with 0%, 5%, 10%, 20%, 30% PBS in order to assess if the decrease in cell viability is originating from bacterial toxins or due to nutrient depravation.

Different cell lines were expected to react differently when exposed to bacterial lysates, and that is the case for the HT-29 cancer-cell line which seems that cannot tolerate the same amount of azurin free bacterial lysate as HeLa CCL-2 lc (Figure 13).

Figure 13. Pictures of HT-29 cell cultures stained with calcein AM after 24-hour incubation with different amount of azurin-free lysate or PBS. The upper row of images depicts the fluorescence stemming from HeLa lc cells incubated with 0%, 5%, 10%, 20%, 30% bacterial lysate, while the lower row depicts the cell samples incubated with 0%, 5%, 10%, 20%, 30% PBS in order to assess if the decrease in cell viability is originating from bacterial toxins or due to nutrient depravation. The sample on the bottom row verifies the functionality of the calcein AM staining of living cells.

CONCLUSION
To minimize the risk of any cytotoxic effect of our bacterial lysate on the HeLa CCL-2 lc, we decided to use 10% (v/v) of the bacterial lysate for the upcoming experiments.

Phase III - Demonstrating azurin's killing ability

Applying Azurin-Rich Lysate to Different Mammalian Cell Lines: Azurin Killing Curve

OBJECTIVE
This was a milestone experiment in our work, as we wanted to assess the cytotoxic potential of azurin-rich bacterial lysates on p53 (+) and p53 (-) cell lines. Literature suggests, that the main mode of action of azurin against tumors is via stabilization od p53 protein [1]. For this reason HeLa CCL-2 lc was a suitable candidate for our experiment as it consists a p53 (+) strain and additionally has been well-examined in our project (good lysate tolerance and reliable cell viability assay). Moreover, the HT-29 p53 (-) negative cell line was also tested using the AM-calcein vaibility assay, in order to examine the effect of azurin against p53 (-) cancer cell - lines.

PROCEDURE
For the two different cell lines (HeLa CCL-2 lc and HT-29), cells were seeded in triplicate samples in a 96-well plates with initial concentration of 100'000 cells/mL. After 24-hour growth the cells are incubated with increasing amounts of azurin-rich lysate in order to assess the effect of the different azurin concentrations on the cancer cell growth. The concentration gradient of azurin-rich lysate was achieved by mixing azurin-free and azurin-rich bacterial lysate in different ratios. Simultaneously, cell samples were also incubated with 10% of PBS and also with 10% of azurin-free lysate. These negative control samples allowed us to check for any other cytotoxic effect stemming either from nutrient depravaion or native bacterial proteins. Moreover, cell samples were also incubated with 1 mM of 5-fluouracil allowing us to test the functionality of our staining assay. After 24 hours the cells, apart from the HeLa CCL-2 lc, were incubated with calcein AM solution. The viability of the cells was assessed via citrine fluorescence for the HeLa CCL-2 lc cells and calcein fluorescence for the HT-29 cells in a spectrophotometer. A detailed protocol is available in Protocols.

RESULTS The cytotoxic effect of our azurin-rich lysate was verified by the decrease in the citrine fluorescence, following an increase in the azurin content. From the following graph (Figure 14) even 3.2% of azurin lysate (~2 nM) is enough to detect a cytotoxic effect.

Figure 14. Fluorescence data from HeLa CCL-2 lc cell cultures incubated with different amount of azurin-rich lysate. Decrease in citrine fluorescence is only detected in samples were cells were incubated with bacterial lysate containing azurin, while in ony other case (azurin free lysate and PBS) no cytotoxicity was detectable. The functionality of the viability assay was successfully evaluated by the detection of cytotoxicity on cells incubated with a lethal concnentration of 5-fluoruracil (1 mM).

Azurin's cytotoxic effect cannot be detected against HT-29 cells according to the results of the calcein AM viability assay (Figure 15).

Figure 15. Fluorescence data from HT-29 cell cultures incubated with different amount of azurin-rich lysate. No obvious decrease in fluorescence is detecteed in samples incubated with bacterial lysates. A small fluorescence decrease is detected in the sample incubated with a lethal concnentration of 5-fluoruracil (1 mM).

CONCLUSION
The cytotoxic effect of azurin on the p53 + HeLa CCL-2 lc cancer cell line was detected. An effective azurin concentration of 1.8 nM could be deduced from this experiment and it was used as an input for the estimation of the killing area our system could achieve.

Determination of azurin absolute effective concentration

OBJECTIVE
Quantification of the azurin content in our bacterial lysate.

PROCEDURE
Freshly prepared bacterial lysate from cultures expressing azurin/gfp after induction with AHL, were analyzed via SDS - PAGE together with a concentration gradient of purified His₈-azurin provided by Freiburg iGEM Team. The standard protocol for SDS-PAGE was followed. Quantitative analysis of the SDS-PAGE gel was done with a LI-COR Odyssey CLx scanner.

RESULTS
The result of the SDS-PAGE verified the presence of the co-expressed azurin and sfGFP in the produced lysate (Figure 16). The same gel was used to quantify the amount of azurin via a concentration standard curve of purified azurin (Figure 17). Based on the OD of of the bacterial liquid culture before lysis (OD = 5) we could deduce that our lysate contained 7 mg/L of azurin (it constituted 12% of the total protein content). Considering the molecular weight of azurin (14.8 kDa), this amounts to a concentration of 570 nM in the bacterial lysate.

We also want to know what was the intracellular concentration of azurin to assess the maximimum concentration that we could achieve in our bacteria. For this we used the "rule of thumb" proportionality relationship between the OD and the number of cells (8.10⁸ cells/OD) and the approximate volume of E. coli: 1 µm³. We found an intracellular concentration of azurin of 140 µM, which is consistent with the 12% figure of protein content.

Figure 16. SDS-PAGE analysis of different bacterial lysates and concnetration gradient of purified His₈-azurin. The samples analyzed were equipped with a vector expressing azurin/sfGFP, or an empty vector. A detectable band corresponding to azurin's molecular weight only appears in the induced sample (AHL +) while it is absent in the mock-induced sample (AHL -). The purified azurin's molecular weight appears slightly higher due to the additional His-tag in the protein sequence.

Figure 17. Quantitative SDS-PAGE analysis of different bacterial lysates and concnetration gradient of purified His₈-azurin. The total protein content of the lysates was, 2.4 μg for all the bacterial lysate samples. The concentration gradient was consisting of 2.4 μg, 1.92 μg, 1.44 μg, 1.2 μg, 0.96 μg, 0.72 μg, 0.24 μg purified His₈-azurin.

With this data at hand, we went back to the experiments where we assessed the cytotoxicity of azurin on the HeLa CCL-2 lc cells. The effective azurin concentration was then determined as 2 nM.

CONCLUSION
We manage to quantify the amount of azurin in our bacerial lysate and thus to defone the cytotoxic azurin concetration (2 nM) against p53 + cancer cells (HeLa CCL-2 lc).

References

1. Jiang, Jian-Xin, et al. "Overexpression of microRNA-125b sensitizes human hepatocellular carcinoma cells to 5-fluorouracil through inhibition of glycolysis by targeting hexokinase II." Molecular medicine reports 10.2 (2014): 995-1002. doi: 10.3892/mmr.2014.2271
2. Yamada, Tohru, et al. "A peptide fragment of azurin induces a p53-mediated cell cycle arrest in human breast cancer cells." Molecular cancer therapeutics 8.10 (2009): 2947-2958. doi: 10.1158/1535-7163.MCT-09-0444
3. Choi, Moonsung, and Victor L. Davidson. "Cupredoxins—a study of how proteins may evolve to use metals for bioenergetic processes." Metallomics 3.2 (2011): 140-151. doi: 10.1039/c0mt00061b

Top

• 
• 
• 

Sponsors

• 
• 
• 
• 
• 
• 
• 
• 
• 
•