Intro

This experiment is a follow up to several other attempted experiments to perturb the DNA bound phenazines in colony biofilms. We tried to treat biofilms with DNase 1 in he underlying agar to chop up the eDNA into smaller pieces. This was somewhat successful when DNase was not incubated in its NEB buffer, which killed the cells. However, the effects were modest, so I tried repeating the experiment with DNase in PBS supplemented with Mg2+, which should further activate the enzyme. Further, I wanted to try the experiment on WT and ∆pelB colonies, since we think that Pel may be binding eDNA and blocking PYO binding sites it may also interfere with the activity of the DNase…?

I have also tried perturbing colony retained phenazines with a competitive intercalator ethidium (also tried propidium). This experiment has worked reasonably well multiple times, although it is complicated to displace WT phenazines, since the cells just seem to make more. Recently I tried exposing ∆phz* to phenazine and an etbr competitor simultaneously and it worked well. However, I only had duplicate measurements and I did not test PCA (only PYO and PCN). Therefore I wanted triplicates this time and also expose to PCA.

Methods

Colonies were grown as normal in 6 well plates on 0.2um filter membranes atop agar, incubated in the dark at room temperature. Images were taken on the keyence on days 3 and 4.

Results

First let’s look at the colony volumes to see if there are large differences in colony sizes that may skew our results. Remember these data were acquired simply by pipetting up the entire solution and measuring the volume crudely with a pipette. That means the measurements were extremely susceptible to bubbles that rise to the top of the solution and displace volume that wasn’t due to liquid.

df_vol <- read_csv("data/2019_10_22_colony_volumes.csv")

df_vol %>% kable() %>% kable_styling() %>% scroll_box(height = "300px")

First let’s look at the DNase treated colonies:

ggplot(df_vol %>% filter(treatment != "etbr"), aes(x = strain, 
    y = Measured_Volume, fill = treatment)) + geom_jitter(shape = 21, 
    width = 0.1)

You can see that there’s a lot of overlap between the treated and untreated colonies, however, the WT colonies may be slightly larger overall. Doesn’t seem like volume will be an issue, which makes sense since the colonies look pretty identical.

Now let’s look at the ethidium treated colonies:

df_vol$etbr_added_uM = factor(df_vol$etbr_added_uM)

levels(df_vol$etbr_added_uM) = c("0uM", "10uM", "100uM", "500uM")

ggplot(df_vol %>% filter(treatment == "etbr"), aes(x = etbr_added_uM, 
    y = Measured_Volume, fill = etbr_added_uM)) + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(~phz_added, scales = "free") + 
    scale_fill_viridis_d(guide = F)

Here we can see a few things.

First, the variation in the measurements between groups is pretty high compared to the DNased colonies. One thing to note is that the DNased colonies were measured with a p200, while these colonies were measured with the p1000 and there may just be an accuracy difference involved in the pipette itself or in the process of filling the tip with liquid (e.g. more bubbles?). The reason I used a different pipette was because the etbr treated colonies were relatively hard to resuspend, so I ended up sort of shearing the biofilm material by pipetting with the p1000 and I was able to suck up larger chunks using the larger tips.

Second, it looks like the 500uM etbr condition had smaller measured volumes for all three of the phenazines, although there’s some variability. This sort of makes sense based on morphology, the etbr colonies all had a compact morphology, but they did seem to be quite thick. However, I think there’s something else going on.

One other reason these colonies may have had consistently smaller volumes, is that they were so hard to get off the membrane that I ended up removing the biofilm / membrane post vortex and transferring to a petri dish. Then I physically removed the biofilm from the membrane using sterilized tweezers. When removing the biofilm and membrane I think a significant amount of liquid was lost. Now the question is - would that skew the LC-MS measurement? Possibly.

If most of the phenazine was already extracted upon vortexing and some of that liquid were lost then the measurement would not be affected. If very little of the phenazine was already extracted and most of it was still in the intact colony, then perhaps the loss of volume would actually skew the concentration slightly higher…I think that both of these effects should be pretty small, but importantly I would say it’s inconclusive whether or not the 500uM etbr colonies actually contained smaller biofilm volumes than the other colonies. As I thought before, I think we don’t really have the resolution to say. These measurements are just not very sensitive…also note that all these volumes are below the initial 800uL, because a significant volume of liquid is lost on all the membranes upon removal (even ones that come out clean). It’s unclear if that loss component is even constant…

Let’s move forward to the LC-MS measurements with this in the back of our minds, but ultimately few conclusions drawn.

df_dnase <- read_csv("data/2019_10_22_colony_dnase_data.csv")

df_dnase %>% kable() %>% kable_styling() %>% scroll_box(height = "300px")

# Convert to concentrations

df_dnase_conc <- df_dnase %>% filter(material != "blank") %>% 
    mutate(phz_conc = case_when(material == "cells" ~ Amount * 
        2 * 800/60, material == "agar" ~ Amount * 2 * 8/5))

# formatting levels for plotting

# df_dnase_conc$treatment <- factor(df_dnase_conc$treatment,
# levels = c('none','dnase') ) df_dnase_conc$strain <-
# factor(df_dnase_conc$strain, levels = c('parWT','dPEL'))


# Calculate retentions ratios
df_dnase_ratio <- left_join(df_dnase_conc %>% filter(material == 
    "cells"), df_dnase_conc %>% filter(material == "agar"), by = c("measured_phenazine", 
    "strain", "treatment", "day", "rep"), suffix = c("_cells", 
    "_agar")) %>% mutate(ret_ratio = phz_conc_cells/phz_conc_agar)

# ggplot(df_dnase %>% filter(strain == 'parWT'), aes(x =
# treatment, y = Amount, fill = treatment)) +
# geom_jitter(shape = 21, width = 0.1) +
# facet_wrap(material~measured_phenazine, scales = 'free') +
# ylim(0,NA)

ggplot(df_dnase_conc %>% filter(strain == "parWT"), aes(x = treatment, 
    y = phz_conc, fill = treatment)) + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(material ~ measured_phenazine, 
    scales = "free") + ylim(0, NA)

ggplot(df_dnase_ratio %>% filter(strain == "parWT"), aes(x = treatment, 
    y = ret_ratio, fill = treatment)) + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(~measured_phenazine, scales = "free") + 
    ylim(0, NA)

ggplot(df_dnase_conc %>% filter(strain == "dPEL"), aes(x = treatment, 
    y = phz_conc, fill = treatment)) + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(material ~ measured_phenazine, 
    scales = "free") + ylim(0, NA)

ggplot(df_dnase_ratio %>% filter(strain == "dPEL"), aes(x = treatment, 
    y = ret_ratio, fill = treatment)) + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(~measured_phenazine, scales = "free") + 
    ylim(0, NA)

ggplot(df_dnase_conc %>% filter(treatment == "none"), aes(x = strain, 
    y = Amount, fill = treatment)) + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(material ~ measured_phenazine, 
    scales = "free") + ylim(0, NA)

ggplot(df_dnase_ratio %>% filter(treatment == "none"), aes(x = strain, 
    y = ret_ratio, fill = treatment)) + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(~measured_phenazine, scales = "free") + 
    ylim(0, NA)

df_etbr <- read_csv("data/2019_10_22_colony_etbr_data.csv")

<<<<<<< HEAD
df_etbr %>% kable() %>% kable_styling() %>% scroll_box(height = "300px")

ggplot(df_etbr, aes(x = factor(etbr_added_int), y = Amount, fill = factor(etbr_added_int))) + 
    geom_jitter(shape = 21, width = 0.1) + facet_wrap(~phz_added, 
    scales = "free") + scale_fill_viridis_d()

df_etbr_conc <- df_etbr %>% filter(phz_added != "blank") %>% 
    filter(measured_phenazine == phz_added) %>% mutate(phz_conc = Amount * 
    2 * 800/60) %>% group_by(measured_phenazine, phz_added, etbr_added_uM) %>% 
    mutate(mean = mean(phz_conc))

ggplot(df_etbr_conc, aes(x = factor(etbr_added_int), y = phz_conc, 
    fill = factor(etbr_added_int))) + geom_col(data = . %>% filter(rep == 
<<<<<<< HEAD
    "1"), aes(y = mean), fill = "light gray") + geom_jitter(shape = 21, 
    width = 0.1) + facet_wrap(~phz_added, scales = "free") + 
    scale_fill_viridis_d() + ylim(0, NA)

Conclusions

Very little about these experimental results conformed to my expectations. It’s unclear whether this was due to changes in experimental setup or true variability in the results. It may be worth repeating these experiments again the way that they were done previously.

Specific to the EtBr experiment, it may also be worth trying etbr concentrations lower than 500 or even 100. It’s not ideal that the colonies have a dramatically different morphology / texture…could suggest the etbr is doing something unintended.

If you repeat the etbr experiment, perhaps try to resuspend the tough etbr colonies using the p1000, while the colony is still on the membrane in the tube. This would prevent any difference in processing involved in taking the colony / membrane out of the tube.

sessionInfo()

## R version 3.5.2 (2018-12-20)
## Platform: x86_64-apple-darwin15.6.0 (64-bit)
## Running under: macOS Mojave 10.14.6
## 
## Matrix products: default
## BLAS: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRblas.0.dylib
## LAPACK: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRlapack.dylib
## 
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
##  [1] kableExtra_1.0.1  knitr_1.23        viridis_0.5.1    
##  [4] viridisLite_0.3.0 cowplot_0.9.4     forcats_0.3.0    
##  [7] stringr_1.3.1     dplyr_0.8.1       purrr_0.2.5      
## [10] readr_1.3.1       tidyr_0.8.2       tibble_2.1.3     
## [13] ggplot2_3.2.0     tidyverse_1.2.1  
## 
## loaded via a namespace (and not attached):
##  [1] tidyselect_0.2.5 xfun_0.7         haven_2.0.0      lattice_0.20-38 
##  [5] colorspace_1.4-0 generics_0.0.2   htmltools_0.3.6  yaml_2.2.0      
##  [9] rlang_0.4.0      pillar_1.3.1     glue_1.3.1       withr_2.1.2     
## [13] modelr_0.1.2     readxl_1.2.0     munsell_0.5.0    gtable_0.2.0    
## [17] cellranger_1.1.0 rvest_0.3.2      evaluate_0.14    labeling_0.3    
## [21] highr_0.7        broom_0.5.1      Rcpp_1.0.1       scales_1.0.0    
## [25] backports_1.1.3  formatR_1.5      webshot_0.5.1    jsonlite_1.6    
## [29] gridExtra_2.3    hms_0.4.2        digest_0.6.18    stringi_1.2.4   
## [33] grid_3.5.2       cli_1.1.0        tools_3.5.2      magrittr_1.5    
## [37] lazyeval_0.2.1   crayon_1.3.4     pkgconfig_2.0.2  xml2_1.2.0      
## [41] lubridate_1.7.4  assertthat_0.2.1 rmarkdown_1.13   httr_1.4.0      
## [45] rstudioapi_0.9.0 R6_2.4.0         nlme_3.1-140     compiler_3.5.2