14/08/18

The affect of antibiotics on the ratio of CAP+

At first, the cells grew in KB w/o FeSO4. The cultures treated with Nalidixic acid showed large peak on GFP. Rifampicin seemed to select strongly on CAP-.
hist12.jpghist24.jpg

Pyoverdin could interfere GFP signal. In order to eliminate interference from pyoverdin, cells grew in KB with FeSO4 in the following figures.
This time the ratio of second peak of Nal treated cultures is almost the same as control. Rif again shows a strong selection on CAP-.
antibiotics8hr.jpg
antibiotics16hr.jpg
antibiotics24hr.jpg

p_gfp.jpg

14/08/14 -

Calibration data

Ratio of Cap+ measured from 3 methods: microscope, plating, FACS
Microscope: 1.9% (11/569)
FACS: 10.4%
Plating: IW4 has 4.6% Cap+ colonies, IW4 GFP has 6.1%

Data extracting code for FACS (Matlab)



14/08/12

FACS CSV data

Here's the data on the selection of Cap+/Cap- via centrifugation.
Samples were made from overnight KB culture.
1W4_1 : non GFP
1W4_3 : GFP
3_1~3_6 are centrifugal treated according to the following protocol:
Centrifuge 1ml 1W4_3 at 4K rpm for 1min, gently pipette up and down in the supernate to wash off the Cap+ cells and transfer all the liquid into another eppendorf, labeled as 3_1. The rest of the precipitate (mostly Cap-) is diluted in 1ml fresh medium and labeled as 3_2.
Do the same to 3_1 and 3_2 respectively. The disrupted supernate of 3_1 is labeled as 3_3, diluted precipitate 3_4. And the disrupted supernate of 3_2 is labeled as 3_5, precipitate 3_6.
EXPECTATION: better isolation of Cap- in 3_6 than 3_2, and that of Cap+ in 3_3 than 3_1.
Result: 3_6 as expected. 3_3 still contains ~50% Cap-.



Ideas

1 Different methods to measure switching rates or n+/n-
-- Bubbles
-- Time lapse or snapshot under microscope
-- FACS
-- Plating
2 Environmental effects
-- Temperature and [uracil]
-- The 'capsule'
-- Antibiotics: using Tecan or plating on agar with antibiotics
3 Competition between different strains or CAP+ vs. CAP-



Effect of temperature and uracil on colony structure


Both the concentration of uracil in the medium and the temperature affect the proportion of capsulated cells observed, which is a zero-order proxy for the probability of the CAP- ➤ CAP+ transition. How is the colony structure affected by such parameters? Can we get data about that?

From previous studies it is known that:
  • CAP+ cells grow ~5% faster than the CAP- on agar plates;
  • the percentage of CAP+ cells negatively correlates with the amount of uracil available;
  • the fraction of opaque or sectored colonies negatively correlates with temperature;
  • strain Re1.4 has a higher percentage of CAP+ cells than the 1w4 strain.

Data we are interested in:
  • 1st sector appearance time
  • number of sectors N = N(t)
  • distribution of sectors' length, area and maximum width
  • closed/open sectors ratio

Protocol

3 strains: SBW25 (wild type), 1w4xGFP, Re1.4xGFP
4 uracil concentrations (0, 0.4, 0.8, 1.6 mM, directly on the agar plates)
2 different temperatures (30, 25 °C)
3 replicates

Dilute the night culture 1:10, measure OD (x 5 10^8), that must be between 0.1 to 1 to get significative measurements
Calculate the dilution to have 100 µL, ~50-100 colonies/plate (glass beads)
Incubate the plates at 30° C and at 25° C

Image analysis with Ilastik.

Results

In the following table we present the first counts on the Petri dishes used in the experiment. A first observation about the combined effect of high uracil and high temperature: uracil at 1.6 mM seems to be toxic for 1w4 and Re1.4, especially at 30° C.
Strain
Dilution plated
[uracil]
CFU at 25° C
CFU at 30° C
SBW25
-7
0
90
78
SBW25
-7
0,4
87
96
SBW25
-7
0,8
86
70
SBW25
-7
1,6
90
83
1w4xGFP
-6
0
18
118
1w4xGFP
-6
0,4
40
1
1w4xGFP
-6
0,8
110
1
1w4xGFP
-6
1,6
90
4
Re1.4xGFP
-6
0
28
28
Re1.4xGFP
-6
0,4
22
54
Re1.4xGFP
-6
0,8
62
26
Re1.4xGFP
-6
1,6
15
1

To get insights about the effect on the switching behaviour we took pictures of ~5 colonies/condition for both 1w4 and Re1.4.
40.jpg
1w4xGFP - 25° C - 0

23.jpg
1w4xGFP - 25° C - 0.4 mM

35.jpg
1w4xGFP - 25° C - 0.8 mM

28.jpg
1w4xGFP - 25° C - 1.6 mM


The first qualitative considerations are the following:
  • number, area and length of the CAP+ sectors anticorrelate with the concentration of uracil;
  • colonies are rounder at 30° C than at 25° C: a possible explanation could be in a lower variability in the growth rate;
  • Re1.4 is switches more with respect to 1w4.

The photos were then segmentated with Ilastik and the sector length and width distributions analysed with MATLAB.
28.png
1w4xGFP - 25° C - 1.6 mM uracil

45.png
Re1.4xGFP - 30° C - 0 mM uracil


sector_distribution.png
Distribution of sector length and width over 5 colonies of Re1.4xGFP at 30° C


To improve the quality of the analysis, we could take fluorescent images of the colonies with the Nikon microscope (4x objective).
3-yfp.png
Fluorescent imaging of a colony of 1w4xGFP grown at 30° C

To test this procedure and understand whether it could be a better way to study the system during next weeks, ten colonies were imaged and segmentated with the Nikon software.


nsect.dat.png
Distribution of the number of CAP+ sectors in colonies (N = 13)

results-4.dat.png
Distribution of the fraction of the total area covered by CAP+ sectors (N = 75)


Spatial model

g_rate_0.51_on_rate_0.001.png
g_rate0=0.51, g_rate1=0.49, on_rate=0.001, off_rate=0.001, start=0.95

g_rate_0.505_on_rate_0.0001.png
g_rate0=0.505, g_rate1=0.495, on_rate=0.0001, off_rate=0.001, start=0.8

g_rate_0.5_on_rate_0.001.png
g_rate_0.504_on_rate_0.001.png











Model expectations: effect of parameters on distribution of sector length and width

summary.png
Distribution of sector length and width as a function of off rate, on rate, initial cap- proportion, relative growth rate


Model expectations: effect of parameters on average number of sectors

number_sectors_start_0.9_g_0.5.png
start=0.9, g_rate=1
number_sectors_start_0.9_g_0.6.png
start=0.9, g_rate=1.5
number_sectors_start_1_g_0.5.png
start=1,g_rate=1


Powerpoint presentation



Transient dynamics of switchers


Strain 1w4 switches between a CAP- phenotype (no capsule) and a CAP+ phenotype (thick capsule). One possible way of testing this could be to measure the fraction of cells f(t) over some time from an initial condition f(0) and use the transients observed to inform the model.

Protocol

  1. Grow O/N of 1W4 strain at 30C to minimize number of CAP+ cells. Take concentration of cells and quantify fraction using FACS
  2. Freeze down 20 of these in 1ml tubes
  3. Seed 10^7 cells / ml in 50ml Kings B switcher media in 250ml conical flask
  4. Wait 45 mins. Until ~2 x 10^7 cells / ml
  5. FACS the cells. Measure both fraction of CAP+ and the total concentration
  6. Take enough cells to seed a new flask with 10^7 cells / ml

Repeat 10 times, grow O/N cultures in King's B switcher medium.

Switching and growth rates


The parameters of the model (s, p, rA, rB) can be measured through single-cell methods. Data on GFP fluorescent versions of the 7 switching strains can be obtained via time-lapse microscopy or through the droplets millifluidics device.

Fluctuation test with colonies


MIC growth curves


Two strains, E. coli B and an unknown strain containing a GFP marker, were tested in serial dilutions of four antibiotics: ampicillin, gentamycin, chloramphenicol and nalidixic acid. MICs were estimated and growth curves obtained at two sub-MIC antibiotic concentrations for the latter three antibiotics (both strains turned out ampicillin resistant, which is suspicious for now).

Super-MIC effect on phenotypic switching


The switching P. fluorescens strains are resistant to gentamycin via a gene in their chromosomal DNA. Anyway, very high concentrations of that AB are lethal. We aim at exploring the effect of high concentrations of gentamycin on cell growth and switching.


Articles


Leiber & Kussell, PNAS (2010)
Norman et al., Nature (2013)

Useful protocols

  • simple protocol for live cell imaging
  • King's B (KB) medium recipe
  • night cultures: 10 µg of gentamycin for 1 mL of final volume