Live-Cell Forward Genetic Approach to Identify and Isolate Developmental Mutants in Chlamydia trachomatis

All Python scripts used in this protocol are available on Github https://github.com/SGrasshopper/Live-cell-data-processing

1. Mutagenize Reporter  Chlamydia

NOTE: Ctr-L2-hctBprom-mKate2/euoprom-Clover EBs were directly mutagenized using ethyl methanesulfonate (EMS) in the axenic media CIP-1 as this media supports EB metabolism and maintenance of EB infectivity¹².

1. Thaw a chlamydial stock on ice containing ~3 x 10⁷ EBs transformed with the p2TK2-hctBprom-mKate2/euoprom-Clover reporter plasmid and pellet at >14,000 x g for 30 min at 4 ˚C.
   NOTE: Chlamydia organisms used for these experiments were 30% renografin density purified and frozen at -80 °C in 1x sucrose-phosphate-glutamate buffer (SPG).
2. Discard the supernatant and resuspend the EB pellet in 100 µL of CIP-1 buffer with sonication on ice at 10% power for 10 s. Divide the 100 µL of EB suspension into two 50 µL aliquots for mutagenized and mock treated samples.
3. Prepare 20 mg/mL of EMS-CIP-1 solution in a separate 1.5 mL microcentrifuge tube. To do so, add 6.8 µL of EMS in 375 µL total volume.
4. Add 50 µL of the EMS-CIP-1 solution into one of the chlamydial aliquots for mutagenesis and 50 µL of CIP-1 only to the other chlamydial aliquot for mock mutagenesis.
   NOTE: Final EMS concentration is 10 mg/mL. The chlamydial titer, EMS concentration, and the time of exposure used in this protocol lead to approximately a 60-80% reduction in infectious progeny. This level of reduction corresponds to ~5-20 DNA lesions per chlamydial genome⁸.
5. Incubate for 20 min at room temperature. The mutagenized EBs will be used directly to infect monolayers in section 2.
   CAUTION: EMS is a known carcinogen. All equipment and materials that come in contact with EMS must be soaked in 1 M NaOH for 24 h before disposal, gloves should be used at all times during the protocol and cleanup of EMS materials.

2. Imaging of mutant Ctr

1. Host cell culture for imaging and isolation of mutagenized Ctr
   1. Seed a 6 well glass bottom plate with 6 x 10⁵ Cos-7 cells (ATCC) per well in 2 mL of complete media (RPMI-1640 supplemented with 10% fetal bovine serum and 10 mg/mL gentamicin). Use this glass bottom plate for imaging of mutagenized Ctr.
   2. Seed a 24 well polystyrene plate with 1 x 10⁵ Cos-7 cells (ATCC) per well in 1 mL complete media. Use this polystyrene plate for reinfection of isolated Chlamydia of interest.
   3. Incubate both the plates at 5% CO₂, 37 °C for approximately 18 h. Once cells reach confluency, replace media with complete media supplemented with 1 µg/mL of cycloheximide and incubate overnight.
2. Infecting the host cell culture with mutagenized Ctr
   1. Infect 5 wells of the glass bottom plate with ~6 x 10⁵ of mutagenized EBs in 1.5 mL/well ice cold HBSS. This will result in the MOI of ~ 0.3 as ~70% mortality rate is expected due to mutagenesis.
   2. Infect the remaining well with ~2 x 10⁵ mock mutagenized EBs in 1.5 mL/well ice cold HBSS. Without mutagenesis, expect less mortality, thus one-third of the inoculum is used to achieve the MOI of ~0.3.
      NOTE: MOI of ~0.3 ensures that host cells are infected by a single EB and allows for the enough separation between infected cells for clonal isolation.
   3. Incubate the plate for 15 min, with rocking, at 37 ˚C.
   4. Wash the infected host cells with prewarmed (37 ˚C) HBSS containing 1 mg/mL heparin followed immediately by an HBSS rinse. Repeat heparin wash, immediately rinsing 2x with HBSS to ensure the heparin is removed.
      NOTE: Heparin inhibits and can reverse the early electrostatic interactions between the host cell and EBs¹³. The heparin washes remove EBs that have yet to enter the host cells, synchronizing the infection. When washing cells do so gently to prevent dislodging the cells from the surface of the wells. HBSS and heparin solutions contain residual EMS and should be placed in a beaker containing 1 M NaOH for 24 h before disposal.
   5. Replace HBSS with 4 mL/well of prewarmed (37 ˚C) imaging media (complete media, 1 µg/mL cycloheximide, 20 mM HEPES, and no phenol red).
   6. Fill the interwell spaces with prewarmed (37 ˚C) deionized H₂O to aid in the temperature control and reduce evaporation. Incubate the plate at 37 °C incubator with 5% CO₂ for 10 h.
3. Microscope set up and imaging
   NOTE: Multicolor multiposition automated live-cell fluorescent imaging is used to collect time-lapse images to identify chlamydial mutants that differ in the developmental gene expression dynamics. This protocol utilizes the open source µManager software package for automated microscope control¹⁴.
   1. Begin the microscope setup 10 h post infection. Set the microscope stage incubator to 5% CO₂, 37 ˚C. Place the infected 6 well glass bottom plate into the stage incubator and insert the sample thermistor into the interwell H₂O.
   2. Calibrate the XY stage using the High Content Screening (HCS) plugin. Click Plugins | Acquisition Tools | HCS Site Generator in the µManager microscope control software (JoVE61365_screenfile1, JoVE61365_screenfile2).
   3. Select the 6-well plate template and generate an imaging position list consisting of 12 fields of view (FOV) per well within the HCS plugin (JoVE61365_screenfile3, JoVE61365_screenfile4). Open the Stage Position List and manually focus and set the initial Z position for each FOV using the Stage Control plugin (JoVE61365_screenfile5, JoVE61365_screenfile6). Adjust the XY coordinates of any FOV that has missing cells or does not contain a uniform monolayer.
      NOTE: Due to the time it takes for each image to be captured, a maximum number of 72 FOV can be taken per 30 min interval.
   4. Use a 20x objective lens for imaging. This magnification allows ~8 inclusions to be imaged per FOV while still providing the desired resolution.
   5. Save the positions list as this will be used to locate the inclusions of interest after data analysis (JoVE61365_screenfile7).
   6. Use the Auto Focus option in the imaging software, to set the focus for automated imaging (JoVE61365_screenfile8).
   7. Use the following selections and values to produce the most consistent focus results using image based autofocus in µManager. In the Autofocus properties window select OughtaFocus from the drop-down menu and use the following settings. OughtaFocus-SearchRange_µm: 350, OughtaFocus-Tolerance_µm: 0.5, OughtaFocusCropFactor: 0.3, OughtaFocus-Exposure: 20, OughtaFocus-FFTLowerCutoff(%): 2.5, OughtaFocus-FFTUpperCutoff(%): 14, OughtaFocus-ShowImages: Yes, OughtaFocus-Maximize: SharpEdges, OughtaFocus-Channel: DIC.
      NOTE: Reducing the autofocus imaging window by decreasing the crop factor allows for more consistent auto focusing. Selecting Yes for OughtaFocus-ShowImages allows the user to view the autofocus image.
   8. Capture the kinetics of the developmental cycle by imaging for 24 h with 30 min time intervals (JoVE61365_screenfile9). Imaging between 12-36 HPI ensures that Chlamydia completes the developmental cycle but does not lyse the host cell.
   9. Image the cell monolayers with a 250 ms exposure at 4% and 18% intensity in the GFP and RFP channels, respectively (JoVE61365_screenfile10). Detect the Clover (GFP) signal by excitation at 470 nm with a 514/30 nm bandpass emissions filter. Detect the mKate2 (RFP) signal by excitation at 595 nm and with a 641/75 nm bandpass emissions filter.
      NOTE: Minimizing the excitation intensity is critical for minimizing fluorophore photobleaching and phototoxicity to Chlamydia. The minimum excitation intensity to generate a resolved image with a 200-300 ms exposure should be determined empirically in pilot studies.
   10. Capture multiple Z-slices with a range of focus that ends on either side of the in-focus slice. In this experiment, at 20x magnification, this was achieved with 4 slices at 10 µm steps (JoVE61365_screenfile11).
   11. Select Relative Z for imaging multiple slices in the acquisition window. The relative Z option uses the Z plane location saved in the imaging position list as the starting point for the next time interval. Input the appropriate Z-offset values for fluorescence imaging channels (JoVE61365_screenfile12).
       NOTE: The image based focusing system is imperfect and over a 24 h imaging period this leads to focus drift. It was found that by capturing 3-4 Z focal planes for each time point an in-focus image was maintained. Z-offset is needed to correct for the differences in focal planes between the fluorescent image channels and the DIC channel. Empiric determination of this Z-offset will be needed.
   12. Save images by selecting the root directory and naming the experiment. Use the µManager image stack file option to save images as tiff stack files (JoVE61365_screenfile13).
   13. Record the experimental details in the Acquisitions Comments box in the Multi-D Acquisition window (JoVE61365_screenfile13). i.e., Well A1: Untreated control. Well A2-3 and B1-3: EMS Mutants. Imaging: 12-36HPI. Start the image acquisition at 12 HPI.
       NOTE: If using µManager, leave the program, experimental setup, and microscope hardware running after the experiment is complete. The imaging sites will be revisited for inclusion isolation once analysis of the mutagenized population is completed.

3. Identify and isolate mutagenized  Chlamydia with altered developmental phenotypes

1. Creating an in-focus image stack
   1. Extract the most in-focus image from the Z-stacks using the saved image data which contains 4 Z slices per time point. Use the kurtosis measurement option (Analyze | Measure) in ImageJ/FIJI to automatically identify the most in focus Z slice (highest kurtosis score) and create a new image stack with just these in-focus images. A Python script is included as a supplementary file (Reduce_Z_kertosis_2ch_JOVE.py) to automate this process.
2. Quantify fluorescence expression in individual inclusions
   NOTE: To quantify the expression kinetics of the two reporters use the open source image analysis application ImageJ/FIJI and the plugin Trackmate¹⁵. Trackmate identifies ‘spots’ (corresponding to inclusions in this case) and follows them through a time-lapse image stack recording the X,Y location and signal intensity for each inclusion over time (JoVE61365_screenfile14). This information is saved as a CSV file and will be imported into a custom Python notebook for analysis.
   1. Open the in-focus Z-reduced image stack in ImageJ/FIJI using Bio-formats Importer by clicking Plugins | Bio-Formats | Bio-formats Importer. Select Hyperstack and Composite (JoVE61365_screenfile15, JoVE61365_screenfile16).
   2. Subtract the image background by clicking on Process | Subtract Background using a Rolling ball radius of 50.0 pixels and enhance the image contrast to 0.3% for Saturated pixels (Process | Enhance Contrast). Multiply the image values by 10.0 (Process | Math | Multiply) (JoVE61365_screenfile17 – JoVE61365_screenfile22).
   3. In Trackmate (Plugins | Tracking | Trackmate), select an estimated blob diameter of 48 pixels (empirically determined based on the size of the inclusion at the end of imaging). Produce non-fragmented inclusion tracks by selecting a Linking max distance and Gap-closing max distance of 8.0 pixels and Gap-closing max frame gap of 1 (JoVE61365_screenfile23 – JoVE61365_screenfile25).
      NOTE: To improve Trackmate’s ability to identify and track inclusions over the entire cycle a separate image channel is created by adding the euoprom and hctBprom channels together using the image math function in ImageJ/FIJI. This channel is then used by Trackmate to identify and follow inclusions over time. The fluorescent values of the euoprom and hctBprom channels are then recorded in channels 2 and 3.
   4. Record tracks that meet a minimum continuous duration: Duration of track: 20 (JoVE61365_screenfile26). Analyze the tracks and save the Spots in track statistics as a CSV file (JoVE61365_screenfile27).
      NOTE: This process has been automated using a custom Python script that is provided in the supplemental data (TrackMate_Zreduced_JOVE.py).
3. Identify inclusion tracks with altered developmental profiles
   NOTE: To identify inclusions containing Chlamydia with altered developmental profiles, each inclusion track was visualized using Python notebook. These visualizations allow for the identification of inclusions with kinetic gene expression profiles that differed from the mock-treated population. The Python notebook used for identification of inclusions with altered developmental programming is provided in the supplemental data (EMS_Screen-Markdown).
   1. Import the inclusion track data from the Spots in tracks statistics CSV files into Pandas data frame using the Import cell in the EMS_Screen-Markdown Python Notebook.
   2. Baseline correct each track by subtracting the minimum value of each track from the rest of the track values using the Baseline Subtract cells. Save the resulting values as a pickle file using the Save as Pickle cell. This will permanently save the channel values after baseline subtraction for later retrieval.
      NOTE: Baseline subtraction sets the starting fluorescent intensity of every inclusion to zero.
   3. Eliminate traces from inclusions near the edges of the FOV using the Filter Edges cell as their fluorescent profiles may not be fully captured; these traces may produce false positive developmental profiles.
   4. Calibrate the frame (totalFrames) values from the image slices to time (startTime, interval) values with the Time-lapse Calibration cell using the experimental start time and imaging time interval. Exclude partial inclusion reads by filtering out traces that do not extend over the last 20 h of the experiment (16-36 HPI) using the Track Duration Filter cell.
   5. Separate tracks into individual data frames by experimental condition with the Assign Treatment cell. Filter for inclusions that exhibit sufficient growth using the Filter for Growth cell. In the Filter for Growth cell, empirically set the fluorescence intensity threshold for the euoprom channel by changing the values within the last two lines of code. This will filter out Chlamydia that did not grow.
      NOTE: Be cautious when setting threshold filters, if the filter is too low the resulting scatter plots will be noisy, yet if filtering is set too high important mutants may be eliminated.
   6. Calculate the percent mortality caused by EMS by dividing the number of mutagenized tracks/well by the number of mock-treated tracks/well. Multiply the number of mock-treated tracks by the initial dilution factor of 3 to calculate the number of mock-treated tracks/well. Use the Count Inclusion Tracks and Calculate Percent Mortality cells to perform this task.
   7. Calculate the time to half-maximal expression for both early and late reporters for each track using the Calculate cell. These values will be used to compare mutagenized and mock populations to identify developmental mutants.
   8. Within the Half-Max Plot cell use the bokeh plotting package to visualize the time to half-max expression of each promoter, graphing the euoprom time to half-maximal expression against that of hctBprom. Identify inclusions from the mutant population that fall outside the mock-treated scatter cloud using the bokeh interactive track ID explorer. Make note of the FOV and XY coordinates of the inclusions of interest (Figure 2).
      NOTE: Pick candidate inclusions that visually fall outside of the control cloud as verification and statistical evaluation of each clone will be performed subsequently.
   9. Within the Animated Plot cell visualize changes in promoter expression kinetics dynamically through time by graphing the expression intensities of euoprom against hctBprom using the plotting tool Plotly¹⁶. The scatter function of Plotly is used to animate the gene expression over time (Video 1).
   10. Visualize a snapshot from the Plotly animated graph in the Inclusion Locator cell, plotting euoprom and hctBprom expression at a specific time point (i.e. 28 HPI) using the bokeh package (Figure 3). Identify inclusions from the mutant population as described in step 3.3.8.
       NOTE: This analysis needs to be performed quickly (<4 h) as the inclusions are still expanding and will start to lyse host cells ~48 h after infection.
4. Isolate developmental mutants from inclusions of interest
   NOTE: To isolate Chlamydia from the inclusions that were determined to display altered gene regulation a micromanipulator with capillary needles was employed. The Well ID, FOV and X,Y coordinates of inclusions of interest were determined using the data visualization in section 3.3.
   1. Prepare capillary needles by holding the center of the capillary tube in a flame and pulling both ends of the capillary tube until it has separated. Create an opening in the pulled capillary needle by breaking the pulled tip on a microscope slide. To break the needle, place the closed tip on the frosted portion of the microscope slide at an angle and apply pressure.
      NOTE: Capillary tube specifications were 1.0 mm O.D., 0.5 mm I.D.
   2. Check that the needle opening is approximately the size of an inclusion under a microscope, 20x objective.
   3. Prepull ~25-30 capillary needles for isolation of candidate inclusions (one needle per inclusion).
   4. Fill the microinjector with mineral oil ensuring that no air bubbles are present.
   5. Attach a glass capillary needle to the microinjector and expel oil to the tip of the needle, expunging any air bubbles. Place the capillary needle in complete media and draw media up halfway. Filling the capillary needle with media prevents oil contamination in the well.
   6. Using the saved position list in µManager from step 2.3.5, migrate to the well and FOV of an inclusion of interest identified in the Python visualization notebook.
   7. Use the joystick of the micromanipulator to localize the capillary needle to the XY coordinates of the inclusion of interest.
   8. Use the 595 nm excitation channel to visualize EBs for extraction and the phase/DIC white light channel for needle visualization. Maneuver the capillary needle to the inclusion, rupture the inclusion and then draw the EBs into the capillary needle using the microinjector.
   9. Expel the EBs from the capillary needle into a single well of the prepared 24 well polystyrene plate prepared in step 2.1.2. Remove the capillary needle and replace with a fresh capillary needle for next inclusion extraction. Repeat section 3.4 for all candidate inclusions.
      NOTE: For expansion and to ensure a high enough titer for re-imaging, incubate mutant isolates in a 5% CO₂, 37 ˚C incubator until the majority of the host cells are infected (~1 week). Wells should be monitored closely as different isolates may exhibit different growth rates.
5. Harvest mutant isolates
   1. On ice, disrupt the infected monolayer by scraping with a 1 mL micropipette tip. Transfer the media, cell debris, and released Chlamydia into a 1.5 mL microcentrifuge tube.
   2. Pellet Chlamydia by centrifugation for 30 min at 4 ˚C, >14,000 x g. Remove the supernatant and resuspend pellet in 75 µL of ice cold 1x SPG. Aliquot into three 1.5 mL screw-cap microcentrifuge tubes. Store at -80 ˚C.

4. Verification of mutant isolate phenotypes

1. Host cell culture for imaging mutagenized isolates
   1. Seed a 96 well glass bottom plate with 1.6 x 10⁴ Cos-7 cells (ATCC) per well in 100 µL of complete media. Incubate at 5% CO₂, 37 °C. Cells should reach confluency in approximately 24 h. After cells are confluent, replace media with complete media supplemented with 1 µg/mL cycloheximide, incubate overnight.
2. Infect cells with candidate isolates for phenotypic verification
   1. Thaw mutant clones and wildtype Chlamydia on ice.
   2. In the prepared 96 well plate, perform a two-fold serial dilution of mutant isolates, using one column per isolate (11 columns). Start with an initial dilution of 1:20 in 100 µl HBSS.
      NOTE: Serial dilution of Chlamydia is performed to ensure mutant samples are imaged at an MOI < 1.
   3. Infect the remaining (12th) column with wildtype Chlamydia at MOI ~0.5.
      NOTE: Wildtype Chlamydia are used as a control for comparison against mutagenized isolates.
   4. Incubate for 15 min rocking at 37 °C.
   5. Wash infected host cells with prewarmed (37 °C) HBSS with 1 mg/mL of heparin and HBSS as specified in section 2.2.
   6. Replace with 200 µL per well of prewarmed (37 °C) imaging media.
   7. Fill the interwell spaces with prewarmed (37 °C) deionized H₂O.
   8. Incubate at 5% CO₂, 37 °C for 10 h.
3. Microscope setup
   NOTE: Refer to section 2.3 for microscope setup, this section will only contain the required setup modifications.
   1. Select the 96 well plate template from the HCS plugin.
   2. Empirically determine wells corresponding to an MOI < 1 for each mutant isolate. Clover expression under control of the euo promoter is observable at ~10 HPI making early visualization of inclusions possible (Figure 1B).
   3. Select three wells per mutant isolate that correspond to an MOI < 1 and generate an imaging position list consisting of two FOV per well.
      NOTE: Only 72 images can be taken per time interval due to hardware constraints, this equates to three dilutions (wells) per strain using two imaging sites per well if 12 samples are imaged.
   4. Record the developmental cycle of each mutant isolate for 36 h at 30 min time intervals starting at 12 HPI.
   5. Record the experimental details in the Acquisitions Comments box. i. e., Well ABC1: wildtype control. Well ABC2: Mutant strain 1, ABC3: Mutant strain 2, etc… Imaging: 12-48 HPI.
   6. Start the image acquisition at 12 HPI.

5. Data analysis for isolate verification

1. Create in-focus image stacks and quantify fluorescence expression in individual inclusions
   1. Generate fluorescent intensity traces for each inclusion as specified in section 3.1 – 3.2.
2. Verify Ctr mutagenized isolates
   NOTE: To verify the altered developmental profiles of mutant isolates, their expression profiles are compared to the wildtype expression profile using Python notebook. The Python notebook used for verification of mutant clones with altered developmental programming is provided in the supplemental data (clone_check-Markdown).
   1. Import and filter the inclusion trace data in the clone_check-Markdown Python notebook as done in section 3.3.
   2. Calculate the mean and standard deviation (STD) from the traces of each isolate and wildtype control population using the Calculate Mean & STD cell.
   3. With the Graph Iso vs WT cell plot the mean and standard error of the mean (SEM) of each mutant clone against the wildtype control to determine if the mutant expression kinetics are divergent from the wildtype sample (Figure 4).
   4. Determine if the isolated mutant population is clonal by plotting the mutant traces and comparing them to wildtype inclusion traces using a scatter plot as done in section 3.3 (steps 3.3.7-3.3.10) (Figure 2, Figure 3). If the isolate is a mixed population the plot will show one population overlaying with wildtype and a second distinct population outside of the wildtype scatter cloud. If the population looks mixed the mutant can be re-isolated using the original procedure described in section 3.4.
      NOTE: To determine if the developmental profile of an isolate is statistically different from wildtype the curves for each isolate should be compared to wildtype using ANOVA.